The Purple Phototrophic Bacteria
Advances in Photosynthesis and Respiration VOLUME 28 Series Editor : GOVINDJEE University of Illinois, Urbana, Illinois, U.S.A.
Consulting Editors: Julian EATON-RYE, Dunedin, New Zealand Christine H. FOYER, Newcastle upon Tyne, U.K. David B. KNAFF, Lubbock, Texas, U.S.A. Anthony L. MOORE, Brighton, U.K. Sabeeha MERCHANT, Los Angeles, California, U.S.A. Krishna NIYOGI, Berkeley, California, U.S.A. William PARSON, Seatle, Washington, U.S.A. Agepati RAGHAVENDRA, Hyderabad, India Gernot RENGER, Berlin, Germany
The scope of our series, beginning with volume 11, reflects the concept that photosynthesis and respiration are intertwined with respect to both the protein complexes involved and to the entire bioenergetic machinery of all life. Advances in Photosynthesis and Respiration is a book series that provides a comprehensive and state-of-the-art account of research in photosynthesis and respiration. Photosynthesis is the process by which higher plants, algae, and certain species of bacteria transform and store solar energy in the form of energy-rich organic molecules. These compounds are in turn used as the energy source for all growth and reproduction in these and almost all other organisms. As such, virtually all life on the planet ultimately depends on photosynthetic energy conversion. Respiration, which occurs in mitochondrial and bacterial membranes, utilizes energy present in organic molecules to fuel a wide range of metabolic reactions critical for cell growth and development. In addition, many photosynthetic organisms engage in energetically wasteful photorespiration that begins in the chloroplast with an oxygenation reaction catalyzed by the same enzyme responsible for capturing carbon dioxide in photosynthesis. This series of books spans topics from physics to agronomy and medicine, from femtosecond processes to season long production, from the photophysics of reaction centers, through the electrochemistry of intermediate electron transfer, to the physiology of whole organisms, and from X-ray crystallography of proteins to the morphology or organelles and intact organisms. The goal of the series is to offer beginning researchers, advanced undergraduate students, graduate students, and even research specialists, a comprehensive, up-to-date picture of the remarkable advances across the full scope of research on photosynthesis, respiration and related processes. For other titles published in this series, go to www.springer.com/series/5599
The Purple Phototrophic Bacteria Edited by
C. Neil Hunter University of Sheffield, United Kingdom
Fevzi Daldal University of Pennsylvania, USA
Marion C. Thurnauer Argonne National Laboratory, USA and
J. Thomas Beatty University of British Columbia, Canada
Library of Congress Control Number: 2008932524
ISBN 978-1-4020-8814-8 (HB) ISBN 978-1-4020-8815-5 (e-book) Published by Springer, P.O. Box 17, 3300 AA Dordrecht, The Netherlands. www.springer.com
Cover: Four aspects of purple phototrophic bacteria, from one of their habitats through to atomic resolution structures, are superimposed on a map derived from the genome sequence of Rhodopseudomonas palustris CGA009 supplied by Professor Caroline Harwood, University of Washington, Seattle, USA. Top left. Purple sulfur bacteria (Amoebobacter purpureus) on the shoreline of Mahoney Lake, British Columbia, Canada. Image from Professor J.T. Beatty. Top right. Rhodobacter capsulatus streaked out on an agar plate. Image from Professor J.T. Beatty. Bottom right. Model of a spherical chromatophore vesicle from Rhodobacter sphaeroides constructed by the in silico combination of atomic force microscopy, linear dichroism, electron microscopy, and X-ray crystallography data. Image from Dr. Melih Sener and Professor Klaus Schulten, prepared using VMD (Humphrey et al. (1996) J Mol Graphics 14: 33–38). Bottom left. Structure of the reaction center complex from Rhodobacter sphaeroides showing the subunits and the pathway of electron transfer between cofactors. See Fig. 1, Chapter 20. Image from Professor Colin Wraight, prepared using VMD. The camera ready text was prepared by Lawrence A. Orr, Center for Bioenergy & Photosynthesis, Arizona State University, Tempe, Arizona 85287-1604, USA.
Printed on acid-free paper
All Rights Reserved © 2009 Springer Science + Business Media B.V. No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work.
This book is dedicated to Roderick K. Clayton, a pioneer of photosynthesis research
From the Series Editor Advances in Photosynthesis and Respiration Volume 28: The Purple Phototrophic Bacteria I am delighted to announce the publication, in the Advances in Photosynthesis and Respiration (AIPH) Series, of The Purple Phototrophic Bacteria. Four distinguished authorities from three countries (UK, USA and Canada) have edited this Volume: C. Neil Hunter, the Chief Editor, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty. This book is produced as a sequel to Volume 2 of the Series (Anoxygenic Photosynthetic Bacteria), published in 1995, and edited by Robert E. Blankenship, Michael T. Madigan and Carl E. Bauer. Published Volumes (2008–1994) • Volume 27 (2008) Sulfur Metabolism in Phototrophic Organisms, edited by Christiane Dahl, Rüdiger Hell, David Knaff and Thomas Leustek, from Germany and USA. 24 Chapters, 551 pp, Hardcover. ISBN: 978-4020-6862-1 • Volume 26 (2008): Biophysical Techniques In Photosynthesis, Volume II, edited by Thijs Aartsma and Jörg Matysik, both from The Netherlands. 24 Chapters, 548 pp, Hardcover. ISBN: 978-1-4020-8249-8 • Volume 25 (2006): Chlorophylls and Bacteriochlorophylls: Biochemistry, Biophysics, Functions and Applications, edited by Bernhard Grimm, Robert J. Porra, Wolfhart Rüdiger, and Hugo Scheer, from Germany and Australia. 37 Chapters, 603 pp, Hardcover. ISBN: 978-140204515-8 • Volume 24 (2006): Photosystem I: The LightDriven Plastocyanin: Ferredoxin Oxidoreductase, edited by John H. Golbeck, from USA. 40 Chapters, 716 pp, Hardcover. ISBN: 978-140204255-3 • Volume 23 (2006): The Structure and Function of Plastids, edited by Robert R. Wise and J. Kenneth Wise, from USA. 27 Chapters, 575 pp, Softcover: ISBN: 978-1-4020-6570-6, Hardcover, ISBN: 978-1-4020-4060-3
• Volume 22 (2005): Photosystem II: The LightDriven Water:Plastoquinone Oxidoreductase, edited by Thomas J. Wydrzynski and Kimiyuki Satoh, from Australia and Japan. 34 Chapters, 786 pp, Hardcover. ISBN: 978-1-4020-4249-2 • Volume 21 (2005): Photoprotection, Photoinhibition, Gene Regulation, and Environment, edited by Barbara Demmig-Adams, William W. III Adams and Autar K. Mattoo, from USA. 21 Chapters, 380 pp, Hardcover. ISBN: 97814020-3564-7 • Volume 20 (2006): Discoveries in Photosynthesis, edited by Govindjee, J. Thomas Beatty, Howard Gest and John F. Allen, from USA, Canada and UK. 111 Chapters, 1304 pp, Hardcover. ISBN: 978-1-4020-3323-0 • Volume 19 (2004): Chlorophyll a Fluorescence: A Signature of Photosynthesis, edited by George C. Papageorgiou and Govindjee, from Greece and USA. 31 Chapters, 820 pp, Hardcover. ISBN: 978-1-4020-3217-2 • Volume 18 (2005): Plant Respiration: From Cell to Ecosystem, edited by Hans Lambers and Miquel Ribas-Carbo, from Australia and Spain. 13 Chapters, 250 pp, Hardcover. ISBN: 978-14020-3588-3 • Volume 17 (2004): Plant Mitochondria: From Genome to Function, edited by David Day, A. Harvey Millar and James Whelan, from Australia. 14 Chapters, 325 pp, Hardcover. ISBN: 978-14020-2399-6 • Volume 16 (2004): Respiration in Archaea and Bacteria:Diversity of Prokaryotic Respiratory Systems, edited by Davide Zannoni, from Italy. 13 Chapters, 310 pp, Hardcover. ISBN: 97814020-2002-5 • Volume 15 (2004): Respiration in Archaea and Bacteria: Diversity of Prokaryotic Electron Transport Carriers, edited by Davide Zannoni, from Italy. 13 Chapters, 350 pp, Hardcover. ISBN: 978-1-4020-2001-8 • Volume 14 (2004): Photosynthesis in Algae,
edited by Anthony W. Larkum, Susan Douglas and John A. Raven, from Australia, Canada and UK. 19 Chapters, 500 pp, Hardcover. ISBN: 978-0-7923-6333-0 • Volume 13 (2003): Light-Harvesting Antennas in Photosynthesis, edited by Beverley R. Green and William W. Parson, from Canada and USA. 17 Chapters, 544 pp, Hardcover. ISBN: 97807923-6335-4 • Volume 12 (2003): Photosynthetic Nitrogen Assimilation and Associated Carbon and Respiratory Metabolism, edited by Christine H. Foyer and Graham Noctor, from UK and France. 16 Chapters, 304 pp, Hardcover. ISBN: 978-07923-6336-1 • Volume 11 (2001): Regulation of Photosynthesis, edited by Eva-Mari Aro and Bertil Andersson, from Finland and Sweden. 32 Chapters, 640 pp, Hardcover. ISBN: 978-0-7923-6332-3 • Volume 10 (2001): Photosynthesis: Photobiochemistry and Photobiophysics, by Bacon Ke, from USA. 36 Chapters, 792 pp, Softcover: ISBN: 978-0-7923-6791-8. Hardcover: ISBN: 978-0-7923-6334-7 • Volume 9 (2000): Photosynthesis: Physiology and Metabolism, edited by Richard C. Leegood, Thomas D. Sharkey and Susanne von Caemmerer, from UK, USA and Australia. 24 Chapters, 644 pp, Hardcover. ISBN: 978-0-7923-6143-5 • Volume 8 (1999): The Photochemistry of Carotenoids, edited by Harry A. Frank, Andrew J. Young, George Britton and Richard J. Cogdell, from UK and USA. 20 Chapters, 420 pp, Hardcover. ISBN: 978-0-7923-5942-5 • Volume 7 (1998): The Molecular Biology of Chloroplasts and Mitochondria in Chlamydomonas, edited by Jean David Rochaix, Michel Goldschmidt-Clermont and Sabeeha Merchant, from Switzerland and USA. 36 Chapters, 760 pp, Hardcover. ISBN: 978-0-7923-5174-0 • Volume 6 (1998): Lipids in Photosynthesis: Structure, Function and Genetics, edited by Paul-André Siegenthaler and Norio Murata, from Switzerland and Japan. 15 Chapters, 332 pp, Hardcover. ISBN: 978-0-7923-5173-3 • Volume 5 (1997): Photosynthesis and the Environment, edited by Neil R. Baker, from UK. 20 Chapters, 508 pp, Hardcover. ISBN: 97807923-4316-5 • Volume 4 (1996): Oxygenic Photosynthesis: The Light Reactions, edited by Donald R. Ort,
and Charles F. Yocum, from USA. 34 Chapters, 696 pp, Softcover: ISBN: 978-0-7923-3684-6. Hardcover: ISBN: 978-0-7923-3683-9 • Volume 3 (1996): Biophysical Techniques in Photosynthesis, edited by JanAmesz and Arnold J. Hoff, from The Netherlands. 24 Chapters, 426 pp, Hardcover. ISBN: 978-0-7923-3642-6 • Volume 2 (1995): Anoxygenic Photosynthetic Bacteria, edited by Robert E. Blankenship, Michael T. Madigan and Carl E. Bauer, from USA. 62 Chapters, 1331 pp, Hardcover. ISBN: 978-0-7923-3682-8 • Volume 1 (1994): The Molecular Biology of Cyanobacteria, edited by Donald R. Bryant, from USA. 28 Chapters, 916 pp, Hardcover. ISBN: 978-0-7923-3222-0 Further information on these books and ordering instructions can be found at
under the Book Series ‘Advances in Photosynthesis and Respiration.’ Table of Contents of Volumes 1–25 can be found at . Special discounts are available to members of the International Society of Photosynthesis Research, ISPR (). (See http://www.springer.com/ispr for more information.) About Volume 28: The Purple Phototrophic Bacteria The Purple Phototrophic Bacteria has 48 authoritative Chapters, and is authored by 116 international authorities from 13 countries (Australia; Canada; Czech Republic; France; Germany; Israel; Italy; Japan; Netherlands; Poland; Russia; UK; and USA). It is a truly international book and the Chief Editor of this volume, C. Neil Hunter, and his three coeditors, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty, deserve our thanks and our congratulations for compiling this updated survey of these interesting and important organisms. The Purple Phototrophic Bacteria is a comprehensive survey of all aspects of these fascinating bacteria, most metabolically versatile organisms on Earth. This volume is organized into the following sections: Physiology, Evolution and Ecology; Biosynthesis of Pigments, Cofactors and Lipids; Antenna Complexes: Structure, Function and Orgaviii
nization; Reaction Centre Structure and Function; Cyclic Electron Transfer Components and Energy Coupling Reactions; Metabolic Processes; Genomics, Regulation and Signaling; and New Applications and Techniques. This book is a compilation of 48 chapters, written by leading experts who highlight the huge progress made in spectroscopic, structural and genetic studies of these bacteria since 1995, when the last such book was published (Anoxygenic Photosynthetic Bacteria, Volume 2 in the Advances in Photosynthesis and Respiration Series, edited by Robert E. Blankenship, Michael T. Madigan and Carl E. Bauer; Kluwer Academic Publishers (now Springer), Dordrecht). This new volume is similarly intended to be the definitive text on these bacteria for many years to come, and it will be a valuable resource for experienced researchers, Ph.D. students, and advanced undergraduates in the fields of ecology, microbiology, biochemistry and biophysics. Scientists interested in future applications of these bacteria which could harness their potential for nanotechnology, solar energy research, bioremediation, or as cell factories, will also find the book useful. The readers can easily find the titles and the authors of the individual chapters in the Table of Content of this book. Instead of repeating this information here, I prefer to thank each and every author by name (listed in alphabetical order) that reads like a ‘Who’s Who’ in the field of purple phototrophic bacteria:
Loach; Chris Mackenzie; Chris Madigan; Michael T. Madigan; Pier Luigi Martelli; Bernd Masepohl; Shinji Masuda; Donna L. Mielke; Osamu Miyashita; Mamoru Nango; Robert A. Niederman; Wolfgang Nitschke; Vladimir I. Novoderezhkin; Dror Noy; U. Mirian Obiozo; Melvin Okamura; Ozlem Onder; Miroslav Papiz; Pamela S. ParkesLoach; William W. Parson; Marcela Ávila Pérez; Tomáš Polívka; Oleg G. Poluektov; Pu Qian; Jason Raymond; Bruno Robert; Brigitte R. Robinson; Simona Romagnoli; Johannes Sander; Carsten Sanders; Hugo Scheer; Simon Scheuring; Barbara Schoepp-Cothenet; Klaus Schulten; Melih K. Şener; Aaron Setterdahl; James P. Shapleigh; James N. Sturgis; Wesley D. Swingley; F. Robert Tabita; Shinichi Takaichi; Banita Tamot; Serdar Turkarslan; Lisa M. Utschig; Rienk van Grondelle; Michael A. van der Horst; Luuk J. van Wilderen; André Verméglio; Paulette M. Vignais; Martin J. Warren; Arieh Warshel; JoAnn C. Williams; Robert D. Willows; Colin A. Wraight; Jiang Wu; Vladimir Yurkov; Davide Zannoni; and Jill Helen Zeilstra-Ryalls. It is a pleasure to note that the following 27 participants in Volume 2 have also contributed to Volume 28; they are (alphabetically): James P. Allen; Judith P. Armitage; Carl E. Bauer; J. Thomas Beatty; Robert E. Blankenship; Richard J. Cogdell; Fevzi Daldal; Harry A. Frank; Caroline S. Harwood; C. Neil Hunter; J. Baz Jackson; Pierre Joliot; Gabriele Klug; Robert G. Kranz; Paul A. Loach; Michael T. Madigan; Melvin Okamura; Pamela S. Parkes-Loach; William W. Parson; Hugo Scheer; F. Robert Tabita; Rienk van Grondelle; André Verméglio; Paulette M. Vignais; Arieh Warshel; JoAnn C. Williams; and Davide Zannoni. It is remarkable that all three editors of Volume 2 (Carl E. Bauer, Robert E. Blankenship (Chief Editor), and Michael T. Madigan) are authors in Volume 28; and 3 of the 4 editors of Volume 28 (J. Thomas Beatty; Fevzi Daldal; and C. Neil Hunter (Chief Editor)) were authors in Volume 2. As Volume 28 is a sequel to Volume 2, it is beneficial for the readers of the new volume to consult and cite chapters in the earlier volume; I present below authors, titles of chapters and page numbers of all the chapters in that book. Please note that this volume was published by Kluwer Academic Publishers which was later acquired by Springer, the publishers of the current volume.
Maxime T.A. Alexandre; James P. Allen; Judith P. Armitage; Herbert Axelrod; Carl E. Bauer; Christoph Benning; Edward A. Berry; Robert E. Blankenship; Francesca Borsetti; Paula Braun; Per A. Bullough; Rita Casadio; Madhusudan Choudhary; Toh Kee Chua; Richard J. Cogdell; Jason W. Cooley; Julius T. Csotony; Christiane Dahl; Fevzi Daldal; Evelyne Deery; Takehisa Dewa; Timothy J. Donohue; Katie Evans; Boris A. Feniouk; Leszek Fiedor; Anthony Fordham-Skelton; Harry A. Frank; Elaine R. Frawley; Mads Gabrielsen; Alastair T. Gardiner; Toni Georgiou; Marie-Louise Groot; Marilyn R. Gunner; Wolfgang Haehnel; Deborah K. Hanson; Caroline S. Harwood; Klaas J. Hellingwerf; Johnny Hendriks; Theresa Hillon; Jonathan Hosler; Li-Shar Huang; C. Neil Hunter; Kouji Iida; J. Baz Jackson; Pierre Joliot; Michael R. Jones; Deborah O. Jung; Wolfgang Junge; Samuel Kaplan; John T. M. Kennis; Gabriele Klug; Hans Georg Koch; Jürgen Köhler; David M. Kramer; Robert G. Kranz; Alison M. Kriegel; Philip D. Laible; Jérôme Lavergne; Dong-Woo Lee; Paul A. ix
Complete List of Chapters in Anoxygenic Photosynthetic Bacteria, edited by R.E. Blankenship, M.T. Madigan and C.E. Bauer, Kluwer Academic Publishers, 1995
in purple bacteria, pp 349–372 Chapter 18: H.A. Frank and R.L. Christensen (1995) Singlet energy transfer from carotenoids to bacteriochlorophylls, pp 373–384 Chapter 19: A. Freiberg (1995) Coupling of antennas to reaction centers, pp 385–398 Chapter 20: R.E. Blankenship, J.M. Olson and M. Mette (1995) Antenna complexes from green photosynthetic bacteria, pp 399–435 Chapter 21: P.A. Loach and P.S. Parkes-Loach (1995) Structure-function relationships in core light-harvesting complexes (LHI) as determined by characterization of the structural subunit and by reconstitution experiments, pp 437–471 Chapter 22: C.N. Hunter (1995) Genetic manipulation of the antenna complexes of purple bacteria, pp 473–501 Chapter 23: C.R.D. Lancaster, U. Ermler and H. Michel (1995) The structures of photosynthetic reaction centers from purple bacteria as revealed by X-ray crystallography, pp 503–526 Chapter 24: N.W. Woodbury and J.P. Allen (1995) The pathway, kinetics and thermodynamics of electron transfer in wild type and mutant reaction centers of purple nonsulfur bacteria, pp 527–557 Chapter 25: W.W. Parson and A. Warshel (1995) Theoretical analyses of electron-transfer reactions, pp 559–575 Chapter 26: M.Y. Okamura and G. Feher (1995) Proton-coupled electron transfer reactions of QB in reaction centers from photosynthetic bacteria, pp 577–594 Chapter 27: M. Volk, A. Ogrodnik and M.E. Michel-Beyerle (1995) The recombination dynamics of the radical pair P+H– in external magnetic and electric fields, pp 595–626 Chapter 28: W. Mäntele (1995) Infrared vibrational spectroscopy of reaction centers, pp 627–647 Chapter 29: H. Scheer and G. Hartwich (1995) Bacterial reaction centers with modified tetrapyrrole chromophores, pp 649–663 Chapter 30: U. Feiler and G. Hauska (1995) The reaction center from green sulfur bacteria, pp 665–685 Chapter 31: J. Amesz (1995) The antenna-reaction center complex of heliobacteria, pp 687–697 Chapter 32: R. Feick, J.A. Shiozawa and A. Ertlmaier (1995) Biochemical and spectroscopic properties of the reaction center of the green filamentous bacterium, Chloroflexus aurantiacus, pp 699–708 Chapter 33: R.G. Kranz and D.L. Beckman (1995) Cytochrome biogenesis, pp 709–723
Chapter 1: J.F. Imhoff (1995) Taxonomy and physiology of phototrophic purple bacteria and green sulfur bacteria, pp 1–15 Chapter 2: M.T. Madigan and J.G. Ormerod (1995) Taxonomy, physiology and ecology of heliobacteria, pp 17–30 Chapter 3: B.K. Pierson and R.W. Castenholz (1995) Taxonomy and physiology of filamentous anoxygenic phototrophs, pp 31–47 Chapter 4: H. Van Gemerden and J. Mas (1995) Ecology of phototrophic sulfur bacteria, pp 49–85 Chapter 5: R.W. Castenholz and B.K. Pierson (1995) Ecology of thermophilic anoxygenic phototrophs, pp 87–103 Chapter 6: K. Shimada (1995) Aerobic anoxygenic phototrophs, pp 105–122 Chapter 7: D.E. Fleischman, W.R. Evans and I.M. Miller (1995) Bacteriochlorophyll-containing rhizobium species, 123–136 Chapter 8: M.O. Senge and K.M. Smith (1995) Biosynthesis and structures of the bacteriochlorophylls, pp 137–151 Chapter 9: S.I. Beale (1995) Biosynthesis and structures of porphyrins and hemes, pp 153–177 Chapter 10: J.F. Imhoff and U. Bias-Imhoff (1995) Lipids, quinones and fatty acids of anoxygenic phototrophic bacteria, pp 179–205 Chapter 11: J. Weckesser, H. Mayer and G. Schulz (1995) Anoxygenic phototrophic bacteria: Model organisms for studies on cell wall macromolecules, pp 207–230 Chapter 12: G. Drews and J.R. Golecki (1995) Structure, molecular organization, and biosynthesis of membranes of purple bacteria, pp 231–257 Chapter 13: J. Oelze and J.R. Golecki (1995) Membranes and chlorosomes of green bacteria: Structure, composition and development, pp 259–278 Chapter 14: A. Verméglio, P. Joliot and A. Joliot (1995) Organization of electron transfer components and supercomplexes, pp 279–295 Chapter 15: W.S. Struve (1995) Theory of electronic energy transfer, pp 297–313 Chapter 16: H. Zuber and R.J. Cogdell (1995) Structure and organization of purple bacterial antenna complexes, pp 315–348 Chapter 17: V. Sundström and R. Van Grondelle (1995) Kinetics of excitation transfer and trapping x
Chapter 34: T.E. Meyer and T.J. Donohue (1995) Cytochromes, iron-sulfur, and copper proteins mediating electron transfer from the Cyt bc1 complex to photosynthetic reaction center complexes, pp 725–745 Chapter 35: K.A. Gray and F. Daldal (1995) Mutational studies of the cytochrome bc1 complexes, pp 747–774 Chapter 36: W. Nitschke and S.M. Dracheva (1995) Reaction center associated cytochromes, pp 775–805 Chapter 37: Z. Gromet-Elhanan (1995) The protontranslocating F0F1 ATP synthase-ATPase complex, pp 807–830 Chapter 38: J.B. Jackson (1995) Proton-translocating transhydrogenase and NADH dehydrogenase in anoxygenic photosynthetic bacteria, pp 831–845 Chapter 39: D.C. Brune (1995) Sulfur compounds as photosynthetic electron donors, pp 847–870 Chapter 40: R. Sirevåg (1995) Carbon metabolism in green bacteria, pp 871–883 Chapter 41: F.R. Tabita (1995) The biochemistry and metabolic regulation of carbon metabolism and CO2 fixation in purple bacteria, pp 885–914 Chapter 42: M.T. Madigan (1995) Microbiology of nitrogen fixation by anoxygenic photosynthetic bacteria, pp 915–928 Chapter 43: P.W. Ludden and G.P. Roberts (1995) The biochemistry and genetics of nitrogen fixation by photosynthetic bacteria, pp 929–947 Chapter 44: D. Zannoni (1995) Aerobic and anaerobic electron transport chains in anoxygenic phototrophic bacteria, pp 949–971 Chapter 45: J. Mas and H. Van Gemerden (1995) Storage products in purple and green sulfur bacteria, pp 973–990 Chapter 46: J. Gibson and C.S. Harwood (1995) Degradation of aromatic compounds by nonsulfur purple bacteria, pp 991–1003 Chapter 47: J.P. Armitage, D.J. Kelly and R.E. Sockett (1995) Flagellate motility, behavioral responses and active transport in purple non-sulfur bacteria, pp 1005–1028 Chapter 48: J.C. Williams and A.K.W. Taguchi (1995) Genetic manipulation of purple photosynthetic bacteria, pp 1029–1065 Chapter 49: M. Fonstein and R. Haselkorn (1995) Physical mapping of Rhodobacter capsulatus: Cosmid encyclopedia and high resolution genetic map, pp 1067–1081 Chapter 50: M. Alberti, D.H. Burke and J.E. Hearst (1995) Structure and sequence of the photosynthesis
gene cluster, pp 1083–1106 Chapter 51: J.L. Gibson (1995) Genetic analysis of CO2 fixation genes, pp 1107–1124 Chapter 52: A.J. Biel (1995) Genetic analysis and regulation of bacteriochlorophyll biosynthesis, pp 1125–1134 Chapter 53: G.A. Armstrong (1995) Genetic analysis and regulation of carotenoid biosynthesis: Structure and fuction of the crt genes and gene products, pp 1135–1157 Chapter 54: J.A. Shiozawa (1995) A foundation for the genetic analysis of green sulfur, green filamentous and heliobacteria, pp 1159–1173 Chapter 55: P.M. Vignais, B. Toussaint and A. Colbeau (1995) Regulation of hydrogenase gene expression, pp 1175–1190 Chapter 56: R.G. Kranz and P.J. Cullen (1995) Regulation of nitrogen fixation genes, pp 1191–1208 Chapter 57: J.T. Beatty (1995) Organization of photosynthesis gene transcripts, pp 1209–1219 Chapter 58: C.E. Bauer (1995) Regulation of photosynthesis gene expression, pp 1221–1234 Chapter 59: G. Klug (1995) Post-transcriptional control of photosynthesis gene expression, pp 1235–1244 Chapter 60: R.C. Fuller (1995) Polyesters and photosynthetic bacteria: from lipid cellular inclusions to microbial thermoplastics, pp 1245–1256 Chapter 61: E.R. Goldman and D.C. Youvan (1995) Imaging Spectroscopy and combinatorial mutagenesis of the reaction center and light harvesting II antenna, pp 1257–1268 Chapter 62: Michiharu Kobayashi and Michihiko Kobayashi (1995) Waste remediation and treatment using anoxygenic phototrophic bacteria, pp 1269–1282
Future AIPH and Other Related Books The readers of the current series are encouraged to watch for the publication of the forthcoming books (not necessarily arranged in the order of future appearance): • C-4 Photosynthesis and Related CO2 Concentrating Mechanisms (Editors:Agepati S. Raghavendra and Rowan Sage); • Photosynthesis: Perspectives on Plastid Biology, Energy Conversion and Carbon Metabolism (Editors: Julian Eaton-Rye and Baishnab Tripathy); xi
• Abiotic Stress Adaptation in Plants: Physiological, Molecular and Genomic Foundation (Editors: Ashwani Pareek, Sudhir K. Sopory, Hans J. Bohnert and Govindjee); • The Chloroplast: Biochemistry, Molecular Biology and Bioengineering (Editors: Constantin Rebeiz, Hans Bohnert, Christoph Benning, Henry Daniell, Beverley R. Green, J. Kenneth Hoober, Hartmut Lichtenthaler, Archie R. Portis and Baishnab C. Tripathy); • Photosynthesis In Silico: Understanding Complexity from Molecules to Ecosystems (Editors: Agu Laisk, Ladislav Nedbal and Govindjee); and • Lipids in Photosynthesis: Essential and Regulatory Function (Editors: Hajime Wada and Norio Murata).
Readers are encouraged to send their suggestions for these and future Volumes (topics, names of future editors, and of future authors) to me by E-mail (gov@ illinois.edu) or fax (1-217-244-7246). In view of the interdisciplinary character of research in photosynthesis and respiration, it is my earnest hope that this series of books will be used in educating students and researchers not only in Plant Sciences, Molecular and Cell Biology, Integrative Biology, Biotechnology, Agricultural Sciences, Microbiology, Biochemistry, and Biophysics, but also in Bioengineering, Chemistry, and Physics. I take this opportunity to thank and congratulate C. Neil Hunter, Fevzi Daldal, Marion Thurnauer and Thomas Beatty for their editorial work. My special thanks go to C. Neil Hunter, the Chief Editor of Volume 28, for his painstaking work not only in editing, but also in organizing this book for Springer, and for his highly professional dealing with the typesetting process and his help in preparing this editorial. I particularly thank all the 116 authors (see the list above) of this book: without their authoritative chapters, there would be no such Volume. I give special thanks to Larry Orr for typesetting this book: his expertise has been crucial in bringing this book to completion. We owe Jacco Flipsen, Noeline Gibson and André Tournois (of Springer) thanks for their friendly working relation with us that led to the production of this book. Thanks are also due to Jeff Haas (Director of Information Technology, Life Sciences, University of Illinois at Urbana-Champaign, UIUC), Evan DeLucia (Head, Department of Plant Biology, UIUC) and my dear wife Rajni Govindjee for constant support.
In addition to these contracted books, the following topics, among others, are under consideration: • • • • • • • • • • •
Cyanobacteria Genomics, Proteomics and Evolution Biohydrogen Production ATP Synthase and Proton Translocation Interactions between Photosynthesis and other Metabolic Processes Carotenoids II Green Bacteria and Heliobacteria Ecophysiology Photosynthesis, Biomass and Bioenergy Global Aspects of Photosynthesis Artificial Photosynthesis
May 14, 2008 Govindjee Series Editor, Advances in Photosynthesis and Respiration University of Illinois at Urbana-Champaign, Department of Plant Biology, 265 Morrill Hall, 505 South Goodwin Avenue, Urbana, IL 61801-3707, USA E-mail: [email protected]; URL: http://www.life.uiuc.edu/govindjee
xii
Govindjee, Series Editor
Govindjee, born in 1932, obtained his B.Sc. (Chemistry, Biology) and M.Sc. (Botany, Plant Physiology) in 1952 and 1954, from the University of Allahabad, India, and his Ph.D. (Biophysics, under Eugene Rabinowitch), in 1960, from the University of Illinois at Urbana-Champaign (UIUC), IL, U.S.A. He is best known for his research on excitation energy transfer, light emission, primary photochemistry and electron transfer in Photosystem II (PS II). His research, with many collaborators, has included the discovery of a short-wavelength form of chlorophyll (Chl) a functioning in the Chl b-containing system, now called PS II, and of the two-light effects in Chl a fluorescence and in NADP (nicotinamide adenine dinucleotide phosphate) reduction in chloroplasts (Emerson Enhancement). Further, he has worked on the existence of different spectral fluorescing forms of Chl a and the temperature dependence of excitation energy transfer down to 4 K; basic relationships between Chl a fluorescence and photosynthetic reactions; the unique role of bicarbonate on the acceptor side of PS II, particularly in protonation events involving the QB binding region; the theory of thermoluminescence in plants; picosecond measurements on the primary photochemistry of PS II; and the use of Fluorescence Lifetime Imaging Microscopy (FLIM) of Chl a fluorescence in understanding photoprotection against excess light. His research on photosynthetic bacteria
included the observation of the absence of the Emerson Enhancement Effect (1960s); measurements on the lifetime of bacteriochlorophyll fluorescence (1970s); and the use of the bacterial reaction center structure in homology modeling of Photosystem II, particularly on its electron acceptor side (1990s). His current focus is on the ‘History of Photosynthesis Research,’ in ‘Photosynthesis Education’, and in the ‘Possible Existence of Extraterrestrial Life.’ He has served on the faculty of the UIUC for ~40 years. Since 1999, he has been Professor Emeritus of Biochemistry, Biophysics and Plant Biology at the same institution. His honors include: Fellow of the American Association of Advancement of Science; Distinguished Lecturer of the School of Life Sciences, UIUC; Fellow and Lifetime member of the National Academy of Sciences (India); President of the American Society for Photobiology (1980–1981); Fulbright Scholar and Fulbright Senior Lecturer; Honorary President of the 2004 International Photosynthesis Congress (Montréal, Canada); the 2006 Recipient of the Lifetime Achievement Award from the Rebeiz Foundation for Basic Biology; and the 2007 Recipient of the ‘Communication Award’ of the International Society of Photosynthesis Research (ISPR), presented to him at the 14th International Congress on Photosynthesis, held in Glasgow, Scotland, U.K.
xiii
Contents From the Series Editor
vii
Contents
xv
Preface
xxxi
Author Index
xxxvii
Color Plates
CP1–CP16
Part 1: Physiology, Ecology and Evolution 1
An Overview of Purple Bacteria: Systematics, Physiology, and Habitats Michael T. Madigan and Deborah O. Jung Summary I. Introduction II. Systematics of Purple Bacteria III. Physiology of Purple Bacteria IV. Habitats of Purple Bacteria V. Purple Bacteria in Extreme Environments VI. Final Remarks Acknowledgments References
2
1–15 2 2 3 4 7 9 12 12 12
Evolutionary Relationships Among Purple Photosynthetic Bacteria and the Origin of Proteobacterial Photosynthetic Systems 17–29 Wesley D. Swingley, Robert E. Blankenship and Jason Raymond Summary I. Introduction II. The Alphaproteobacteria III. Aerobic Purple Bacteria IV. The Photosynthesis Gene Cluster and its Role in Evolution V. Proteobacterial Comparative Genomics: Photosynthetic versus NonPhotosynthetic Proteins VI. Origin and Evolution of Proteobacterial Phototrophy VII. Origin and Evolution of Proteobacterial Carbon-fixation VIII. Future Directions: High-Throughput Sequencing and Metagenomics Acknowledgments References
xv
17 18 18 19 20 21 22 24 27 28 28
3
New Light on Aerobic Anoxygenic Phototrophs Vladimir Yurkov and Julius T. Csotonyi
31–55
Summary 31 I. Introduction 32 II. Morphological Diversity, Taxonomic Nuances, Phylogeny and Evolution 34 III. Nutritional Versatility and Peculiarities of Carbon Metabolism 40 IV. Photosynthetic Pigment Composition and Synthesis Reveal Surprises 41 V. The Mysterious Photosynthetic Apparatus of Aerobic Anoxygenic Phototrophs 44 VI. Speculation on Ecological Roles 47 VII. Concluding Remarks and Perspectives 51 Acknowledgments 52 References 52
Part 2: Molecular Structure and Biosynthesis of Pigments and Cofactors 4
Biosynthesis of Bacteriochlorophylls in Purple Bacteria Robert D. Willows and Alison M. Kriegel
57–79
Summary I. Introduction II. δ-Aminolevulinate to Protoporphyrin IX III. Protoporphyrin IX to Bacteriochlorophyll a and b IV. Concluding Remarks Acknowledgments References
5
Vitamin B12 (Cobalamin) Biosynthesis in the Purple Bacteria Martin J. Warren and Evelyne Deery
57 58 59 65 75 75 75
81–95
Summary I. Background II. Function of Cobalamin III. B12 Biosynthesis in Rhodobacter capsulatus and Rhodobacter sphaeroides IV. Summary of Events Required for Cobalamin Biosynthesis V. Biosynthesis of Precorrin-2 and Its Onward Route Towards Siroheme and Cobalamin VI. Control and Regulation of Cobalamin Biosynthesis Acknowledgments References
6
Distribution and Biosynthesis of Carotenoids Shinichi Takaichi Summary I. Introduction II. Carotenogenesis III. Carotenoids in Purple Bacteria Acknowledgments References
xvi
81 81 82 83 84 86 92 92 92
97–117 97 98 101 111 114 114
7
Membrane Lipid Biosynthesis in Purple Bacteria Banita Tamot and Christoph Benning Summary I. Introduction II. Fatty Acids III. Phosphoglycerolipids IV. Glycoglycerolipids V. Betaine Lipid VI. Ornithine Lipid VII. Lipid Function VIII. Perspectives Acknowledgments References
119–134 119 120 122 123 126 128 129 130 131 131 132
Part 3: Antenna Complexes: Structure, Function and Organization 8
Peripheral Complexes of Purple Bacteria Mads Gabrielsen, Alastair T. Gardiner and Richard J. Cogdell Summary I. Introduction II. Structure III. The Biology of Purple Bacterial Antenna Complexes IV. Final Remarks Acknowledgments References
9
Reaction Center-Light-Harvesting Core Complexes of Purple Bacteria Per A. Bullough, Pu Qian and C. Neil Hunter Summary I. Introduction II. The Building Blocks: The α and β Polypeptides of Light-Harvesting Complex 1, and PufX III. Circles, Arcs and Ellipses — The Light-Harvesting 1 Complex IV. Monomeric Reaction Center-Light-Harvesting 1 Complexes V. Monomeric Reaction Center-Light-Harvesting 1-PufX Complexes VI. Dimeric Reaction Center-Light-Harvesting 1-PufX Complexes VII. The Biogenesis of Core Complexes Acknowledgments References
xvii
135–153 135 136 136 146 151 151 151
155–179 155 156 156 160 162 166 168 174 175 175
10 Structure-Function Relationships in Bacterial Light-Harvesting Complexes Investigated by Reconstitution Techniques Paul A. Loach and Pamela S. Parkes-Loach
181–198
Summary I. Introduction II. Reversible Dissociation of Core Light-Harvesting 1 Complexes to a Subunit Form (B820) III. Reversible Dissociation of B820 to its Fundamental Components IV. Cofactor Requirements V. Cofactor – Protein Interactions VI. Cofactor – Cofactor Interactions VII. Protein–Protein Interactions VIII. Effect of PufX on Reconstitution of Light-Harvesting 1 Complexes IX. In vitro versus In vivo Assembly of Complexes X. Reconstitution of the Reaction Center Acknowledgments References
11 Spectroscopic Properties of Antenna Complexes from Purple Bacteria Bruno Robert
181 182 182 183 184 184 188 188 191 192 193 195 195
199–212
Summary 1. Introduction II. The Different Spectral Forms of Antenna Proteins from Purple Bacteria III. Antenna Proteins from Purple Bacteria: Variations Around a Structural Theme IV. The Role of Ground State Interactions in Tuning the Antenna Absorption Transition V. Excitonic Interactions and Disorder in Light-Harvesting Complexes VI. Chromophore Interactions in Light-Harvesting Proteins: Additional Effects VII. Conclusions Acknowledgments References
12 Energy Transfer from Carotenoids to Bacteriochlorophylls Harry A. Frank and Tomáš Polívka
201 203 205 207 208 209 209
213–230
Summary I. Introduction II. Carotenoid Excited States III. Energy Transfer in Light-Harvesting 2 Complexes IV. Energy Transfer in Light-Harvesting 1 Complexes and Reaction Centers V. Outlook Acknowledgments References
xviii
199 200 200
213 214 215 216 226 227 227 227
13 Spectroscopy and Dynamics of Excitation Transfer and Trapping in Purple Bacteria 231–252 Rienk van Grondelle and Vladimir I. Novoderezhkin Summary 232 I. Introduction 232 II. Structure and Exciton Spectra of Light-Harvesting 1 and 2 Bacterial Antenna Complexes 235 III. Equilibration Dynamics 239 IV. Competition of Intraband B800-800 and Interband B800-850 Energy Transfer in the Light-Harvesting 2 Complex 241 V. Energy Trapping in the Core Reaction Center-Light-Harvesting 1 Complex 243 VI. Slow Conformational Motions and Excitation Dynamics in the B850-LightHarvesting 2 Complex 244 VII. Concluding Remarks 247 Acknowledgments 248 References 248
14 Organization and Assembly of Light-Harvesting Complexes in the Purple Bacterial Membrane 253–273 James N. Sturgis and Robert A. Niederman Summary I. Introduction II. Themes and Variations — Structural Variability of Complexes in Native Membranes III. Principles of Photosynthetic Unit Organization IV. Proposals for the Functional Organization of Photosynthetic Units V. In Vivo Assembly of Light-Harvesting Complexes VI. Perspectives for the Next Ten Years References
15 From Atomic-Level Structure to Supramolecular Organization in the Photosynthetic Unit of Purple Bacteria Melih K. ùener and Klaus Schulten
254 254 255 257 263 267 269 270
275–294
Summary 275 I. Introduction 276 II. Components of the Photosynthetic Unit 278 III. Quantum Physics of Light-Harvesting and Excitation Energy Transfer 279 IV. Effect of Thermal Disorder on the Spectra of Light-Harvesting Complexes 282 V. Physical Constraints Shaping the Structure of Individual Light-Harvesting Complexes 284 VI. Supramolecular Organization of the Photosynthetic Unit 286 VII. An Atomic-level Structural Model for a Photosynthetic Membrane Vesicle 287 VIII. Light-Harvesting and Excitation Transfer across a Photosynthetic Membrane Vesicle 287 IX. Conclusions 289 Acknowledgments 290 References 290
xix
Part 4: Reaction Center Structure and Function 16 Structural Plasticity of Reaction Centers from Purple Bacteria Michael R. Jones
295–321
Summary I. The Plastic Purple Reaction Center II. Biochemical and Genetic Alteration of Polypeptide Composition III. Cofactor Exclusion IV. Cofactor Replacement V. Helix Symmetrization VI. Water and Other Unexpected Things in Electron Density Maps VII. Building New Functionality VIII. Conclusions Acknowledgments References
295 296 297 299 302 307 310 312 313 314 314
17 Structure and Function of the Cytochrome c2:Reaction Center Complex from Rhodobacter sphaeroides Herbert Axelrod, Osamu Miyashita and Melvin Okamura Summary I. Introduction II. History III. Structure of the Cytochrome c2:Reaction Center Complex IV. Electron Transfer Reactions V. Effects of Mutation VI. Mechanism of Inter-Protein Electron Transfer Acknowledgments References
18 Directed Modification of Reaction Centers from Purple Bacteria JoAnn C. Williams and James P. Allen Summary I. Introduction II. Properties of the Cofactors III. Electron Transfer Concepts IV. Pathways of Electron Transfer V. Conclusions Acknowledgments References
xx
323–336 323 324 324 325 327 329 332 333 333
337–353 337 338 338 343 346 349 349 349
19 Mechanism of Charge Separation in Purple Bacterial Reaction Centers William W. Parson and Arieh Warshel Summary I. The Reaction Sequence and Kinetics II. Energies of the Radical-Pair Intermediates III. Unusual Features of the Charge-Separation Reactions IV. Theories of Electron-Transfer Reactions Acknowledgments References
20 The Acceptor Quinones of Purple Photosynthetic Bacteria— Structure and Spectroscopy Colin A. Wraight and Marilyn R. Gunner Summary I. Introduction II. The Acceptor Quinone Reactions III. Structural Features of the Acceptor Quinone Binding Sites IV. Spectroscopy of the Acceptor Quinones V. Functionality of the Two Quinone Positions in the QB Site VI. Conclusions Note Added in Proof Acknowledgments References
355–377 355 356 357 360 363 370 370
379–405 379 380 382 383 389 396 398 398 398 399
Part 5: Cyclic Electron Transfer Components and Energy Coupling Reactions 21 Biogenesis of c-type Cytochromes and Cytochrome Complexes Carsten Sanders, Serdar Turkarslan, Ozlem Onder, Elaine R. Frawley, Robert G. Kranz, Hans Georg Koch and Fevzi Daldal Summary I. Introduction II. Maturation of c-type Cytochromes: Ccm-system I and Ccs-system II III. Biogenesis of Cytochrome Complexes Acknowledgments References
xxi
407–423
407 408 409 415 421 421
22 Structural and Mutational Studies of the Cytochrome bc1 Complex 425–450 Edward A. Berry, Dong-Woo Lee, Li-Shar Huang and Fevzi Daldal Summary I. Introduction II. Structural and Mutational Studies III. Conclusions and Perspectives Acknowledgments References
23 The Cytochrome bc1 and Related bc Complexes: The Rieske/ Cytochrome b Complex as the Functional Core of a Central Electron/Proton Transfer Complex David M. Kramer, Wolfgang Nitschke and Jason W. Cooley
425 426 427 446 447 447
451–473
Summary I. Introduction II. Structures of the Cytochrome bc1 and Related Rieske/Cytochrome b Complexes III. Catalysis in the Rieske/Cytochrome b Complexes: The General Q-cycle Framework IV. Phylogeny and Evolution V. The ‘Third’ Redox Subunit is a Phylogenetic Marker VI. The Rieske/Cytochrome b Complex, a Primordial Enzyme VII. The Molecular Mechanism of the Qo Site: Avoiding Q-Cycle Short Circuits VIII. The Quinone Reduction Site, Qi of the Cytochrome bc1 Complexes IX. The Quinone Reduction Site,Qi of the Cytochrome b6 f and Related Complexes X. The Functional Mechanism of the Rieske/Cytochrome b Complexes is Conserved Acknowledgments References
452 452 453 455 456 458 459 460 467 468 469 469 469
24 Proton Translocation and ATP Synthesis by the FoF1-ATPase of Purple Bacteria Boris A. Feniouk and Wolfgang Junge
475–493
Summary I. Introduction II. Structure and Rotary Catalysis III. Proton Translocation and its Coupling to ATP Synthesis/Hydrolysis III. Role of Proton Translocation in the Regulation of ATP Synthase Acknowledgments References
475 476 476 478 486 487 488
xxii
25 Proton-Translocating Transhydrogenase in Photosynthetic Bacteria J. Baz Jackson and U. Mirian Obiozo
495–508
Summary I. Introduction II. An Overview of the Main Structural Features of Transhydrogenase III. Distribution of Transhydrogenase Among Species IV. Phylogenetic Relationships between Transhydrogenases from Different Species V. The Function of Transhydrogenase in Photosynthetic Bacteria VI. The Mechanism of Coupling Between Hydride Transfer and Proton Translocation in Transhydrogenase VII. Conformational Changes in the Coupling Reactions of Transhydrogenase Acknowledgments References
495 496 496 496 498 500 501 504 505 506
26 Functional Coupling Between Reaction Centers and Cytochrome bc1 Complexes 509–536 Jérôme Lavergne, André Verméglio and Pierre Joliot Summary I. Introduction II. Structure of the Protein Complexes III. The Electron Donors to the Reaction Center IV. Kinetics of P+ Reduction by Mobile Cytochromes V. Donor Side Shuttling and Turnover of the Cytochrome bc1 Complex VI. Quinone Reactions VII. Supramolecular Organization in Rhodobacter sphaeroides and Rhodobacter capsulatus VIII. Quinone Confinement in Rhodobacter sphaeroides IX. Quinone Traffic in the PufX– Mutant of Rhodobacter sphaeroides X. The Supercomplex Model: Difficulties and Alternative Possibilities XI. Mitochondrial Supercomplexes XII. Diffusion and Confinement of Cytochrome c2: Possible Mechanisms XIII. Diffusion and Confinement of Quinones: Possible Mechanisms XIV. Conclusions References
509 510 512 514 515 517 518 519 522 522 524 526 527 528 529 530
Part 6: Metabolic Processes 27 Respiration and Respiratory Complexes 537–561 Davide Zannoni, Barbara Schoepp-Cothenet and Jonathan Hosler Summary I. Aerobic Respiration II. Respiration Utilizing Substrates other than Oxygen III. Respiration versus Photosynthesis: Which One Came First? IV. Respiration and Photosynthesis are Intermingled Acknowledgments References
xxiii
538 538 546 553 553 555 555
28 Carbon Dioxide Metabolism and its Regulation in Nonsulfur Purple Photosynthetic Bacteria Simona Romagnoli and F. Robert Tabita
563–576
Summary I. Introduction II. Regulation of cbb Gene Expression Acknowledgments References
29 Degradation of Aromatic Compounds by Purple Nonsulfur Bacteria Caroline S. Harwood
563 564 564 575 575
577–594
Summary I. Introduction II. Biochemical Themes III. Species that Degrade Aromatic Compounds IV. Aerobic Degradation of Aromatic Compounds by Rhodopseudomonas palustris V. Anaerobic Benzoate Degradation VI. The Molecular Regulation of Anaerobic Aromatic Compound Degradation VII. Comparative Aspects Acknowledgments References
30 Metabolism of Inorganic Sulfur Compounds in Purple Bacteria Johannes Sander and Christiane Dahl Summary I. Introduction II. Sulfur Oxidation Capabilities of Purple Bacteria III. Sulfur Oxidation Pathways IV. Sulfate Assimilation V. Conclusions Acknowledgments References
31 Dissimilatory and Assimilatory Nitrate Reduction in the Purple Photosynthetic Bacteria James P. Shapleigh Summary I. Introduction II. Denitrification III. Assimilation of Nitrogen IV. Conclusion Acknowledgments References
577 578 578 579 579 580 589 590 591 591
595–622 596 596 596 602 612 615 616 616
623–642 623 624 624 636 638 639 639
xxiv
32 Swimming and Behavior in Purple Non-Sulfur Bacteria Judith P. Armitage
643–654
Summary I. Introduction II. Swimming III. Behavioral Responses Acknowledgments References
643 644 644 646 653 653
33 Metals and Metalloids in Photosynthetic Bacteria: Interactions, Resistance and Putative Homeostasis Revealed by Genome Analysis Francesca Borsetti, Pier Luigi Martelli, Rita Casadio and Davide Zannoni
655–689
Summary I. Introduction II. Classification of Metals and Metalloids by their Toxicity or Essentiality to Bacterial Cells (Groups I, II and III). III. Metal Toxicity, Tolerance and Resistance: Generalities IV. General Features on Microbial Metal Resistance/Tolerance Mechanisms V. On the Bacterial Interactions with Metals and Metalloids VI. Metal(loid)s Homeostasis in Phototrophs as Revealed by Genome Analysis VII. Concluding Remarks Acknowledgments References
656 656 656 657 658 666 675 679 682 682
Part 7: Genomics, Regulation and Signaling 34 Purple Bacterial Genomics 691–706 Madhusudan Choudhary, Chris Mackenzie, Timothy J. Donohue and Samuel Kaplan I. II. III. IV.
Introduction Genome Architecture and Characteristics Gene Homologs and Metabolic Versatility Variation in Transcriptional Regulation and Adaptation to Changing Environments V. Transposons and Genomic Rearrangements VI. Circadian Clock and Gas Vesicle Proteins VII. Inorganic Compounds as Reducing Power VIII. Genomic Insights into the Photosynthetic Lifestyle Acknowledgments References
xxv
692 692 696 701 702 702 703 703 704 704
35 Regulation of Gene Expression in Response to Oxygen Tension 707–725 Carl E. Bauer, Aaron Setterdahl, Jiang Wu and Brigitte R. Robinson Summary I. Introduction II. RegB/RegA Two-Component Signal Transduction System III. Aerobic repression by CrtJ IV. Regulation by Fnr Acknowledgments References
36 Regulation of Genes by Light Gabriele Klug and Shinji Masuda
707 707 708 716 721 722 722
727–741
Summary I. Introduction II. Photoreceptors in Purple Photosynthetic Bacteria III. Light-Dependent Responses that Do Not Depend on Photoreceptors IV. Concluding Remarks Acknowledgments References
37 Regulation of Hydrogenase Gene Expression Paulette M. Vignais
727 728 728 735 737 737 737
743–757
Summary I. Introduction II. Regulation of Hydrogenase Gene Expression: Signaling and Transcription Control References
38 Regulation of Nitrogen Fixation Bernd Masepohl and Robert G. Kranz
743 744 744 755
759–775
Summary I. Nitrogen Fixation in Purple Nonsulfur Bacteria II. Three Regulatory Levels of Nitrogen Fixation and Molecular Mechanisms Studied in Rhodobacter capsulatus III. Other Factors that Feed into the Nitrogen Regulatory Circuitry IV. Regulation in Other Purple Photosynthetic Bacteria V. Future Perspectives Acknowledgments References
xxvi
760 760 762 768 770 771 772 772
39 Regulation of the Tetrapyrrole Biosynthetic Pathway Jill Helen Zeilstra-Ryalls
777–798
Summary I. Introduction II. Tetrapyrrole Biosynthesis Genes III. Comparing and Contrasting Oxygen Control of Tetrapyrrole Biosynthesis Genes in Species of Rhodobacter IV. Other Aspects of Transcriptional Regulation of Tetrapyrrole Biosynthesis Genes in Rhodobacter Species V. A Genomics Perspective on the Regulation of Tetrapyrrole Biosynthesis in Other Purple Anoxygenic Photosynthetic Bacteria Note Added in Proof Acknowledgments References
40 Bacteriophytochromes Control Photosynthesis in Rhodopseudomonas palustris Katie Evans, Toni Georgiou, Theresa Hillon, Anthony Fordham-Skelton and Miroslav Papiz I. II.
777 778 778 779 789 789 795 795 795
799–809
Introduction Bacteriophytochrome Gene Organization and Regulation of Photosynthesis in Rhodopseudomonas palustris III. Phytochrome Domain Organization IV. Bilin Chromophore Photo-conversion V. Chromophore Binding Domain of Deinococcus radiodurans Bacteriophytochrome VI. Small Angle X-ray Scattering Solution Structure of Bph4 from Rhodopseudomonas palustris VII. Conclusions Acknowledgments References
800 800 803 804 805 805 807 807 807
41 Photoreceptor Proteins from Purple Bacteria 811–837 Johnny Hendriks, Michael A. van der Horst, Toh Kee Chua, Marcela Ávila Pérez, Luuk J. van Wilderen, Maxime T.A. Alexandre, Marie-Louise Groot, John T. M. Kennis and Klaas J. Hellingwerf Summary I. Introduction II. Light, Oxygen, or Voltage Domains III. The BLUF Domain Containing Family of Photoreceptors IV. Comparison Between LOV and BLUF Domains V. The Xanthopsins VI. Bacteriophytochromes VII. Concluding Remarks Acknowledgments References
xxvii
811 812 813 816 822 823 827 831 832 832
Part 8: New Applications and Techniques 42 Foreign Gene Expression in Photosynthetic Bacteria Philip D. Laible, Donna L. Mielke and Deborah K. Hanson Summary I. Introduction II. Design of a Rhodobacter-Based System for the Expression of Membrane Proteins for Structural and Functional Studies III. Production of Foreign Membrane Proteins in Rhodobacter IV. Optimization and Generalization of Heterologous Expression in Rhodobacter V. Advantages Afforded by Rhodobacter VI. Perspectives Acknowledgments References
43 Assembly of Bacterial Light-Harvesting Complexes on Solid Substrates Kouji Iida, Takehisa Dewa and Mamoru Nango Summary I. Introduction II. Atomic Force Microscopy Imaging of Reassociated Bacterial LightHarvesting Complex 1 on a Mica Substrate III. Conductivity of the Bacterial Reaction Center on Chemically Modified Gold Substrates Using Conductive Atomic Force Microscopy IV. Molecular Assembly of Photosynthetic Antenna Core Complex on an Amino-terminated Indium Tin Oxide Electrode V. Concluding Remarks Acknowledgments References
44 Optical Spectroscopy of Individual Light-Harvesting Complexes from Purple Bacteria Jürgen Köhler Summary I. Introduction II. The Experimental Setup III. Energy Transfer, Excitons, Strong and Weak Coupling IV. Spectroscopy of Single Light-Harvesting 2 Antenna Complexes V. Energy Transfer in a Single Photosynthetic Unit Acknowledgements References
xxviii
839–860 840 840 841 847 850 852 856 856 856
861–875 861 862 864 865 869 873 873 873
877–894 877 877 879 880 882 889 891 891
45 De novo Designed Bacteriochlorophyll-Binding Helix-Bundle Proteins Wolfgang Haehnel, Dror Noy and Hugo Scheer
895–912
Summary I. Introduction II. Chlorophyll Structures and Interactions with Natural Proteins III. Challenges in Designing de novo Chlorophyll- and Bacteriochlorophyllbinding Proteins IV. Modular Organized Chlorophyll Proteins Based on Branched Fourhelix Bundle Proteins V. Incorporating Chlorophylls and Bacteriochlorophylls into Self-assembling Protein Maquettes VI. From Water-soluble to Amphiphilic Chlorophyll- and Bacteriochlorophyllprotein Maquettes Acknowledgments References
46 Design and Assembly of Functional Light-Harvesting Complexes Paula Braun and Leszek Fiedor Summary I. Introduction II. Design of Model Light-Harvesting Proteins III. Assembly of Functional Light-Harvesting 1 Complexes IV. Conclusions and Prospects Acknowledgments References
895 896 897 899 901 904 905 907 907
913–940 914 914 916 924 935 935 936
47 The Supramolecular Assembly of the Photosynthetic Apparatus of Purple Bacteria Investigated by High-Resolution Atomic Force Microscopy 941–952 Simon Scheuring Summary I. Introduction II. Atomic Force Microscopy Analysis of the Complexes of the Bacterial Photosynthetic Apparatus III. Conclusions IV. Feasibilities, Limitations and Outlook Acknowledgments References
xxix
941 942 943 949 950 951 951
48 Protein Environments and Electron Transfer Processes Probed with High-Frequency ENDOR Oleg G. Poluektov and Lisa M. Utschig
953–973
Summary I. Introduction II. Low Temperature Interquinone Electron Transfer in the Photosynthetic Reaction Center. Characterization of QB– States III. Electron Transfer Pathways and Protein Response to Charge Separation IV. Concluding Remarks Acknowledgments References
Index
953 954 956 963 968 969 969
975–1013
xxx
Preface This book follows the tradition initiated in 1963, then subsequently extended in 1978 and 1995, of summarizing our knowledge of the photototrophic bacteria in a single volume. The first book was Bacterial Photosynthesis (Howard Gest, Anthony San Pietro and Leo P.Vernon, eds), 1963, Antioch Press, Yellow Springs, Ohio. Fifteen years later, Roderick K. Clayton and William R. Sistrom edited The Photosynthetic Bacteria (1978, Plenum Press, New York), an indispensable book for any scientist in the field at that time. In 1995, Robert Blankenship, Michael Madigan and Carl Bauer edited the most complete survey of the subject, Anoxygenic Photosynthetic Bacteria. By that time, the book had been taken under the wing of the Advances in Photosynthesis Series, initiated by Govindjee, as Volume 2 (Kluwer Academic Publishers, Dordrecht), and now, in 2008, we come to volume 28, entitled The Purple Phototrophic Bacteria (Springer, Dordrecht). The word “phototrophic” is used because it has become clear that many purple bacteria are incapable of photosynthesis (synthesis of all cellular carbon components from CO2), although they are capable of obtaining energy in the form of ATP from light (using light-harvesting and reaction center complexes that are homologous to those of purple photosynthetic bacteria). This latest survey of the field is restricted to the purple bacteria, and does not attempt to cover, for example, the green sulfur bacteria. There have been so many exciting developments since 1995 that the editors felt that it was sufficiently challenging to summarize thirteen years of research on the purple bacteria. This proved to be the case, and 48 chapters and more than 1000 pages were necessary to encompass the depth and breadth of the progress made in studying these fascinating organisms. Since 1995, there has been an explosion of information available from genome sequencing and related projects. The first 3-D structure of the bacterial cytochrome bc1 complex has been determined, and more structural information has been obtained for light-harvesting and reaction center membrane-protein complexes. Site-directed modifications of light-harvesting and reaction center complexes have been correlated with altered spectroscopic properties, even on a femtosecond timescale. Spectroscopic methods in general have advanced considerably since the last volume in 1995; an important example is the development of single
molecule approaches. New theoretical frameworks have had to keep pace with such technical developments. Through the use of atomic force microscopy it has been possible to examine the organization of clusters of individual light-harvesting and reaction center complexes in their native membranes. We are now beginning to see how the properties of native and modified photosynthetic complexes can be harnessed on the nanoscale for the design of biologically-inspired energy and electron transfer devices. In addition to these advances, we are reminded that many new phototrophic bacteria are still being discovered, and although it has been 45 years since the first book on these bacteria appeared, there is much that we do not know or understand. Recent publications indicate that purple phototrophic bacteria are ubiquitous on Earth, and raise questions about their contributions to global cycles of elements. Certainly, the extraordinary metabolic versatility of the purple bacteria, and their amenability to investigation by genetic, biochemical and biophysical techniques, will ensure that despite the inevitably cyclical and variable nature of science funding, there will always be compelling reasons to carry out research on the purple bacteria. Since 1995, it has become ever easier to obtain information online, and it could be argued that the need for a book on this topic is not as compelling as it might have been thirty years ago. However, there is still much value in having a hard copy of the diverse collection of information represented by the 48 chapters of this book, compiled at this point in time and which can be held in the hand. We have attempted to impart some coherence to this project, a process helped by using the organization of the 1995 volume as a starting point. The editors have adopted a pragmatic approach to the issue of taxonomy and, in the light of the ever-changing nature of specific bacterial names, a rigid policy of using only the most recent ones has not been enforced. Thus, Rhodopseudomonas acidophila is now Rhodoblastus acidophilus, for example, but the former name still predominates in the book. Helpful lists of purple phototrophic bacteria, as well as lists of genes, enzymes, pathways, and many more attributes of these bacteria, appear throughout this book. In addition, a section is included at the end of the present volume on new applications and techniques, with the hope that perhaps some of these will form the basis of a xxxi
fifth book, several years from now. This new volume is intended to be a resource for present and future researchers in the fields of ecology, microbiology, biochemistry and biophysics, some of whom might be interested in harnessing the potential of these bacteria as cell factories, or for bioremediation, nanotechnology or solar energy research. We hope that this book will help to attract a new generation of scientists to this field. We thank the authors of all the chapters for entering into the spirit of this project, which is intended to create a lasting work of reference, and a milestone in the field, a staging post on a journey that still has a long way to run.
We are grateful to various individuals who have offered advice, including Govindjee, (the Series Editor), John Golbeck (Editor of the book on Photosystem I, Volume 24 in the Series), Robert Blankenship (one of the Editors of Volume 2 in the Series) and Hugo Scheer (one of the Editors of Volume 25 that focused on Chlorophylls and Bacteriochlophylls), but above all we thank Larry Orr, whose guidance has underpinned the progress of this book. Finally the editors thank their respective families and members of their laboratories for their patience throughout the editing process. May 14, 2008 C. Neil Hunter Krebs Institute for Biomolecular Research Department of Molecular Biology and Biotechnology University of Sheffield Firth Court, Western Bank Sheffield S10 2TN U.K. Email: c.n.hunter@sheffield.ac.uk Fevzi Daldal Department of Biology University of Pennsylvania Philadelphia, PA 19104 U.S.A. Email: [email protected] Marion C. Thurnauer Argonne National Laboratory Chemistry Division, E125 9700 S. Cass Avenue Argonne, Illinois 60439 U.S.A. Email: [email protected] J. Thomas Beatty Dept. of Microbiology & Immunology University of British Columbia Rm. 4556, 2350 Health Sciences Mall Vancouver, BC, V6T 1Z3 Canada Email: [email protected]
xxxii
C. Neil Hunter was born in 1954 in Yorkshire, U.K., and is a Professor in the Department of Molecular Biology and Biotechnology at the University of Sheffield, U.K. He obtained his B.Sc. Degree at the University of Leicester in 1975, where the lectures of Professor Hans Kornberg and a laboratory project with Professor Peter Henderson inspired a lifelong interest in metabolism and bioenergetics. Wanting to learn about photosynthesis, Hunter moved to Bristol University, where the laboratories of Professors Trevor Griffiths, Owen Jones and Tony Crofts were studying many aspects of this topic, from pigment biosynthesis through to electron transport. Hunter studied for his Ph.D. under the guidance of Professor Owen Jones on the topic of membrane assembly in bacterial photosynthesis, and a sabbatical visit of Professor Robert Niederman to the Jones laboratory in Bristol led, in 1978, to a postdoctoral fellowship at Rutgers University, New Jersey. Hunter’s interests in light-harvesting complexes and membrane assembly developed further at this stage, but research in this
area was hampered by a lack of molecular genetic tools to study Rhodobacter sphaeroides. After returning to Bristol to learn about molecular biology in the laboratory of Professor Geoff Turner, it was possible eventually to use transposon Tn5 mutagenesis to gain access to the genes encoding the enzymes for bacteriochlorophyll and carotenoid biosynthesis and to develop a toolkit for site directed mutagenesis of photosynthetic complexes. In 1984, Hunter was appointed to a Lectureship at Imperial College, London, attached to the group of Professor James Barber. He returned to his native Yorkshire in 1988 to a Senior Lectureship at Sheffield University. In 1996 Hunter was awarded a D. Sc. by Bristol University. He is now the Krebs Professor of Biochemistry at Sheffield, and continues to apply a combination of molecular genetic, biochemical, structural and spectroscopic approaches to dissect the pathways for bacteriochlorophyll and carotenoid biosynthesis, and to investigate the assembly, structure, function and organization of bacterial photosynthetic membranes.
xxxiii
Fevzi Daldal is a Professor in the Department of Biology, University of Pennsylvania. He obtained his BS/MS degrees in 1974 at the Institut National des Sciences Appliquées (INSA) de Lyon, France, where he was first introduced into bacterial research in François Stoeber’s group. He followed his graduate studies on the genetics of Escherichia coli cell division at the Université Louis Pasteur, Strasbourg, France, under the guidance of Drs. Raymond Minck (Director of the Microbiology Institute ) and Maxime Schwartz (who was at the Pasteur Institute in Paris, France), obtaining his doctorate in 1977. From 1969 to 1977, his undergraduate and graduate studies were supported by scholarships from the French Government. In 1978, he joined, as a postdoctoral fellow, Dr. Dan G. Fraenkel at Harvard Medical School, Microbiology and Molecular Genetics Department to work on E. coli intermediary metabolism and on the phosphofructokinase II enzyme. After joining the Cold Spring Harbor Laboratory as a Scientist in 1983, Daldal started his studies with purple non sulfur bacteria Rhodobacter species, focusing on cy-
xxxiv
tochromes, which led him to discover the membrane attached electron carrier cytochrome cy, to co-discover the cytochrome cbb3 oxidase as well as novel genes involved in cytochrome biogenesis. His main work on the cytochrome bc1 complex included many studies, extending from the isolation of the structural genes to the crystallization of the enzyme complex to address detailed structure-function aspects using molecular genetics, biochemistry and biophysical approaches with Rhodobacter. Daldal’s current research is centered on the structure, function, regulation, proteomics and biogenesis of membrane cytochromes in Rhodobacter species. His contributions have been recognized by his election as a Member to the Turkish Academy of Science (2000), the American Academy of Microbiology (2005), the American Association for the Advancement of Science (2006) and Chair of the Gordon Conference on Molecular and Cellular Bioenergetics (2007). Daldal was an Editor of the Archives of Microbiology (2000–2006), and currently is a member of the Editorial Board of the Journal of Bacteriology, and a reviewer for many journals.
Marion C. Thurnauer received a B.A.(1968), M.S.(1969),and Ph.D. (1974) in Chemistry, all from the University of Chicago, Illinois, USA. Her thesis research, under the guidance of Gerhard Closs, was in the area of photochemistry; in particular, she studied magnetic interactions in radical pairs using electron paramagnetic resonance (EPR) spectroscopy. In 1974, she became ‘hooked’ on photosynthesis research after accepting a postdoctoral position at the Argonne National Laboratory (ANL), Lemont, IL, USA, to study magnetic resonance properties of excited triplet states of chlorophylls in photosynthetic units and in low temperature solutions. During that period, she and her mentors James R. Norris and Joseph J. Katz demonstrated that the unique electron spin polarization observed in the EPR spectrum of the triplet state of the primary electron donor in the reaction center was due to radical pair induced intersystem crossing. In 1977, she was promoted to staff member at ANL where she served as Group Leader of the Photosynthesis Research Group (1993–1995) and Director of the ANL Chemistry Division (1995–2003). Currently, she is Argonne Distinguished Fellow, Emeritus. Her research involved studies of sequential electron transfer in natural photosynthetic systems of photosynthetic bacteria and oxygenic photosynthesis, using primarily time-resolved EPR methods. She and her
colleagues demonstrated the spin correlated radical pair nature of the transient oxidized primary electron donor and reduced quinone acceptor in photosystem I, and, by comparison in purple photosynthetic bacteria. Her work (together with her colleagues) also extended to development of time-resolved magnetic resonance techniques, particularly for application to study photochemical energy conversion. Collaborating with Tijana Rajh, her research also included EPR studies of photoexcited surface modified nanocrystalline metal oxide colloids to mimic the energy transduction of natural photosynthesis. She served as Chair of the Gordon Research Conference on Photosynthesis: Biophysical Aspects (1994); she has organized several symposia and workshops on magnetic resonance and photosynthesis. During 2002–2003, she served on the Editorial Board of Biophysical Journal. In 2002, she was awarded the 2002 Francis P. Garvan-John M. Olin Medal by the American Chemical Society. Other honors include: the University of Chicago Award for Distinguished Performance at Argonne in 1991; the Agnes Fay Morgan Research Award in 1987; elected as a Fellow of the American Association for the Advancement of Science in 1998. She was the first recipient of the University of Chicago-Argonne Pinnacle of Education Award in 2007.
xxxv
J. Thomas Beatty is a Professor in the Department of Microbiology and Immunology at the University of British Columbia, Canada. He obtained the B.S. degree at the University of Washington in 1976, where a research project under the supervision of James T. Staley sparked Beatty’s interest in phototrophic bacteria. Beatty went on to graduate studies (M.A. in 1978; Ph.D. in 1980) under the guidance of Howard Gest at Indiana University, Bloomington, IN, USA, where he studied the metabolism of purple and green bacteria. Beatty’s postdoctoral research (1980–1983), in Stanley N. Cohen’s laboratory in the Department of Genetics, at the Stanford University School of Medicine, California, USA, included the discovery of differential degradation of light-harvesting 1 (LH1) and reaction center (RC) mRNA segments in the purple bacterium Rhodobacter capsulatus as a process that underlies the relative amounts of LH1 and RC complexes in the cell membrane. After taking up a faculty position at the University of British Columbia in 1983, Beatty has contributed to diverse areas of photosynthesis research: transcrip-
xxxvi
tional regulation of photosynthesis gene expression and elucidation of ‘superoperons’; discovery of the PufX protein and its role in quinone translocation as part of the RC/LH1/PufX core complex; isolation and characterization of new species of phototrophic bacteria; mechanisms of proton translocation into the RC; assembly of photosynthetic complexes; genomics and proteomics of purple bacteria. Beatty’s current research continues on the problems mentioned above, as well as on the ‘gene transfer agent’ (GTA) of Rba. capsulatus. His contributions have been recognized by a prize from the American Society for Microbiology (2002) and a Killam Research Fellowship from the Canada Council for the Arts (2008); he has been featured in popular publications such as the National Geographic and in several newspaper articles. Beatty was an Editor of FEMS Microbiology Letters, and a member of the Editorial Boards of Applied and Environmental Microbiology and the Journal of Bacteriology for many years. He is at present an Editor of Photosynthesis Research.
Author Index Alexandre, Maxime T. A. 811–837 Allen, James P. 337–353 Armitage, Judith P. 643–654 Axelrod, Herbert 323–336
Hillon, Theresa 799–809 Hosler, Jonathan 537–561 Huang, Li-Shar 425–450 Hunter, C. Neil 155–179
Bauer, Carl E. 707–725 Benning, Christoph 119–134 Berry, Edward A. 425–450 Blankenship, Robert E. 17–29 Borsetti, Francesca 655–689 Braun, Paula 913–940 Bullough, Per A. 155–179
Iida, Kouji 861–875
Casadio, Rita 655–689 Choudhary, Madhusudan 691–706 Chua, Toh Kee 811–837 Cogdell, Richard J. 135–153 Cooley, Jason W. 451–473 Csotony, Julius T. 31–55
Kaplan, Samuel 691–706 Kennis, John T. M. 811–837 Klug, Gabriele 727–741 Koch, Hans Georg 407–423 Köhler, Jürgen 877–894 Kramer, David M. 451–473 Kranz, Robert G. 407–423; 759–775 Kriegel, Alison M. 57–79
Dahl, Christiane 595–622 Daldal, Fevzi 407–423; 425–450 Deery, Evelyne 81–95 Dewa, Takehisa 861–875 Donohue, Timothy J. 691–706
Jackson, J. Baz 495–493 Joliot, Pierre 509–536 Jones, Michael R. 295–321 Jung, Deborah O. 1–15 Junge, Wolfgang 475–493
Laible, Philip D. 839–860 Lavergne, Jérôme 509–536 Lee, Dong-Woo 425–450 Loach, Paul A. 181–198
Evans, Katie 799–809 Feniouk, Boris A. 475–493 Fiedor, Leszek 913–940 Fordham-Skelton, Anthony 799–809 Frank, Harry A. 213–230 Frawley, Elaine R. 407–423 Gabrielsen, Mads 135–153 Gardiner, Alastair T. 135–153 Georgiou, Toni 799–809 Groot, Marie-Louise 811–837 Gunner, Marilyn R. 379–405 Haehnel, Wolfgang 895–912 Hanson, Deborah K. 839–860 Harwood, Caroline S. 577–594 Hellingwerf, Klaas J. 811–837 Hendriks, Johnny 811–837
Mackenzie, Chris 691–706 Madigan, Michael T. 1–15 Martelli, Pier Luigi 655–689 Masepohl, Bernd 759–775 Masuda, Shinji 727–741 Mielke, Donna L. 839–860 Miyashita, Osamu 323–336 Nango, Mamoru 861–875 Niederman, Robert A. 253–273 Nitschke, Wolfgang 451–473 Novoderezhkin, Vladimir I. 231–252 Noy, Dror 895–912 Obiozo, U. Mirian 495–493 Okamura, Melvin 323–336 Onder, Ozlem 407–423
xxxvii
Papiz, Miroslav 799–809 Parkes-Loach, Pamela S. 181–198 Parson, William W. 355–377 Pérez, Marcela Ávila 811–837 Polívka, Tomáš 213–230 Poluektov, Oleg G. 953–973
Tabita, F. Robert 563–576 Takaichi, Shinichi 97–117 Tamot, Banita 119–134 Turkarslan, Serdar 407–423
Qian, Pu 155–179
van Grondelle, Rienk 231–252 van der Horst, Michael A. 811–837 van Wilderen, Luuk J. 811–837 Verméglio, André 509–536 Vignais, Paulette M. 743–757
Raymond, Jason 17–29 Robert, Bruno 199–212 Robinson, Brigitte R. 707–725 Romagnoli, Simona 563–576 Sander, Johannes 595–622 Sanders, Carsten 407–423 Scheer, Hugo 895–912 Scheuring, Simon 941–952 Schoepp-Cothenet, Barbara 537–561 Schulten, Klaus 275–294 Şener, Melih K. 275–294 Setterdahl, Aaron 707–725 Shapleigh, James P. 623–642 Sturgis, James N. 253–273 Swingley, Wesley D. 17–29
Utschig, Lisa M. 953–973
Warren, Martin J. 81–95 Warshel, Arieh 355–377 Williams, JoAnn C. 337–353 Willows, Robert D. 57–79 Wraight, Colin A. 379–405 Wu, Jiang 707–725 Yurkov, Vladimir 31–55 Zannoni, Davide 537–561; 655–689 Zeilstra-Ryalls, Jill Helen 777–798
xxxviii
Color Plates
Fig. 1. Schematic comparison of photosynthesis gene clusters (PGC) from various purple bacteria. Lines represent transpositions of syntenous gene regions between two PGC. Numbers annotated above the Rhodospirillum rubrum PGC represent the location on the genome of each separate PGC segment. Genes colored in green represent BChl biosynthesis genes, orange represent carotenoid biosynthesis genes, red represent structural genes of the photosynthetic apparatus, blue represent regulatory genes and other colors represent unique genes. The black and blue lines with arrows represent inversion of the genes contained between the lines and the red lines represent shifts in position without inversions. See Chapter 2, p. 20.
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. CP1–CP16. © 2009 Springer Science + Business Media B.V.
Color Plates
Fig. 2. Photosynthetic and habitat diversity of aerobic anoxygenic phototrophs (AAP) A–E. In vivo absorbance spectra of cells grown under different physiological conditions, with numerals above peaks denoting LH and RC wavelengths. Insets, photographs of liquid cultures. A. Roseicyclus mahoneyensis, illuminated (dashed line) and dark (solid line). B,C. Dashed line, rich organic medium (3 g·l–1 organics); solid line, minimal acetate (1 g·l–1) medium (B. Roseococcus thiosulfatophilus, C. Erythrobacter litoralis). D. Erythromicrobium ramosum, rich organic medium, dark (dashed line) and oligotrophic medium, light:dark regimen of 12h:12h, illuminated with diffuse ambient sunlight (solid line). E. Citromicrobium bathyomarinum strain JF1, rich organic medium (dashed line), minimal glucose medium (dotted line), compared with Erythrobacter litoralis (solid line). F. Hypothetical electron transfer system of aerobic anoxygenic photosynthesis, showing electron flow through major carriers: P870, special pair of BChl in RC (photoexcited state indicated by ‘*’); BChl-BPh, accessory BChl and bacteriophaeophytin in the RC; QA, quinone primary electron acceptor; Cyt bc1, cytochrome bc1 complex; Cyt c2, cytochrome c2. The symbol ‘+’ indicates that the midpoint potential of QA in all tested AAP is positive and higher than in anaerobic phototrophs. G–I. Extreme environment habitats of AAP. G. Hydrothermal vent field in Eastern Pacific Ocean, showing smoker chimney. H. Hypersaline spring system (East German Creek) in Manitoba, Canada: spring pool in foreground and playa in background, with white patches of salt precipitates. I. Cyanobacterial-Thiothrix mat development in thermal springs of Banff National Park, Alberta, Canada. See Chapter 3, p. 39.
CP2
Color Plates
Fig. 3. Representations of the β/α-protomer building blocks that comprise LH2. The β-chains are shown in blue and the α-chains in green. The BChls are shown in red and the carotenoid in pale blue. a) A view of the α/β-protomer with the pigments removed. The main BChl liganding amino acid side chains are shown. b) A view of how two adjacent β/α-protomers interact with each other, again with the pigments removed. c) Similar to b) but looking down onto the two adjacent α/β-protomers from the C-terminal side of the complex. d) The same view as a) but with the pigments included. e) A view of the α/β-protomer from Phaeospirillum (Phs.) molischianum. The main differences between d) and e) are the altered orientation of the B800 bacteriochlorin ring and the different organization of the C-terminal region of the α-apoprotein. f) A view of the α/β-protomer from Rps. acidophila strain 7050. Note the different amino acids shown at the C-terminal end of the α-apoprotein that are responsible for the altered absorption properties of this B800-820 complex. The structures shown in a–d are representations of the structure of the B800-850, LH2 complex from Rps. acidophila strain 10050. See Chapter 8, p. 138.
Fig. 4. The LH2 complex from Rhodopseudomonas (Rps.) acidophila strain 10050. a) A view looking down on the top of the complex from the presumed periplasmic surface of the membrane. (b) A side view looking from within the presumed photosynthetic membrane with the periplasmic side of the complex uppermost. In both panels the LH2α polypeptide is green, LH2β is blue, B850 BChls are purple, B800 BChls are brown and the carotenoids are red. See Chapter 8, p. 140.
CP3
Color Plates
Fig. 5. (A) Dynamics of the transient absorption (TA) spectrum measured upon 1017 nm excitation of the LH1 of Blc. viridis at 77 K (Monshouwer et al., 1998). (B) Measured (points) and calculated (solid lines) red-shift of the TA spectrum between 0 and 400 fs upon 1017 nm excitation of the LH1 of Blc. viridis at 77K (Novoderezhkin and van Grondelle, 2002). (C) Measured (points) and modeled (solid line) TA kinetics at 1050 nm traces upon 1055 nm excitation of the LH1 of Blc. viridis at 77K (Novoderezhkin et al., 2000). See Chapter 13, p. 240.
Fig. 6. Coherent dynamics of the density matrix for a single LH2 complex, i.e., calculated without relaxation of populations and coherences for one realization of the disorder corresponding to a blue-shifted (top), an intermediate (middle), and a red-shifted (bottom) FL spectrum (Novoderezhkin et al., 2006). The initial population corresponds to the thermal equilibrium at room temperature (shown in the inserts by blue, green, and red for the blue-shifted, intermediate, and red-shifted spectra, respectively), initial coherences have arbitrary fixed phases. The figure shows the diagonal elements of the density matrix in the site representation ρ(n,n) as a function of time. The color scale is used to indicate the absolute values of the density matrix from zero (blue) to the maximal value (red). See Chapter 13, p. 247.
CP4
Color Plates
Fig. 7. Structural models of monomeric and dimeric core complexes. In each case a side view and a space-filling representation is used. For the RC the complete structure is present (shown in blue), but for the LH1 (red) and helix W/PufX components (green) only the transmembrane helices are shown. a. The monomeric RC-LH1-W core complex from Rhodopseudomonas palustris (Roszak et al., 2003). b. The dimeric RC-LH1-PufX core complex of Rhodobacter (Rba.) sphaeroides. The coordinates of the LH1 αβ pairs were adopted from Roszak et al. (2003), fitted into the tilted 3-D dimer structure from single particle analysis (Chapter 9, Bullough et al.), and adjusted according to the cryo-EM projection map in Qian et al. (2005). The transmembrane domain of PufX was taken from Tunnicliffe et al. (2006). See Chapter 9, p. 172.
Fig. 8. Organization of the photosynthetic unit (PSU) from the supramolecular architecture to individual chlorophylls and the energy transfer network. The high-light adapted vesicle model for Rba. sphaeroides (Şener et al., 2007) is shown. The three different constituent proteins and their BChls are colored as follows: LH2: green, LH1: red, RC: blue. From left to right three different rendering styles are featured. On the left, the light-harvesting proteins are shown in the tube representation (backbone only, no pigments); in the middle, the proteins are rendered transparently while the BChls are visible represented by their porphyrin rings; on the right, only the BChls are shown but this time together with their respective electronic couplings (Eq. 1). For simplicity only the strongest couplings are shown. The vesicle construction depicted here is based on atomic force microscopy data on intracytoplasmic membranes (Bahatyrova et al., 2004). Figure made with VMD (Humphrey et al., 1996). See Chapter 15, p. 288.
CP5
Color Plates
Fig. 9. Structure of the Cyt c2 : RC complex. A) Side view of the complex. The Cyt c2 is bound with heme positioned directly over the BChl2. A small region of close contact is indicated by the circle. A conformational change, indicated by the arrow, is due to crystal packing interactions with an adjacent bound Cyt. B) View of the structure in the region between the cofactors. The heme and BChl are separated by the planar aromatic ring of Tyr (L162). The closest distance between the conjugated rings in the cofactors is 14.2 Å as shown by the long arrow. A strong tunneling pathway between the two cofactors is indicated by the short arrows. C) Open book view of the interface between the complexes. The two proteins are separated by rotating around the center line. The central region of shortrange contacts (indicated by the circle) is surrounded by charged residues involved in long-range electrostatic interactions. Interacting residues are color coded. The contact between the heme (green) and the ‘hot spot’ Tyr L162 (green) is shown by the arrow. Hydrophobic residues Leu (M191), Val (M192) (black) on the RC interact with Phe (C102) on the Cyt (black). Hydrogen bonds (cyan) are formed between atoms on RC and Cyt. A cation-pi interaction is formed between Tyr (M295) and Arg (C32). Electrostatic interactions occur between negatively charged groups on the RC (red) and positively charged groups on the Cyt (violet). Other groups in van der Waals contact are shown in yellow. Modified from Axelrod et al. ( 2002). See Chapter 17, p. 326.
CP6
Color Plates
Fig. 10. The cytoplasmic surface of a model LM preparation of the Rba. sphaeroides RC. (a) The LM heterodimer is shown as a solid object, with the L- and M-polypeptides shown in yellow and white, respectively, and the quinones shown as spheres with green carbons and red oxygens. The view is along the symmetry axis, and parts of the QA quinone are visible from the outside of the complex. (b) View as in (a) but with amino acids overlaying the quinone sites shown in stick format, revealing the underlying quinones (carbons in yellow (L) or white (M), oxygens in red and nitrogens in blue). In both panels the dotted ovals show the approximate positions of the QA and QB quinones. See Chapter 16, p. 298.
Fig. 11. Region around the HB site in the wild-type and AM149W RC complexes from Rba. sphaeroides. The protein is shown as a solid object, with the L- and M-polypeptides shown in yellow and white, respectively, and bound waters shown in cyan. The BB and QB cofactors are shown as green and beige solid objects, respectively. In (a) the HB cofactor is shown as blue sticks and the phytol chain of PB as red sticks. In (b) the phytol chain of PB in the normal conformation is shown as red sticks and the phytol chain of PB in the alternate conformation is shown as orange sticks. See Chapter 16, p. 302.
CP7
Color Plates
Fig. 12. The reaction center (RC) complex from Rhodobacter sphaeroides comprises three subunits, a heterodimer of similar, but non-identical L (yellow) and M (blue) subunits, and subunit H (green), which caps LM on the cytosolic side of the membrane. The LM dimer binds all the cofactors, while subunit H stabilizes the structure and is involved in H+-ion uptake and transfer associated with electron transfer to the quinones. The L and M subunits and all associated cofactors are arranged around a quasi-2-fold rotational symmetry axis, normal to the plane of the membrane and passing through the primary donor (P), the special pair dimer of bacteriochlorophylls (BChl), and a ferrous (Fe2+) iron midway between the two quinones. Electron transfer proceeds from the excited singlet state of the primary donor (P*), via the A-branch of cofactors — monomer BChl (BA) and BPhe (HA), bound to the L subunit — to the primary quinone, QA, which is bound in a fold of the M subunit. From QA– the electron crosses the symmetry axis to the secondary quinone, QB, bound in a similar fold in the L subunit. (Figure prepared in VMD.) See Chapter 20, p. 381.
Fig. 13. The quinone binding sites — QA (top) and QB (bottom) — are related by the two-fold, rotational symmetry axis of the reaction center, which passes through the iron atom of the acceptor quinone complex. Both quinones are hydrogen bonded through the carbonyl oxygens — hydrogen bonds shown in purple for QA (top) and in green for QB (bottom). The binding sites are predominantly hydrophobic, with the notable exception of GluL212, AspL213 and SerL223 in the QB site, which constitute terminal components of the H+ delivery pathway to QB. Note that the orientations of the methoxy groups of QA and QB do not follow the two-fold rotational symmetry. (Figure prepared in VMD.) See Chapter 20, p. 385.
CP8
Color Plates
Fig. 14. The Ccm-system I components for c-type Cyt maturation. All components of the c-type Cyt maturation, except the thiol-disulfide oxidoreductase DsbA, are located in the cytoplasmic membrane. Both apoCyt c and heme follow different routes to the heme ligation core complex, composed of CcmI, CcmH and CcmF. Apocyt c is translocated via the Sec pathway; its cysteine thiols in the conserved CXXCH motif are first oxidized by the DsbA-DsbB pathway and then reduced by the Cyt c maturation specific CcdA-CcmG and/or CcmH thio-reductive pathway. CcmI is involved in delivering apoCyt c to the core heme ligation complex via its different domains. Heme is translocated across the membrane, possibly via ABC–type transporter CcmABCD and is covalently attached to the conserved His residue of the heme chaperone CcmE. CcmC is involved in attaching heme to CcmE and CcmD enhances holo-CcmE production. CcmA and CcmB promote release of holo-CcmE from CcmC and CcmD. Upon formation of the thioether bonds between the apoCyt c and the heme vinyls, catalyzed by the CcmH-CcmI-CcmF complex, mature holoCyt c is released. See Chapter 21, p. 410.
Fig. 15. The Ccs-system II components for c-type Cyt maturation. Similarly to system I, all components of the system II c-type Cyt maturation pathway are located in the cytoplasmic membrane, with the exception of DsbA. The apoCyt c is translocated across the membrane by the Sec machinery and is oxidized by DsbA, then reduced by CcsX. The transmembrane proteins CcsB and CcsA form a complex that represents the system II synthetase (heme ligation core complex). The reduced apoCyt c is likely bound by the periplasmic domain of CcsB. Heme is translocated by the CcsBA complex, probably through CcsA and binds to the CcsA periplasmic WWD domain before ligation to the reduced apoCyt c. Following covalent ligation of the heme vinyl groups to the reduced cysteines of the apoCyt c CXXCH motif, the holoCyt c folds into its mature form with release from the complex. See Chapter 21, p. 412.
CP9
Color Plates
Fig. 16. Overall structure of the Rhodobacter (Rba.) capsulatus Cyt bc1. Dimeric Cyt bc1 is depicted in the membrane, indicating the likely positions of the boundaries of the membrane lipid bilayer (A). The lower panels show a monomeric Cyt bc1 with the Fe/S protein ED in the b (B) or c1 (C) positions. The substrate Cyt c2 (1C2R.pdb) is modeled into its likely reactive position based on the yeast co-crystal structure (1KYO.pdb). The regions of the Cyt bc1 structure that differ significantly from the mitochondrial enzyme are colored in red, otherwise Cyt b is shown in different shades of blue, Cyt c1 in green, the Fe/S protein in yellow, and Cyt c2 in magenta. Hemes are shown in red, and the [2Fe2S] cluster is shown as a spacefilled structure in yellow and orange. The Qo site inhibitor stigmatellin locks the Fe/S protein ED in the b position. See Chapter 22, p. 427.
Fig. 17. Stereoview of structural details of Q sites in the Rba. capsulatus Cyt bc1. Panel A, helices α-a, A, D, and E making the Qi pocket are colored in light pink, light pink, wine, and green, respectively. Heme bH is shown in gray and quinone bound to the Qi site in yellow as ball-and-stick models. Panels B and C, the surface of Cyt b, showing interaction of the Fe/S protein ED with the ‘binding crater’ at the Qo site when the ED is in the ‘b’ (B) or ‘c1’ (C) positions. Cyt b is depicted using space-filling representation with different regions colored as described below. The white outline demarks the position of the Fe/S protein ED. Four regions surrounding the binding crater, the cd1-cd2 helices, ef-loop, gh loop and α-ef2 portion are indicated in green, magenta, yellow and cyan, respectively. At the bottom-center of the binding crater stigmatellin is shown in red. The side chains of the H135 and H156 residues of the Fe/S protein subunit are shown as ball-and-stick models. Atoms of Cyt b that are in contact (i. e., less than 3.9 Å) with the Fe/S protein ED are colored in gray, and Y302 (light green, with gray OH atom) in the α-ef helix and K329 (lime green with gray Ce and Nz atoms) between the α-ef 2 and F helices are indicated by arrows. See Chapter 22, p. 434.
CP10
Color Plates
Fig. 18. Similarities and differences in the electron transfer chains and core structures of the cyt bc1 and cyt b6 f complexes. The gross functional monomeric structures of the cyt b (blue), cyt c (red) and the ISP (yellow) coordinating portions of the cyt bc1- (left) or cyt b6 ftype (right) complexes are illustrated side-by-side as matching ribbon structures with the cofactors coordinated in each drawn as space filled spheres (colored as the individual subunits to which they belong). The glaring differences of the presence of an extra c-type heme (heme ci) adjacent to the Qi (or QN) site of the cyt b6 f, as well as the presence of the non-redox active chlorophyll molecule (dark green) in this same complex, are readily seen in these side-by-side illustrations. However, in each case, a similar high (red) and low potential (blue) chain of cofactors facilitating bifurcated electron flow following QH2 oxidation is easily envisioned. Both types of complex are shown with the inhibitor, stigmatellin (light green), bound at the Qo (or Qp) site as well as amino acids, pictured as stick models (blue), thought to be important for Qi (or QN) site substrate binding to aid in visually identifying the two spatially distinct Q binding sites. The illustrations have been worked up from the Rhodobacter (Rba.) capsulatus and Mastigocladus laminosus derived atomic coordinate files 1ZRT.pdb and 1VF5.pdb, respectively. See Chapter 23, p. 454.
CP11
Color Plates
Fig. 19. Side reactions that bypass the Q-cycle. Reactions shown are those that differ from the Q-cycle in the fate of the SQ generated at the Qo site. Details are given in the text. See Chapter 23, p. 461.
(a)
(b)
Fig. 20. (a) The hydride-transfer site in 2OO5.pdb. The dI(B) component is shown in blue and dIII in green. The nucleotides are in standard atom colors (except that the C4 atoms are in yellow and are linked by the pink dashed line). H-bonds are show as green dashed lines. (b) The loop E ‘lid’ of dIII and the ‘mobile loop’ of dI closed down over the hydride transfer site of 2OO5.pdb. The nucleotides (NADP+ in dIII, and the NADH analog, H2NADH, in dI) are shown in a space-filling format. The dI(B) polypeptide is shown in green (with its mobile loop in cyan), and the dI(A) polypeptide in yellow. The dIII polypeptide is shown in pink (with its loop E in magenta). Also shown are the RQD loop of dI and helix D/loop D of dIII (see text). See Chapter 25, p. 504.
CP12
Color Plates
Fig. 21. In response to a short flash of light photosynthetic reaction centers generate transmembrane voltage and release quinol (QH2); subsequent oxidation of QH2 by the cytochrome bc1 complex is coupled to proton pumping into the chromatophore lumen. The resultant ~ + drives protons through the ATP synthase and powers ATP synthesis. See Chapter 24 (inset from Fig 3A), page 483. ∆µ H
CP13
Color Plates (i) (a)
(ii)
(iii)
(b)
(c)
(i)
(ii)
Fig. 22. Localization of cytoplasmic chemosensory proteins and loss of localization and position in PpfA mutant. (a) Rba. sphaeroides with a soluble chemoreceptor TlpC-GFP fusion and McpG-GFP, expressed from genomic replacements. (b) Fluorescence images of cells of Rba. sphaeroides expressing both (i) CheW2-YFP and (ii) CheW4-CFP from genomic replacements showing separation of the two chemotaxis pathways; (iii) overlay image. (c)(i) Cephalexin treated Rba. sphaeroides with genomically expressed TlpT-CFP showing localization at single cell distances through the filament. (ii) Cephalexin treated Rba. sphaeroides ∆PpfA showing loss of localization of the chemosensory clusters. The scale bars are 2 µM. See Chapter 32, p. 650.
Fig. 23. BLUF photocycle and active site structure. Panel a: active site receptor state. Panel b: active site signaling state. Panel c: typical photocycle of BLUF-domains. Panels a and b were prepared using the program PyMOL (http://www.pymol.org). Color coding for panels a and b: Grey, Carbon; Blue, Nitrogen; Red, Oxygen; Orange, Phosphorus; Green, hydrogen bond.The structure coordinate file of the BLUF domain of AppA from Rba. sphaeroides (PDB ID: 1YRX) (Anderson et al., 2005) was used to create both panels. Note that for panel b the positions of the side-chain nitrogen and oxygen of Q63 were manually switched to indicate the Q-flip. See Chapter 41, p. 818.
CP14
Color Plates
Fig. 24. The X-ray structure of the chromophore binding domain (CBD) from Deinococcus (D.) radiodurans (Wagner et al., 2005). The PAS-2 (blue) and GAF (yellow) domains are aligned, relative to one another, by a trefoil knot formed from a loop of the GAF domain (yellow) and residues 1-35 (green) N-terminal to PAS-2. Biliverdin IXα (BV-purple) is covalently bound to Cys24 and is located in a pocket within the GAF domain formed from a β-sheet and two α-helices. See Chapter 40, p. 806.
Fig. 25. A small angle X-ray scattering (SAXS) ab-initio envelope (transparent orange) of Bph4 from Rhodopseudomonas palustris (Evans et al., 2006). The structure is a homodimer and is modelled with the crystal structures of the CBD (green) (Wagner et al., 2005) and histidine kinase (HK, cyan) (Marina et al., 2005). The positions of the BV (purple), phosphorylating histidine (red) and ATP (orange) are shown. There is no high resolution structure of a PHY domain that can be fitted, but it is assumed to occupy the envelope between CBD and the HK domains. See Chapter 40, p. 807.
CP15
Color Plates
Fig. 26. Top:Helical net representation of shielding helices Si and binding helices Bj. The amino acids that were varied (Xk for Si and Zl for Bj) are emphasized in gray, the ligating histidine of the binding helices in green. The tables contain the amino acids of the individual helices at the varied positions. Long, solid arrows indicate the N→C direction of the peptide chains, dashed arrows the dipole moments of salt bridges. Bottom: photograph of a cellulose membrane carrying 38 modular proteins, after incubation with Ni-BPheide in buffer/ DMF and washing in buffer. Columns 2-7 correspond to identical binding helices Bj (j = 7 – 12), rows 1 - 6 to identical shielding helices Si (i = 1 - 6). The empty spots in columns 1 and 8 correspond to modular proteins that have been punched out for mass spectrometry; the two isolated spots in the rightmost column were not used for the analyses. See Chapter 45, p. 902.
Fig. 27. Stages of LH1 formation in the presence of carotenoids in vitro. See Chapter 46, page 934.
CP16
Chapter 1 An Overview of Purple Bacteria: Systematics, Physiology, and Habitats Michael T. Madigan* and Deborah O. Jung Department of Microbiology, Southern Illinois University, Carbondale, IL 62901, U.S.A.
Summary .................................................................................................................................................................. 2 I. Introduction......................................................................................................................................................... 2 II. Systematics of Purple Bacteria .......................................................................................................................... 3 A. Purple Sulfur Bacteria .......................................................................................................................... 4 B. Purple Nonsulfur Bacteria .................................................................................................................... 4 III. Physiology of Purple Bacteria ............................................................................................................................ 4 A. Purple Sulfur Bacteria .......................................................................................................................... 4 B. Purple Nonsulfur Bacteria .................................................................................................................... 6 1. Photoheterotrophy....................................................................................................................... 6 2. Dark Growth ................................................................................................................................ 7 3. Nitrogen Fixation ......................................................................................................................... 7 IV. Habitats of Purple Bacteria................................................................................................................................. 7 A. Purple Sulfur Bacteria .......................................................................................................................... 7 1. Blooms in Stratified Lakes........................................................................................................... 7 2. Microbial Mats ............................................................................................................................. 8 B. Purple Nonsulfur Bacteria .................................................................................................................... 9 1. Sewage ....................................................................................................................................... 9 2. Purple Nonsulfur Bacteria in Waste Lagoons ............................................................................. 9 V. Purple Bacteria in Extreme Environments.......................................................................................................... 9 A. Thermophilic Purple Bacteria ............................................................................................................. 10 1. Thermochromatium tepidum .................................................................................................... 10 2. Other Thermophilic Purple Bacteria .......................................................................................... 10 B. Halophilic and Alkaliphilic Purple Bacteria ......................................................................................... 10 C. Acidophilic Purple Bacteria ................................................................................................................ 11 D. Purple Bacteria from Permanently Cold Habitats............................................................................... 11 E. Environmental Limits to Photosynthesis in Purple Bacteria ............................................................... 12 VI. Final Remarks .................................................................................................................................................. 12 Acknowledgments ................................................................................................................................................... 12 References .............................................................................................................................................................. 12
*Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 1–15. © 2009 Springer Science + Business Media B.V.
2
Michael T. Madigan and Deborah O. Jung
Summary Anoxygenic phototrophic purple bacteria are a major group of photosynthetic microorganisms widely distributed in nature, primarily in aquatic habitats. Nearly 50 genera of these organisms are known and some have become prime model systems for the experimental dissection of photosynthesis. Purple sulfur bacteria differ from purple nonsulfur bacteria on both metabolic and phylogenetic grounds, but species of the two major groups often coexist in illuminated anoxic habitats in nature. Purple sulfur bacteria are strong photoautotrophs and capable of limited photoheterotrophy, but they are poorly equipped for metabolism and growth in the dark. By contrast, purple nonsulfur bacteria, nature’s preeminent photoheterotrophs, are capable of photoautotrophy, and possess diverse capacities for dark metabolism and growth. Several purple bacteria inhabit extreme environments, including extremes of temperature, pH, and salinity. Collectively, purple bacteria are important phototrophs because they (1) consume a toxic substance, H2S, and contribute organic matter to anoxic environments by their autotrophic capacities; (2) consume organic compounds, primarily non-fermentable organic compounds, in their roles as photoheterotrophs; and (3) offer scientists in the photosynthesis community a smörgasbord of molecular diversity for the study of photosynthesis. I. Introduction Anoxygenic phototrophic purple bacteria are a major group of phototrophic microorganisms that inhabit aquatic and terrestrial environments. Purple bacteria that inhabit oxic habitats and which carry out photosynthesis only aerobically are called ‘aerobic anoxygenic phototrophs’ and are covered in Chapter 3 (Yurkov and Csotonyi). The current chapter covers only the classical purple bacteria: purple sulfur bacteria and purple nonsulfur bacteria. Purple bacteria are photosynthetic gram-negative prokaryotes that convert light energy into chemical energy by the process of anoxygenic photosynthesis. Purple bacteria contain photosynthetic pigments–bacteriochlorophylls and carotenoids — and can grow autotrophically with CO2 as sole carbon source. Many genera of purple bacteria are known and the organisms share many basic properties with their nonphototrophic relatives. Some general characteristics of purple bacteria are listed in Table 1. Purple bacteria share with oxygenic phototrophic prokaryotes — the cyanobacteria — the ability to conserve energy by photophosphorylation. However, unlike cyanobacteria and aerobic anoxygenic phototrophs, photosynthesis in purple bacteria only occurs under anoxic (O2-free) conditions. This is also true of the other classical anoxygenic phototrophs: green sulfur bacteria, green nonsulfur bacteria, and Abbreviations: BChl – bacteriochlorophyll; LH – light-harvesting; Rba. – Rhodobacter; Rcy. – Rhodocyclus; Rfx. – Rhodoferax; Rps. – Rhodopseudomonas; Rsp. – Rhodospirillum; Tch. – Thermochromatium
the heliobacteria (Blankenship et al., 1995). Purple bacteria require anoxic conditions for phototrophic growth because pigment synthesis in these organisms is repressed by molecular oxygen (Cohen-Bazire et al., 1957). Thus, the competitive success of purple bacteria in nature requires both light and anoxic conditions. This combination is most commonly found in lakes, ponds, estuaries, and other aquatic environments where H2S is present (Pfennig, 1967, 1978a, 1989). Once these general conditions are met, the exact physiochemical nature of the habitat (sulfide concentration, pH, light quality and intensity, temperature) controls the abundance and diversity of purple bacteria that develop there (Pfennig, 1978a, 1989; Madigan, 1988). Purple bacteria participate in the anoxic cycling of carbon both as primary producers (CO2 fixation, photoautotrophy) and as light-stimulated consumers of reduced organic compounds (photoheterotrophy). In certain habitats particularly favorable for their development, purple bacteria have been shown to be significant primary producers (Czeczuga, 1968; Takahashi and Ichimura, 1968; Overmann et al., 1994, 1996, 1999). However, in most illuminated sulfidic habitats the role of purple bacteria as H2S consumers is probably more important than any contribution they make to primary production; H2S is a highly poisonous substance for plants and animals and also for many bacteria. The oxidation of sulfide by purple bacteria yields nontoxic forms of sulfur, such as elemental sulfur (S0) and sulfate (SO42–). Sulfide oxidation thus allows the upper waters of a lake to remain oxic and suitable for plants, animals, and aerobic bacteria.
Chapter 1
Overview of Purple Bacteria
3
Table 1. General properties of anoxygenic purple phototrophic bacteriaa Property
Examples
Groups/phylogeny
Purple sulfur bacteria (gammaproteobacteria); purple nonsulfur bacteria (alphaor betaproteobacteria) Allochromatium vinosum and Thiocapsa roseopersicina (purple sulfur bacteria); Rhodobacter capsulatus, Rhodobacter sphaeroides, Rhodospirillum rubrum, and Rhodopseudomonas palustris (purple nonsulfur bacteria)
Major species studied
Pigments/color of dense cell suspensions
BChl a or b; major carotenoids include spirilloxanthin, spheroidene, lycopene, and rhodopsin, and their derivatives; cell suspensions purple, purple-red, purpleviolet, red, orange, brown, or yellow brown (BChl a-containing species); green or yellow (BChl b-containing species)
Location of pigments in cells
Within intracytoplasmic membranes arranged as membrane vesicles, tubes, bundled tubes, or in stacks resembling lamellae
Absorption maxima of living cells
BChl a-containing species: near 800 nm, and anywhere from 815–960 nm; BChl b-containing species: 835–850 nm and 1010–1040 nm
Electron donors/sulfur globulesb
H2S, S0, S2O32–, H2, Fe2+; if S0 is produced from the oxidation of sulfide, the S0 is stored intracellularly only in certain purple sulfur bacteria (see Fig. 1a)
Photoheterotrophy/dark respiratory growth
Purple sulfur bacteria limited on both accounts; purple nonsulfur bacteria typically diverse on both accounts
a
All purple bacteria are gram-negative prokaryotes. All species contain peptidoglycan and an outer membrane containing lipopolysaccharide. bVirtually all purple bacteria are capable of autotrophic growth. When growing autotrophically, the Calvin cycle (reductive pentose phosphate cycle) is used as the mechanism for CO2 fixation
II. Systematics of Purple Bacteria Purple sulfur bacteria and purple nonsulfur bacteria were originally distinguished on physiological grounds based on their tolerance and utilization of sulfide. Purple sulfur bacteria were species that tolerated millimolar levels of sulfide and oxidized sulfide to sulfur globules stored intracellularly (Fig. 1a), while purple nonsulfur bacteria were species that did neither (van Niel, 1932, 1944). However, classic chemostat experiments by Hansen and van Gemerden (1972) showed that these criteria for classifying purple bacteria were not absolute. At low levels of sulfide, typically less than 0.5 mM, most species of purple nonsulfur bacteria will grow and in so doing, oxidize sulfide to S0, S4O62–, or SO42–. Nevertheless, an important distinction in the sulfide metabolism of purple sulfur and purple nonsulfur bacteria remains: any S0 formed by purple nonsulfur bacteria is not stored intracellularly, but instead is deposited outside the cell (Hansen and van Gemerden, 1972; Brune, 1995) (species of Ectothiorhodospiraceae are an exception here). Thus, when grown on sulfide, it is easy to differentiate a purple sulfur from a purple nonsulfur bacterium because of the microscopically obvious globules of S0 formed (Fig. 1a). Subsequent isolations of purple nonsulfur bacteria from highly sulfidic habitats have shown that
Fig. 1. Phase photomicrographs of phototrophic purple bacteria. (a) Cells of a strain of the purple sulfur bacterium Thermochromatium tepidum isolated from a New Mexico (USA) hot spring. Note the bright refractile intracellular sulfur globules (arrows). (b) Cells of Rhodobaca bogoriensis, an alkaliphilic purple nonsulfur bacterium isolated from Lake Bogoria (Kenya). If purple nonsulfur bacteria oxidize sulfide, any S0 they produce remains outside the cell. Cells of Tch. tepidum are about 1.5 µm wide and cells of Rbc. bogoriensis are about 0.8 µm wide.
many species of this group are actually quite sulfide tolerant. For example, both the marine species Rhodobacter (originally Rhodopseudomonas, Rps.) sulfidophilus and the cold-active species Rhodoferax antarcticus (isolated from the sulfidic bottom waters
4 of the permanently frozen freshwater Lake Fryxell, McMurdo Dry Valleys) can tolerate over 4 mM sulfide (Hansen and Veldkamp, 1973; Jung et al., 2004). This concentration of sulfide is toxic to many purple sulfur bacteria (Pfennig, 1967, 1978a, 1989)! The original classification of purple bacteria on the basis of sulfide metabolism has been supported by molecular criteria. Phylogenetic analyses of purple bacteria based on comparative 16S rRNA sequencing have shown that purple sulfur bacteria are species of gammaproteobacteria while purple nonsulfur bacteria are either alpha- or betaproteobacteria (Imhoff et al., 2005) (Tables 1 and 2).
Michael T. Madigan and Deborah O. Jung phic species (Imhoff et al., 2005). If one considers the fact that the pigments and photocomplexes in the different species of purple nonsulfur bacteria are very similar, this suggests that the acquisition of phototrophic capacity in purple nonsulfur bacteria has occurred by lateral gene transfer. Sequence analyses of photocomplex proteins have confirmed this (Nagashima et al., 1997). Table 2 also lists the three letter genus name abbreviations for both purple sulfur and nonsulfur bacteria; these abbreviations will be used throughout this book. III. Physiology of Purple Bacteria
A. Purple Sulfur Bacteria Over 25 genera of purple sulfur bacteria are now recognized, consisting of a variety of morphological types (Table 2). Purple sulfur bacteria include both species that store S0 inside the cell (family Chromatiaceae, see Fig. 1a), and those that produce extracellular S0 (Ectothiorhodospiraceae) (Table 1). [It should be noted that research on the mechanism of sulfide oxidation by Allochromatium vinosum has shown that the ‘intracellular sulfur’ produced by this organism actually accumulates in the periplasm rather than in the cytoplasm (Pattaragulwanit et al., 1998); this is likely true for all species of Chromatiaceae as well]. Most laboratory studies of purple sulfur bacteria have focused on Allochromatium and Thiocapsa species (Table 1) since these are the most easily grown. Many species of purple sulfur bacteria are ‘extremophilic’ species, including in particular, species that grow best at high salt and/or pH (Table 2). B. Purple Nonsulfur Bacteria Twenty genera of purple nonsulfur bacteria are now recognized (Table 2). Species of Rhodobacter and Rhodopseudomonas have been the workhorses for laboratory studies of anoxygenic photosynthesis (Table 1). But many other interesting species, some of which have one or more unusual metabolic features, are also known. For example, extremophilic species inhabiting hot, cold, salty, alkaline (Fig. 1b), and acidic environments have been isolated and will be discussed in a later section of this chapter (see section V). As shown in Table 2, all purple nonsulfur bacteria are proteobacteria, and phylogenetic trees show the various species to be closely related to nonphototro-
Purple bacteria are relatively easy to grow in laboratory culture; in most cases all that is needed is an anoxic mineral medium supplemented with either sulfide plus bicarbonate (photoautotrophic growth) or an organic compound (photoheterotrophic growth). Because of this, and because anoxygenic photosynthesis is a simpler form of photosynthesis than the oxygenic process, purple bacteria have emerged as ideal model systems for dissecting the physiology, biochemistry and molecular biology of photosynthesis. Moreover, anoxygenic photosynthesis preceded oxygenic photosynthesis on Earth by billions of years. Thus, studies of purple and other anoxygenic phototrophs have contributed in major ways to our understanding of the evolution of photosynthesis (Raymond et al., 2003; Chapter 2, Swingley et al.). A. Purple Sulfur Bacteria The physiology of purple sulfur bacteria is intimately linked to sulfide, and large populations of purple sulfur bacteria are observed in nature only in illuminated environments where sulfide is present (Pfennig, 1967, 1978a, 1989). This implies that the growth of purple sulfur bacteria in nature is primarily phototrophic. If growth is photoautotrophic, sulfide, thiosulfate or H2 are used as photosynthetic electron donors (Trüper and Fischer, 1982; Madigan, 1988; Brune, 1995). A few species can also use ferrous (Fe2+) iron as an electron donor, oxidizing it to ferric (Fe3+) iron (Ehrenreich and Widdel, 1994), and at least one species, a strain of Thiocapsa, can use nitrite (NO2–) as photosynthetic electron donor, oxidizing it to nitrate (NO3–. In addition to autotrophic growth, a few organic carbon sources are photoassimilated by purple sulfur
Chapter 1
Overview of Purple Bacteria
5
Table 2. Genera of anoxygenic phototrophic purple bacteria Taxonomy/Phylogeny Purple Nonsulfur Bacteria Alphaproteobacteria
Betaproteobacteria
Purple Sulfur Bacteria Gammaproteobacteria Family Chromatiaceae b
Family Ectothiorhodospiraceae b
Genus
Genus abbreviation a
Morphology
Rhodobaca c Rhodobacter Rhodovulum Rhodopseudomonas c Rhodoblastusc Blastochloris Rhodomicrobium Rhodobium Rhodoplanes Rhodocistac Rhodospirillum Phaeospirillum Rhodopila c Rhodospira Rhodovibrio c Rhodothallasium c Roseospira Roseospirillum Rhodocyclus Rhodoferaxc Rubrivivax
Rca. Rba. Rdv. Rps. Rbl. Blc. Rmi. Rbi. Rpl. Rcs. Rsp. Phs. Rpi. Rsa. Rhv. Rts. Ros. Rss. Rcy. Rfx. Rvi.
Cocci to short rods Rods Rods-Cocci Budding rods Budding rods Budding rods Budding rods Rods Rods Spirilla Spirilla Spirilla Cocci Spirilla Vibrio Spirilla Spirilla Spirilla Curled vibrios Rods, vibrios Rods, curved rods
Allochromatium Amoebobacter Chromatium Halochromatium c Isochromatium Lamprobacter Lamprocystis Marichromatium Rhabdochromatium Thermochromatium c Thioalkalicoccus c Thiobaca Thiocapsa Thiococcus Thiocystis Thiodictyon Thioflavicoccus Thiohalocapsac Thiolamprovum Thiopedia Thiorhodococcus Thiorhodovibrio Thiospirillum
Alc. Amb. Chr. Hch. Isc. Lpb. Lpc. Mch. Rbc. Tch. Tac. Tba. Tca. Tco. Tcs. Tdc. Tfc. Thc. Tlp. Tpd. Trc. Trv. Tsp.
Rods Cocci in plates or clumps Rods Rods Rods Rods Cocci in clusters Rods Rods Rods Cocci Rods Cocci Cocci Cocci to short rods Rods forming aggregates Cocci Cocci Cocci Cocci, often in plates Cocci Vibrios to spirilla Spirilla
Ectothiorhodospira c Ect. Vibrios to spirilla Hlr. Vibrios to spirilla Halorhodospira c Trs. Vibrios to spirilla Thiorhodospira c Ectothiorhodosinus Ets. Rods a Abbreviations in accordance with Imhoff and Madigan (2004). bSpecies of Chromatiaceae store sulfur from the oxidation of sulfide intracellularly (see Fig. 1A); species of Ectothiorhodospiraceae do not. c Contain one or more extremophilic species growing at an extreme of temperature, pH, or salinity greater than marine salinity
6 bacteria. Organic acids and fatty acids are the preferred substrates, but short-chain alcohols and even carbohydrates are used by certain species (Sojka, 1978). Photoheterotrophic growth of Allochromatium vinosum and other Allochromatium species that are capable of assimilatory sulfate reduction does not require sulfide. However, some purple sulfur bacteria will not grow without sulfide and are also nutritionally quite restricted. These include the large-celled Chromatium species such as Chromatium okenii and Chromatium weissei, and Thiospirillum (Trüper, 1978), as well as the thermophilic species Thermochromatium (Tch.) tepidum (Madigan, 1986) (Fig. 1a). Sulfide is required for growth of these species and the only organic compounds that are photoassimilated are acetate and pyruvate (Trüper 1981; Madigan, 1986). Dark growth of some purple sulfur bacteria is possible. For example, certain Chromatiaceae, including species of Allochromatium, Thiocystis, Amoebobacter, and Thiocapsa, can grow in darkness as either chemoorganotrophs or chemolithotrophs when the oxygen concentration is significantly reduced [microaerobic growth; Kämpf and Pfennig, 1980]. Thiocapsa roseopersicina and Thiocystis violacea are the most oxygen tolerant purple sulfur bacteria (Kondratieva et al., 1976; Kämpf and Pfennig, 1980); however, respiratory growth of these species is very slow compared with phototrophic growth. If one considers that dark growth of purple sulfur bacteria in nature puts them in direct competition with nonphototrophic bacteria as well as with purple nonsulfur bacteria, the ecological significance of dark metabolism by purple sulfur bacteria is probably minor. It is more likely that dark energy metabolism by purple sulfur bacteria helps these organisms survive intermittently oxygenated environments or is used as a means to generate ATP at night, rather than being a major means of supporting extended growth in nature (van Gemerden, 1968). B. Purple Nonsulfur Bacteria Purple nonsulfur bacteria are a physiologically versatile group of purple bacteria that can grow well both phototrophically and in darkness. Growth of some purple nonsulfur bacteria, for example, Rhodobacter (Rba.) capsulatus, is possible under phototrophic conditions with either CO2 or organic carbon, or in darkness by respiration, fermentation, or chemolithotrophy. This makes Rba. capsulatus probably the
Michael T. Madigan and Deborah O. Jung most metabolically versatile of all known bacteria (Madigan and Gest, 1979). Carbon metabolism in purple nonsulfur bacteria has been summarized in the excellent reviews by Tabita (1995) and by Gibson and Harwood (1995); see also Chapter 28, Romagnoli andTabita; and Chapter 29, Harwood. 1. Photoheterotrophy Under phototrophic (anoxic/light) conditions, typical purple nonsulfur bacteria can grow photoautotrophically with H2 or low levels of sulfide as electron donors; a few species can use S2O32– or Fe2+ as photosynthetic electron donors (Ehrenreich and Widdel, 1994; Brune, 1995). However, most purple nonsulfur bacteria grow best as photoheterotrophs in media containing a readily useable organic compound, such as malate or pyruvate, and ammonia as nitrogen source (Sojka, 1978). Yeast extract is a common addition to media formulated for purple nonsulfur bacteria (Biebl and Pfennig, 1981). Yeast extract is a source of B-vitamins, one or more of which are required by the majority of recognized species of purple nonsulfur bacteria. Requirements for thiamine, nicotinic acid, biotin, and p-aminobenzoic acid are the most common. Requirements for B-complex vitamins have never been observed in purple sulfur bacteria, although many species require vitamin B12, a growth factor required by only a handful of purple nonsulfur bacteria (Pfennig, 1978b; Siefert and Koppenhagen, 1982). However, beyond its role as a source of vitamins, yeast extract also stimulates the growth of purple nonsulfur bacteria because of its assortment of organic compounds that can fuel photoheterotrophic growth. Several individual organic compounds support photoheterotrophic growth of purple nonsulfur bacteria. Organic acids, amino acids, fatty acids, alcohols, carbohydrates, and even C-1 compounds are metabolized by different species (Sojka, 1978; Trüper and Pfennig, 1981). With minor exceptions, the citric acid cycle intermediates malate, succinate, and fumarate are universally used, as are pyruvate and acetate; many species also use ethanol, lactate, and propionate (Sojka, 1978; Trüper and Pfennig, 1981). A few purple nonsulfur bacteria photoassimilate aromatic compounds such as benzoate, hydroxy derivatives of benzoate, and cyclohexane carboxylate (Gibson and Harwood, 1995). Enrichment cultures employing benzoate as carbon source typically yield strains of Phaeospirillum (formerly Rhodospirillum)
Chapter 1
Overview of Purple Bacteria
fulvum or Rps. palustris (Gibson and Harwood, 1995). Benzene is not utilized by these or other purple nonsulfur bacteria, but at least one aromatic hydrocarbon, toluene, supports photoheterotrophic growth of certain strains of Blastochloris sulfoviridis (Zengler et al., 1999). Growth of purple nonsulfur bacteria on aliphatic hydrocarbons has not been described. 2. Dark Growth Many of the same organic compounds that are photoassimilated by purple nonsulfur bacteria can also be used as electron donors and carbon sources for dark respiratory growth. Oxygen tolerances for respiratory growth vary among species, but some, such as Rhodobacter species, can be grown with vigorous aeration (Madigan, 1988). Certain purple nonsulfur bacteria can grow under anoxic dark conditions by either fermentation or anaerobic respiration. For example, pyruvate (Uffen and Wolfe, 1970; Gurgen et al., 1976) and certain sugars (Madigan and Gest, 1978; Schultz and Weaver, 1982) support fermentative growth of some purple nonsulfur bacteria, most notably Rhodospirillum (Rsp.) rubrum and Rba. capsulatus. Extensive fermentative growth of Rba. capsulatus requires addition of an accessory oxidant such as dimethyl sulfoxide or trimethylamine-N-oxide (Madigan and Gest, 1978; Schultz and Weaver, 1982). Rba. sphaeroides is capable of true denitrification, reducing NO3– to N2 using nonfermentable carbon sources as electron donors (Satoh et al., 1976). Dark chemolithotrophic growth of certain species of purple nonsulfur bacteria is possible using H2 or S2O32– as electron donors. In Rba. capsulatus, chemolithotrophic growth on H2 occurs and the organism can be grown in a synthetic medium supplied with the gases H2, O2, and CO2 as electron donor, electron acceptor, and carbon source, respectively (Madigan and Gest, 1979). Whether chemolithotrophy is a significant growth strategy for purple bacteria in nature is unknown, but it is likely that the ability to conserve energy from the oxidation of inorganic electron donors gives purple bacteria an added physiological dimension in competition with nonphototrophic bacteria. 3. Nitrogen Fixation With only a couple of known exceptions, purple nonsulfur bacteria can fix nitrogen (N2 + 8H → 2NH3 + H2) (Madigan, 1995). The Rhodobacter species Rba.
7 capsulatus and Rba. sphaeroides grow most rapidly with N2 as sole nitrogen source and show the highest rates of in vivo nitrogenase activity (Madigan et al., 1984). Consequently, Rhodobacter species tend to dominate enrichment cultures for purple nonsulfur bacteria that employ nitrogen fixation as a selective condition. Because in general purple nonsulfur bacteria are excellent nitrogen-fixing bacteria (Madigan et al., 1984), it is likely that the capacity for diazotrophy confers a significant competitive advantage on them in anoxic environments that are limited in fixed nitrogen. IV. Habitats of Purple Bacteria A. Purple Sulfur Bacteria A detailed description of the major habitats of purple bacteria including a wealth of specific examples can be found in the reviews of Madigan (1988), Pfennig (1967, 1978a, 1989), and van Gemerden and Mas (1995). Large masses (blooms) of purple sulfur bacteria often develop in sulfidic aquatic ecosystems exposed to light. Although blooms of phototrophic sulfur bacteria may occur in shallow lagoons polluted by sewage (which triggers the activities of sulfate-reducing bacteria), densely stratified ‘plates’ of purple bacteria form only in the deep waters of lakes protected from excessive wind mixing and which contain sufficient sulfate to support sulfate reduction in the sediments. 1. Blooms in Stratified Lakes Intense microbial activity occurs in the sediments of productive stratified lakes. Organic material reaching the bottom waters is catabolized by fermentation, which releases a variety of reduced organic products, including lactate, ethanol, and fatty acids. In freshwater lakes containing low levels of sulfate, these fermentation products can be photoassimilated by purple nonsulfur bacteria or converted to methane by the cooperative interactions of syntrophic bacteria and methanogenic Archaea. Alternatively, if electron acceptors such as Fe3+ or NO3– are available, the fermentation products will fuel anaerobic respirations supported by these electron acceptors before extensive occurs. If sulfate is present, sulfate-reducing bacteria
8 will be active in the sediments forming sulfide; the sulfide diffuses upwards from the sediments into the water column forming a gradient. Sulfide triggers the growth of purple sulfur bacteria, which develop in specific zones of the water column where light and sulfide are optimal (Pfennig, 1967, 1975, 1978a). If cell numbers are sufficiently high, the lake water itself will become pigmented red, purple, or reddishbrown (Pfennig 1978a, 1989; Overmann et al., 1994, 1996, 1999). When this occurs, it is often possible to identify the major genera of purple bacteria present by simple microscopic examination. In many stratified lakes the bloom of purple bacteria consists of a mixture of species (Caldwell and Tiedje, 1975a,b), while in others, the bloom may contain only a single species (Overmann et al., 1994, 1996, 1999). A nice example of the layering of phototrophic purple bacteria in stratified lakes can be found in the work of Caldwell and Tiedje (1975a,b). These workers examined water samples collected at 1 m intervals from two eutrophic lakes — Wintergreen and Burke — located in southwest Michigan (USA). In these studies, done before the days of molecular microbial ecology, the purple sulfur bacteria could be adequately identified by their characteristic morphologies. Both lakes contained species of Thiopedia, Thiospirillum, Thiocystis, and Chromatium. However, in Burke Lake the Thiospirillum population dominated while in Wintergreen Lake, the Thiopedia population dominated. Although physiochemical profiles of the two lakes were not performed, the dominant population in each case was likely selected by the major characteristics — light, sulfide, pH, dissolved organic carbon, and the like — that defined each lake (Caldwell and Tiedje, 1975a,b). Both Burke and Wintergreen lakes also contained green sulfur bacteria in the waters beneath the purple sulfur bacteria. The green bacterial population contained phototrophic consortia (Overmann and Schubert, 2002) such as ‘Chlorochromatium’ (Caldwell and Tiedje, 1975a,b). Green bacteria can outcompete purple bacteria at low light intensities because they possess large antenna pigment structures called chlorosomes (Kimble and Madigan, 2002; Frigaard and Bryant, 2004). Green bacteria also use different regions of the spectrum than do purple bacteria and are typically much more sulfide tolerant. Thus, green bacteria can exist beneath a layer of purple bacteria in a water column, and this pattern is common in stratified lakes (Pfennig, 1967, 1975, 1978a, 1989).
Michael T. Madigan and Deborah O. Jung In hypersulfidic shallow lakes, such as those in the karstic Banyoles area of northern Spain, millimolar levels of sulfide are present and phototrophic sulfur bacteria bloom throughout the water column. In Lake Cisó, the best studied of these lakes, the entire water column is anoxic, and the lake turns bright red during an active bloom (Guerrero et al., 1985). Both green and purple sulfur bacteria are present in Lake Cisó but the lake is dominated by a Chromatium (probably Allochromatium) species during periods of the most extensive bloom. 2. Microbial Mats Purple sulfur bacteria are also common in microbial mats, including mats that form in marine or hypersaline environments (van Gemerden and Mas, 1995) and in the effluents of thermal springs (Castenholz and Pierson, 1995). Microbial mats are laminated organo-sedimentary structures composed primarily of filamentous cyanobacteria and anoxygenic phototrophs, such as Chloroflexus, but often contain purple bacteria as well. As new growth occurs from the top of the mat, the lower mat layers decompose and sulfate-reduction typically occurs. This supplies the sulfide necessary to trigger growth of purple bacteria, usually purple sulfur bacteria. Mat thickness can vary considerably. In siliceous alkaline hot spring microbial mats, mats can be 4–5 cm thick. Mats containing only purple bacteria, such as those of Tch. tepidum that form in the Mammoth hot springs of Yellowstone are much thinner, up to 0.5 cm in thickness (Ward et al., 1989). Thiocapsa and Allochromatium species are common inhabitants of marine microbial mats. These purple sulfur bacteria often form a dense pigmented layer between the cyanobacteria and the lower layers of the mat. In this niche purple bacteria oxidize sulfide that diffuses upwards from below before it reaches the cyanobacterial layers (van Gemerden and Mas, 1995). Thiocapsa roseopersicina, in particular, is very common in marine mats, probably because of its metabolic versatility. Besides its photoautotrophic and photoheterotrophic capacities, this purple sulfur bacterium can grow in darkness by heterotrophic and chemolithotrophic means (Kondratieva et al., 1976). This versatility allows Thiocapsa roseopersicina to take full advantage of the variable growth conditions that characterize different layers of microbial mats (van Gemerden and Mas, 1995).
Chapter 1
Overview of Purple Bacteria
B. Purple Nonsulfur Bacteria Purple nonsulfur bacteria occasionally form dense blooms in habitats where levels of sulfide are either low or undetectable. Purple nonsulfur bacteria are usually present in only low numbers in blooms of purple sulfur bacteria, probably because of their sulfide sensitivity. Instead of photoautotrophy, purple nonsulfur bacteria specialize in photoheterotrophy. Although this puts them in competition with heterotrophs for organic compounds, photoheterotrophic capacity likely confers a significant selective advantage on purple nonsulfur bacteria. This is because unlike heterotrophs, phototrophs do not need to conserve energy from the carbon sources they photoassimilate; carbon goes almost quantitatively into cell material. However, no known purple nonsulfur bacteria can hydrolyze major polymeric substances such as cellulose or starch, and so ultimately, the phototrophs depend on the heterotrophs to generate the lowmolecular-weight compounds they photoassimilate (Pfennig, 1978a). 1. Sewage Purple nonsulfur bacteria are present in sewage (Holm and Vennes 1970; Siefert et al., 1978). In a detailed study by Siefert et al. (1978), it was shown that the mean number of purple nonsulfur bacteria in a sewage plant in Göttingen, Germany was highest in the activated sludge stage of the treatment process; counts fluctuated between 105 and 106 cells ml–1 but were never greater than 106 cells ml–1 (as measured using plate counting techniques). Purple sulfur bacteria, on the other hand, were quantitatively insignificant in sewage and were detectable by culture only from activated sludge (~103 cells ml–1). A variety of purple nonsulfur bacteria were identified in the sewage plant, including Rba. sphaeroides and Rba. capsulatus, Rps. palustris and Rps. (now Blastochloris) viridis, Rhodocyclus (now Rubrivivax) gelatinosus and Rhodocyclus (Rcy.) tenuis and Rsp. photometricum. Rba. sphaeroides, Rubrivivax gelatinosus, Rps. palustris, and Rba. capsulatus made up the bulk of the purple bacteria present. Despite these relatively high numbers, it was concluded that purple nonsulfur bacteria probably played only a minor role in organic matter transformations in sewage compared with heterotrophic bacteria, which were present at 108 to 109 cells ml–1. Purple nonsulfur bacteria were also detectable in the strictly anaerobic (and dark)
9 sewage sludge digestor, but these likely represented only transient cells traveling through the system (Siefert et al., 1978). 2. Purple Nonsulfur Bacteria in Waste Lagoons Waste lagoons offer excellent conditions for growth of purple nonsulfur bacteria (Jones, 1956; Cooper et al., 1975; Kobayashi, 1975). For example, a pigmented bloom of purple nonsulfur bacteria was reported from the waste lagoon of a vegetable canning plant in Minnesota (USA); prolific growth leading to intensely red-colored lagoons occurred and the bloom was associated with a significant reduction in odor (Cooper et al., 1975). Rba. sphaeroides, Rba. capsulatus, and Rps. palustris were the key species in this bloom, and it is likely that their consumption of volatile fatty acids produced by fermentation led to the odor reduction observed. The morphologically and phylogenetically unique purple nonsulfur bacterium Rcy. purpureus, isolated by Norbert Pfennig over 30 years ago, was the dominant phototroph in a swine waste lagoon in Iowa (USA) (Pfennig, 1978b). This organism, of which only a single strain has ever been isolated, probably thrived on the combination of organic constituents present in the waste materials. However, it is of interest that Rcy. purpureus, one of the only purple nonsulfur bacteria to lack a nitrogenase system (Madigan et al., 1984; Madigan, 1995), was the dominant purple bacterium in this particular habitat and that it has never been reported from elsewhere. One would expect a swine waste lagoon to be high in amines and thus that nitrogen fixation would be unnecessary. It is thus possible that Rcy. purpureus is somehow selected for in otherwise suitable habitats for purple nonsulfur bacteria that are very high in ammonia and volatile amines. Since Rcy. purpureus is easily cultured, an enrichment study of its distribution in nature using media containing elevated levels of ammonia and amines could yield insight on its ecology. V. Purple Bacteria in Extreme Environments Purple bacteria have been isolated from extreme environments, including hot, cold, acidic, alkaline, and hypersaline (Madigan, 2003). Unfortunately, with a few exceptions, these inherently interesting ‘ex-
10 tremophilic’ purple bacteria have been little utilized in the study of photosynthesis thus far. However, the success of these unusual purple bacteria in their harsh habitats implies that they have evolved important solutions to photosynthesis under stress conditions. We can therefore learn much from studying them. For example, molecular adaptations linked to photosynthesis under extreme conditions should be relatively easy to identify in extremophilic purple bacteria by applying a bioinformatics/structural biology approach to their genomes. A. Thermophilic Purple Bacteria The first extremophilic purple bacteria were discovered in the 1960s and were either halophiles or acidophiles, including extremely halophilic species of the genus Ectothiorhodospira. Into the late 1970s, several new halophilic and haloalkaliphilic purple bacteria were discovered. In the 1980s, the first thermophilic purple bacterium was isolated, Tch. tepidum. Since then, a large diversity of alkaliphilic and halophilic purple bacteria has been isolated. Psychrophilic phototrophs have been described only very recently, with two representatives currently in culture. 1. Thermochromatium tepidum Purple bacteria were first identified in Yellowstone microbial mats over 30 years ago (Castenholz, 1977). But it was not until the 1980s that the purple sulfur bacterium Thermochromatium (originally Chromatium) tepidum was isolated in pure culture (Madigan, 1984, 1986). Tch. tepidum (Fig. 1a) is thermophilic (optimumtemperature ~50 ºC, maximum temperature 57 ºC) and produces a novel light-harvesting (LH) 1 photopigment complex that absorbs maximally near 920 nm (Garcia et al., 1986; Nozawa et al., 1986). The Tch. tepidum LH1 (core) antenna complex has been studied in connection with the mechanism of energy transfer to the reaction center. A biophysical conundrum exists with this photocomplex in that its absorption maximum is 50 nm to the red of that of the reaction center. Nevertheless, because there is a small overlap between spectra of the two components, efficient energy transfer occurs from the Tch. tepidum LH1 complex to the reaction center (Kramer and Amesz, 1996). Examples of long-wavelength-absorbing core antenna complexes even more spectacular than that of Tch. tepidum have been discovered, indicating that in purple bacteria, LH1 complexes
Michael T. Madigan and Deborah O. Jung that absorb very far to the red (963 nm) can still transfer energy to the reaction center (Permentier et al., 2001). The photosynthetic reaction center of Tch. tepidum is similar in most respects to that of other purple bacteria, except for its increased thermal stability (Nozawa and Madigan, 1991). To probe the mechanism behind this, the Tch. tepidum reaction center was crystallized. From this work, key substitutions were identified in the L and M subunits of the Tch. tepidum reaction center that are likely responsible for the thermostability of this photocomplex (Nogi et al., 2000). In addition to thermal stable photocomplexes, a thermophilic ribulose bisphosphate carboxylase (a key enzyme of the Calvin cycle) of the green plant type was characterized from Tch. tepidum and shown to be stable to at least 60 ºC (Heda and Madigan, 1988. 1989). 2. Other Thermophilic Purple Bacteria Other mildly thermophilic purple bacteria (optimum growth temperature ~40 ºC) have been cultured from hot spring microbial mats. These include the BChl b-containing species Rhodopseudomonas sp. strain GI, isolated from a New Mexico hot spring (Resnick and Madigan, 1989), and Rps. cryptolactis (Statwald-Demchick et al., 1990) and Rsp. centenum (Rhodocista centenaria) (Favinger et al., 1989), both isolated from a Thermopolis (Wyoming, USA) hot spring. Rhodocista centenaria in particular has been useful as a model organism for biochemical/genetic research on phototaxis and related issues of motility (see, for example McClain et al., 2002), and its genome has recently been sequenced (C. E. Bauer, personal communication). B. Halophilic and Alkaliphilic Purple Bacteria Several extremophilic purple bacteria are halophilic or haloalkaliphilic (Imhoff et al., 1978, 1979). These include purple sulfur bacteria such as Ectothiorhodospira, Halorhodospira, Halochromatium, Marichromatium, and Thiohalocapsa, and purple nonsulfur bacteria such as Rhodovibrio, Rhodothalassium, Rhodobium, Rhodovulum, and Roseospira. Collectively, these purple bacteria have salt optima that range from seawater salinities to over 20% NaCl (Imhoff, 2001). Interestingly, the Dead Sea purple nonsulfur bacterium Rhodovibrio sodomensis shows a distinct intermediate level salt requirement (optimum
Chapter 1
Overview of Purple Bacteria
at 8–11% NaCl) (Mack et al., 1993), which is very near that of its habitat. Some purple sulfur bacteria can grow in saturated salt solutions, making them the most halophilic of all known phototrophic bacteria (Imhoff, 2001). In the 1990s several new purple bacteria were isolated from low salinity soda lakes. Most of these differed dramatically from known halophilic or haloalkaliphilic species in that they required little if any NaCl for growth. These isolates are, however, strongly alkaliphilic (pH optima near 9) and phylogenetically distinct. These include purple nonsulfur bacteria such as Rhodobaca (Milford et al., 2000), and purple sulfur bacteria such as Thioalkalicoccus (Bryantseva et al., 2000), and Thiorhodospira (Bryantseva et al., 1999). Rhodobaca (Fig. 1b) is of particular interest because it lacks a peripheral (LH2) antenna complex, a rarity among purple nonsulfur bacteria (Glaeser and Overmann, 1999), and it produces several unusual carotenoids that render phototrophic cultures yellow in color (Takaichi et al., 2001). Rhodobaca shows various metabolic peculiarities as well, including an inability to fix N2 and to grow photoautotrophically (Milford et al., 2000); both of these properties are hallmarks of purple nonsulfur bacteria (Madigan, 1988). C. Acidophilic Purple Bacteria The list of acid-loving anoxygenic phototrophs is short, as only two genera (three species) are known. Rhodoblastus acidophilus (formerly Rhodopseudomonas acidophila) is common in mildly acidic environments, such as bogs, marshes, and acidic lakes. In the original description of this organism, several strains were described, some containing orange/brown carotenoids and others purple/red carotenoids. The strains are otherwise similar, and all show a lower limit for growth near pH 4 (Pfennig, 1969). A very similar organism is the species Rhodoblastus sphagnolica (Kulichevskaya et al., 2006). Rhodopila globiformis was isolated from acidic warm sulfide springs (pH 3.5–4) that flow out along the Gibbon River in Yellowstone National Park (Pfennig, 1974). Rhodpila globiformis-like organisms have also been obtained from springs that feed into Nymph Lake, a warm, acidic and sulfidic lake adjacent to the Gibbon River; the pH of the Nymph Lake springs is 3 (Madigan, 2003). Rhodopila globiformis has a low pH optimum for growth similar to that of Rhodoblastus species, but, if carefully tested, would likely be more
11 acid tolerant than Rhodoblastus species because of the more strongly acidic nature of its habitat. Phylogenetically, Rhodopila and Rhodoblastus are quite distinct. No acidophilic purple sulfur bacteria are known and this is likely because at acid pH, sulfide would exist exclusively as H2S, its most toxic form. D. Purple Bacteria from Permanently Cold Habitats Permanently cold environments are habitats for cold-active purple bacteria (Burke and Burton, 1988; Madigan, 1998). The Madigan laboratory has been studying phototrophs that inhabit lakes in the McMurdo Dry Valleys of Antarctica. These are closed basin lakes with a biology that is exclusively microbial, and they remain permanently frozen with ice covers of 4–7 m. Purple nonsulfur bacteria have been isolated from samples of microbial mats that develop along the edge of the lakes as well as from water under the ice. Molecular studies using pufM as a measure of the biodiversity of purple bacteria suggest that species related to known purple nonsulfur bacteria reside in these lakes (Karr et al., 2003). The purple bacterium Rhodoferax antarcticus inhabits both microbial mats and the water column of Lake Fryxell (77º S latitude), a lake supporting major sulfur cycling activities (Sattley and Madigan, 2006). Lake Fryxell is unmixed and weakly stratified, with saline bottom waters overlain by freshwater. A gradient of sulfide is present in Lake Fryxell from micromolar levels at a depth of 9 m to nearly 1.5 millimolar near the sediments (~18 m). Despite these nearly perfect conditions for the development of phototrophic sulfur bacteria, no evidence for purple sulfur bacteria has emerged from enrichment culture or nucleic acid probing studies using pufM (Achenbach et al., 2001; Karr et al., 2003). Instead, Rfx. antarcticus, a very sulfide tolerant purple nonsulfur bacterium (Jung et al., 2004), seems to dominate anoxygenic photosynthesis in Lake Fryxell. Although Rfx. antarcticus is not strongly psychrophilic (optimal growth occurs at 18 ºC and no growth occurs above 24 ºC), it is the first anoxygenic phototroph to show a distinct cold adaptation and growth at 0 ºC (Madigan et al., 2000). Also in the water column of Lake Fryxell, gas vesiculate purple nonsulfur bacteria are present, including a morphologically unique strain of Rfx. antarcticus (betaproteobacteria) and a rod-shaped organism related to Rhodobacter species (alphaproteobacteria). These gas vesiculate
12 purple bacteria position themselves in zones of the water column where photosynthesis can occur optimally (Karr et al. 2003; Jung et al., 2004). Biomass measurements have shown that they localize in the water column at a depth of 10 m and that cell numbers drop off sharply below this depth and are undetectable above this depth (Madigan, unpublished). E. Environmental Limits to Photosynthesis in Purple Bacteria Work with extremophilic purple bacteria has given a good indication of the limits to which photosynthesis can be pushed. Photosynthesis in purple bacteria can occur at temperatures up to at least 57 ºC and down to 0 ºC, pH values as low as 3 or as high as 11, and at salinities up to and including saturated solutions of NaCl (~32%). Although these are probably not the absolute physiochemical limits to photosynthesis in anoxygenic phototrophs, they are likely to be very close to the limits. It is notable that the green nonsulfur bacterium Chloroflexus aurantiacus, which contains a purple bacterial-type photosynthetic reaction center (Achenbach et al., 2001), can grow up to 70 °C (Castenholz and Pierson, 1995). This holds out hope that purple bacteria capable of growth at temperatures above 57 °C (the upper limit for Tch. tepidum, Madigan, 1986) may exist in nature. Further exploration for new purple bacteria in sulfidic hot springs with temperatures above 60 °C should answer this question. VI. Final Remarks Our understanding of purple bacteria goes far beyond what has been discussed here. Through the years purple bacteria have become increasingly important as research tools for the study of basic problems in photosynthesis and have contributed in major ways to our understanding of the biochemistry, genetics and evolution of photosynthesis. And importantly, the beautiful colors and metabolic versatility of purple bacteria continue to attract talented young people to the field. The reader will see the fruits of their labors as well as those of the more seasoned investigators in the following chapters where the biology of purple bacteria will unfold in a spectacular way.
Michael T. Madigan and Deborah O. Jung Acknowledgments Research in the Madigan laboratory is supported by the United States National Science Foundation, most recently by grant MCB0237576. MTM thanks all of his former students and postdoctoral colleagues who did research on purple bacteria in his laboratory. These include Rich Masters, Rick Stegeman, Glenn Wright, Carletta Ooten, Erin Mack, Joseph Mayers, Ghanshyam Heda,Vasiliki Karayiannis, Sol Resnick, Chad Rubin, Gerrit Hoogewerf, Ike Pantazopoulous, Amy Milford, Linda Kimble, Amy Stevenson, Terry Locke, Tiffany Full, Jen Carey, Elizabeth Karr, Jill Crespi, Mahmoud Tayah, Tom Wahlund, Matt Sattley, and Marie Asao. References Achenbach LA, Carey JR and Madigan MT (2001) Photosynthetic and phylogenetic primers for detection of anoxygenic phototrophs in natural environments. Appl Environ Microbiol 67: 2922–2926 Biebl H and Pfennig N (1981) Isolation of members of the family Rhodospirillaceae. In: Starr MP, Stolp H, Trüper HG, Balows A and Schlegel HG (eds) The Prokaryotes — a Handbook on Habitats, Isolation and Identification of Bacteria, pp 267–273. Springer-Verlag, New York Blankenship RE, Madigan MT and Bauer CE (1995) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2). Kluwer Academic Publishers, Dordrecht Brune DC (1995) Sulfur compounds as photosynthetic electron donors. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 847–870. Kluwer Academic Publishers, Dordrecht Bryantseva IA, Gorlenko VM, Kompantseva EI and Imhoff JF (2000) Thioalkalicococcus limnaeus gen. nov., sp. nov., a new alkaliphilic purple sulfur bacterium with bacteriochlorophyll b. Int J Syst Bacteriol 50: 2157–2163 Bryantseva IA, Gorlenko VM, Kompantseva EI, Imhoff JF, Süling J and Mityushina L (1999) Thiorhodospira sibirica gen. nov., sp. nov., a new alkaliphilic purple sulfur bacterium from a Siberian soda lake. Int J Syst Bacteriol 49: 697–703 Burke CM and Burton HR (1988) Photosynthetic bacteria in meromictic lakes and stratified fjords of the Vestfold Hills, Antarctica. Hydrobiologia 165: 13–23 Caldwell DE and Tiedje JM (1975a) A morphological study of anaerobic bacteria from the hypolimnia of two Michigan lakes. Can J Microbiol 21: 362–376 Caldwell DE and Tiedje JM (1975b) The structure of anaerobic bacterial communities in the hypolimnia of several Michigan lakes. Can J Microbiol 21: 377–385 Castenholz RW (1977) The effect of sulfide on the blue-green algae of hot springs II. Yellowstone National Park. Microbial Ecology 3: 79–105 Castenholz RW and Pierson BK (1995) Ecology of thermophilic
Chapter 1
Overview of Purple Bacteria
anoxygenic phototrophs. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Phototrophic Bacteria, pp 87–103. Kluwer Academic Publishers, Dordrecht Cohen-Bazire G, Sistrom WR and Stanier RY (1957) Kinetic studies of pigment synthesis by non-sulfur purple bacteria. J Cell Comp Physiol 49: 25–68 Cooper DE, Rands MB and Woo C-P (1975) Sulfide reduction in fellmongery effluent by red sulfur bacteria. J Water Pollution Control Fed 47: 2088–2100 Czeczuga B (1968) Primary production of the purple sulphuric bacteria, Thiopedia rosea Winogr. (Thiorhodaceae). Photosynthetica 2: 161–166 Ehrenreich A and Widdel F (1994) Anaerobic oxidation of ferrous iron by purple bacteria, a new type of phototrophic metabolism. Appl Environ Microbiol 60: 4517–4526 Favinger J, Stadtwald R and Gest H (1989) Rhodospirillum centenum, sp. nov., a thermotolerant cyst-forming anoxygenic photosynthetic bacterium. Ant van Leeuwenhoek 55: 291–296 Frigaard NU and Bryant DA (2004) Seeing green bacteria in a new light: Genomics-enabled studies of the photosynthetic apparatus in green sulfur bacteria and filamentous anoxygenic phototrophic bacteria. Arch Microbiol 182: 265–276 Garcia D, Parot P, Verméglio A and Madigan MT (1986) The light-harvesting complexes of a thermophilic purple sulfur photosynthetic bacterium Chromatium tepidum. Biochim Biophys Acta 850: 390–395 Gibson J and Harwood CS (1995) Degradation of aromatic compounds by nonsulfur purple bacteria. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 991–1003. Kluwer Academic Publishers, Dordrecht Glaeser J and Overmann J (1999) Selective enrichment and characterization of Roseospirillum parvum, gen. nov. and sp. nov., a new purple nonsulfur bacterium with unusual light absorption properties. Arch Microbiol 171: 405–416 Griffin BM, Schott J and Schink B (2007) Nitrite, an electron donor for anoxygenic photosynthesis. Science 316: 1870 Guerrero R, Montesinos E, Pedrós-Alió C, Esteve I, Mas J, van Gemerden H, Hofman PAG and Bakker JF (1985) Phototrophic sulfur bacteria in two Spanish lakes: Vertical distribution and limiting factors. Limnol Oceanogr 30: 919–931 Gurgen V, Kirchner G and Pfennig N (1976) Fermentation of pyruvate by 7 species of phototrophic purple bacteria. Z Allg Mikrobiolo 16: 573–586 Hansen TA and van Gemerden H (1972) Sulfide utilization by purple nonsulfur bacteria. Arch Mikrobiol 86: 49–56 Hansen TA and Veldkamp H (1973) Rhodopseudomonas sulfidophila, nov. spec., a new species of the purple nonsulfur bacteria. Arch Mikrobiol 92: 45–58 Heda GD and Madigan MT (1988) Thermal properties and oxygenase activity of ribulose-1,5-bisphosphate carboxylase from the thermophilic purple bacterium, Chromatium tepidum. FEMS Microbiol Lett 51: 45–50 Heda GD and Madigan MT (1989) Purification and characterization of the thermostable ribulose-1,5-bisphosphate carboxylase/ oxygenase from the thermophilic purple bacterium Chromatium tepidum. Eur J Biochem 184: 313–319 Holm HW and Vennes JW (1970) Occurrence of purple sulfur bacteria in a sewage treatment lagoon. Appl Microbiol 19: 988–996 Imhoff JF (2001) True marine and halophilic anoxygenic photo-
13 trophic bacteria. Arch Microbiol 176: 243–254 Imhoff JF and Madigan MT (2004) International Committee on Systematics of Prokaryotes Subcommitteee on the taxonomy of phototrophic bacteria. Minutes of the meetings, 27 August 2003, Tokyo, Japan. Int J Syst Evol Microbiol 54: 1001–1003 Imhoff JF, Hashwa F and Trüper HG (1978) Isolation of extremely halophilic phototrophic bacteria from the alkaline Wadi Natrun, Egypt. Arch Hydrobiol 84: 381–388 Imhoff JF, Sahl HG, Soliman GSH and Trüper HG (1979) The Wadi Natrun: chemical composition and microbial mass developments in alkaline brines of eutrophic desert lakes. Geomicrobiol J 1: 219–234 Imhoff JF, Hiraishi A and Süling J (2005) Anoxygenic phototrophic bacteria. In: Brenner DJ, Krieg NR and Staley JT (eds) Bergey’s Manual of Systematic Bacteriology, 2nd ed, Vol 2, part A, pp 119–132. Springer, New York Jones BR (1956) Studies of pigmented non-sulfur purple bacteria in relation to cannery waste lagoon odors. Sewage Ind Wastes 28: 883–893 Jung DO, Achenbach LA, Karr EA, Takaichi S and Madigan MT (2004) A gas vesiculate planktonic strain of the purple non-sulfur bacterium Rhodoferax antarcticus isolated from Lake Fryxell, Dry Valleys, Antarctica. Arch Microbiol 182: 236–243 Kämpf C and Pfennig N (1980) Capacity of Chromatiaceae for chemotrophic growth. Specific respiration rates of Thiocystis violacea and Chromatium vinosum. Arch Microbiol 127: 125–135 Karr EL, Sattley WM, Jung DO, Madigan MT and Achenbach LA (2003) Remarkable diversity of phototrophic purple bacteria in a permanently frozen Antarctic lake. Appl Environ Microbiol 69: 4910–4914 Kimble-Long LK and Madigan MT (2002) Irradiance effects on growth and bacteriochlorophyll content of phototrophic heliobacteria, purple and green photosynthetic bacteria. Photosynthetica 40: 629-632 Kobayashi M (1975) Role of photosynthetic bacteria in foul water purification. Prog Water Technol 7: 309–315 Kondratieva EN, Zhukov VG, Ivanovsky RN, Petushkova YP and Monosov EZ (1976) The capacity of phototrophic sulfur bacterium Thiocapsa roseopersicina for chemosynthesis. Arch Microbiol 108: 287–292 Kramer H and Amesz J (1996) Antenna organization in the purple sulfur bacteria Chromatium tepidum and Chromatium vinosum. Photosynth Res 49: 237–244 Kulichevskaya IS, Guzev VS, Gorlenko VM, Liesack W and Dedysh SN (2006) Rhodoblastus sphagnicola sp. nov., a novel acidophilic purple non-sulfur bacterium from Sphagnum peat bog. Intl J Syst Evol Microbiol 56: 1397–1402 Mack EE, Mandelco L, Woese CR and Madigan MT (1993) Rhodospirillum sodomense, sp. nov., a Dead Sea Rhodospirillum species. Arch Microbiol 160: 363–371 Madigan MT (1984) A novel photosynthetic purple bacterium isolated from a Yellowstone hot spring. Science 225: 313–315 Madigan MT (1986) Chromatium tepidum sp. nov., a thermophilic photosynthetic bacterium of the family Chromatiaceae. Int J Syst Bacteriol 36: 222–227 Madigan MT (1988) Microbiology, physiology, and ecology of phototrophic bacteria. In: AJB Zehnder (ed) Biology of Anaerobic Microorganisms, pp 39–111, John Wiley & Sons, New York
14 Madigan MT (1995) Microbiology of nitrogen fixation by anoxygenic photosynthetic bacteria. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 915–928. Kluwer Academic Publishers, Dordrecht Madigan MT (1998) Isolation and characterization of psychrophilic purple bacteria from Antarctica. In: Peschek GA, Löffelhardt W and Schmetterer G (eds) The Phototrophic Prokaryotes, pp 699–706. Plenum, New York Madigan MT (2003) Anoxygenic phototrophic bacteria from extreme environments. Photosynth Res 76: 157–171 Madigan MT and Gest H (1978) Growth of a photosynthetic bacterium anaerobically in darkness, supported by ‘oxidant-dependent’ sugar fermentation. Arch Microbiol 117: 119–122 Madigan MT and Gest H (1979) Growth of the photosynthetic bacterium Rhodopseudomonas capsulata chemoautotrophically in darkness with H2 as the energy source. J Bacteriol 137: 524–530 Madigan MT, Cox SS and Stegeman RA (1984) Nitrogen fixation and nitrogenase activities in members of the family Rhodospirillaceae. J Bacteriol 157: 73–78 Madigan MT, Jung DO, Woese CR and Achenbach LA (2000) Rhodoferax antarcticus sp. nov., a moderately psychrophilic purple nonsulfur bacterium isolated from an Antarctic microbial mat. Arch. Microbiol. 173: 269–277 McClain J, Rollo DR, Rushing BG and Bauer CE (2002) Rhodospirillum centenum utilizes separate motor and switch components to control lateral and polar flagellum rotation. J Bacteriol 184: 2429–2438 Milford AD, Achenbach LA, Jung DO and Madigan MT (2000) Rhodobaca bogoriensis gen. nov. and sp. nov., an alkaphilic purple nonsulfur bacterium from African Rift Valley soda lakes. Arch Microbiol 174: 18–27 Nagashima KVP, Hiraishi A, Shimada K and Matsuura K (1997) Horizontal transfer of genes coding for the photosynthetic reaction centers of purple bacteria. J Mol Evol 45: 131–136 Nogi T, Fathir I, Kobayashi M, Nozawa T and Miki K (2000) Crystal structures of photosynthetic reaction center and highpotential iron-sulfur protein from Thermochromatium tepidum: Thermostability and electron transfer. Proc Natl Acad Sci USA 97: 13561–13566 Nozawa T and Madigan MT (1991) Temperature and solvent effects on reaction centers from Chloroflexus aurantiacus and Chromatium tepidum. J Biochem 110: 588–594 Nozawa T, Fukada T, Hatano M and Madigan MT (1986) Organization of intracytoplasmic membranes in a novel thermophilic purple photosynthetic bacterium as revealed from absorption, circular dichroism, and emission spectra. Biochim Biophys Acta 852: 191–197 Overmann J and Schubert K (2002) Phototrophic consortia: Model systems for symbiotic interrelations between prokaryotes. Arch Microbiol 177: 201–208 Overmann J, Beatty JT and Hall KJ (1994) Photosynthetic activity and population dynamics of Amoebobacter purpureus in a meromictic saline lake. FEMS Microbiol Ecol 15: 309–320 Overmann J, Beatty JT and Hall KJ (1996) Purple sulfur bacteria control the growth of aerobic heterotrophic bacterioplankton in a meromictic salt lake. Appl Environ Microbiol 62: 3251–3258 Overmann J, Hall KJ, Northcote TG and Beatty JT (1999) Grazing of the copepod Diaptomus connexus on purple sulphur bacteria
Michael T. Madigan and Deborah O. Jung in a meromictic salt lake. Environ Microbiol 1: 213–221 Pattaragulwanit K, Brune DC, Trüper HG and Dahl C (1998) Molecular genetic evidence for extracytoplasmic localization of sulfur globules in Chromatium vinosum. Arch Microbiol 169: 434–444 Permentier HP, Neerken S, Overmann J and Amesz J (2001) A bacteriochlorophyll a antenna complex from purple bacteria absorbing at 963 nm. Biochemistry 40: 5573–5578 Pfennig N (1967) Photosynthetic bacteria. Ann Rev Microbiol 21: 285–324 Pfennig N (1969) Rhodopseudomonas acidophila, sp. n., a new species of the budding purple nonsulfur bacteria. J Bacteriol 99: 597–602 Pfennig N (1974) Rhodopseudomonas globiformis, sp. n., a new species of the Rhodospirillaceae. Arch Microbiol 100: 197–206 Pfennig N (1975) The phototrophic bacteria and their role in the sulfur cycle. Plant Soil 43: 1–16 Pfennig N (1978a) General physiology and ecology of photosynthetic bacteria. In: Clayton RK and Sistrom WR (eds) The Photosynthetic Bacteria, pp 3–18. Plenum Press, New York Pfennig N (1978b) Rhodocyclus purpureus gen. nov. and sp. nov., a ring-shape, vitamin B12-requiring member of the family Rhodospirillaceae. Int J Syst Bacteriol 28: 283–288 Pfennig N (1989) Ecology of phototrophic purple and green sulfur bacteria. In: Schlegel HG and Bowien B (eds) Autotrophic Bacteria, pp 97–116. Springer-Verlag, Heidelberg Raymond J, Zhaxybayeva O, Gogarten JP and Blankenship RE (2003) Evolution of photosynthetic prokaryotes: A maximumlikelihood mapping approach. Phil Tran Roy Soc Lond B Biol Sci 358: 223–230 Resnick SM and Madigan MT (1989) Isolation and characterization of a mildly thermophilic nonsulfur purple bacterium containing bacteriochlorophyll b. FEMS Microbiol Lett 65: 165–170 Satoh T, Hoshino Y and Kitamura H (1976) Rhodopseudomonas sphaeroides forma sp. denitrificans, a denitrifying strain as a subspecies of Rhodopseudomonas sphaeroides. Arch Microbiol 108: 265–269 Sattley WM and Madigan MT (2006) Isolation, characterization and ecology of cold-active, chemolithotrophic sulfur-oxidizing bacteria from perennially ice-covered Lake Fryxell, Antarctica. Appl Environ Microbiol 72: 5562–5568 Schultz JE and Weaver PF (1982) Fermentation and anaerobic respiration by Rhodospirillum rubrum and Rhodopseudomonas capsulata. J Bacteriol 149: 181–190 Siefert E and Koppenhagen VB (1982) Studies on the vitamin B12 auxotrophy of Rhodocyclus purpureus and two other vitamin B12-requiring purple nonsulfur bacteria. Arch Microbiol 132: 173–178 Siefert E, Irgens RL and Pfennig N (1978) Phototrophic purple and green bacteria in a sewage treatment plant. Appl Environ Microbiol 35: 38–44 Sojka GA (1978) Metabolism of nonaromatic organic compounds. In: Clayton RK and Sistrom WR (eds) The Photosynthetic Bacteria, pp 707–718. Plenum Press, New York Stadtwald-Demchick R, Turner FR and Gest H (1990) Rhodopseudomonas cryptolactis, sp. nov., a new thermotolerant species of budding phototrophic purple bacteria. FEMS Microbiol Lett 71: 117–121 Tabita FR (1995) The biochemistry and metabolic regulation of
Chapter 1
Overview of Purple Bacteria
carbon metabolism and CO2 fixation in purple bacteria. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 885–914. Kluwer Academic Publishers, Dordrecht Takahashi M and Ichimura S (1968) Vertical distribution and organic matter production of photosynthetic sulfur bacteria in Japanese lakes. Limnol Oceanog 13: 644–655 Takaichi S, Jung DO and Madigan MT (2001) Accumulation of unusual carotenoids in the spheroidene pathway, demethylspheroidene and demethylspheroidenone, in an alkaliphilic purple nonsulfur bacterium Rhodobaca bogoriensis. Photosynth Res 67:207–214 Trüper HG (1978) Sulfur metabolism. In: Clayton RK and Sistrom WR (eds) The Photosynthetic Bacteria, pp 677–690. Plenum Press, New York Trüper HG (1981) Versatility of carbon metabolism in phototrophic bacteria. In: Dalton H (ed) Microbial Growth on C1 Compounds, pp 116–121. Heyden, London Trüper HG and Fischer U (1982) Anaerobic oxidation of sulphur compounds as electron donors for bacterial photosynthesis. Phil Trans Roy Soc Lond B 298: 529–542 Trüper HG and Pfennig N (1981) Characterization and identification of the anoxygenic phototrophic bacteria. In: Starr MP, Stolp H, Trüper HG, Balows A and Schlegel HG (eds.) The Prokaryotes, a Handbook on Habitatss, Isolation, and Identification of
15 Bacteria, pp 299–312. Springer-Verlag, New York Uffen RL and Wolfe RS (1970) Anaerobic growth of purple nonsulfur bacteria under dark conditions. J Bacteriol 104: 462–472 van Gemerden H (1968) On the ATP generation by Chromatium in darkness. Arch Mikrobiol 64: 118–124 van Gemerden H and Mas J (1995) Ecology of phototrophic sulfur bacteria. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 50–85. Kluwer Academic Publishers, Dordrecht van Niel CB (1932) On the morphology and physiology of the purple and green sulphur bacteria. Arch Mikrobiol 3: 1–112 van Niel CB (1944) The culture, general physiology, morphology, and classification of the non-sulfur purple and brown bacteria. Bacteriol Rev 8: 1–118 Ward DM, Weller R, Shiea J, Castenholz RW and Cohen Y (1989) Hot spring microbial mats: anoxygenic and oxygenic mats of possible evolutionary significance. In: Cohen Y and Rosenberg E (eds) Microbial Mats: Physiological Ecology of Benthic Microbial Communities, pp 3–15. American Society for Microbiology, Washington, DC Zengler K, Heider J, Rossello-Mora R and Widdel F (1999) Phototrophic utilization of toluene under anoxic conditions by a new strain of Blastochloris sulfoviridis. Arch Microbiol 172: 204–212
Chapter 2 Evolutionary Relationships Among Purple Photosynthetic Bacteria and the Origin of Proteobacterial Photosynthetic Systems Wesley D. Swingley Institute of Low Temperature Science, Hokkaido University, N19W8, Sapporo 060-0819, Japan
Robert E. Blankenship Departments of Biology and Chemistry, Washington University in St. Louis, Campus Box 1337, St. Louis, MO 63130, U.S.A.
Jason Raymond* School of Natural Sciences, University of California-Merced, Merced, CA 95344, U.S.A.
Summary ................................................................................................................................................................. 17 I. Introduction....................................................................................................................................................... 18 II. The Alphaproteobacteria .................................................................................................................................. 18 III. Aerobic Purple Bacteria.................................................................................................................................... 19 IV. The Photosynthesis Gene Cluster and its Role in Evolution ............................................................................ 20 V. Proteobacterial Comparative Genomics: Photosynthetic versus non-Photosynthetic Proteins ....................... 21 VI. Origin and Evolution of Proteobacterial Phototrophy ....................................................................................... 22 VII. Origin and Evolution of Proteobacterial Carbon-fixation .................................................................................. 24 VIII. Future Directions: High-Throughput Sequencing and Metagenomics.............................................................. 27 Acknowledgments ................................................................................................................................................... 28 References .............................................................................................................................................................. 28
Summary The purple bacteria occupy a unique position among photosynthetic bacteria. Nested within the various proteobacterial lineages, the origin and evolution of purple bacterial photosynthesis has been the topic of innumerable debates. Attempts to reconstruct the evolutionary history of individual photosynthetic protein families have further fueled debate over lateral vs. vertical transfer of genetic elements. In this era of high-throughput sequencing we can begin to distance ourselves from this dependency on single-gene and single-protein phylogenies. Here we present automated comparative genomics-based methods useful for reconstructing the genomic history of not only the purple bacterial lineage, but the proteobacterial lineage as a whole. These reconstructions integrate phylogenetic data inferred from 200 to more than 1000 protein families common to all or part of the proteobacterial lineage. This framework allows us to reconstruct the evolutionary history of *Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 17–29. © 2009 Springer Science + Business Media B.V.
18
Wesley D. Swingley, Robert E. Blankenship and Jason Raymond
each proteobacterial class and parse out the finer relationships among photosynthetic species. By telescoping inward on protein families of interest, we can delve deeper than ever before into the convoluted evolutionary origin of the primary photosynthetic traits, phototrophy and autotrophy. While these full-genome comparisons clarify the nature of many poorly understood phylogenetic relationships, they do not yet serve to resolve the entire mystery surrounding the history of proteobacterial phototrophy. I. Introduction Among six phyla of photosynthetic prokaryotes, the purple proteobacteria are the most metabolically diverse. Unlike some other groups of phototrophs such as cyanobacteria or green sulfur bacteria, in which all members are capable of photosynthesis, the proteobacteria have a remarkable range of metabolic lifestyles. Within this group are organisms that grow aerobically and anaerobically, photoautotrophically and photoheterotrophically, and many that utilize a diverse number of metabolic pathways for energy generation, carbon assimilation, as well as nitrogen, sulfur, and phosphorous metabolism (Imhoff et al., 2005; Chapter 1, Madigan and Jung). This diversity reflects the widespread nature of these organisms throughout the proteobacterial lineage. In many cases, phototrophic proteobacteria share genus-level relationships with non-phototrophs. This fact could be explained as either a loss of ancestral phototrophy or a lateral gain of the genes necessary to produce photosynthetic pigments and proteins. Although the proteobacteria are the most sequenced bacterial phylum to date, with more than 500 genomes complete or in progress (according to NCBI), the purple bacteria have been far less sampled, with fewer than 30 genome projects as of this writing, fewer than half of which are completed. This disparity is, in part, due to the overwhelming medical interest in proteobacteria, but it speaks volumes that the first purple bacterial sequence, Rhodopseudomonas palustris CGA009, was not published until 2004 (Larimer et al., 2004), nine years after the first complete bacterial genome sequence (Fleischmann et al., 1995). Notwithstanding, preliminary sequence data from Rhodobacter capsulatus and Rhodobacter sphaeroides 2.4.1 have been available since 2001 (http://www. integrated genomics.com/genomereleases.html; Mackenzie et al., 2001).
Early purple bacteria sequencing projects focused on model organisms, a common trend throughout the genome sequencing arena, completing the heavily studied species Rhodobacter sphaeroides (Choudhary et al., 2007), Rhodopseudomonas palustris (Larimer et al., 2004), Rhodospirillum rubrum (NCBI), and both phototrophic and nonphototrophic Bradyrhizobia (Kaneko et al., 2002; Giraud et al., 2007). Recently, the focus has begun to shift to those purple bacteria with a significant involvement in oceanic carbon and nutrient cycling, with the completed sequence of Roseobacter denitrificans Och 114 (Swingley et al., 2007) leading the way to a tremendous number of marine organismal genome sequences in progress or recently completed . While all of the above sequences are from the alpha class of proteobacteria, few other phototrophic proteobacteria have been targeted for sequencing. Three of these belong to the gammaproteobacteria, Halorhodospira halophila SL1, Congregibacter litoralis KT71, and the undescribed marine strain HTCC2080 (NCBI). Of these, Halorhodospira and Congregibacter have been characterized as phototrophic (Imhoff and Süling, 1996; Fuchs et al., 2007), but HTCC2080, which carries the genetic capability for phototrophy, has not yet been shown to produce photosynthetic pigments. Two major clades of phototrophic betaproteobacteria, genera Rubrivivax and Rhodoferax, have yet to be sequenced, as the only sequenced members, the former Rubrivivax gelatinosus PM1 (now Methylibium petroleiphilum PM1) and Rhodoferax ferrireducens T118, are non-phototrophic strains. M. petroleiphilum was only recently disambiguated from a cluster of closely-related species (both phototrophic and nonphototrophic) lumped under a relatively small number of genera (Nakatsu et al., 2006). II. The Alphaproteobacteria
Abbreviations: AAP – aerobic anoxygenic phototrophs; BChl – bacteriochlorophyll; GOS – Global Ocean Survey; kb – kilobase(s); PGC – photosynthesis gene cluster; rubisco– ribulose 1,5-bisphosphate carboxylase
The vast majority of phototrophic proteobacteria are spread throughout the alpha lineage. However, phototrophy is by no means ubiquitous in alphapro-
Chapter 2
Evolution of Purple Bacteria
teobacteria. Less than 22% of all alphaproteobacteria currently sequenced bear signatures of phototrophy (presence of chlorophyll biosynthesis or reaction center genes). While the number of sequenced phototrophic taxa is low, they are widespread throughout the alpha cluster (Fig. 1), which raises important questions on the origin of their phototrophy. Furthermore, a sequencing bias for the large number of agriculturally-important alphaproteobacteria could skew the aforementioned ratio. In fact, 18 of 39 currently available ‘in progress’ proteobacteria genomes at NCBI are potentially phototrophic. Unlike phototrophy, carbon-fixation is not common among alphaproteobacteria. The typical method of autotrophic carbon-fixation in all proteobacteria is through the Calvin Cycle using the key enzyme ribulose 1,5-bisphosphate carboxylase (rubisco). Variations of this enzyme are found throughout the tree of life (Tabita, 1999), but it is most famously associated with chlorophyll and photoautotrophy in photosynthetic organisms. However, a unique class of alphaproteobacteria (discussed below) grows photoheterotrophically rather than photoautotrophically. As a result, a relatively low proportion of alphaproteobacteria are photoautotrophic when compared to the beta- and gammaproteobacteria, where phototrophy is more rare. III. Aerobic Purple Bacteria Canonically speaking, purple bacteria grow photoautotrophically only at low oxygen levels, while they down-regulate their photosynthetic apparatus under higher oxygen levels and instead grow chemotrophically using organic carbon (Harashima et al., 1982; Yurkov and van Gemerden, 1993; Bauer et al., 2003; Yurkov and Csotonyi, 2003). Recent oceanic metagenomic surveys have created a burgeoning interest in a related class of purple bacteria that grow primarily under aerobic conditions. Dubbed the aerobic anoxygenic phototrophs (AAP or AAnP) or aerobic phototrophic bacteria (APB), these purple bacteria produce photosynthetic pigments and carry out phototrophic metabolism only under aerobic conditions (Shiba et al., 1979; Shiba and Harashima, 1986; Yurkov and van Gemerden, 1993;Yurkov and Beatty, 1998; Chapter 3, Yurkov and Csotonyi). However, they produce bacteriochlorophyll (BChl) a only in the dark, possibly to avoid the potential for oxidative damage caused by the excitation of chlorophyll
19
Fig. 1. 16S rRNA tree of the Proteobacteria, based on sequences from the Ribosomal Database [http://rdp.cme.msu.edu]. To facilitate visualization, 16S rRNA gene data are filtered so that only lineages at 97% or less 16S rRNA gene distance are shown (i.e. filtered at roughly the species level). Bold lineages are those proteobacterial species with one or more genome-sequenced members. Bold, numbered lineages indicate the positions of sequenced purple/phototrophic proteobacteria in the tree. Other lineages have 16S rRNA data available from isolates or environmental samples, but do not have genome projects in progress as of this writing. Proteobacterial classes are shown in Greek letters. Numbered purple proteobacteria are: 1. Rhodospirillum rubrum, 2. Rhodospirillum centenum, 3. Erythrobacter NAP1, 4. Roseobacter denitrificans (multiple strains), 5. Roseovarius sp. (multiple strains), 6. Loktanella, 7. Jannaschia sp. CCS1, 8. Rhodobacter sphaeroides (incl. R. capsulatus), 9. Fulvimarina, 10. Methylobacterium, 11. Bradyrhizobium (multiple species), 12. Rhodopseudomonas palustris (multiple strains), 13. Thermochromatium tepidum (proprietary sequence), 14. Halorhodospira sp. SL1, 15. Rubrivivas gelatinosus (photosynthetic gene cluster only, no genome)
intermediates (Beatty, 2002). While these AAP have long been overlooked in oceanic studies, recent data shows that they are widely distributed in the oceans and, consequently, their contribution to the global carbon cycle could be significant (Kolber et al., 2001; Buchan et al., 2005; Yutin et al., 2007). Historically, the origin of aerobiosis in the AAP has been a puzzling problem, especially in regards to the regulation of photosynthetic gene expression. The largest cluster of AAP, the Roseobacter clade, fall within the order Rhodobacterales and these species are closely related to the model species Rhodobacter sphaeroides and Rhodobacter capsulatus (Yurkov and
20
Wesley D. Swingley, Robert E. Blankenship and Jason Raymond
Fig. 2. Schematic comparison of photosynthesis gene clusters (PGC) from various purple bacteria. Lines represent transpositions of syntenous gene regions between two PGC. Numbers annotated above the Rhodospirillum rubrum PGC represent the location on the genome of each separate PGC segment. The lines with arrows represent inversion of the genes contained between the lines and the other lines represent shifts in position without inversions. See also Color Plate 1, Fig. 1, for a color version with additional information.
Beatty, 1998). Even within this group of organisms the common proteobacterial trend holds true in that not all members are phototrophic. Thus far, nearly all sequenced species in this cluster lack rubisco and thus the ability to fix carbon autotrophically (Swingley et al., 2007). The reason for the seemingly detrimental absence of autotrophic carbon-fixation is unclear. Distantly related families of AAP, such as some Erythrobacter and Congregibacter species, also lack rubisco (Swingley et al., 2007; Yutin et al., 2007). One possibility is that the presence of rubisco with a poor specificity for CO2 vs. O2 is a liability in aerobic conditions, which would lead to a large amount of photorespiration — which has led to multiple rubisco gene losses throughout evolutionary history. The only AAP shown to exhibit photoautotrophic carbon-fixation are several members of the plant endosymbiont genus Bradyrhizobia (Giraud et al., 2007). IV. The Photosynthesis Gene Cluster and its Role in Evolution In all known cases, the majority of the genetic information needed to build the photosynthetic apparatus
in purple bacteria is clustered in large groups of genes known as the Photosynthesis Gene Cluster (PGC). In most organisms, a single cluster of 40–50 kb contains all the structural genes for the photosystem, as well as genes for the latter stages of BChl and carotenoid biosynthesis (Fig. 2). Exceptions to this pattern are Rhodospirillum rubrum and Methylobacterium species, which contain three separate PGC clusters that are widely dispersed in the chromosome. Figure 2 shows gene neighborhoods for the PGC in several purple bacterial species. It is apparent that while the genes are organized into one or a few large clusters, the precise gene organization within the cluster is highly variable. It is possible to produce the various gene synteny arrangements found in PGCs by inversions and rearrangements, as shown in Fig. 2. The dramatic clustering of nearly all genes involved in phototrophy has provided a perfect framework for the lateral spread of phototrophy. It is intriguing that there is not a similar clustering of photosynthesisrelated genes in other groups of phototrophs, with the exception of the heliobacteria (Xiong et al., 1998). The gene organization of the PGC in heliobacteria is not generally similar to that found in the purple bacteria, with the exception of some of the operons
Chapter 2
Evolution of Purple Bacteria
for multisubunit enzyme complexes involved in BChl biosynthesis. This suggests that the assembly of genes into the PGC in the purple bacteria and heliobacteria may have been independent evolutionary events. The heliobacteria seem to have little in common with the purple bacteria in that they have a type I reaction center, are strict anaerobes and are incapable of photoautotrophic metabolism. Why these two very different groups of phototrophs have both assembled the genetic information needed to do photosynthesis into large gene clusters while other groups of phototrophs have not done so is a question for which no answer is yet available. The possible functional role of such extreme clustering of genes that code for functionally related proteins has a long history in biology, starting with the lac operon of Escherichia coli (Jacob and Monod, 1961). The utility of clustering and co-expression of genes that code for parts of a multiprotein complex is easy to appreciate. A single mRNA encoding each component of a complex, which is then translated into proteins that then associate with each other, permits precise control of the expression levels and spatial proximity of the components. However, what is not so clear is the benefit of a large gene cluster such as the PGC, which consists of more than thirty genes. This cluster has genes that are divergently transcribed, so they cannot be part of a single operon, which is confirmed by transcriptional analysis (Kiley and Kaplan, 1988; Wellington and Beatty, 1991). Furthermore, the dosage requirements of the various photosynthetic complexes and biosynthetic enzymes are vastly different. Regulatory elements can easily coordinate gene expression in genes that are distant on the chromosome, so there seems to be little functional need for physical clustering. The ‘selfish operon’ theory as formulated by Lawrence and Roth (Lawrence and Roth, 1996; Lawrence, 2003) proposes that operons assemble to facilitate horizontal transfer of an operon and its easy replacement if it happens to be lost. This theory has been questioned and some predictions of the selfish operon theory have not been observed (Pál and Hurst, 2004; Price et al., 2005). Nagashima et al. (1997) analyzed the gene sequences of the L and M reaction center proteins and concluded that significant amounts of horizontal gene transfer has taken place among purple bacteria. This same conclusion was extended to the PGC in a later study (Igarashi et al., 2001). However, this issue has not been addressed in detail since the advent of large-scale genome sequencing
21 of photosynthetic bacteria and it represents an area that needs to be investigated in more detail. V. Proteobacterial Comparative Genomics: Photosynthetic versus Non-Photosynthetic Proteins The foremost challenge in using a wealth of genomic data to reconstruct evolutionary events is how to integrate thousands of genes worth of (often discordant) phylogenies. One approach to this end involves concatenating multiple aligned gene or protein sequences and inferring a single phylogeny for this concatenated dataset. The underlying assumption for this approach is that the incongruence in single phylogenies, caused by data that are poorly fit by evolutionary models or by horizontal gene transfer, will be overtaken by the reinforcing signal of vertical evolution. Simulation studies have illustrated the robustness of concatenated datasets while raising the important caveat that, under some circumstances, concatenation only serves to reinforce the wrong tree (Gadagkar et al., 2005). No doubt the approach will continue to improve with more rigorous methods for determining which genes are suitable for concatenation. Concatenated phylogenies represent an averaged evolutionary overview of a set of genes or proteins, providing a potentially useful broad brush history but also masking potentially interesting details. It is important that this, and other approaches, be considered along with — not in lieu of — rigorous, single gene or protein phylogenies. Keeping these assumptions and ideas in mind, it is useful to examine the photosynthesis gene phylogeny for a large set of sequenced proteobacteria and contrast this with the 16S rRNA tree. This method involves the concatenation of specific PGC proteins to compare the evolutionary history of proteobacterial photosynthesis with the history of the phylum. Figure 3 (left) shows the phylogeny inferred from concatenating 208 orthologous proteins found in 29 proteobacterial genomes and several non-proteobacterial phototrophic bacteria, yielding an alignment of nearly 69,000 positions. Figure 3 (right) shows the concatenation of 38 proteins from the PGC of Rhodobacter sphaeroides with their homologs in other bacterial phototrophs, which resulted in an alignment of nearly 16,400 positions. The 208-protein (all-protein) concatenation (Fig. 3, left) is generally consistent with the 16S rRNA tree in the distinct
22
Wesley D. Swingley, Robert E. Blankenship and Jason Raymond
Fig. 3. A comparison of purple bacterial phylogeny based on whole genome proteins versus only photosynthesis-related proteins. On the left is the phylogeny based on concatenation of 208 proteins found in all, or all but one, genome from representative proteobacteria and several other phototrophic phyla. The tree on the right shows the phylogeny based on the concatenation of 38 proteins from the photosynthetic gene cluster (PGC) in purple proteobacteria and their homologs in other phototrophic phyla. Taxa in gray are present in only one of the two trees, whereas taxa in black are present in both trees. The lineage line-width is proportional to topological (branching order) differences between the two trees. Proteobacterial classes are indicated by Greek letters and surrounded by dashed lines. PGC protein homologs included in the concatenation are: LhaA, Lhb1, PpaA, PpsR, PucC, PufL, PufM, PufQ, PufX, PuhH, AcsF, BchB, BchC, BchD, BchE, BchF, BchG, BchH, BchI, BchJ, BchL, BchM, BchN, BchO, BchP, BchX, BchY, BchZ, CrtA, CrtB, CrtC, CrtD, CrtE, CrtF, CrtI, CrtK, Idi, and Lha1.
separation of the proteobacterial classes as well as with the branching order of species within these classes. In contrast, the 38-protein PGC concatenation shows no congruence at the class level with either the 16S rRNA tree or the all-protein concatenation. Identical branching is observed at lower taxonomic levels, as seen for instance in the topology of order Rhodobacteriales (Rhodobacter sphaeroides through Roseobacter denitrificans), and also in conservation of the Rhodopseudomonas/Bradyrhizobium clade. At face value, the PGC tree is consistent with the idea of an alpha origin of proteobacterial phototrophy, followed by subsequent horizontal gene transfer into members of the beta- and gammaproteobacterial classes. Note that the position of the non-proteobacterial phototrophs as outgroups on both trees is based on midpoint rooting on the longest branch in the tree; in principle, one could use additional lines of evidence to root the trees along other lineages.
VI. Origin and Evolution of Proteobacterial Phototrophy The scattered distribution of phototrophic organisms throughout the proteobacteria has presented a longstanding puzzle. On one hand, the sheer number of lateral transfer events necessary to accommodate such a distribution seems overwhelming. Conversely, the number of deletion events required seems to rule out a phototrophic proteobacterial ancestor. However, by breaking down the problem into manageable pieces, some conclusions can be drawn. Because the alphaproteobacteria represent the largest repository for phototrophy, they are the best group for studying the elusive origin of this trait. One thing made clear by the growing library of sequence data is that alphaproteobacterial phototrophy is evenly dispersed across evolutionary space (Fig. 1), indicating that vertical transfer coupled with gene loss may be more important than lateral gene transfer in this group. However, it is clear that the earliest
Chapter 2
Evolution of Purple Bacteria
phototrophic ancestor would likely occur after the divergence of the Rickettsiales — none of which are phototrophic — although the highly reduced genomes of these intracellular parasites could have selectively lost such an extraneous metabolic activity. In fact, the next earliest-branching clade includes the genus Acidiphilium, which contains phototrophic species (Hiraishi and Shimada, 2001). One of the largest and best studied alphaproteobacterial lineages is the Rhodobacteraceae. Over half of the in progress or complete phototrophic alpha genomes belong to this clade. However, even though the Rhodobacteraceae is a monophyletic, terminal clade, it contains a large number of non-phototrophic species. The conservation of phototrophy in this cluster suggests that the loss of phototrophy (and indeed photosynthesis as a whole) in these species may be quite common. One possible reason for such a phenomenon stems from the highly diverse metabolic library of Rhodobacteraceae species (Buchan et al., 2005). It is possible that such species modulate their metabolism based on availability of nutrients and/or light availability in their ecological niche. Given that bacterial genomes are generally streamlined, the loss of photosynthesis is likely in environments where photosynthesis is not a practical metabolic option. Most of the other sequenced phototrophic proteobacteria are found in the alphaproteobacterial family Bradyrhizobiaceae. This family is composed of three major genera, Bradyrhizobium, Rhodopseudomonas, and Nitrobacter, of which the former two both contain phototrophic species. Internal membranes in Nitrobacter species that are reminiscent of purple bacterial photosynthetic membranes suggest that this genus may have originated from a phototrophic ancestor. While all known Rhodopseudomonas species are phototrophic, there is only one monophyletic cluster of phototrophic Bradyrhizobium species, typified by the first discovered phototrophic strain Bradyrhizobium sp. BTAi1. Because Bradyrhizobium strains have typically been isolated from legume root nodules, it has been speculated that selective pressure has caused the loss of photosynthetic activity in many species (Giraud et al., 2007). This hypothesis is supported by the fact that all photosynthetic strains have been isolated from light-accessible stem nodules rather than light-inaccessible root nodules (Willems et al., 2003). Naturally, a ‘chicken or egg’ argument can be posed for any case of photosynthesis gene loss vs. gain. In cases such as the Rhodobacteraceae where a vast
23 majority of polyphyletic species are phototrophic, it is likely that phototrophy is a vertically inherited trait. However, in the case of Bradyrhizobium, the monophyletic phototrophic cluster could be the result of a single lateral transfer. While the presence of the closely related phototrophic Rhodopseudomonas and remnant ‘photosynthetic’ membranes in Nitrobacter support a phototrophic ancestry, they are not conclusive proof. A wealth of sequenced Bradyrhizobia, both phototrophic (strains BTAi1 and ORS278) and nonphototrophic (strain USDA110), and other Bradyrhyzobiaceae (five Rhodopseudomonas and three Nitrobacter species) allow for thorough analysis of this problem. One striking case is the similarity of gene orientation (synteny) between the phototrophic and non-phototrophic Bradyrhizobia (Fig. 4). While the number of homologous genes (E-value greater than 10–50) is greater between BTAi1 and ORS278, the overall gene order is more conserved between either of the two phototrophic strains and USDA110. This suggests two possible conclusions: a) genome rearrangement is a common occurrence and syntenic homology between distant strains occurs by coincidence or lower rates of recombination; or b) phototrophic and non-phototrophic strains are more closely related to each other than previous work has shown (Willems et al., 2003). It is very difficult to support the second case, which would contradict multiple phylogenetic reconstructions from 16S rRNA and rRNA ITS sequences as well as that presented in this chapter. The immense library of literature discussing genome rearrangement strongly supports the first conclusion. A high rate of genome rearrangement also leads to an increase in the rate of gene duplication and loss. This could have greater implications on the scattered nature of phototrophy throughout the proteobacterial clade, where a high rate of gene loss could account for numerous closely-related strains with and without phototrophic ability. Recent work by Haffa et al. (A.L. Haffa, personal communication) suggests that viral invasion has had a great influence on both random and directed gene losses in proteobacterial species, a fact that further convolutes evolutionary history in these species. Expanding beyond the phototrophic alphaproteobacteria, we see only a few select phototrophic clades throughout the rest of the proteobacterial tree. Methylibium petroleiphilum strain PM1and Rhodoferax ferrireducens T118 are the only sequenced members from phototrophic betaproteobacterial
24
Wesley D. Swingley, Robert E. Blankenship and Jason Raymond
Fig. 4. Dot-plot comparison representing a genome alignment between Bradyrhizobium species. Dots represent aligned nucleotide regions with higher than 75% identity. Only alignments with a BLAST score better than E=10-50 are shown. Axes represent sequence location in megabases. Dots in sequence along the 1:1 axis represent laterally homologous gene regions. Dots in sequence along the -1:1 axis represent contiguous translocations.
clades. Unfortunately, neither is a phototrophic species. However, Nagashima et al. (1993) sequenced the entire PGC from a phototrophic member of the same family, Rubrivivax gelatinosus, providing the most detailed glimpse so far into betaproteobacterial phototrophy. Halorhodospira halophilum SL1 is the only completely sequenced phototrophic gam-
maproteobacterium. Sequence data for two new phototrophic gammas, Congregibacter litoralis KT71 and unnamed marine strain HTCC2080, are currently available, but as yet incomplete. The sparse distribution of phototrophic beta- and gammaproteobacteria suggests that phototrophy did not originate in these clades. This idea is supported by the evolutionary history of their photosynthetic proteins (Fig. 3, right). Rubrivivax phototrophy appears to originate early on the Bradyrhizobiaceae branch, before the divergence of Rhodopseudomonas and Bradyrhizobium. The lack of conservation in the gene order between Rubrivivax and either alpha PGC suggests that this represents a relatively old lateral transfer event. On the other hand, the PGC gene order in Congregibacter and gamma HTCC2080 shares a great deal of similarity to their apparent origin from a close relative of Fulvimarina pelagi HTCC2506. The lesser homology between Halorhodospira and Fulvimarina illustrates the longer phylogenetic distance between the PGC in these two species; however, the origin is still quite near the Fulvimarina cluster. The precise origin of phototrophy in individual families of beta and gamma proteobacteria will not be resolved until further taxa are sequenced (if not full genomes, then at least PGC sequences). However, this chapter shows that only a few lateral transfer events could account for all non-alpha proteobacterial phototrophy. Less clear is the origin of the few chlorophyll biosynthesis genes that are not always contained within the PGC, such as the aerobic and anaerobic monomethyl ester cyclases (acsF and bchE, respectively) and divinyl reductase (gene not yet identified for all purple bacteria). These genes are not always present in the PGC of beta and gamma phototrophs, implying that they were either transferred independently by a later event or were originally in the PGC and have since translocated away. Such an outward translocation seems in direct opposition to our current understanding of bacterial gene clusters/operons (as discussed above). VII. Origin and Evolution of Proteobacterial Carbon-fixation Carbon-fixation is an inextricable component of photosynthesis. However, when studying purple bacterial photosynthesis we must be careful to extract and analyze the two portions independently. As discussed in the preceding section, phototrophy is
Chapter 2
Evolution of Purple Bacteria
unevenly distributed throughout the proteobacteria and the same is true for carbon-fixation. Nevertheless, these two traits do not appear to be linked in any clear fashion. Certainly there are many species that grow photoautotrophically using Calvin Cycle carbon-fixation, but just as many grow photoheterotrophically. On the other hand, many proteobacterial species fix CO2 without the use of light energy. Unlike photosynthetic proteins, the subunits of rubisco, RbcL/CbbL and RbcS/CbbS, are evolutionarily split into several classes (Watson and Tabita, 1997; Tabita, 1999; Ashida et al., 2005; Tabita et al., 2007). Therefore, while carbon-fixation may appear conserved in any given monophyletic cluster, the actual genes may have very different origins. This
25 is most striking in the Rhodobacteraceae, where some AAP species contain a rubisco that shares a close phylogenetic origin with Type IV rubisco-like proteins (Swingley et al., 2007) (Fig. 5 and Table 1). The role of the Type IV enzymes has not been clearly described; however, they have not been shown to fix CO2 (Hanson and Tabita, 2001; Imker et al., 2007; Tabita et al., 2007). There are also many cases where members of a single genus carry copies of rubisco from different family-types. Most notable among these are the Rhodopseudomonas species — which contain up to four different types of rubisco — and some Rhodobacter species that code for a number of rubisco variations (Table 1). Even with the likelihood of widespread rubisco
Fig. 5. Phylogenetic representation of purple bacterial ribulose 1,5-bisphosphate (rubisco) sequences. Roman numeral labels identify the recognized distinct ‘Form’ family of rubisco or rubisco-like protein. This analysis revealed three divergent groups of Form IV proteins that are further split into types A, B, and C, which are analogous to those described by Tabita (2007): A = IV-DeepYkr, B = IV-Photo, C = IV-NonPhoto.
Wesley D. Swingley, Robert E. Blankenship and Jason Raymond
26
Table 1. Ribulose 1,5-bisphosphate (rubisco) content of the purple bacteria and non-phototrophic members of genera that contain known phototrophs. rubisco Form-types are labeled as shown in Fig. 5 Family
α
Organism Bradyrhizobium japonicum USDA 110 Bradyrhizobium sp. BTAi1 Bradyrhizobium sp. ORS278 Dinoroseobacter shibae DFL 12 Erythrobacter litoralis HTCC2594 Erythrobacter sp. NAP1 Erythrobacter sp. SD-21 Fulvimarina pelagi HTCC2506 Jannaschia sp. CCS1 Loktanella vestfoldensis SKA53 Magnetospirillum magneticum AMB-1 Magnetospirillum magnetotacticum MS-1 Methylobacterium chloromethanicum CM4 Methylobacterium extorquens PA1 Methylobacterium sp. 4-46 Oceanicola batsensis HTCC2597 Oceanicola granulosus HTCC2516 Rhodobacter capsulatus Rhodobacter sphaeroides 2.4.1 Rhodobacter sphaeroides ATCC 17025 Rhodobacter sphaeroides ATCC 17029 Rhodocista centenaria Rhodopseudomonas palustris BisA53 Rhodopseudomonas palustris BisB18 Rhodopseudomonas palustris BisB5 Rhodopseudomonas palustris CGA009 Rhodopseudomonas palustris HaA2 Rhodospirillum rubrum ATCC 11170 Roseobacter denitrificans OCh 114 Roseobacter sp. AzwK-3b Roseobacter sp. CCS2 Roseobacter sp. MED193 Roseobacter sp. SK209-2-6 Roseovarius nubinhibens ISM Roseovarius sp. 217 Roseovarius sp. HTCC2601 Roseovarius sp. TM1035 Sulfitobacter sp. EE-36 Sulfitobacter sp. NAS-14.1
β
Methylibium petroleiphilum PM1 Rhodoferax ferrireducens T118
Form I ID IA, ID (x2) IA, ID ––––-
Form II
Form IV
IVC IVC –II II (×2) ––IVC –IVA IA IC IA, IC IC IA, ID ID ID IA, ID ID ID
II II II II II II II II II II
IVA, IVB IVA, IVB IVA, IVB IVA, IVB IVA, IVB IVA
–––IVC –––ID –––-
IVC (×2)
IC (x2) II
Chapter 2
Evolution of Purple Bacteria
27
Table 1. Continued Family
Organism
Form I
γ
Congregibacter litoralis KT71 Halorhodospira halophila SL1 marine gamma proteobacterium HTCC2080
–IA –-
Chlorobium tepidum TLS Chloroflexus aurantiacus J-10-fl Clostridium acetobutylicum ATCC 824 Desulfitobacterium hafniense Y51 Pelodictyon luteolum DSM 273 Prochlorococcus marinus MIT 9313 Roseiflexus sp. RS-1 Synechococcus sp. JA-2-3B´a(2-13)
lateral gene transfer, some phylogenetic information can be drawn from the proteobacterial genome tree. The aforementioned Rhodopseudomonas and Bradyrhizobium species likely share a Form ID rubiscocontaining ancestor with the related non-phototrophic genus Nitrobacter. However, the presence of many proteins from unrelated species in this phylogenetic cluster indicates that lateral transfer has likely occurred on multiple occasions. The presence of Form II in all Rhodobacter species also indicates that it is ancestral to this genus. Given the lack of autotrophic CO2-fixation in the AAP, Form II rubisco may not be ancestral to the Rhodobacteraceae in general. Nevertheless, the presence of Form IV-like proteins in AAPs suggests its possible ancestry to this group. Although Form IB rubisco, representing the traditional plant and cyanobacterial class of enzymes, is not present in any purple bacteria, other Form I enzymes are common throughout the proteobacteria. Like phototrophy, the origin and evolution of carbon-fixation in purple bacteria is difficult to clarify. The apparently rampant lateral transfer of Calvin Cycle enzymes in proteobacteria (and bacteria/Archaea as a whole) creates a high level of background noise that is difficult to separate from phylogenetic signals. Even with such a confusing situation, carbon-fixation is clearly an ancestral trait for the Bradyrhizobiaceae and likely many other lineages. However, given the current level of sequence coverage, we cannot make any further conclusions on the origin of purple bacterial — or proteobacterial for that matter — carbon-fixation.
Form II
Form IV IVA IVB
–––IVB IA –IB
VIII. Future Directions: High-Throughput Sequencing and Metagenomics One of the exciting recent advances in microbiology is the application of genomics to complex microbial communities taken directly from their natural environments. Environmental genomics, or metagenomics, provides an opportunity to sidestep the so-called cultivation bias: the fact that only a few percent of microbes from natural communities have been grown and studied in a laboratory. Indeed the recent discovery of phototrophy among the phylum Acidobacteria (Bryant et al., 2007) was founded on analysis of a Yellowstone metagenome sequencing project. Environmental sequencing was also used to demonstrate the abundance of rhodopsin-based phototrophy in the oceans (Béjà et al., 2000). This has been borne out in large-scale metagenomics of plankton in the Sargasso Sea (Venter et al., 2004) and other marine samples taken as part of the ongoing Global Ocean Survey (GOS) (Yooseph et al., 2007). These data underscore the remarkable phenotypic diversity that has yet to be characterized in the laboratory. For instance, the GOS dataset contains several hundred homologs of genes encoding the purple reaction center subunits. Homologs to known marine strains are particularly abundant, but the reaction center proteins of nearly every phototrophic proteobacterium discussed in this chapter has multiple close homologs in the GOS data. Beyond the enormous repository of marine metagenomes, a growing number of projects should provide additional insight into the genetic diversity of
28
Wesley D. Swingley, Robert E. Blankenship and Jason Raymond
phototrophy. These include samples from hot springs, wastewater treatment ponds, hypersaline microbial mats, even soil and sediment samples that receive some solar input. While exciting, these approaches also demand new ways of thinking about experimental validation and follow-through. In many cases, clone libraries are built from small inserts of DNA, typically 3–10 kb in length. If and when truly novel discoveries are sifted out of these enormous datasets, tracing the origin of sequences in these data will be a serious challenge. Small insert libraries will provide, at most, a handful of contextual genes that provide almost no insight into the organism’s physiology. Isolating the genome sequence of a single, specific organism from a complex community is incredibly difficult — though it is amenable with genetic ‘clues’ that would come from higher levels of targeted sequencing. This underscores the importance of either sequencing, or at the very least archiving, large insert (>100 kb) libraries that can provide a more substantial glimpse into the neighborhoods surrounding genes of interest. Ultimately, extrapolating useful information from enormous repositories of ‘letters on a page’ will require unprecedented synergism of genomics with clever new microbiological techniques. Acknowledgments WDS is supported by a Postdoctoral Fellowship for Foreign Researchers from the Japan Society for the Promotion of Science. Work on the evolution of photosynthesis by REB has been supported by a grant from the Exobiology program from NASA. References Ashida H, Danchin A and Yokota A (2005) Was photosynthetic rubisco recruited by acquisitive evolution from rubisco-like proteins involved in sulfur metabolism? Res Microbiol 156: 611–618 Bauer CE, Elsen S, Swem LR, Swem DL and Masuda S (2003) Redox and light regulation of gene expression in photosynthetic prokaryotes. Philos Trans R Soc Lond B Biol Sci 358: 147–154 Beatty JT (2002) On the natural selection and evolution of the aerobic phototrophic bacteria. Photosynth Res 73: 109–114 Béjà O, Aravind L, Koonin EV, Suzuki MT, Hadd A, Nguyen LP, Jovanovich SB, Gates CM, Feldman RA, Spudich JL, Spudich EN and DeLong EF (2000) Bacterial rhodopsin: Evidence for a new type of phototrophy in the sea. Science 289: 1902–1906 Bryant DA, Garcia Costas AM, Maresca JA, Chew AGM,
Klatt CG, Bateson MM, Tallon LJ, Hostetler J, Nelson WC, Heidelberg JF and Ward DM (2007) Candidatus Chloracidobacterium thermophilum: An aerobic phototrophic acidobacterium. Science 317: 523–526 Buchan A, Gonzalez J and Moran MA (2005) Overview of the marine Roseobacter lineage. Appl Environ Microbiol 71: 5665–5677 Choudhary M, Zanhua X, Fu YX and Kaplan S (2007) Genome analyses of three strains of Rhodobacter sphaeroides: Evidence of rapid evolution of chromosome II. J Bacteriol 189: 1914–1921 Fleischmann RD, Adams MD, White O, Clayton RA, Kirkness EF, Kerlavage AR, Bult CJ, Tomb JF, Dougherty BA, Merrick JM, McKenney K, Sutton G, Fitzhugh W, Fields C, Gocayne JD, Scott J, Shirley R, Liu LI, Glodek A, Kelley JM, Weidman JF, Phillips CA, Spriggs T, Hedblom E, Cotton MD, Utterback TR, Hanna MC, Nguyen DT, Saudek DM, Brandon RC, Fine LD, Fritchman JL, Fuhrmann JL, Geoghagen NSM, Gnehm CL, McDonald LA, Small KV, Fraser CM, Smith HO and Venter JC (1995) Whole-genome random sequencing and assembly of Haemophilus influenzae Rd. Science 269: 496–512 Fuchs BM, Spring S, Teeling H, Quast C, Wulf J, Schattenhofer M, Yan S, Ferriera S, Johnson J, Glockner FO and Amann R (2007) Characterization of a marine gammaproteobacterium capable of aerobic anoxygenic photosynthesis. Proc Natl Acad Sci USA 104: 2891–2896 Gadagkar SR, Rosenberg MS and Kumar S (2005) Inferring species phylogenies from multiple genes: Concatenated sequence tree versus consensus gene tree. J Exp Zoolog B Mol Dev Evol 304B: 64–74 Giraud E, Moulin L, Vallenet D, Barbe V, Cytryn E, Avarre J-C, Jaubert M, Simon D, Cartieaux F, Prin Y, Bena G, Hannibal L, Fardoux J, Kojadinovic M, Vuillet L, Lajus A, Cruveiller S, Rouy Z, Mangenot S, Segurens B, Dossat C, Franck WL, Chang W-S, Saunders E, Bruce D, Richardson P, Normand P, Dreyfus B, Pignol D, Stacey G, Emerich D, Verméglio A, Medigue C and Sadowsky M (2007) Legumes symbioses: Absence of nod genes in photosynthetic Bradyrhizobia. Science 316: 1307–1312 Hanson TE and Tabita FR (2001) A ribulose-1,5-bisphosphate carboxylase/oxygenase (RubisCO)-like protein from Chlorobium tepidum that is involved with sulfur metabolism and the response to oxidative stress. Proc Natl Acad Sci USA 98: 4397–4402 Harashima K, Nakagawa M and Murata N (1982) Photochemical activities of bacteriochlorophyll in aerobically grown cells of aerobic heterotrophs, Erythrobacter species (Och 114) and Erythrobacter longus (Och 101). Plant Cell Physiol 23: 185–193 Hiraishi A and Shimada K (2001) Aerobic anoxygenic photosynthetic bacteria with zinc-bacteriochlorophyll. J Gen Appl Microbiol 47: 161–180 Igarashi N, Harada J, Nagashima S, Matsuura K, Shimada K and Nagashima KVP (2001) Horizontal transfer of the photosynthesis gene cluster and operon rearrangement in purple bacteria. J Mol Evol 52: 333–341 Imhoff JF, Hiraishi A and Suling J (2005) Anoxygenic phototrophic purple bacteria. In: Brenner DJ, Krieg NR and Staley JT (eds) Bergey’s Manual of Systematic Bacteriology, pp 119–132. Springer, New York Imhoff JF and Süling J (1996) The phylogenetic relationship
Chapter 2
Evolution of Purple Bacteria
among Ectothiorhodospiraceae: A reevaluation of their taxonomy on the basis of 16S rDNA analyses. Arch Microbiol 165: 106–113 Imker HJ, Fedorov AA, Fedorov EV, Almo SC and Gerlt JA (2007) Mechanistic diversity in the rubisco superfamily: The ‘enolase’ in the methionine salvage pathway in Geobacillus kaustophilus. Biochemistry 46: 4077–4089 Jacob F and Monod J (1961) On regulation of gene activity. Cold Spring Harb Symp Quant Biol 26: 193 Kaneko T, Nakamura Y, Sato S, Minamisawa K, Uchiumi T, Sasamoto S, Watanabe A, Idesawa K, Iriguchi M, Kawashima K, Kohara M, Matsumoto M, Shimpo S, Tsuruoka H, Wada T, Yamada M and Tabata S (2002) Complete genomic sequence of nitrogen-fixing symbiotic bacterium Bradyrhizobium japonicum USDA110. DNA Res 9: 189–197 Kiley PJ and Kaplan S (1988) Molecular genetics of photosynthetic membrane biosynthesis in Rhodobacter sphaeroides. Microbiol Mol Biol Rev 52: 50–69 Kolber ZS, Plumley FG, Lang AS, Beatty JT, Blankenship RE, VanDover CL, Vetriani C, Koblizek M, Rathgeber C and Falkowski PG (2001) Contribution of aerobic photoheterotrophic bacteria to the carbon cycle in the ocean. Science 292: 2492–2495 Larimer FW, Chain P, Hauser L, Lamerdin J, Malfatti S, Do L, Land ML, Pelletier DA, Beatty JT, Lang AS, Tabita FR, Gibson JL, Hanson TE, Bobst C, Torres J, Peres C, Harrison FH, Gibson J and Harwood CS (2004) Complete genome sequence of the metabolically versatile photosynthetic bacterium Rhodopseudomonas palustris. Nat Biotechnol 22: 55–61 Lawrence JG (2003) Gene organization: Selection, selfishness, and serendipity. Annu Rev Microbiol 57: 419–440 Lawrence JG and Roth JR (1996) Selfish operons: Horizontal transfer may drive the evolution of gene clusters. Genetics 143: 1843–1860 Mackenzie C, Choudhary M, Larimer F, Predki P, Stilwagen S, Armitage J, Barber R, Donohue T, Hosler J, Newman J, Shapleigh J, Sockett R, Zeilstra-Ryalls J and Kaplan S (2001) The home stretch, a first analysis of the nearly completed genome of Rhodobacter sphaeroides 2.4.1. Photosynth Res 70: 19–41 Nagashima KVP, Shimada K and Matsuura K (1993) Phylogenetic analysis of photosynthetic genes of Rhodocyclus gelatinosus — possibility of horizontal gene-transfer in purple bacteria. Photosynth Res 36: 185–191 Nagashima KVP, Hiraishi A, Shimada K and Matsuura K (1997) Horizontal transfer of genes coding for the photosynthetic reaction centers of purple bacteria. J Mol Evol 45: 131–136 Nakatsu CH, Hristova K, Hanada S, Meng X-Y, Hanson JR, Scow KM and Kamagata Y (2006) Methylibium petroleiphilum gen. nov., sp. nov., a novel methyl tert-butyl ether-degrading methylotroph of the Betaproteobacteria. Int J Syst Evol Microbiol 56: 983–989 Pál C and Hurst LD (2004) Evidence against the selfish operon theory. Trends Genet 20: 232–234 Price MN, Huang KH, Arkin AP and Alm EJ (2005) Operon formation is driven by co-regulation and not by horizontal gene transfer. Genome Res 15: 809–819 Shiba T and Harashima K (1986) Aerobic photosynthetic bacteria. Microbiol Sci 3: 376–378 Shiba T, Simidu U and Taga N (1979) Distribution of aerobic bacteria which contain bacteriochlorophyll a. Appl Environ Microbiol 38: 43–45
29 Swingley WD, Sadekar S, Mastrian SD, Matthies HJ, Hao J, Ramos H, Acharya CR, Conrad AL, Taylor HL, Dejesa LC, Shah MK, O’Huallachain ME, Lince MT, Blankenship RE, Beatty JT and Touchman JW (2007) The complete genome sequence of Roseobacter denitrificans reveals a mixotrophic rather than photosynthetic metabolism. J Bacteriol 189: 683–690 Tabita FR (1999) Microbial ribulose 1,5-bisphosphate carboxylase/oxygenase: A different perspective. Photosynth Res 60: 1–28 Tabita FR, Hanson TE, Li H, Satagopan S, Singh J and Chan S (2007) Function, structure, and evolution of the RubisCO-like proteins and their RubisCO homologs. Microbiol Mol Biol Rev 71: 576–599 Venter JC, Remington K, Heidelberg JF, Halpern AL, Rusch D, Eisen JA, Wu D, Paulsen I, Nelson KE, Nelson W, Fouts DE, Levy S, Knap AH, Lomas MW, Nealson K, White O, Peterson J, Hoffman J, Parsons R, Baden-Tillson H, Pfannkoch C, Rogers Y-H and Smith HO (2004) Environmental genome shotgun sequencing of the Sargasso Sea. Science 304: 66–74 Watson GMF and Tabita FR (1997) Microbial ribulose 1,5bisphosphate carboxylase/oxygenase: A molecule for phylogenetic and enzymological investigation. FEMS Microbiol Lett 146: 13–22 Wellington CL and Beatty JT (1991) Overlapping mRNA transcripts of photosynthesis gene operons in Rhodobacter capsulatus. J Bacteriol 173: 1432–1443 Willems A, Munive A, de Lajudie P and Gillis M (2003) In most Bradyrhizobium groups sequence comparison of 16S-23S rDNA internal transcribed spacer regions corroborates DNADNA hybridizations. Syst Appl Microbiol 26: 203–210 Xiong J, Inoue K and Bauer CE (1998) Tracking molecular evolution of photosynthesis by characterization of a major photosynthesis gene cluster from Heliobacillus mobilis. Proc Natl Acad Sci USA 95: 14851–14856 Yooseph S, Sutton G, Rusch DB, Halpern AL, Williamson SJ, Remington K, Eisen JA, Heidelberg KB, Manning G, Li W, Jaroszewski L, Cieplak P, Miller CS, Li H, Mashiyama ST, Joachimiak MP, van Belle C, Chandonia J-M, Soergel DA, Zhai Y, Natarajan K, Lee S, Raphael BJ, Bafna V, Friedman R, Brenner SE, Godzik A, Eisenberg D, Dixon JE, Taylor SS, Strausberg RL, Frazier M and Venter JC (2007) The Sorcerer II Global Ocean Sampling expedition: Expanding the universe of protein families. PLoS Biol 5: e16 Yurkov V and van Gemerden H (1993) Impact of light/dark regime on growth rate, biomass formation and bacteriochlorophyll synthesis in Erythromicrobium hydrolyticum. Arch Microbiol 159: 84–89 Yurkov VV and Beatty JT (1998) Aerobic anoxygenic phototrophic bacteria. Microbiol Mol Biol Rev 62: 695–724 Yurkov VV and Csotonyi JT (2003) Aerobic anoxygenic phototrophs and heavy metal reducers from extreme environments. In: Pandalai SG (ed) Recent Research Developments in Bacteriology, pp 247–300. Transworld Research Network, Trivandrum, India Yutin N, Suzuki MT, Teeling H, Weber M, Venter JC, Rusch DB and Béjà O (2007) Assessing diversity and biogeography of aerobic anoxygenic phototrophic bacteria in surface waters of the Atlantic and Pacific Oceans using the Global Ocean Sampling expedition metagenomes. Environ Microbiol 9: 1464–1475
Chapter 3 New Light on Aerobic Anoxygenic Phototrophs Vladimir Yurkov* and Julius T. Csotonyi Department of Microbiology, University of Manitoba, Winnipeg, Manitoba, Canada R3T 2N2
Summary ................................................................................................................................................................. 31 I. Introduction....................................................................................................................................................... 32 II. Morphological Diversity, Taxonomic Nuances, Phylogeny and Evolution ........................................................ 34 A. Morphological Richness ..................................................................................................................... 34 B. Taxonomic Diversity ........................................................................................................................... 34 C. Persistent Taxonomic-Phylogenetic Tangles ..................................................................................... 37 D. Evolution ............................................................................................................................................ 38 III. Nutritional Versatility and Peculiarities of Carbon Metabolism ......................................................................... 40 A. Carbon Sources ................................................................................................................................. 40 B. Alternative Electron Donors and Acceptors........................................................................................ 41 IV. Photosynthetic Pigment Composition and Synthesis Reveal Surprises........................................................... 41 A. Carotenoids ........................................................................................................................................ 41 B. Bacteriochlorophyll ............................................................................................................................. 42 C. Highly Evolved Regulation of Pigment Synthesis .............................................................................. 42 V. The Mysterious Photosynthetic Apparatus of Aerobic Anoxygenic Phototrophs.............................................. 44 A. The Reaction Center of Aerobic Anoxygenic Phototrophs Has Purple Bacterial Roots ..................... 44 B. Light-Harvesting Complexes: Ordinary and Extraordinary ................................................................. 45 C. The Riddle of Aerobic Anoxygenic Phototrophy................................................................................. 46 VI. Speculation on Ecological Roles ...................................................................................................................... 47 A. Technical Challenges of Enumerating Aerobic Anoxygenic Phototrophs .......................................... 47 B. Ocean Surface ................................................................................................................................... 47 C. Vertical Distribution of Deep Ocean Aerobic Anoxygenic Phototrophs .............................................. 50 D. Aerobic Anoxygenic Phototrophs Thrive in Extreme Environments ................................................... 51 E. Soil and Freshwater ........................................................................................................................... 51 VII. Concluding Remarks and Perspectives ........................................................................................................... 51 Acknowledgments ................................................................................................................................................... 52 References .............................................................................................................................................................. 52
Summary Discovered 30 years ago, aerobic anoxygenic phototrophs (AAP) represent an entirely new bacterial functional group that was surprisingly found to constitute nearly 10% of microbial cells in the world’s biggest surface ecosystem, the ocean. These intriguing and colorful descendents of anaerobic anoxygenic phototrophs possess a fully functional photosynthetic apparatus that is paradoxically operative only under oxic conditions. An obviously ancient group, the AAP display numerous extensive evolutionary modifications to their photosynthetic machinery from that of their ancestors, such as different suites of light-harvesting 2 complexes and, in some *Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 31–55. © 2009 Springer Science + Business Media B.V.
32
Vladimir Yurkov and Julius T. Csotonyi
species, the only zinc-based chlorophyll pigments found anywhere in nature. Whereas AAP are incapable of photoautotrophy and rely on heterotrophy for 80% or more of their cellular energetics, sunlight can double organic carbon assimilatory efficiency over that of strict heterotrophs, making AAP key players in the marine carbon cycle. The AAP inhabit not just soil, rivers and oceans, but also hypersaline waters, thermal springs and even the dark realm of deep ocean hydrothermal vents. Ubiquity and atypical photosynthetic nature has inspired an ever-increasing scientific interest in the AAP, for which there are more exceptions than rules. I. Introduction It is astonishing that a bacterial functional group constituting nearly a tenth of microorganisms in the earth’s largest surface ecosystem, the ocean, could have been overlooked until 1978. Yet, this unlikely tale describes obligately aerobic anoxygenic phototrophs (AAP), which have long lain hidden beneath the shortsighted assumption that anoxygenic photosynthesis, originally evolving to generate metabolic energy from light in reducing Proterozoic environments, has for the past 3.5 billion years remained an exclusively anoxic process. Oxygen does generally repress synthesis of the bacteriochlorophyll (BChl)-containing photosynthetic apparatus in conventional anoxygenic phototrophs, relegating purple, green and heliobacterial light-harvesting to illuminated zones of anaerobic habitats such as stratified lakes and sulfide springs (Pfennig, 1978). However, O2 release by cyanobacteria ~2.5 GYa exerted sufficient selection pressure for an aerobic counterpart to evolve (Beatty, 2002). Indeed, Japanese scientists’ intuitive leap to search for such organisms was met with immediate success (Shiba et al., 1979). Since then, description of new species has increased in pace (Fig. 1A). AAP have turned up in practically every environment probed for them, including rivers, soil, acidic mine drainage, hypersaline springs, nutrient-rich microbial mats, oligotrophic ocean surface waters, and even near marine hydrothermal vents (Yurkov and Csotonyi, Abbreviations: AAP – aerobic anoxygenic phototrophs; BChl – bacteriochlorophyll; Cmi. – Citromicrobium; Cyt – cytochrome; D – Dinoroseobacter; DMSP – dimethylsulfoniopropionate; E. – Erythromicrobium; Erb. – Erythrobacter; GYa – 109 years ago; H. – Hoeflea; IREM – infrared epifluorescence microscopy; IRFRR – infrared fast repetition rate fluorometry; LH – lightharvesting; PEP – phosphoenol pyruvate; PSU – photosynthetic unit; QPCR – quantitative polymerase chain reaction; Qy – electronic transition of BChl a from ground state to lowest excited singlet state; R. – Roseicyclus; Rba. – Rhodobacter; RC – reaction center; Rps. – Rhodopseudomonas; Rsc. – Roseococcus; Rst. – Roseateles; Rubisco – ribulose-1,5-bisphosphate carboxylase/oxygenase; Rva – Roseovarius; S. – Stappia; Srb. – Sandaracinobacter; Stl. – Staleya; TMAO – trimethylamine N-oxide
2003). Comprehensive reviews have already set the groundwork for an introduction to the AAP (Yurkov and Beatty, 1998; Hiraishi and Shimada, 2001; Yurkov, 2006; Yurkov and Csotonyi, 2003; Rathgeber et al., 2004). However, the veritable explosion of research on this group during the last four years justifies both a synthesis of exciting new findings and suggestion of new investigative directions. Two other branches of bacteria produce BChl exclusively aerobically: photosynthetic methylotrophs such as Methylobacterium extorquens, and photosynthetic rhizobia like Bradyrhizobium denitrificans (van Berkum et al., 2006). Distinct metabolic strategies or specialized plant symbioses impart distinct functional group status on these organisms, and they will not be treated further in this chapter. Interested readers are directed to reviews by Komagata (1989) and Giraud and Fleischman (2004). The AAP are primarily chemoheterotrophic Proteobacteria that nonetheless produce a functional purple bacterial photosynthetic reaction center (RC) and one or more peripheral light-harvesting (LH) complexes that can amend heterotrophic energy generation by up to 20% (Kolber et al., 2001; Yurkov and van Gemerden, 1993). As their name implies, they evolve no O2 photosynthetically, because they do not use water for reducing power. Five overarching traits distinguish AAP from classical anoxygenic phototrophs: (1) requirement of O2 for photosynthesis, granting escape from restrictive anaerobic illuminated environments; (2) inhibition, paradoxically, of BChl synthesis by light; (3) absence of the Calvin cycle and inability to subsist on inorganic carbon; (4) much lower number of photosynthetic units (PSU) per cell; but (5) great abundance of carotenoids (Yurkov and Csotonyi, 2003; Rathgeber et al., 2004). With such a peculiar mixture of characteristics, where do the mysterious yet functionally and numerically significant AAP fit into our conception of the microbial world? On one hand, they are typical proteobacterial cells, constructed from the same building blocks as their relatives, expressing no earth-shatteringly novel metabolic modes, possess-
Chapter 3
Aerobic Anoxygenic Phototrophs
33
Fig. 1. Taxonomic richness and phylogeny of AAP. A. Cumulative growth in number of described genera over 24 years. B. Neighborjoining phylogenetic tree of AAP and select nonphototrophs (marked with asterisks) based on 16S rDNA sequences. Scale bar, 10% substitution rate. Bootstrap values at branch points are based on 500 resamplings. Phylogenetic tree was prepared by E. Stackebrandt and J. Swiderski.
34 ing typical cell walls, synthesizing mostly trivial pigments and electron carriers, etc. On the other hand, evolutionary forces have combined some of these common components in chimeric ways, permitting anoxygenic phototrophy to function under conditions contrary to those for which it originally evolved, thus opening for exploitation huge niches not available to their ancestors. This principal evolutionary development necessarily caused cascades of corollary adaptations, such as major changes to the regulation of photosynthetic pigment synthesis, and proliferation of photoprotective carotenoids. In the end, the peculiar AAP display more exceptions than rules in form and function. II. Morphological Diversity, Taxonomic Nuances, Phylogeny and Evolution A. Morphological Richness The AAP are morphologically as well rounded as are their anaerobic relatives, even expressing properties unique among bacteria (Fig. 2). Pleomorphism is common in the Erythrobacter clade. Erythromicrobium (E.) ramosum was named for its propensity to produce branches (Fig. 2A), while Citromicrobium (Cmi.) bathyomarinum forms unusual Y-cells during trinary division (Yurkov and Csotonyi, 2003) (Fig. 2B). Rathgeber et al. (2005) reported the first AAP that can form cyclical cells, Roseicyclus (R.) mahoneyensis (Fig. 2C), reminiscent of Rhodocyclus purpureus (Pfennig, 1978), to which it is only distantly related. Uncultured spirilloid cells ascribed to AAP have turned up in the Sargasso Sea (Sieracki et al., 2006). Cell accessories and intercellular connective structures provoke further interest, especially ecologically. Pleomorphic Porphyrobacter neustonensis and Porphyrobacter meromictius produce appendages, perhaps enhancing intake of nutrients by maximizing surface-to-volume ratio (Fig. 2D) (Yurkova et al., 2002; Rathgeber et al., 2007). More interestingly, cell chains are connected via bubble-like formations with a central tubular structure, an arrangement heretofore unobserved in the domain Bacteria (Fig. 2E) (Yurkova et al., 2002; Rathgeber et al., 2007). Bubble-like membranous connective structures are also seen between cells of Cmi. bathyomarinum (Yurkov and Beatty, 1998). The relevance of these structures is unknown. Perhaps they facilitate mate-
Vladimir Yurkov and Julius T. Csotonyi rial exchange between cells. Several members of the α-2, α-3 and α-4 Proteobacteria, including Staleya (Stl.) guttiformis, Stappia (S.) marina, Roseovarius (Rva.) mucosus, Roseisalinus antarcticus and deep ocean strain C8 (Fig. 2F) form star-shaped rosettes of cells, joined by unknown means (Biebl et al., 2005a; Labrenz et al., 2005; Kim et al., 2006). The taxonomically unassigned meromictic lake strains BL7 and BL14, however, form the most impressive rosettes, 10 µm or more in diameter and resembling dandelion heads (Fig. 2G) or even brain coral (Fig. 2H), in which cells are linked by a polar hook-like structure (Fig. 2I) (Rathgeber et al., 2004). Interestingly, all of these rosette-forming species except S. marina and C8 are meromictic lake isolates, suggesting that aggregation may bestow an ecological advantage, such as buoyancy regulation, a useful feature in lake habitats where salinity varies with depth. Alternately, the convoluted surface of rosettes may assist in trapping particulate nutrients. B. Taxonomic Diversity AAP are dispersed throughout the Alphaproteobacteria, with one known betaproteobacterial representative, Roseateles (Rst.) depolymerans (Fig. 1B) (Yurkov and Csotonyi, 2003) and a recently described gammaproteobacterial species, Congregibacter litoralis (Fuchs et al., 2007). Thus far, alphaproteobacterial species occur in five cohesive clusters, in all four subclasses. Soil, hot spring and acidophilic isolates define the α-1 cluster, consisting of Acidiphilium, Acidisphaera, Craurococcus, Geminicoccus, Paracraurococcus, Roseococcus and Rubritepida. The exclusively marine α-2 species form two unrelated branches, the first containing Labrenzia, Roseibium and S. marina, the second Hoeflea (H.) phototrophica. The oceanic or halophilic α-3 Roseobacter clade, allied closely with purple nonsulfur Rhodobacter and Rhodovulum, includes Dinoroseobacter, Roseibacterium, Roseicyclus, Roseinatronobacter, Roseisalinus, Roseivivax, Roseobacter, Roseovarius, Rubrimonas, Staleya and Thalassobacter. Finally, Blastomonas natatoria, Citromicrobium, Erythrobacter, Erythromicrobium, Erythromonas, Sandaracinobacter and Sandarakinorhabdus represent the α-4 AAP, hailing from diverse freshwater and marine habitats. Citromicrobium, Erythrobacter, Erythromicrobium and Porphyrobacter form a clade so cohesive that 16S rDNA sequence analysis is often insufficient to place species into proper genera, generic assignment rely-
Chapter 3
Aerobic Anoxygenic Phototrophs
35
Fig. 2. Morphological diversity of AAP. A. Branching in Erythromicrobium ramosum. B. Y-cell of Citromicrobium bathyomarinum. C. Tightly curved vibroid cells of Roseicyclus mahoneyensis. D. Appendage (arrow) of Porphyrobacter meromictius. E. Tube-in-bubble type of intercellular connections of Porphyrobacter meromictius. F. Rosettes of cells in strain C8. G. Dandelion head-like rosettes of strain BL7. H. Brain coral-like rosettes of strain BL14. I. Hooked appendage of strain BL14. C, F, G, H. Phase contrast micrographs. A, B, C (inset), D, E, I. Electron micrographs. Scale bars 0.25 µm (C inset); 0.5 µm (D, I); 1 µm (A, B, E); 2.5 µm (G, H); 5 µm (C, F).
ing heavily on phenotypic characteristics. Rainey et al. (2003) suggested unification of this cluster under a single genus. Instead of taxonomic abridgement, it is advisable to utilize the phenotypic diversity of this group (e.g., presence/absence of the unique B798-832 LH2 in Erythromicrobium and appendage formation in Porphyrobacter) to assign generic and specific status. Perhaps Porphyrobacter dokdonensis, possessing an 835-nm LH2 absorption peak and lacking appendages (Yoon et al., 2006), should have been placed within Erythromicrobium. Additional powerful phylogenetic analysis techniques, such as 23S rRNA sequence analysis (Peplies et al., 2004) or internal transcribed spacer analysis (Brown and Fuhrman, 2005), may be more phyletically infor-
mative for tight clusters than traditional 16S rDNA sequence analysis. The number of known AAP (currently 52 species in 33 genera) is waxing rapidly (Fig. 1A), with 18 new species described in the last four years (Table 1). H. phototrophica represents the first AAP from the α-2 proteobacterial Phyllobacteriaceae, which also contains rhizobia (Biebl et al., 2006). Meanwhile, R. mahoneyensis is the first inland Roseobacter-clade representative (Rathgeber et al., 2005). Description of the nearly completely genome-sequenced C. litoralis is extraordinary because not since 1999 has a new proteobacterial subclass been added to the short list that contains AAP (Fuchs et al., 2007). Gathering interest in this group, use of new high-throughput
36
Table 1. Determinative characteristics of novel species of aerobic anoxygenic phototrophs described since 2003 1. Habitat, site of isolation
Phylogenetic affiliation 2
Color
Cell shape
Congregibacter litoralis 3 (Fuchs et al., 2007)
sea water, North Sea
NOR5/OM60 clade (γ)
pink
pleomorphic
In vivo BChl peaks (nm) N.D.4
Dinoroseobacter shibae (Biebl et al., 2005b)
dinoflagellate culture, North Sea Rhodobacteraceae (α-3) Jannaschia helgolandensis, 94.1%
beige to light red
cocci or ovoid
804, 868
Geminicoccus roseus (Foesel et al., 2007)
marine aquaculture biofilter, Israel
diplococci
N.D.
Hoeflea phototrophica (Biebl et al., 2006)
dinoflagellate culture, North Sea Phyllobacteriaceae (α-2) Hoeflea marina, 98.5%
colorless to pink
short rod with capsule
N.D.
Labrenzia alexandrii (Biebl et al., 2007)
dinoflagellate culture, North Sea Rhodobacteraceae (α-2) Stappia alba and Stappia marina, 98.0%
faint pink
rod, uneven ends
801, 865
Porphyrobacter cryptus (Rainey et al., 2003)
hot spring water, Portugal
Sphingomonadaceae (α-4) Porphyrobacter tepidarius, 98.3%
reddish-orange
short rod
800, 870
Porphyrobacter dokdonensis (Yoon et al., 2006)
sea water, Dokdo, Korea
Sphingomonadaceae (α-4) Porphyrobacter cryptus, 98.7%
reddish-orange
pleomorphic: cocci, ovoid 800, 835, 862 or rod
Porphyrobacter donghaensis (Yoon et al., 2004b)
sea water, East Sea, Korea
Sphingomonadaceae (α-4) Erythromicrobium ramosum, 99.0%
reddish-orange
pleomorphic: cocci, ovoid 808, 867 or rod
Porphyrobacter meromictius (Rathgeber et al., 2007)
meromictic lake water, Canada
Sphingomonadaceae (α-4) Erythrobacter aquamaris, 98.1%
reddish-brown
short rod or pleomorphic
806-808, 866-867
Roseibacterium elongatum (Suzuki et al., 2006)
sand, Shark Bay, Australia
Rhodobacteraceae (α-3) Marinosulfonomonas methylotropha, 93%,
pink
rod, variable length
800, 879
Roseicyclus mahoneyensis (Rathgeber et al., 2005)
meromictic lake water, Canada
Rhodobacteraceae (α-3) Ketogulonicigenium vulgare, 92.6%
pinkish-purple to purple
pleomorphic: ovoid, long rod, vibrioid, cyclical rod
805-806, 870-871
Roseinatronobacter monicus (Boldareva et al., 2007)
hypersaline lake, USA
Rhodobacteraceae (α-3) Roseinatronobacter thiooxidans, 98.1%
pink
ovoid
804, 870
Roseisalinus antarcticus (Labrenz et al., 2005)
meromictic lake water, Antarctica
Rhodobacteraceae (α-3) Jannaschia helgolandensis, 94%
red
rod
800-801, 870
Novel clade (α−1) whitish-grey to pink Candidatus “Alysiosphaera europaea,” 90.8%
Vladimir Yurkov and Julius T. Csotonyi
Species (described by)
1 New taxa described before 2003 are covered by Yurkov and Beatty (1998), Yurkov and Csotonyi (2003) and Rathgeber et al., (2004). 2 family (nomenclature after Garrity et al., 2005), proteobacterial subclass (in parentheses), nearest 16S rDNA relative, phylogenetic distance to nearest relative; 3 names in bold indicate novel genera. 4 N.D., not determined.
N.D. ovoid to irregular rod salmon-pink Rhodobacteraceae (α-3) Jannaschia helgolandensis, 95.7% Thalassobacter stenotrophicus (Macián et al., 2005)
sea water, Mediterranean Sea
N.D. N.D. rod Rhodobacteraceae (α-2) Stappia aggregata, 98.5% Stappia marina (Kim et al., 2006)
tidal flat, Korea
N.D. rod orange-red soil, South Korea Sphingomonas kaistensis (Kim et al., 2007)
Sphingomonadaceae (α-4) Sphingomonas oligophenolica, 95.8%
800, 837, 865 rod Sandarakinorhabdus limnophila freshwater lake water, Germany Sphingomonadaceae (α-4) Sandaracinobacter sibiricus, 92.8% (Gich and Overmann, 2006)
orange-red
ovoid to rod whitish to faint pink dinoflagellate culture, North Sea Rhodobacteraceae (α-3) Roseovarius tolerans, 96.4% Roseovarius mucosus (Biebl et al., 2005a)
In vivo BChl peaks (nm) N.D. Cell shape Phylogenetic affiliation 2 Habitat, site of isolation Species (described by)
Table 1. Continued
Aerobic Anoxygenic Phototrophs
Color
Chapter 3
37 isolation methods, and the application of phylogenetic techniques to detect photosynthesis genes are responsible for this taxonomical flurry (Gich et al., 2005). Our current delineation of AAP diversity is by no means the end of the taxonomic tale. Sequencing marine photosynthesis genes or their mRNA, Béjà et al. (2002) and Allgaier et al. (2003) detected much higher AAP diversity than suggested by culture-dependent studies. Description of novel lotic β-proteobacterial species is soon anticipated (Waidner and Kirchman, 2005), while procedures facilitating recent isolation of previously uncultivable SAR11 marine bacterioplankton (Rappé et al., 2002) may yield additional elusive γ-proteobacterial AAP. C. Persistent Taxonomic-Phylogenetic Tangles Embedding a thorn in taxonomists’ sides, 16S ribosomal genes of AAP often ally more closely with non-phototrophs than with purple nonsulfur bacteria. This phyletic interspersion still baffles enough researchers that, regrettably, an old argument must be revisited regarding the confusion of taxonomy with phylogeny in the classification of AAP. Taxonomic convention recognizes major metabolic capabilities such as phototrophy as markers to distinguish genera. After all, in spite of the deceptive visual homogeneity of bacteria harboring divergent metabolic strategies, the difference made by possession of a photosynthetic apparatus is as significant as would be the discovery of a fish breathing nitrate rather than oxygen. Yet, since 2002, no less than nine species of BChl-lacking bacteria have been proposed as members of phototrophic genera (Erythrobacter aquamaris, Erythrobacter citreus, Erythrobacter flavus, Erythrobacter gaetbuli, Erythrobacter luteolus, Erythrobacter vulgaris, Roseovarius crassostreae, Roseovarius nubinhibens) (Denner et al., 2002; Yoon et al., 2003, 2004a, 2005a,b; González et al., 2003; Boettcher et al., 2005; Ivanova et al., 2005). Conversely, and despite repeated appeals to the contrary since 1998 (Yurkov and Beatty, 1998), BChl-producing species such as H. phototrophica and S. marina are included in definitively non-phototrophic genera. Ironically, the very phylogenetic analyses sparking original misclassifications, upon closer inspection, support genus-level distinction of BChl-free isolates from true AAP. New non-phototrophic erythrobacters were sufficiently unrelated to phototrophic congener-
38 ics (16S rDNA sequence similarities as low as 94.1% and DNA-DNA reassociation levels often below 10%) to justify erection of new genera (Yoon et al., 2003, 2004a, 2005a). Similarly, non-phototrophic Rva. nubinhibens and Rva. crassostreae can clearly be differentiated from the phototrophic Roseovarius cluster by specific signature nucleotides at several positions of the secondary structure of the 16S rRNA gene (Biebl et al., 2005a). Judicious phylogenetic analysis supports the repeated insistence by the International Committee on Systematic Bacteriology that phenotypic traits should remain important criteria in delineation of genera (Wayne et al., 1987; Murray et al., 1990; Stackebrandt et al., 2002). Phylogenetic analysis is not sufficiently advanced for consensus on the use of a standard suite of purely genetic techniques for resolution of both macro- and microdiversity, and thus it remains inappropriate to use phylogenetic relatedness as a sole or overriding determinative factor in assigning taxonomic status. Errors are occasionally corrected by reclassifying miscategorized species — see Yurkov and Csotonyi (2003) for examples — but lamentably, some new papers still contain the original misnomers and cite erroneous papers as major references. For instance, Boettcher et al. (2005) alluded to ‘Roseobacter gallaciensis,’ which was long ago reclassified as Ruegeria gallaciensis (Uchino et al., 1998). Clearly, reviewers and editors of premier bacterial systematics journals must exercise greater discrimination when considering proposals of taxonomic placement to avoid subsequent confusion from rearrangements necessary to correct taxonomic faux pas. D. Evolution We must distinguish evolution of aerobic anoxygenic phototrophy as a physiological process from the evolutionary history of its phylogenetic distribution among AAP clades. In hindsight, it is not surprising that aerobic anoxygenic phototrophy arose from anaerobic phototrophy after global oxygenation; given life’s rapid niche-occupying tendency, such a transition ought to occur over 2 billion years after the earth’s atmosphere became oxygenated (Rye and Holland, 1998). This question of how such evolution played out still deserves much research effort, but speculation is fruitful. Similarity of the AAP photosynthetic apparatus to that of purple nonsulfur bacteria implicates the latter as ancestors. Probably
Vladimir Yurkov and Julius T. Csotonyi the most serious hurdle for evolution to clear was aerobic phototoxicity (Beatty, 2002). Exposure of BChl to light in the presence of O2 generates highly reactive chemical species called triplet BChl and singlet oxygen, which can cause considerable oxidative damage to cells (Krinsky, 1971; Beatty, 2002). Hence, repression of aerobic BChl synthesis by anaerobic phototrophs in response to global O2 pollution served as much a protective as a resource conservation role. Interestingly, a logical solution to phototoxicity conveniently explains three definitive features of AAP. First, low PSU abundance minimizes generation of triplet BChl and singlet O2 (Beatty, 2002). Second, BChl is only generated during dark periods, when it is safe to do so (Beatty, 2002). This paradoxical trait seems counterintuitive only until photochemistry is considered. Third, synthesis of light-harvesting photoactive carotenoids by anaerobic phototrophs also preadapted their descendents for an aerobic existence. Not only do carotenoids with more than nine conjugated double bonds filter out high-energy radiation, but also their low lying first excited triplet state enables them to efficiently quench the destructive energy of singlet O2 and triplet BChl (Krinsky, 1971; Fraser et al., 2001). AAP needed only to upregulate carotenoid synthesis to achieve better protection from the dangerous combination of light, O2 and BChl — which is exactly what we see, especially for the particularly photoprotective carotenoid sulfates (Krinsky, 1971; Yurkov and Beatty, 1998). In light of the inherent self-limitation of aerobic phototrophic energy generation, AAP have understandably lost the energetically expensive Calvin cycle for carbon fixation, relying instead on heterotrophy. Beyond explaining their appearance as a functional group, however, phylogenetic interspersion of AAP with non-phototrophs seriously challenges evolutionary microbiologists wishing to trace the phylogenetic roots of each clade. Unfortunately, genetic research has found that many of the pieces of this evolutionary puzzle do not fit snugly. There is little doubt that AAP are polyphyletic at the level of the 16S rRNA gene. However a battle ensues between two opposing views: (1) that phototrophy is ancestral in Proteobacteria, with non-phototrophs such as Paracoccus or Ketogulonicigenium arising through loss of photosynthesis, and (2) that aerobic photosynthetic genes were transferred laterally to unrelated non-phototrophs (Yurkov and Beatty, 1998). The argument hinges on the relative topologies of ribosomal and photosyn-
Aerobic Anoxygenic Phototrophs C
B Absorbance
A
39
Absorbance
Chapter 3
869
857
804
799
400
500
600
700
800
900
1000
400
500
D
E
600
700
800
900
1000
Wavelength (nm)
Wavelength (nm)
P870*
F
Absorbance
BChl BPh
+
QA
E0'
868 836
cyt bc1
799
cyt c2 400
500
600
700
800
900
1000
P870
Wavelength (nm)
G
H
I
Fig. 3. Photosynthetic and habitat diversity of aerobic anoxygenic phototrophs (AAP). A–E. In vivo absorbance spectra of cells grown under different physiological conditions, with numerals above peaks denoting LH and RC wavelengths. Insets, photographs of liquid cultures. A. Roseicyclus mahoneyensis, illuminated (dashed line) and dark (solid line). B,C. Dashed line, rich organic medium (3 g·l–1 organics); solid line, minimal acetate (1 g·l–1) medium (B. Roseococcus thiosulfatophilus, C. Erythrobacter litoralis). D. Erythromicrobium ramosum, rich organic medium, dark (dashed line) and oligotrophic medium, light:dark regimen of 12h:12h, illuminated with diffuse ambient sunlight (solid line). E. Citromicrobium bathyomarinum strain JF1, rich organic medium (dashed line), minimal glucose medium (dotted line), compared with Erythrobacter litoralis (solid line). F. Hypothetical electron transfer system of aerobic anoxygenic photosynthesis, showing electron flow through major carriers: P870, special pair of BChl in RC (photoexcited state indicated by ‘*’); BChl-BPh, accessory BChl and bacteriopheophytin in RC; QA, quinone primary electron acceptor; cyt bc1, cytochrome bc1 complex; cyt c2, cytochrome c2. The symbol ‘+’ indicates that the midpoint potential of QA in all tested AAP is positive and higher than in anaerobic phototrophs. G–I. Extreme environment habitats of AAP. G. Hydrothermal vent field in Eastern Pacific Ocean, showing smoker chimney. H. Hypersaline spring system (East German Creek) in Manitoba, Canada: spring pool in foreground and playa in background, with white patches of salt precipitates. I. Cyanobacterial-Thiothrix mat development in thermal springs of Banff National Park, Alberta, Canada. See also Color Plate 2, Fig. 2.
thetic phylogenetic trees, and on the diversity of the aerobic photosynthetic apparatus. The answer probably lies between opposing schools of thought: loss of phototrophy may account for close non-phototrophic relatives, and lateral gene transfer for the rise of a few
AAP clades. Ribosomal and pufL (the gene encoding the L subunit of the photosynthetic reaction center) cladistic trees agree topologically at the highest level of similarity (such as the Erythrobacter clade), but a few deep incongruities exist (Allgaier et al., 2003).
40 Both α-3 Staleya and β-proteobacterial Roseateles possess pufL genes characteristic of α-4 Erythrobacter (Allgaier et al., 2003). Even more surprising are three marine strains, HTCC2080, HTCC2148 and HTCC2246, which, despite forming a tight 16S rDNA gammaproteobacterial cluster in the NOR5/OM60 clade, possess pufL genes of gammaproteobacterial, α-1 and α-2 proteobacterial affiliation, respectively (Cho et al., 2007). Other studies (Béjà et al., 2002; Oz et al., 2005) have underscored the observed interspersion of photosynthesis genes among different clades, but the emerging picture is complicated, the pattern of interspersion depending on the gene that is sequenced. For example, unlike pufL, the pufM gene of Rst. depolymerans clusters more closely with those of other Betaproteobacteria rather than with those of Erythrobacter (Béjà et al., 2002). It remains to be explained how such a linked gene pair as pufL and pufM can exhibit such phylogenetic topological variety. Furthermore, although the plasmid pBLM2 was used to mobilize the entire photosynthesis gene cluster of Rhodobacter capsulatus, pBLM2 did not induce functional pigment expression in Escherichia coli (Marrs, 1981). Thus, lateral photosynthetic gene transfer might occur successfully only between bacteria that are already phototrophic, possibly because specific membrane configurations are prerequisite for functional photosynthetic machinery (Marrs, 1981). Adding a further twist, the apparent α-proteobacterial origin of the entire 37-kb photosynthetic gene cluster of the β-proteobacterial anoxygenic phototroph Rubrivivax gelatinosus (an anaerobic 16S rDNA close relative of Rst. depolymerans) (Igarashi et al., 2001), implies multiple independent origins of aerobic phototrophy, as Rst. depolymerans probably evolved from a Rubrivivax gelatinosus-like ancestor. Moreover, the difference of LH2 complexes of AAP both from each other and from those of anaerobic phototrophs (see Section V.B.) also implicates multiple origins of aerobic phototrophy even within Alphaproteobacteria. As an attempt to synthesize these arguments into perhaps the most accurate picture of AAP evolution, we suggest that each major AAP clade evolved separately from a different purple nonsulfur bacterial branch. Moderate gene shuffling then occurred only among these phototrophs, and non-phototrophic Proteobacteria arose through loss of phototrophy, as proposed by Woese (1987).
Vladimir Yurkov and Julius T. Csotonyi III. Nutritional Versatility and Peculiarities of Carbon Metabolism A. Carbon Sources Metabolic diversity in AAP is extremely great, as expected for a functional group relying exclusively on heterotrophy for carbon needs. Erythrobacter, Erythromicrobium, Porphyrobacter, Roseicyclus, Roseobacter and Sandaracinobacter are able to grow on most carbon sources tested, including organic acids, carbohydrates, alcohols and complex organics (Yurkov and Csotonyi, 2003). However, some species are more specialized, with a second cluster (deep ocean Citromicrobium; meromictic Mahoney Lake strains ML15, ML17, ML35; Manitoba hypersaline spring strains EG4, EG6 and EG15) growing best on complex organics and on only a few defined carbon sources (Yurkov and Beatty, 1998; Yurkova et al., 2002; Csotonyi and Yurkov, unpublished). Most intriguingly, a third cluster (e.g., Mahoney Lake strains ML37, ML47; Manitoba hypersaline spring strain EG12) requires yeast extract or casamino acids, and subsists on no known defined medium (Yurkova et al., 2002; Csotonyi and Yurkov, unpublished). These species may be adapted to utilizing specific carbon sources produced by organisms or geological processes exclusive to their native environment, and they likely evolved locally by sympatric speciation based on metabolic niche differentiation. Endemism may explain the diversity of AAP in isolated habitats such as meromictic lakes or hypersaline springs. Certain carbon sources exploited by AAP bear ecological significance. Much recent attention focuses on dimethylsulfoniopropionate (DMSP), an osmolyte produced by algae and an important marine store of sulfur. The prominent role of Roseobacter relatives in DMSP consumption by cleavage or demethylation and dethiolation, and the resulting liberation of sulfide or climate-influencing dimethyl sulfide, implicates them as key modulators of the marine sulfur cycle (Wagner-Döbler and Biebl, 2006). The next significant step will be discrimination of BChl-containing obligately aerobic representatives from strict heterotrophs, because phylogenetic and [35S]DMSP-incorporation studies alone cannot distinguish phototrophs from non-phototrophs. Photoautotrophic CO2 fixation is lacking in all physiologically tested AAP, with both enzymological and full genome sequence studies (for Roseobacter
Chapter 3
Aerobic Anoxygenic Phototrophs
(Rsb.) denitrificans and C. litoralis) confirming the absence of the key Calvin cycle enzyme ribulose1,5-bisphosphate carboxylase/oxygenase (Rubisco) (Yurkov and Beatty, 1998; Fuchs et al., 2007; Swingley et al., 2007). However, nonexistence of autotrophy in AAP need not contradict the finding of a low level of CO2 fixation in Acidiphilium rubrum, Erythrobacter longus, Rsb. denitrificans, Sandaracinobacter sibiricus and other AAP (Kolber et al., 2001; Rathgeber et al., 2004). All heterotrophs anaplerotically fix CO2 (at efficiencies of 1–4% of that expressed by autotrophs) via phosphoenolpyruvate (PEP) carboxylase, a key respiratory enzyme that replaces organic acids sequestered from the citric acid cycle by other biochemical pathways. Light-stimulated citric acid cycle reversal may also account for some carbon fixation (Rathgeber et al., 2004). B. Alternative Electron Donors and Acceptors Although most AAP appear to rely on reduced carbon compounds as electron donors, a few species reflect anaerobic phototrophic ancestry by oxidizing reduced sulfur sources. Interestingly, most S-oxidizers belong to the α-1 proteobacterial lineage (e.g., Acidiphillium acidophilum, Roseococcus (Rsc.) thiosulfatophilus and Rubritepida flocculens). The latter two species, as well as Roseinatronobacter thiooxidans and Roseinatronobacter monicus, oxidize thiosulfate (Yurkov and Beatty, 1998; Yurkov and Csotonyi, 2003; Boldareva et al., 2007), whereas Acidiphilium acidophilum oxidizes elemental S via glutathione at a rate nine to fourteen times as high as that of the lithoautotroph, Acidithiobacillus thiooxidans (Rohwerder and Sand, 2003). Most AAP are described as obligately aerobic, but some species are capable of anaerobic chemoheterotrophic growth, substituting O2 with alternate electron acceptors. Dinoroseobacter (D.) shibae can dissimilatorily reduce nitrate, and Rsb. denitrificans reduces nitrate, nitrite and trimethylamine N-oxide (TMAO), a common odoriferous byproduct of decomposing marine organisms (Yurkov and Csotonyi, 2003; Biebl et al., 2005b). Even more interesting, and of bioremediative value, is the pH-elevating effect of dissimilatory Fe3+-reduction by A. acidophilum, perhaps key to facilitating transition of mine drainage systems from acidic to neutral conditions (Marchand and Silverstein, 2003).
41 IV. Photosynthetic Pigment Composition and Synthesis Reveal Surprises A. Carotenoids A ubiquitous feature of AAP is their diverse range of intense colors, conferred by abundant carotenoids. Colors span the spectrum from watermelon red to brown, orange, yellow, pink, purple and intermediates (Fig. 3A-E and Color Plate 2). Most interestingly, in contrast to the light-harvesting function of carotenoids in anaerobic purple phototrophs, the majority of AAP carotenoids are disengaged from energy transduction, and are distributed evenly throughout the cytoplasmic membrane, cytoplasm and cell wall (Yurkov and Beatty, 1998; Koblížek et al., 2003). The primary purpose of this non-photosynthetic carotenoid pool is uncertain, but preventative or remediative amelioration of phototoxicity is a leading suggestion (Yurkov and Beatty, 1998; Fraser et al., 2001; Beatty, 2002). The photoprotection to which AAP appear to have enlisted the bulk of their carotenoids is reflected not only in up-regulation, but also in extensive structural and compositional modification. Whereas AAP LHbound carotenoids resemble those of anaerobic phototrophs, the LH-dissociated pool displays extreme variety (Fig. 3A–E and Color Plate 2), with E. ramosum possessing about 20 types (Yurkov and Beatty, 1998). Moreover, despite predominance of purple nonsulfur-type spheroidenone and spirilloxanthin in some AAP (e.g., Roseobacter and Acidiphilium), the majority in most species are carotenes, which are more widely distributed among oxygenic than anoxygenic phototrophs and give AAP decidedly more orange hues than in purple nonsulfur bacteria (Yurkov and Beatty, 1998) (Fig. 3A–E and Color Plate 2). Some of these carotenoids are structurally unique among phototrophs (Chapter 6, Takaichi). AAP are especially well-endowed with polar carotenoids, such as the short-chain C30 molecule, 4,4´-diapocarotene-4,4´dioate in Rsc. thiosulfatophilus (Fig. 3B and Color Plate 2), or erythroxanthin sulfate in Erythrobacter and Erythromicrobium (Yurkov and Csotonyi, 2003) (Fig. 3C, D and Color Plate 2). Acquisition of the latter carotenoid, which is particularly antioxidative (Krinsky, 1971), is a logical adaptation to illuminated oxic environments.
42 B. Bacteriochlorophyll One of the namesake features of AAP is their possession of BChl, the remarkable bacterial pigment capable of absorbing the electromagnetic energy of light. All known AAP synthesize BChl a esterified to phytol like many purple nonsulfur bacteria (Yurkov and Beatty, 1998). However, members of the genus Acidiphilium are the only organisms known to naturally chelate the BChl tetrapyrrole ring to Zn instead of to Mg, producing a 7-nm blue-shift in the absorbance spectrum of the aberrant BChl a (Hiraishi and Shimada, 2001). Striking similarity in the electrochemistries of Zn and Mg permits this novel pigment to function, while metals such as Ni would produce ineffective bacteriochlorins, exhibiting too short an excited state lifetime to facilitate photosynthetic electron flow (Cogdell et al., 2002). However, a key difference in the acid chemistries of Zn and Mg explains the ecological significance of this pigment in Acidiphilium. Zn-BChl is 106 times more resistant to loss of the metal ligand (pheophytinization) by acids than is Mg-BChl, granting exceptional acid-stability in habitats with a pH as low as 2. Supporting this idea, the ratio of Zn-BChl to Mg-BChl synthesized by A. rubrum increases with acidity (Hiraishi and Shimada, 2001). Regardless of its chemical form, low cellular BChl content is an important determinative trait of the predominantly heterotrophic AAP (Fig. 3A–E and Color Plate 2), which produce at least ten times less BChl than the ~20 nmol/mg dry weight typical of purple nonsulfur bacteria: 2.0 nmol/mg cells in Erb. longus, 1.0 nmol/mg protein in Srb. sibiricus and too low to reliably measure in Rva. mucosus (Yurkov and Csotonyi, 2003; Biebl et al., 2005a). Consequently, absorbance peaks of BChl integrated into RC and LH complexes are about ten-fold lower than those of carotenoids. However, BChl:carotenoid ratios of 1.33:1 (Manitoba hypersaline spring strain EG14; Csotonyi and Yurkov, unpublished) and 1.2:1 (R. mahoneyensis; Rathgeber et al., 2005) (Fig. 3A–E and Color Plate 2) were also found. These resemble anaerobic phototroph ratios, begging two substantial questions. First, how do these AAP species circumvent phototoxicity incurred by aerobic operation of their high levels of BChl? Perhaps unique membrane environments or particularly antioxidant suites of carotenoids nullify this oxidative hazard. Second, does high BChl content indicate recent divergence from anaerobic phototrophy, or does it bespeak an
Vladimir Yurkov and Julius T. Csotonyi even more advanced condition than exists in other AAP, representing the next stage in evolution toward efficient exploitation of aerobic illuminated environments? Whatever the answers to these questions, the elevated BChl content in such strains makes them ideal candidates for investigation of regulatory factors of photosynthetic pigment expression. C. Highly Evolved Regulation of Pigment Synthesis If, as we propose, there was a transition from an anaerobic to an aerobic mode of existence, this necessitated modifications in the regulation of pigment expression. BChl synthesis in AAP is repressed by light (Fig. 3A and Color Plate 2) more extremely than in anaerobic phototrophs (Yurkov and Csotonyi, 2003). However, AAP also exhibit regulatory systems involving nutrients (Fig. 3B–E and Color Plate 2), temperature, pH and salinity (Alarico et al., 2002, Koblížek et al., 2003; Rathgeber et al., 2004; Macián et al., 2005; Biebl et al., 2006). Ongoing research is elucidating the mechanisms driving these regulatory regimes and delineating the range of responses. Ouchane et al. (2004) recently investigated AcsF, a strictly aerobically-active functional equivalent to the anaerobic Mg-protoporphyrin monomethyl ester cyclase, BchE, both of which catalyze formation of a key intermediate in BChl and Chl synthesis (Chapter 4, Willows and Kriegel). AcsF is thought to require O2, whereas BchE is thought to use H2O for the O atom needed for this step in (B)Chl synthesis (Ouchane et al., 2004). Sequencing studies revealed that whereas the acsF gene was absent in strict anaerobes such as green sulfur bacteria, homologs existed in facultative aerobes (purple and green nonsulfur bacteria) and in strict aerobes (cyanobacteria, green algae and plants), with the latter group lacking bchE (Ouchane et al., 2004). Although AAP would be expected to possess acsF and lack bchE, full genome sequencing of Rsb. denitrificans has turned up both genes (Swingley et al., 2007), making it unclear which genetic elements are responsible for restricting BChl synthesis to aerobic conditions in AAP. The final picture of O2 regulation of BChl synthesis is clearly more complex than indicated simply by presence or absence of two complementary genes. Maximal BChl synthesis in obligately aerobic Rst. depolymerans occurs at an atmospheric O2 content of only 2% (Suyama et al., 2002), and R. thiooxidans best expresses BChl microaerophilically (Yurkov and
Chapter 3
Aerobic Anoxygenic Phototrophs
Csotonyi, 2003) and C. litoralis exhibits chemotactic preference for about 10% O2 concentration (Fuchs et al., 2007). What governs species-specific fine-tuning of optimum O2 concentration for pigment synthesis is currently unknown. When complemented with Rsb. denitrificans homologs of the anaerobic phototrophic two-component O2 regulatory system (RegA and RegB) of pigment synthesis, Rhodobacter capsulatus still only expressed O2-deprived pigment synthesis (Masuda et al., 1999), indicating that the AAP system functions as does the Rhodobacter counterpart (Chapter 35, Bauer et al.). The mechanism by which light regulates pigment synthesis in AAP is even cloudier than is the story for O2. Possibly, light repression of pigmentation is an overexpression of a similar but less potent process that purple nonsulfur bacteria employ in response to dangerously intense radiation. In most AAP, such as Erythromicrobium hydrolyticum, even dim light is strongly repressive to BChl synthesis (Yurkov and van Gemerden, 1993). However, Suyama et al. (2002) demonstrated 25% maximal BChl content in Rst. depolymerans at 10 W/m2 light intensity, while D. shibae produced BChl at 15% of its dark rate in the light (Biebl and Wagner-Döbler, 2006). Most interesting is a recent investigation of pigment synthesis by D. shibae under light-dark cycles of varying periodicity (Biebl and Wagner-Döbler, 2006), an elaboration on similar work by Yurkov and van Gemerden (1993). When the light:dark regimen was changed from 8h:16h (simulating winter conditions) to 16h:8h (summer conditions), the nightly rate of BChl synthesis increased nearly four-fold, indicating not only sensitivity to instantaneous level of illumination, but also a response to the duration of the previous illuminated interval (Biebl and Wagner-Döbler, 2006). Such regulation would be advantageous in the organism’s native marine environment, preparing more photosynthetic pigment on summer nights in anticipation of longer, warmer days, when higher doubling rates more extensively dilute the cell’s PSU complement. The authors suggest that signal transduction systems such as the recently discovered bacteriophytochrome (Evans et al., 2005; Chapter 40, Evans et al.) might mediate the effects of light in BChl synthesis in AAP (Biebl and Wagner-Döbler, 2006). Such speculation sets the stage for exciting future biochemical research. AAP hold additional fascinating surprises. Pigment expression in Thalassobacter stenotrophicus, H. phototrophica and Cmi. bathyomarinum is salt-sensitive;
43 T. stenotrophicus exhibits no pigmentation at 7% salinity, H. phototrophicum produces maximal BChl at 0.6% salinity and no BChl at 3.5% salinity, and Cmi. bathyomarinum manufactures the most BChl in the absence of NaCl, despite the native marine environment (~3.5% salinity) of all three species (Rathgeber et al., 2004; Macián et al., 2005; Biebl et al., 2006). Indeed, numerous studies have demonstrated that AAP tend to upregulate pigment synthesis under suboptimal growth conditions, implying that photosynthesis proves especially advantageous under stress (Rathgeber et al., 2004). Supporting this idea, thermophilic Rubritepida flocculens synthesizes BChl at 30 °C, but not at its preferred growth temperature of 50 °C (Alarico et al., 2002), and whereas marine Erythrobacter relative NAP1 grew optimally at 32.5 °C, BChl production maximized at 22.5 °C and halted at temperatures above 30 °C (Koblížek et al., 2003). Given the prominence of heterotrophic metabolism in AAP, nutritional status is a paramount regulatory factor of photosynthetic pigment synthesis (Beatty, 2002). Sudden carbon source dilution caused Rst. depolymerans to transcriptionally induce BChl synthesis (Suyama et al., 2002). This response makes sense in light of competition for electron carriers by heterotrophy and phototrophy in AAP: redox agents are maximally available for phototrophy when exogenous carbon sources are limiting (Beatty, 2002). Such is the case for E. ramosum, which produces over twice as much BChl when starved as when amended richly with nutrients (Fig. 3D and Color Plate 2). Most unusually, however, BChl synthesis occurred only in stationary phase cultures of Rst. depolymerans, and illumination only extended the viability of mature cultures instead of enhancing growth rate as in E. hydrolyticum and D. shibae (Yurkov and van Gemerden, 1993; Biebl and Wagner-Döbler, 2006). These observations inspired suggestion of an innovative use for the photosynthetic apparatus in some AAP: maintenance of the electrochemical proton gradient by light-induced cyclic electron flow, a condition that is essential for cellular integrity during starvation conditions (Suyama et al., 2002). In other species, the role of nutrients in regulation of pigment synthesis is more obscure. Erythrobacter litoralis produces more carotenoids but the same amount of BChl when organic carbon is limiting (Fig. 3C and Color Plate 2), and production of BChl and carotenoids in Rsc. thiosulfatophilus are both stimulated by a rich organic medium (Fig. 3B and
44 Color Plate 2), necessitating more research to unravel such mysteries. Even more interestingly, regulatory effects of two factors can interact nonlinearly in unexpected ways, best illustrated by exposure of H. phototrophica and strain DFL-11 to 8-hour-long pulses of light and starvation (Biebl and Wagner-Döbler, 2006). Whereas pulses of light or starvation alone had little or no effect on BChl levels, simultaneous exposure to both factors caused an order-of-magnitude increase in BChl, especially significant in these otherwise poorly pigmented species (Biebl and Wagner-Döbler, 2006). Thus, transient illumination and nutrient deprivation may facilitate the detection of BChl in strains in which its presence is uncertain and therefore aid in AAP identification. In light of this discovery, the tirade about incorrect taxonomic assignment of AAP to non-phototrophic genera in Section II.C deserves the balance of an important counterpoint. Several recent isolates, including Rva. mucosus and H. phototrophica, initially tested spectrophotometrically negative for BChl production and were recognized as phototrophs solely by their pufLM genes (Allgaier et al., 2003). The repeated insistence that photosynthetic apparatus expression is a useful taxonomic marker for AAP because BChl is easily detected (e.g., Yurkov and Csotonyi, 2003) must be tempered with caution. Indeed, Rva. mucosus failed to produce measurable BChl even under the pulse of starvation and illumination that induced such a marked increase in pigmentation in H. phototrophica and strain DFL-11 (Biebl and Wagner-Döbler, 2006). Whereas a negative conclusion of BChl production should only follow a reasonably wide application of culture conditions, it is nonetheless conceivable that some isolates will evade all reasonable attempts to induce BChl synthesis, and would therefore understandably be misclassified as non-phototrophs in the absence of genetic sequencing. Aside from applications in taxonomic hair-splitting, this point inspires a potentially informative investigation. A PCR-based screening of described proteobacterial species for pufLM, acsF and bchE genes could delineate the true phylogenetic distribution of aerobic phototrophy in this group. The pufLM genes would establish phototrophic capability, while presence of acsF and absence of bchE could indicate an obligately aerobic nature. Physiological work on species possessing pufLM should then be performed as a precaution, because Swingley et al. (2007) and Fuchs et al. (2007) have reported bchE in the completely or nearly completely
Vladimir Yurkov and Julius T. Csotonyi sequenced genomes of the AAP R. denitrificans and C. litoralis, repspectively. V. The Mysterious Photosynthetic Apparatus of Aerobic Anoxygenic Phototrophs Demonstration that AAP incorporate BChl into a functional photosynthetic apparatus certainly raised eyebrows in a skeptical research community struggling to accept the idea of an aerobic variant of anoxygenic photosynthesis. Nevertheless, AAP proved to possess a properly transcribed, bonafide purple bacterial-type photosynthetic superoperon, encoding apoproteins of RC, LH1 and often LH2 complexes (Yurkov, 2006). Exactly how a sophisticated set of anaerobically evolved pigment-protein complexes was enlisted to operate only aerobically is still debated, and is grievously neglected in recently published research. The striking genetic and structural similarities of AAP photosynthetic machinery to that of purple nonsulfur bacteria may have stifled some interest in this area. Increasing accessibility of relevant tools and techniques should cause this field to burgeon. A. The Reaction Center of Aerobic Anoxygenic Phototrophs Has Purple Bacterial Roots As in anaerobic purple phototrophs, the heart of the AAP photosynthetic apparatus consists of a type II (quinone-type) RC with three structural subunits (L, M and H), four BChls, two bacteriopheophytins, two ubiquinones, a nonheme high-spin Fe2+, and a carotenoid (Yurkov and Beatty, 1998). In purple bacteria, components are encoded on the ~45 kb photosynthesis gene cluster, and superoperons and polycistronic mRNAs contribute to stoichiometric consistency (Klug, 1993). The structural cornerstone of the RC makes the genes encoding the L and M subunits (pufLM) most frequently utilized in genetic surveys for phototrophs (Yutin et al., 2005). RC gene sequences in AAP are similar to genes from anaerobic anoxygenic phototrophs, illustrating that only a few modifications were required to enable anoxygenic phototrophy to function in the presence of O2. As a prime example, a single key replacement of His L168 with glutamic acid in the RC of Acidiphilium may be largely responsible for stabilization and function
Chapter 3
Aerobic Anoxygenic Phototrophs
of Zn-BChl (Nagashima et al., 1997). However, the flip-side of high structural and functional dependence on fine-scale gene sequence variation is that coarse genetic analysis techniques may fail to resolve much diversity. Indeed, introducing greater variation in two new pufM primers returned 14 times as many phototrophic sequences from Eastern Mediterranean Sea water samples as did probes constructed to amplify the single pufM sequence (Yutin et al., 2005). Some reports of diversity in photosynthetic gene organization are controversial. Exemplifying this, the recent claim by Pradella et al. (2004) that the pufLM genes of Roseobacter litoralis and Stl. guttiformis are not chromosomal but reside on linear plasmids of 91 and 120 kb size, respectively, could provide another ‘smoking gun’ for the idea of lateral photosynthetic gene transfer among AAP. However, these conclusions must be weighed hesitantly, because both partial and complete genome sequencing have established a chromosomal location for the puf operon of Rsb. denitrificans (Liebetanz et al., 1991; Swingley, et al., 2007). It seems unlikely that congeneric organisms (two species of Roseobacter) could possess such vastly different genomic arrangements. If further research rules out the possibility of an artifact, then such a discovery would bestow on the AAP uniquely flexible photosynthesis gene location. Great molecular diversity in AAP exists in Cyt electron carriers that close the photosynthetic electron transfer circuit by returning electrons to photo-oxidized BChl in the RC. In Erythrobacter, Erythromicrobium and Mahoney Lake strain ML31, the RC is reduced by a soluble periplasmic Cyt c2, as in Rhodobacter (Rba.) sphaeroides (Rathgeber et al., 2004). In addition to a soluble Cyt, Cmi. bathyomarinum, Erythromonas ursincola, R. mahoneyensis, Rsb. denitrificans, Rsc. thiosulfatophilus and Srb. sibiricus also possess a tetraheme Cyt c (encoded by pufC in Rsb. denitrificans) tightly bound to the RC, a characteristic of the anaerobic anoxygenic phototroph Blastochloris viridis (Rathgeber et al., 2004; Yutin and Béjà, 2005). Interestingly, Rsb. denitrificans lacks the posttranslational N-terminal cleavage of PufC that anchors the resulting tetraheme Cyt c to the membrane via a diacylglycerol molecule in purple bacterial species such as Blastochloris viridis (Hucke et al., 2003). Instead, the elongated N-terminus fastens Cyt c and contains a transmembrane helix motif resembling the core of another polypeptide, PufX. The Rhodobacter PufX protein affects the orientation and dimerization of the RC-LH1-PufX complex (Francia et al., 1999;
45 Siebert et al., 2004; Chapter 9, Bullough et al.), and the pufX gene is located where pufC sits in other species (Yutin and Béjà, 2005). Roseobacter thus opens a unique window on the evolutionary history of the PufX protein, which may have evolved by incomplete deletion of pufC in ancestral phototrophs switching from a RC-bound to a soluble Cyt c that donates electrons to the RC (Hucke et al., 2003). Although AAP have not been shown to contain pufX, bacterial artificial chromosome (BAC) cloning vectors constructed from marine samples have yielded putative AAP sequences (such as eBAC 60D04) from uncultured Gammaproteobacteria containing a gene strongly resembling pufX (Yutin and Béjà, 2005). However, ascribing uncultured anoxygenic photosynthetic gene sequences from aerobic environments to the AAP is risky, as purple nonsulfur bacterial sequences may masquerade as AAP to unwary eyes. B. Light-Harvesting Complexes: Ordinary and Extraordinary Surrounding the RC, AAP possess LH antenna complexes that sequester additional photonic energy, both by increasing light-harvesting cross-sectional area and by broadening the range of captured wavelengths. The core LH1 complex, existing in all AAP, shows little variation, but accessory LH2 manifests unexpected diversity. Reflecting the conserved nature of LH1 is its close genetic association with the RC: LH1 subunits β and α are encoded by genes (pufBA) cotranscribed with and immediately upstream of pufLMC (Yutin and Béjà, 2005). Carotenoid-rich LH1 forms an intimate ring of about 30 BChls around the RC (Yurkov and Beatty, 1998), facilitating highly efficient transfer of excitons to the RC primary electron donor, P870. In this respect, AAP strongly resemble purple nonsulfur bacteria such as Rba. sphaeroides, sharing similar RC-LH1 near-IR absorption spectra, with maxima at around 800 nm (RC) and 870 nm (LH1) (Yurkov and Csotonyi, 2003). With the exception of Zn-BChl of Acidiphilium, however, a few AAP display further unexplained spectral deviation of LH1, such as blue-shifting (856 nm maximum) in Rsc. thiosulfatophilus (Fig. 3B and Color Plate 2) or red-shifting (879 nm maximum) in Roseovarius tolerans (Yurkov and Beatty, 1998; Yurkov and Csotonyi, 2003). The structural factors responsible for this spectral fine-tuning remain to be elucidated. In contrast to LH1, the peripheral LH2 complex of AAP varies widely in both expression and structure.
46 In Rhodobacter and Rhodopseudomonas anaerobic phototrophic species, the LH2 genes reside in one or more puc operons, presumably enabling more independent transcriptional regulation (Klug, 1993; Larimer et al., 2004). Most tantalizing is the striking spectral differences of LH2 antennas in AAP from those of purple bacteria. When present, the LH2 complexes of anaerobic purple phototrophic bacteria exist in several spectral forms, their nomenclature denoting primary near-IR absorption maxima in nm: e.g., B800-850 (typical Rhodobacter-type), B800-820 (in Rhodopseudomonas (Rps.) acidophila) and the BChl b-containing B800-1020 (in Ectothiorhodospira) (Steiner and Scheer, 1985; Cogdell et al., 2002). The unusual purple nonsulfur bacterium Rps. palustris possesses four different peripheral LH species, including a monomodal LH2 complex with a single absorption peak (Hartigan et al., 2002). Interestingly, and for reasons unknown, the LH2 spectral classes of AAP often differ from those of their anaerobic relatives: B798-832 (in α-4 Erythromicrobium, Sandarakinorhabdus and Porphyrobacter dokdonensis) (Fig. 3D and Color Plate 2), B800-814 (in α-4 Porphyrobacter donghaensis) and B806 (in the α-3 Roseobacter, Roseicyclus and Rubrimonas) (Fig. 3A and Color Plate 2) (Yurkov and Beatty, 1998; Rathgeber et al., 2004; Yoon et al., 2004b, 2006; Gich and Overmann, 2006). The latter is a rare monomodal LH complex, similar to that found in only one anaerobic species, Rps. palustris (Hartigan et al., 2002). However, a 4 nm difference in its absorption maximum calls into question how structurally similar B806 is to the Rps. palustris LH2, which is an octamer with four BChl per αβ peptide instead of the usual three, each arranged radially rather than tangentially (Hartigan et al., 2002). Cogdell et al. (2002) propose that controlling the rotation of the BChl acetyl group relative to the bacteriochlorin ring plane by adjusting the content of hydrogen bonding amino acid residues at key positions in the LH2 apoprotein permits effective modulation of the Qy absorption band shift. Because loss of hydrogen bonding robustly explains the development of B800-820 from B800-850 (Fowler et al., 1992, 1994), a similar transition may have generated B798-832. Curiously, however, Raman spectroscopy indicates the presence of H bonds to the acetyl substituents of BChl in E. ramosum (Yurkov and Beatty, 1998), meaning that the case is definitely not closed on the source of LH2 Qy band shifts in AAP.
Vladimir Yurkov and Julius T. Csotonyi C. The Riddle of Aerobic Anoxygenic Phototrophy Exploring PSU biophysics, biochemistry and molecular genetics should resolve the question addressing the definitive quality of AAP, i.e., what allows their PSU to function aerobically. Unfortunately, similarity in RC architecture between AAP and anaerobic phototrophs (Section V.A.) complicates matters. Further smearing the distinction between aerobic and anaerobic phototrophy is the recent discovery of obligately microaerophilic photosynthesis in R. mahoneyensis (Rathgeber et al., 2004). Still, several studies provide a smattering of reasons that may explain why AAP photosynthesis is obligately aerobic. Energy-conserving photosynthetic electron flow in AAP appears to trace a circuit typical of anaerobic anoxygenic phototrophic bacteria, from the photooxidized special pair of BChl (P870+) to auxiliary BChl, bacteriopheophytin, the ubiquinone primary e– acceptor QA, then exiting the RC to the membrane quinone pool, a membrane bound Cyt bc1 complex, Cyt c2, and back to the RC (Fig. 3F and Color Plate 2). AAP differ subtly from purple nonsulfur bacteria at three principal points, the first involving the redox poise of the primary acceptor, QA. In most AAP, QA has a much higher midpoint potential (+5 to +150 mV) than in anaerobic purple bacteria (usually negative potentials), causing it to become and remain overreduced anoxically and halting electron flow (Fig. 3F and Color Plate 2) (Rathgeber et al., 2004). The structural basis of this functional discrepancy remains a mystery, as homology modeling of quinone binding pockets in the RC of Rsb. denitrificans demonstrates remarkably similar interactions between protein and quinone as compared with Rba. sphaeroides (Kohler et al., 2005). The second distinction of AAP is the lack of an alternative quinol oxidase pathway, thought to maintain the quinone pool at a proper redox state for efficient electron transport under photosynthetic conditions (Yurkov and Beatty, 1998). Finally, failure of soluble Cyts to transfer electrons anoxically from the Cyt bc1 complex to the RC-bound Cyt c in Rsb. denitrificans also implicates these components in arresting anaerobic photosynthetic electron flow in AAP (Schwarze et al., 2000). Hence, properties of one or more electron carriers found in the RC, membrane lipid phase and periplasm may prohibit anaerobic photosynthesis in the AAP, but the relative contribution of each element is presently unknown.
Chapter 3
Aerobic Anoxygenic Phototrophs
Whether scarcity of recent publications is a cause or an effect of the recalcitrance of the central biophysical question encompassing obligately aerobic photosynthesis is uncertain, but hopefully this lull will soon end. VI. Speculation on Ecological Roles Despite strides over the last four years in both culture-based and culture-independent studies, AAP ecological understanding is still in its adolescence. Researchers have discovered experimental tools with great potential, but proper combination of techniques to guarantee maximal analytical power still requires refinement. At the crux of the issue, comprehension of ecological significance requires qualitative description of interactions to go hand in hand with both assessment of biogeochemical cycling and reliable enumeration. Whereas some interesting associations have been described, work on elemental cycling is scant. Also, in spite of its seeming triviality, measuring the abundance of AAP in nature has proven exceptionally difficult. A. Technical Challenges of Enumerating Aerobic Anoxygenic Phototrophs AAP were traditionally enumerated by culturing on rich organic media (Yurkov and Beatty, 1998). This technique is very powerful for assessing the abundance of a physiological subset of the community and offers the added benefit that all enumerated organisms can subsequently be examined in physiological and phylogenetic detail. However, realization of the miniscule proportion of microorganisms amenable to cultivation spurs a growing interest in census techniques that circumvent the need to culture organisms. Regrettably, most culture-independent methods possess significant shortfalls. Phylogenetic surveys based on 16S rRNA genes alone, constituting much of marine microbial research, fail to discriminate between closely related phototrophic and non-phototrophic organisms. Efforts to screen for photosynthesis genes, such as quantitative or real-time PCR (Schwalbach and Fuhrman, 2005; Du et al., 2006), overcome this phylogenetic interspersion issue, but using primers constructed from described organisms inherently limits the diversity of detectable phototrophs (Yutin et al., 2005). Even bulk pigment analysis of ocean water by HPLC, or of individual cells by infrared
47 epifluorescence microscopy (IREM) or its variants (Schwalbach and Fuhrman, 2005; Jiao et al., 2006; Sieracki et al., 2006) are faulty as they measure a pool of BChl produced by both AAP and purple nonsulfur bacteria. IREM can even misidentify cyanobacteria as anoxygenic phototrophs and underestimate Prochlorococcus populations (oxygenic phototrophs particularly numerous in oligotrophic ocean waters) by orders of magnitude (Jiao et al., 2006), although newer studies attempt to correct for this effect (Lami et al., 2007). Finally, the sensitivity of BChl synthesis to a slew of environmental factors (Section IV.C.) generates great variation in pigment content from cell to cell, invalidating estimates of overall AAP abundance based on average cellular BChl concentration (e.g., Kolber et al., 2001; Koblížek et al., 2005). Obviously, measuring the abundance of AAP in nature is no trivial task. Thus far, three preferable alternatives exist: (1) infrared fast repetition rate fluorometry (IRFRR), which unfortunately requires state-of-the-art instruments, but can measure and distinguish between aerobic and anaerobic photosynthetic electron transport; (2) QPCR of pufLM (and of genes that can be used to discriminate between aerobic/anaerobic photosynthesis, such as acsF and bchE); and (3) combination of culture-independent ribosomal and photosynthetic gene surveys with cultivation, followed by 16S rDNA genetic analysis of cultured organisms. To date, only a few studies have applied either the first (Kolber et al., 2000, 2001; Koblížek et al., 2003, 2006) or the third (e.g., Gich et al., 2005; Salka et al., 2006) strategy to AAP, and no studies have yet employed the second. The majority of reported marine AAP enumeration must therefore be judged inconclusive, because abundance estimates represent only an upper limit, pooling AAP with nonphototrophs or anaerobic phototrophs, depending on methods employed. B. Ocean Surface Between 2003 and 2006, nearly 75% of approximately 22 ecologically relevant publications focused on marine environments, from which AAP were originally isolated. Three reports likely instigated this return to the sea. First, AAP were surprisingly recovered from deep ocean hydrothermal vent plumes (Yurkov et al., 1999). Second, although AAP have long been known to constitute up to 10 to 30% of the cultivable bacterial community in hypersaline and hydrothermal environments (Yurkov and Csotonyi, 2003) (Fig. 3G, I and
A, B ND NW Atlantic: Sargasso Sea (11–15 m) Sieracki et al. (2006)
Mar 02
1.3–2.5%; (0.71–1.49 × 104 ml–1); 2.4–3.3% of biomass
IREM
A, B ND NW Atlantic: USA coast to Gulf Stream (11–15 m) Sieracki et al. (2006)
Mar 02
0.8–2%; (0.76–2.1 × 104 ml–1); 1.8–2.5% of biomass
IREM
A, B ND NW Atlantic: Sargasso Sea (11–15 m) Sieracki et al. (2006)
Oct 01
0.8–2.6%; (0.82–1.26 × 104 ml–1); 3.4–5% of biomass
IREM
A, B
A, B NW Atlantic: USA E coast to Gulf Stream (11–15 m) Sieracki et al. (2006)
Oct 01
ND
IREM
A, B
2.3–9.4%; (2.16–9.84 × 104 ml–1); 7–12% of biomass
IREM ND <1–5% (<1–2.5 × 104 ml–1)
IREM 0.3–2.6 0.8–18%
North Pacific Gyre (0–100 m) Cottrell et al. (2006)
Feb 04
Mid Atlantic Bight (0–100 m) Cottrell et al. (2006)
Aug 03
B 0.3–2.2 ND 3 Black Sea Koblížek et al. (2006)
Jun 01
IRFRR
0–2.32 0.1–24.2% (up to 1.94 × 105 ml–1) S Pacific (0–270 m) Lami et al. (2007)
Oct–Dec 04
IREM
IREM, QPCR Delaware Estuary (DE, USA) Waidner and Kirchman (2007)
Aug 02–Mar 06
ND 3
Limitations of Method 2 Method 1 BChl a / Chl a (%)
% of Total Bacterial Population; (Absolute AAP Abundance); % of total bacterial biomass 0.9–34% (1.7–5.1 × 105 ml–1) Month & Year of Sampling Location of Sampling Publication
Table 2. Abundance estimates of marine aerobic anoxygenic phototrophs
Color Plate 2), the extraordinary claim by Kolber et al. (2001) that they represent 11% of the vast marine euphotic microbial community suddenly numbered AAP among the most plentiful organisms on earth, greatly elevating their importance in scientists’ eyes. Third, misinterpretation of anaplerotic CO2 incorporation by marine AAP as photosynthetic CO2 fixation (i.e., photoautotrophic growth) led to inflated estimates of their importance to marine carbon cycling (Kolber et al., 2000). The ensuing half decade was rife with arguments for and against the ecological importance of marine AAP, depending on techniques employed and on time and location of sampling (Table 2). The overall picture materializing is complex, but some resolution of disputes is now becoming possible. Integrating results of several studies yields populations of 104–105 cells ml–1, representing about 1 to 10% of all bacteria (Table 2). Seasonal population cycling is at the heart of much of the variation. In the northwest Atlantic, the numbers of AAP declined nearly ten-fold from autumn to spring, which Sieracki et al. (2006) attributed to temperature sensitivity. Positive correlation with water temperature also explains some ocean basin-wide differences, such as values of 0.01–0.06% of total microbes in Antarctic waters versus 0.72% at Bermuda, both sampled during their respective summers (Schwalbach and Fuhrman, 2005). Exceptions, such as consistently greater numbers of AAP in cold northern European waters — 1.37–10% of bacteria (Koblížek et al., 2003; Schwalbach and Fuhrman, 2005) — indicate influence by other factors, such as primary productivity. High relative abundance in nutrient rich temperate estuaries (10.7–34%) (Schwalbach and Fuhrman, 2005; Waidner and Kirchman, 2007) and algae-rich waters (Cottrell et al., 2006; Sieracki et al., 2006) seems to contradict the proposal that AAP experience greatest competitive success in oligotrophic settings (Beatty, 2002), but a recent report of 24.2% relative abundance in the gyre of the South Pacific Ocean, the most oligotrophic portion of the world ocean, contests this conclusion (Lami et al., 2007). Response to algal blooms, which generate irregular or quasiregular nutrient pulses, may account for long-term fluctuations in AAP populations (Schwalbach and Fuhrman, 2005; Sieracki et al. 2006). Exemplifying the reliance of some AAP on concentrated nutrient sources, Roseobacter-clade species frequently form close associations with dinoflagellates such as Alexandrium (Table 1; Allgaier et al.,
A, B A, C A, B
Vladimir Yurkov and Julius T. Csotonyi
48
Location of Sampling
Month & Year of Sampling
% of Total Bacterial Population; (Absolute AAP Abundance); % of total bacterial biomass
BChl a / Chl a (%)
Method 1
Limitations of Method 2
Koblížek et al. (2005)
Baltic Sea transect (54–60° N)
Aug–Sep 03
3–10%; (0.7–5 × 105 ml–1)
0.12–0.65
KF
A
Schwalbach and Fuhrman (2005) E Pacific (CA, USA)
Jul 01–Aug 02
1.66%
ND
TIREM
A
Schwalbach and Fuhrman (2005) E Pacific (CA, USA)
Jul 01–Aug 02
1.17%
ND
QPCR
A, C
Schwalbach and Fuhrman (2005) NW Atlantic: USA E coast estuaries (2 sites)
Aug 03, Aug 98
10.70%, 18.74%
ND
QPCR
A, C
Schwalbach and Fuhrman (2005) Antarctica (4 sites)
Dec 93–Jan 94
0.01–0.06%
ND
QPCR
A, C
Schwalbach and Fuhrman (2005) Caribbean Sea (Bermuda, Barbados)
May 90, Jun 94
0.12–1.42%, 0.72%
ND
QPCR
A, C
Schwalbach and Fuhrman (2005) W Pacific (Singapore, Philippines, Fiji)
Sep 90, Jul 94, Apr 98
0.30%, 0.71%, 0.92%
ND
QPCR
A, C
Schwalbach and Fuhrman (2005) SW Pacific (New Caledonia, Great Barrier Reef)
Mar 98, Mar 99
0.51%, 0.73%
ND
QPCR
A, C
Schwalbach and Fuhrman (2005) NE Atlantic (France, Norway)
Jul 93, Aug 96
1.37%, 2.28%
ND
QPCR
A, C
Goericke (2002)
E Pacific (CA, USA)
Dec 00–Jul 01
ND
0.1–2.0
HPLC
A
Kolber et al. (2001)
NE Pacific
Jul 00
11%
0.8
IRFRR, IREM, HPLC
B
Aerobic Anoxygenic Phototrophs
Publication
Chapter 3
Table 2. Continued
Kolber et al. (2000) E Pacific (OR, USA) Nov 99 ND 0.7–10 IRFRR B 1 Method codes: HPLC, high performance liquid chromatography; IREM, infrared epifluorescence microscopy; IRFRR, infrared fast repetition rate fluorometry; KF, kinetic fluorometry; QPCR, quantitative polymerase chain reaction; TIREM, time–dependent infrared epifluorescence microscopy. 2 Legend: (A) no distinction of AAP genes or BChl from that of anaerobic phototrophs, (B) potentially underestimates Prochlorococcus, (C) detection of gene sequences limited to only those primers used. 3 ND, not determined.
49
50 2003). Establishing relationships with reliable producers of organic carbon, such as DMSP-producing algae, circumvents the need to diminish cell size as a means to maximize uptake rates of dissolved organic matter. Attachment to motile photoautotrophs such as dinoflagellates also minimizes energy expended on regulating vertical position for optimal light interception. So where do marine AAP reside on the trophic pyramid? Assessments of in situ physiological activity are rare, but an IRFRR study by Kolber et al. (2000) established that AAP can account for 2–5% of photosynthetic electron flux in the upper ocean. Kolber et al. (2001) calculated that light-stimulated CO2-fixation contributed to 1% of the total carbon anabolism in marine AAP cultures grown in 2 to 20 mM organic medium. However, they then reasoned that in the thousand-fold lower concentration of dissolved organic matter in the oligotrophic ocean, CO2-fixation would contribute proportionally more to anabolism. Regrettably, the flawed assumption that anaplerotic carboxylation rates are independent of organic carbon assimilation rates led to the erroneous inference that AAP are capable of photoautotrophy and to a gross overestimate (Fenchel, 2001) of their contribution to the marine carbon budget. Whereas AAP are not primary producers, the two-fold higher efficiency in organic carbon utilization over heterotrophs facilitated by ‘accessory phototrophy’ nevertheless enables them to exert a disproportionately greater influence on marine carbon cycling than their numbers would suggest, signifying keystone consumer status. As Sieracki et al. (2006) elegantly put it, ‘AAP bacteria may act as a sunlight-accelerated shunt of energy between the large pool of oceanic dissolved organic matter and higher trophic levels.’ Three factors additional to standing population size critically influence the AAP trophic significance: cellular pigment content, cell size, and rate of population turnover. Pigment content obviously influences energetics and therefore rate of carbon turnover. Sieracki et al. (2006) calculated cell quotas of BChl for northwest Atlantic AAP in the range of 0.03 to 0.17 fg BChl cell–1, which did not correlate with local marine trophic status even after normalizing pigment content to biomass. This value differs markedly from 0.0098 fg BChl cell–1 for northeastern Pacific AAP (Kolber et al., 2001), underscoring the importance of unknown factors in determining contribution of AAP to euphotic electron flux. One likely factor, cell size, is ecologically significant because it di-
Vladimir Yurkov and Julius T. Csotonyi rectly influences standing crop. Oddly, size spectrum determinations by microscopy indicated that most northwest Atlantic putative AAP cells were 1.3 to 2.6 times larger than other bacteria (Sieracki et al., 2006), and that South Pacific AAP also had two-fold greater biovolumes than other bacteria (Lami et al., 2007), making their relative biomass correspondingly more prominent. Perhaps these AAP placed greater emphasis on nutrient storage than did the rest of the microbial community. Increased size should drive higher turnover rates because of more intense grazing pressure from heterotrophic protists (Sieracki et al., 2006), as observed for AAP in the Baltic Sea, where diel BChl fluctuation translated into mortality rates of 0.7–2.3 day–1 along a latitudinal transect from 54.5° N to 59° N (Koblížek et al., 2005). Follow-up results indicated AAP doubling rates of 1.44–2.13 day–1 from the productive North Atlantic and 0.72–1.03 day–1 in oligotrophic North and South Atlantic gyres, which is ten-fold greater than typically reported for bacteria in oceanic gyres (Koblížek et al., 2007). Rapidly cycling populations represent greater temporally integrated supplies of carbon to higher trophic levels. If research continues to confirm large cell size and high cellular turnover rates of AAP by more direct measures of biomass, then their trophic impact on the ocean is greater than abundance indicates. C. Vertical Distribution of Deep Ocean Aerobic Anoxygenic Phototrophs Undoubtedly the most bizarre aspect of AAP ecology has been the discovery of deep ocean (~2000 m) populations of Cmi. bathyomarinum, comprising 4 to 30% of the microbial community cultured on a rich organic medium (Yurkov and Beatty, 1998) (Fig. 3G and Color Plate 2). Fueled by the hypothesis that near-IR radiation emitted by superheated water from hydrothermal vents is sufficient to support phototrophic growth (possibly explaining the recovery of a green sulfur bacterium from the hydrothermal vent environment; Beatty et al., 2005), a recent study found Cmi. bathyomarinum to possess a completely functional photosynthetic apparatus (Rathgeber et al., unpublished) (Fig. 3E and Color Plate 2). Adding intrigue to the tale, culture-based techniques demonstrated that Cmi. bathyomarinum may be confined to the deep ocean (Rathgeber et al., unpublished). Vertical transects on- and off-axis from the Juan de Fuca Ridge in the eastern Pacific Ocean yielded twelve yellow AAP from within the plume at
Chapter 3
Aerobic Anoxygenic Phototrophs
depths between 500 and 2379 m. Based on 16S rDNA sequences, these strains clustered closely with Cmi. bathyomarinum (99.7–99.8% similarity), whereas a single yellow strain isolated from the ocean surface was 96.9% similar to Erb. litoralis (Rathgeber et al., unpublished) and probably represents a typical member of the euphotic zone AAP (Cottrell et al., 2006; Sieracki et al., 2006). Absence at the surface suggests that Cmi. bathyomarinum may be endemic to deep vent plumes and possibly garners benefit from hydrothermal effluent components. D. Aerobic Anoxygenic Phototrophs Thrive in Extreme Environments The decade preceding the marine ecological campaign stimulated by the startling results of Kolber et al. (2000) was rich in studies of AAP from inland extreme environments (Fig 3H, I and Color Plate 2), because the first non-marine species were isolated from Russian thermal springs in 1990 (Yurkov and Beatty, 1998). Hypersaline ecotopes and acidic coal refuse heaps also yielded numerous novel AAP (Yurkov and Csotonyi, 2003), reflecting the discovery that phototrophy in AAP maximally enhances fitness under suboptimal conditions (Section IV.C.). Perhaps this feature, under pressure to escape competition, has driven many AAP to evolve extremotolerance or extremophily. Halotolerance is especially frequently observed in AAP, and a plethora of species inhabit meromictic lakes (Yurkova et al., 2002; Karr et al., 2003), various salt flats (Yurkov and Csotonyi, 2003) and hypersaline springs (Csotonyi and Yurkov, unpublished) (Fig 3H and Color Plate 2). Yurkov and Csotonyi (2003) comprehensively reviewed the ubiquity of AAP in extreme environments. Aside from enumerating and characterizing novel AAP, though, microbial ecologists have barely scratched the surface of biogeochemical cycling by AAP in extreme environments. E. Soil and Freshwater As microbiologists converge on a rough understanding of marine AAP ecology, attention is turning to environments in which AAP are most likely to interact with humans: soil, freshwater lakes and rivers. Impetus for this shift appears divided between academics and conservation. First, not only were two genera (Craurococcus and Paracraurococcus) isolated from urban soils, but the only known beta-
51 proteobacterial representative (Roseateles) is lotic (Yurkov and Csotonyi, 2003). A fosmid library from the Delaware river recently yielded unique puf (RC gene) sequences in uncultured α-3 and betaproteobacterial AAP, which may be particularly abundant in rivers (Waidner and Kirchman, 2005). The authors conscientiously screened not only pufLM genes, but also acsF and bchE (see Section IV.C.), and the oxygen-dependent oxidase-encoding hemN, to aid in discrimination between aerobic and anaerobic phototrophs using a purely genetic approach. Second, increasing human-induced environmental damage means that enumeration of AAP in ecologically sensitive habitats will valuably contribute to catalogues of biodiversity. A prime Canadian example is the Banff thermal springs (Fig. 3I and Color Plate 2), in which phototrophic bacteria may form an important dietary component of the endangered Banff springs snail (P. johnsoni) (Bilyj, Pacas and Yurkov, unpublished). Ability of AAP such as Blastomonas natatoria, Stl. guttiformis, S. sibiricus, and coastal sand strains 15s. b. and 23s.b. to secrete copious polymers and form multi-layered biofilms (Yurkov and Beatty, 1998; Ivanova et al., 2002; Rickard et al., 2002) may even involve AAP in soil structural development. VII. Concluding Remarks and Perspectives Atmospheric oxygenation ~2 GYa drove convergence of incompatible cellular processes resulting in the AAP, bacteria capable of anoxygenic phototrophy even in the presence of ‘toxic’ O2. Because this abundant cosmopolitan group was discovered only 28 years ago, many questions remain unanswered, and exciting research directions remain to be explored. Regarding photosynthesis gene expression and function, satisfactory answers to why AAP photosynthesis functions only oxically and how these mechanisms vary among clades remain elusive. The molecular rheostats that fine-tune pigment levels in response to levels of light, oxygen, nutrients and other environmental factors are incompletely understood, as is the range of the roles played by photosynthetically disengaged carotenoids. Why AAP possess different suites of LH2 complexes in comparison with some anaerobic anoxygenic bacteria begs an answer, as does the question of whether any AAP have retained autotrophy. More daunting will be delineation of the proportion of currently classified heterotrophs that possess genes for photosynthesis
52 and whether any bacteria possess a nonfunctional or partial photosynthetic gene cluster, because thus far, no such case has been identified. On the contrary, completely functional photosynthetic complexes have turned up in unexpected habitats, such as the deep ocean, where light is not completely absent. Although our familiarity with the ecological role of AAP in marine environments waxes, researchers need to develop reliable standards for enumerating AAP, and they need to place greater emphasis on assessing the role of this functional group in elemental cycling in both ‘middle-of-the-road’ ecotopes and extreme environments. On a more globally relevant scale, a currently less amenable but more significant question considers how AAP may impact global climate through involvement in carbon and sulfur cycles, and conversely, how climate change will affect them via altered water temperature, pH and nutrient status. Finally, in fields of applied science and industry, the tendency of AAP toward extremophily and metallotolerance (Yurkov and Csotonyi, 2003) opens opportunities for bioremediative studies. Recent burgeoning interest in the AAP has uncovered more questions than answers, indicating that we have only begun to raise the floodgates of knowledge on this fascinating and significant group of bacteria, which doubtless hold a wealth of discoveries for the near and long term future. Acknowledgments This work was supported financially by a Natural Sciences and Engineering Research Council grant held by V. Yurkov. We thank E. Stackebrandt and J. Swiderski for generous help with the phylogenetic tree preparation. We thank the following scientists for providing published and unpublished results on AAP research: Drs. O. Béjà, R. Blankenship, I. Wagner-Döbler, G. Drews, V. Gorlenko, D. Green, D. Kirchman, M. Koblížek , M. Labrenz and J. Overmann. References Alarico S, Rainey FA, Empadinhas N, Schumann P, Nobre MF and Da Costa MS (2002) Rubritepida flocculans gen. nov., sp. nov., a new slightly thermophilic member of the α-1 subclass of the Proteobacteria. Syst Appl Microbiol 25: 198–206 Allgaier M, Uphoff H, Felske A and Wagner-Döbler I (2003) Aerobic anoxygenic photosynthesis in Roseobacter clade bacteria from diverse marine habitats. Appl Environ Microbiol 69: 5051–5059
Vladimir Yurkov and Julius T. Csotonyi Beatty JT (2002) On the natural selection and evolution of the aerobic phototrophic bacteria. Photosynth Res 73: 109–114 Beatty JT, Overmann J, Lince MT, Manske AK, Lang AS, Blankenship RE, van Dover CL, Martinson TA and Plumley FG (2005) An obligately photosynthetic bacterial anaerobe from a deep-sea hydrothermal vent. Proc Natl Acad Sci 102: 9306–9310 Béjà O, Suzuki MT, Heidelberg JF, Nelson WC, Preston CM, Hamada T, Elsen JA, Fraser CM and DeLong EF (2002) Unsuspected diversity among marine aerobic anoxygenic phototrophs. Nature 415: 630–633 Biebl H and Wagner-Döbler I (2006) Growth and bacteriochlorophyll a formation in taxonomically diverse aerobic anoxygenic phototrophic bacteria in chemostat culture: Influence of light regimen and starvation. Proc Biochem 41: 2153–2159 Biebl H, Allgaier M, Lünsdorf H, Pukall R, Tindall BJ and WagnerDöbler I (2005a) Roseovarius mucosus sp. nov., a member of the Roseobacter clade with trace amounts of bacteriochlorophyll a. Int J Syst Evol Microbiol 55: 2377–2383 Biebl H, Allgaier M, Tindall, BJ Koblížek M, Lünsdorf H, Pukall R and Wagner-Döbler I (2005b) Dinoroseobacter shibae gen nov., sp. nov., a new aerobic phototrophic bacterium isolated from dinoflagellates. Int J Syst Evol Microbiol 55: 1089–1096 Biebl H, Tindall BJ, Pukall R, Lünsdorf H, Allgaier M and WagnerDöbler I (2006) Hoeflea phototrophica sp. nov., a novel marine aerobic alphaproteobacterium that forms bacteriochlorophyll a. Int J Syst Evol Microbiol 56: 821–826 Biebl H, Pukall R, Lünsdorf H, Schulz S, Allgaier M, Tindall BJ and Wagner-Döbler I (2007) Description of Labrenzia alexandrii gen. nov., sp. nov., a novel alphaproteobacterium containing bacteriochlorophyll a, and a proposal for reclassification of Stappia aggregata as Labrenzia aggregata comb. nov., of Stappia marina as Labrenzia marina comb. nov. and of Stappia alba as Labrenzia alba comb. nov., and emended descriptions of the genera Pannonibacter, Stappia and Roseibium, and of the species Roseibium denhamense and Roseibium hamelinense. Int J Syst Evol Microbiol 57: 1095–1107 Boettcher KJ, Geaghan KK, Maloy AP and Barber BJ (2005) Roseovarius crassostreae sp. nov., a member of the Roseobacter clade and the apparent cause of juvenile oyster disease (JOD) in cultured Eastern oysters. Int J Syst Evol Microbiol 55: 1531–1537 Boldareva EN, Bryantseva IA, Tsapin A, Nelson K, Sorokin DY, Tourova TP, Boichenko VA, Stadnichuk IN and Gorlenko VM (2007) The new alkaliphilic bacteriochlorophyll a-containing bacterium Roseinatronobacter monicus sp. nov. from the hypersaline soda Mono Lake (California, United States). Microbiology 76: 82–92 Brown MV and Fuhrman JA (2005) Marine bacterial microdiversity as revealed by internal transcribed spacer analysis. Aquat Microbial Ecol 41: 15–23 Cho JC, Stapels MD, Morris RM, Vergin MS, Schwalbach MS, Givan SA, Barofsky DF and Giovannoni SJ (2007) Polyphyletic photosynthetic reaction centre genes in oligotrophic marine Gammaproteobacteria. Environ Microbiol 9: 1456–1463 Cogdell RJ, Howard TD, Isaacs NW, McLuskey K and Gardiner AT (2002) Structural factors which control the position of the Qy absorption band of bacteriochlorophyll a in purple bacterial antenna complexes. Photosynth Res 74: 135–141 Cottrell MT, Mannino A and Kirchman DL (2006) Aerobic anoxygenic phototrophic bacteria in the Mid-Atlantic Bight and the North Pacific Gyre. Appl Environ Microbiol 72: 557–564
Chapter 3
Aerobic Anoxygenic Phototrophs
Denner, EBM, Vybiral D, Koblížek M, Kämpfer P, Busse H and Velimirov B (2002) Erythrobacter citreus sp. nov., a yellowpigmented bacterium that lacks bacteriochlorophyll a, isolated from the western Mediterranean Sea. Int J Syst Evol Microbiol 62: 1655–1661 Du H, Jiao N, Hu Y and Zeng Y (2006) Real-time PCR for quantification of aerobic anoxygenic phototrophic bacteria based on pufM gene in marine environment. J Exp Mar Biol Ecol 329: 113–121 Evans K, Fordham-Skelton AP, Mistry H, Reynolds CD, Lawless AM and Papiz MZ (2005) A bacteriophytochrome regulates the synthesis of LH4 complexes in Rhodopseudomonas palustris. Photosynth Res 85: 169–180 Fenchel T (2001) Marine bugs and carbon flow. Science 292: 2444–2445 Fleischman D and Kramer D (1998) Photosynthetic rhizobia. Biochim Biophys Acta 1364: 17–36 Foesel BU, Gößner AS, Drake HL and Schramm A (2007) Geminicoccus roseus gen. nov., sp. nov., an aerobic phototrophic Alphaproteobacterium isolated from a marine aquaculture biofilter. Syst Appl Microbiol 30: 581–586 Fowler GJS, Visschers RW, Grief GG, van Grondelle R and Hunter CN (1992) Genetically modified photosynthetic antenna complexes with blueshifted absorbance bands. Nature 355: 848–850 Fowler GJS, Sockalingum GD, Robert B and Hunter CN (1994) Blue shifts in bacteriochlorophyll absorbance correlate with changed hydrogen bonding patterns in light harvesting LH2 mutants of Rhodobacter sphaeroides with alterations at Tyr44 and Tyr45. Biochem. J. 299: 695–700 Francia F, Wang J, Venturoli G, Melandri BA, Barz WP and Oesterhelt D (1999) The reaction center-LH1 antenna complex of Rhodobacter sphaeroides contains one PufX molecule which is involved in dimerization of this complex. Biochemistry 38: 6834–6845 Fraser NJ, Hashimoto H and Cogdell RJ (2001) Carotenoids and bacterial photosynthesis: The story so far… Photosynth Res 70: 249–256 Fuchs BM, Spring S, Teeling H, Quast C, Wulf J, Schattenhofer M, Yan S, Ferriera S, Johnson J, Glöckner FO and Amann R (2007) Characterization of a marine gammaproteobacterium capable of aerobic anoxygenic photosynthesis. Proc Natl Acad Sci 104: 2891–2896 Gich F and Overmann J (2006) Sandarakinorhabdus limnophila gen. nov., sp. nov., a novel bacteriochlorophyll a-containing, obligately aerobic bacterium isolated from freshwater lakes. Int J Syst Evol Microbiol 56: 847–854 Gich F, Schubert K, Bruns A, Hoffelner H and Overmann J (2005) Specific detection, isolation, and characterization of selected, previously uncultured members of the freshwater bacterioplankton community. Appl Environ Microbiol 71: 5908–5919 Giraud E and Fleischman D (2004) Nitrogen-fixing symbiosis between photosynthetic bacteria and legumes. Photosynth Res 82: 115–130 González JM, Covert JS, Whitman WB, Henriksen JR, Mayer F, Scharf B, Schmitt R, Buchan A, Fuhrman JA, Kiene RP and Moran MA (2003) Silicibacter pomeroyi sp. nov. and Roseovarius nubinhibens sp. nov., dimethylsulfoniopropionatedemethylating bacteria from marine environments. Int J Syst Evol Microbiol 53: 1261–1269 Hartigan N, Tharia HA, Sweeney F, Lawless AM and Papiz MZ
53 (2002) The 7.5-Å electron density and spectroscopic properties of a novel low-light B800 LH2 from Rhodopseudomonas palustris. Biophys J 82: 963–977 Hiraishi A and Shimada K (2001) Aerobic anoxygenic photosynthetic bacteria with zinc-bacteriochlorophyll. J Gen Appl Microbiol 47: 161–180 Hucke O, Schiltz E, Drews G and Labahn A (2003) Sequence analysis reveals new membrane anchor of reaction centrebound cytochromes possibly related to PufX. FEBS Lett 535: 166–170 Igarashi N, Harada J, Nagashima S, Matsuura K, Shimada K and Nagashima KVP (2001) Horizontal transfer of the photosynthesis gene cluster and operon rearrangement in purple bacteria. J Mol Evol 52: 333–341 Ivanova EP, Pham DK, Wright JP and Nicolau DV (2002) Detection of coccoid forms of Sulfitobacter mediterraneus using atomic force microscopy. FEMS Microbiol Lett 214: 177–181 Ivanova EP, Bowman JP, Lysenko AM, Zhukova NV, Gorshkova NM, Kuznetsova TA, Kalinovskaya NI, Shevchenko LS and Mikhailov VV (2005) Erythrobacter vulgaris sp. nov., a novel organism isolated from the marine invertebrates. Syst Appl Microbiol 28: 123–130 Jiao N, Zhang Y and Chen Y (2006) Time series observation based InfraRed Epifluorescence Microscopic (TIREM) approach for accurate enumeration of bacteriochlorophyll-containing microbes in marine environments. J Microbiol Meth 65: 442–452 Karr EA, Sattley WM, Jung DO, Madigan MT and Achenbach LA (2003) Remarkable diversity of phototrophic purple bacteria in a permanently frozen Antarctic lake. Appl Environ Microbiol 69: 4910–4914 Kim B-C, Park JR, Bae J-W, Rhee S-K, Kim K-H, Oh J-W and Park Y-H (2006) Stappia marina sp. nov., a marine bacterium isolated from the Yellow Sea. Int J Syst Evol Microbiol 56: 75–79 Kim MK, Schubert K, Im W-T, Kim K-H, Lee S-T and Overmann J (2007) Sphingomonas kaistensis sp. nov., a novel alphaproteobacterium containing pufLM genes. Int J Syst Microbiol Ecol 57:1527–1534 Klug G (1993) Regulation of expression of photosynthesis genes in anoxygenic photosynthetic bacteria. Arch Microbiol 159: 397–404 Koblížek M, Béjà O, Bidigare RR, Christensen S, BenetizNelson B, Vetriani C, Kolber MK, Falkowski PG and Kolber ZS (2003) Isolation and characterization of Erythrobacter sp. strains from the upper ocean. Arch Microbiol 180: 327–338 Koblížek M, Stoń-Egiert J, Sagan S and Kolber ZS (2005) Diel changes in bacteriochlorophyll a concentration suggest rapid bacterioplankton cycling in the Baltic Sea. FEMS Microbiol Ecol 51: 353–361 Koblížek M, Falkowski PG and Kolber ZS (2006) Diversity and distribution of photosynthetic bacteria in the Black Sea. DeepSea Res II 53: 1934–1944 Koblížek M, Mašín M, Ras J, Poulton AJ and Prášil O (2007) Rapid growth rates of aerobic anoxygenic phototrophs in the ocean. Environ Microbiol 9: 2401–2406 Kohler S, Schwarze C, Herter S, Gad’on N, Drews G and Labahn A (2005) Exploring the structure of the reaction center from the aerobic photosynthetic bacterium Roseobacter denitrificans with fourier transform infrared spectroscopy. In: van der Est
54 A and Bruce D (eds) Photosynthesis: Fundamental Aspects to Global Perspectives, pp 202–204. Alliance Communications Group, Lawrence Kolber ZS, Van Dover CL, Niederman RA and Falkowski PG (2000) Bacterial photosynthesis in surface waters of the open ocean. Nature 407: 177–179 Kolber ZS, Plumley G, Lang AS, Beatty JT, Blankenship RE, Van Dover CL, Vetriani C, Koblížek M, Rathgeber C and Falkowski PG (2001) Contribution of aerobic photoheterotrophic bacteria to the carbon cycle in the ocean. Science 292: 2492–2495 Komagata K (1989) Taxonomy of facultative methylotrophs. In: Harashima K, Shiba T and Murata N (eds) Aerobic Photosynthetic Bacteria, pp 25–38. Japan Scientific Societies Press, Tokyo and Springer-Verlag, Berlin Krinsky (1971) IX. Function. In: Isler O (ed) Carotenoids, pp 670–716. Birkhäuser Verlag, Basel Labrenz M, Lawson PA, Tindall BJ, Collins MD and Hirsch P (2005) Roseisalinus antarcticus gen. nov., sp. nov., a novel aerobic bacteriochlorophyll a-producing α-proteobacterium isolated from hypersaline Ekho Lake, Antarctica. Int J Syst Evol Microbiol 55: 41–47 Lami R, Cottrell MT, Ras J, Ulloa O, Obernosterer I, Claustre H, Kirchman DL and Lebaron P (2007) High abundances of aerobic anoxygenic phototrophic bacteria in the South Pacific Ocean. Appl Environ Microbiol 73: 4198–4205 Larimer FW, Chain P, Hauser L, Lamerdin J, Malfatti S, Do L, Land, ML, Pelletier DA, Beatty JT, Lang AS, Tabita FR, Gibson JL, Hanson TE, Bobst C, Torres JLTy, Peres C, Harrison FH, Gibson J and Harwood CS (2004) Complete genome sequence of the metabolically versatile photosynthetic bacterium Rhodopseudomonas palustris. Nature Biotech 22: 55–61 Liebetanz R, Hornberger U and Drews G (1991) Organization of the genes coding for the reaction center L and M subunits and B870 antenna polypeptides alpha and beta from the aerobic photosynthetic bacterium Erythrobacter sp. OCh114. Mol Microbiol 5: 1459–1468 Macián MC, Arahal DR, Garay E, Ludwig W, Schleifer KH and Pujalte MJ (2005) Thalassobacter stenotrophicus gen. nov., sp. nov., a novel marine α-proteobacterium isolated from Mediterranean sea water. Int J Syst Evol Microbiol 55: 105–110 Marchand EA and Silverstein J (2003) The role of enhanced heterotrophic bacterial growth on iron oxidation by Acidithiobacillus ferrooxidans. Geomicrobiol J 20: 231–244 Marrs B (1981) Mobilization of the genes for photosynthesis from Rhodopseudomonas capsulata by a promiscuous plasmid. J Bacteriol 146: 1003–1012 Masuda S, Matsumoto Y, Nagashima KVP, Shimada K, Inoue K, Bauer CE and Matsuura K (1999) Structural and functional analyses of photosynthetic regulatory genes regA and regB from Rhodovulum sulfidophilum, Roseobacter denitrificans and Rhodobacter capsulatus. J Bacteriol 181: 4205–4215 Murray RGE, Brenner DJ, Colwell RR, De Vos P, Goodfellow M, Grimont PAD, Pfennig N, Stackebrandt E and Zivarzin GA (1990) Report of the ad hoc committee on approaches to taxonomy within Proteobacteria. Int J Syst Bacteriol 40: 213–215 Nagashima KVP, Matsuura K, Wakao N, Hiraishi A and Shimada K (1997) Nucelotide sequences of genes coding for photosynthetic reaction centers and light-harvesting proteins of Acidiphilium rubrum and related aerobic acidophilic bacteria. Plant Cell Physiol 38: 1249–1258
Vladimir Yurkov and Julius T. Csotonyi Ouchane S, Steunou A, Picaud M and Astier C (2004) Aerobic and anaerobic Mg-protoporphyrin monomethyl ester cyclases in purple bacteria. J Biol Chem 279: 6385–6394 Oz A, Sabehi G, Koblížek M, Massana R and Béjà O (2005) Roseobacter-like bacteria in Red and Mediterranean Sea aerobic anoxygenic photosynthetic populations. Appl Environ Microbiol 71: 344–353 Peplies J, Glöckner FO, Amann R and Ludwig W (2004) Comparative sequence analysis and oligonucleotide probe design based on 23S rRNA genes of Alphaproteobacteria from North Sea bacterioplankton. Syst Appl Microbiol 27: 573–580 Pfennig N (1967) Photosynthetic bacteria. Ann Rev Microbiol 21: 285–384 Pfennig N (1978) Rhodocyclus purpureus gen. nov. and sp. nov., a ring-shaped vitamin B12-requiring member of the family Rhodospirillaceae. Int J Syst Bacteriol 28: 283–288 Pradella S, Allgaier M, Hoch C, Päuker O, Stackebrandt E and Wagner-Döbler I (2004) Genome organization and localization of the pufLM genes of the photosynthesis reaction center in phylogenetically diverse marine Alphaproteobacteria. Appl Environ Microbiol 70: 3360–3369 Rainey FA, Silva J, Bobre MF, Silva MT and da Costa MS (2003) Porphyrobacter cryptus sp. nov., a novel slightly thermophilic, aerobic, bacteriochlorophyll a-containing species. Int J Syst Evol Microbiol 53: 35–41 Rappé MS, Connon SA, Vergin KL and Giovannoni SJ (2002) Cultivation of the ubiquitous SAR11 marine bacterioplankton clade. Nature 418: 630–633 Rathgeber C, Beatty JT and Yurkov V (2004) Aerobic phototrophic bacteria: new evidence for the diversity, ecological importance and applied potential of this previously overlooked group. Photosynth Res 81: 113–128 Rathgeber C, Yurkova N, Stackebrandt E, Schumann P, Beatty JT and Yurkov V (2005) Roseicyclus mahoneyensis gen nov., sp. nov., an aerobic phototrophic bacterium isolated from a meromictic lake. Int J Syst Evol Microbiol 55: 1597–1603 Rathgeber C, Yurkova N, Stackebrandt E, Schumann P, Humphrey E, Beatty JT and Yurkov V (2007) Porphyrobacter meromictius sp. nov., an appendaged bacterium, that produces bacteriochlorophyll a. Curr Microbiol 55: 356–361 Rickard AH, Leach SA, Hall LS, Buswell CM, High NJ and Handley PS (2002) Phylogenetic relationships and coaggregation ability of freshwater biofilm bacteria. Appl Environ Microbiol 68: 3644–3650 Rohwerder T and Sand W (2003) The sulfane sulfur of persulfides is the actual substrate of the sulfur-oxidizing enzymes from Acidithiobacillus and Acidiphilium spp. Microbiology 149: 1699–1709 Rye R and Holland HD (1998) Paleosols and the evolution of atmospheric oxygen: A critical review. Am J Sci 298: 621–672 Salka I, Jost G, Jürgens K and Labrenz M (2006) Diversity and distribution of aerobic BChl a-producing bacteria in a depth profile of the central Baltic Sea. In: BIOspectrum Abstracts, VAAM Tagung Jena (PE42), p 142. Spektrum Akademischer Verlag GmBH, Heidelberg Schwalbach MS and Fuhrman JA (2005) Wide-ranging abundances of aerobic anoxygenic phototrophic bacteria in the world ocean revealed by epifluorescence microscopy and quantitative PCR. Limnol Oceanogr 50: 620–628 Schwarze C, Carluccio AV, Venturoli G and Labahn A (2000) Photo-induced cyclic electron transfer involving cytochrome bc1
Chapter 3
Aerobic Anoxygenic Phototrophs
complex and reaction center in the obligate aerobic phototroph Roseobacter denitrificans. Eur J Biochem 267: 422–433 Shiba T, Simidu U and Taga N (1979) Distribution of aerobic bacteria which contain bacteriochlorophyll a. Appl Environ Microbiol 38: 43–45 Siebert CA, Qian P, Fotiadis D, Engel A, Hunter CN and Bullough PA (2004) Molecular architecture of photosynthetic membranes in Rhodobacter sphaeroides: the role of PufX. EMBO J 23: 690–700 Sieracki ME, Gilg IC, Their EC, Poulton NJ and Goericke R (2006) Distribution of planktonic aerobic photoheterotrophic bacteria in the northwest Atlantic. Limnol Oceanogr 51: 38–46 Stackebrandt E, Frederiksen W, Garrity GM, Grimont PAD, Kämpfer P, Maiden MCJ, Nesme X, Roselló-Mora R, Swings J, Trüper HG, Vauterin L, Ward AC and Whitman WB (2002) Report of the ad hoc committee for the re-evaluation of the species definition in bacteriology. Int J Syst Evol Microbiol 52: 1043–1047 Steiner R and Scheer H (1985) Characterization of a B800/1020 antenna from the photosynthetic bacteria Ectothiorhodospia halochloris and Ectothiorhodospia abdelmalekii. Biochim Biophys Acta 3: 278–284 Suyama T, Shigematsu T, Suzuki T, Tokiwa Y, Kanagawa T, Nagashima KVP and Hanada S (2002) Photosynthetic apparatus in Roseateles depolymerans 61A is transcriptionally induced by carbon limitation. Appl Environ Microbiol 68: 1665–1673 Suzuki T, Mori Y and Nishimura Y (2006) Roseibacterium elongatum gen. nov., sp. nov., an aerobic, bacteriochlorophyllcontaining bacterium isolated from the west coast of Australia. Int J Syst Evol Microbiol 56: 417–421 Swingley WD, Gholba S, Mastrian SD, Matthies HJ, Hao J, Ramos H, Acharya CR, Conrad AL, Taylor HL, Dejesa LC, Shah MK, O’Huallachain ME, Lince MT, Blankenship RE, Beatty JT and Touchman JW (2007) The complete genome sequence of Roseobacter denitrificans reveals a mixotrophic as opposed to photosynthetic metabolism. J Bacteriol 189: 683–690 Uchino Y, Hirata A, Yokota A and Sugiyama J (1998) Reclassification of marine Agrobacterium species: Proposals of Stappia stellulata gen. nov., comb. nov., Stappia aggregata, sp. nov., nom. rev., Ruegeria atlantica gen. nov., comb. nov., Ruegeria gelatinovora comb. nov., Ruegeria algicola comb. nov., and Ahrensia kieliense gen. nov., sp. nov., nom. rev. J Gen Appl Microbiol 44: 201–210 van Berkum P, Leibold JM and Eardly BD (2006) Proposal for combining Bradyrhizobium spp. (Aeschynomene indica) with Blastobacter denitrificans and to transfer Blastobacter denitrificans (Hirsch and Muller, 1985) to the genus Bradyrhizobium as Bradyrhizobium denitrificans (comb. nov.). Syst Appl Microbiol 29:207–215 Wagner-Döbler I and Biebl H (2006) Environmental biology of the marine Roseobacter lineage. Ann Rev Microbiol 60: 255–280 Waidner LA and Kirchman DL (2005) Aerobic anoxygenic photosynthesis genes and operons in uncultured bacteria in the Delaware River. Environ Microbiol 7: 1896–1908 Waidner LA and Kirchman DL (2007) Aerobic anoxygenic phototrophic bacteria attached to particles in turbid waters of the Delaware and Chesapeake Estuaries. Appl Environ Microbiol 73: 3936–3944
55 Wayne LG, Brenner DJ, Colwell RR, Grimont PAD, Kandler O, Krichevsky MI, Moore LH, Moore WEC, Murray RGE, Stackebrandt E, Starr MP and Trüper HG (1987) Report of the ad hoc committee on reconciliation of approaches to bacterial systematics. Int J Syst Bacteriol 37: 463–464 Woese CR (1987) Bacterial evolution. Microbiol Rev 51: 221–271 Yoon J-H, Kim H, Kim I-G, Kang KH and Park Y-H (2003) Erythrobacter flavus sp. nov., a slight halophile from the East Sea in Korea. Int J Syst Evol Microbiol 53: 1169–1174 Yoon J-H, Kang KH, Oh T-K and Park Y-H (2004a) Erythrobacter aquimaris sp. nov., isolated from sea water of a tidal flat of the Yellow Sea in Korea. Int J Syst Evol Microbiol 54: 1981–1985 Yoon J-H, Lee M-H and Oh T-K (2004b) Porphyrobacter donghaensis sp. nov., isolated from sea water of the East Sea in Korea. Int J Syst Evol Microbiol 54: 2231–2235 Yoon J-H, Kang KH, Yeo S-H and Oh T-K (2005a) Erythrobacter luteolus sp. nov., isolated from a tidal flat of the Yellow Sea in Korea. Int J Syst Evol Microbiol 55: 1167–1170 Yoon J-H, Oh T-K and Park Y-H (2005b) Erythrobacter seohaensis sp. nov. and Erythrobacter gaetbuli sp. nov., isolated from a tidal flat of the Yellow Sea in Korea. Int J Syst Evol Microbiol 55: 71–75 Yoon J-H, Kang S-J, Lee M-H, Oh HW and Oh T-K (2006) Porphyrobacter dokdonensis sp. nov., isolated from sea water. Int J Syst Evol Microbiol 56: 1079–1083 Yurkov VV (2006) Aerobic phototrophic proteobacteria. In: Dworkin M, Falkow S, Rosenberg E, Schleifer K-H and Stackebrandt E (eds) Prokaryotes (3rd Edition), Vol 5, pp 562–584. Springer-Verlag, New York Yurkov VV and Beatty JT (1998) Aerobic anoxygenic phototrophic bacteria. Microbiol Mol Biol Rev 62: 695–724 Yurkov VV and Csotonyi JT (2003) Aerobic anoxygenic phototrophs and heavy metalloid reducers from extreme environments. In: Pandalai SG (ed) Recent Research Developments in Bacteriology, Vol 1, pp 247–300. Transworld Research Network, Trivandrum Yurkov V and van Gemerden H (1993) Impact of light/dark regimen on growth rate, biomass formation and bacteriochlorophyll synthesis in Erythromicrobium hydrolyticum. Arch Microbiol 159: 84–89 Yurkov VV, Krieger S, Stackebrandt E and Beatty JT (1999) Citromicrobium bathyomarinum, a novel aerobic bacterium isolated from deep-sea hydrothermal vent plume waters that contains photosynthetic pigment-protein complexes. J Bacteriol 181: 4517–4525 Yurkova N, Rathgeber C, Swiderski J, Stackebrandt E, Beatty JT, Hall KJ and Yurkov V (2002) Diversity, distribution and physiology of the aerobic phototrophic bacteria in the mixolimnion of a meromictic lake. FEMS Microbiol Ecol 40: 191–204 Yutin N and Béjà O (2005) Putative novel photosynthetic reaction centre organizations in marine aerobic anoxygenic photosynthetic bacteria: insights from metagenomics and environmental genomics. Environ Microbiol 7: 2027–2033 Yutin N, Suzuki MT and Béjà O (2005) Novel primers reveal wider diversity among marine aerobic anoxygenic phototrophs. Appl Environ Microbiol 71: 8958–8962
Chapter 4 Biosynthesis of Bacteriochlorophylls in Purple Bacteria Robert D. Willows* and Alison M. Kriegel Department of Chemistry and Biomolecular Sciences, Macquarie University, North Ryde, 2109, Australia
Summary ................................................................................................................................................................. 57 I. Introduction....................................................................................................................................................... 58 II. δ-Aminolevulinate to Protoporphyrin IX ............................................................................................................ 59 A. δ-Aminolevulinic acid Synthase.......................................................................................................... 60 B. δ-Aminolevulinic acid Dehydratase ................................................................................................... 62 C Hydroxymethylbilane Synthase ........................................................................................................... 63 D. Uroporphyrinogen III Synthase .......................................................................................................... 63 E. Uroporphyrinogen III Decarboxylase .................................................................................................. 64 F. Coproporphyrinogen Oxidase............................................................................................................. 64 G. Protoporphyrinogen Oxidase ............................................................................................................. 65 III. Protoporphyrin IX to Bacteriochlorophyll a and b ............................................................................................ 65 A. Magnesium Chelatase........................................................................................................................ 66 B. S-Adenosyl-L-methionine:Magnesium Protoporphyrin IX-O-methyltransferase................................. 71 C. Magnesium-protoporphyrin IX Monomethylester Oxidative Cyclase ................................................. 71 D. Modification of Vinyl Groups on Rings I and II ................................................................................... 72 E. Protochlorophyllide Oxidoreductase and Chlorin Reductase ............................................................. 73 F. Esterification to the Propionate on Ring IV ........................................................................................ 74 IV. Concluding Remarks ........................................................................................................................................ 75 Acknowledgments ................................................................................................................................................... 75 References .............................................................................................................................................................. 75
Summary The purple bacteria make bacteriochlorophylls for the photosynthetic mode of growth. These pigments are made from the simple precursors glycine and succinyl CoA and the initial steps in the pathway of bacteriochlorophyll biosynthesis are shared with the vitamin B12 and heme biosynthetic pathways. This chapter concentrates on the biochemical properties of the enzymes involved in each step of the pathway and the discovery and assignment of the genes encoding these enzymes. The characterization of purple bacterial enzymes involved in these steps has been crucial in understanding similar enzymes from other sources. The characterization of the early steps in the pathway within purple bacteria, such as δ-aminolevulinate synthase, contributed significantly to the understanding of the mammalian enzymes in the 1950s and 1960s. More recently the study of the purple bacterial enzymes toward the end of the pathway has been instrumental in identifying and characterizing the orthologous enzymes from cyanobacteria and plants. In this review we present the details of the properties of these enzymes from the purple bacteria, such as purification methods and kinetic analyses from the early literature, through to more recent studies using recombinant purple bacterial enzymes. *Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 57–79. © 2009 Springer Science + Business Media B.V.
58 I. Introduction The purple photosynthetic bacteria are able to derive their cellular energy from light, organic compounds, or inorganic compounds depending on the chemical and physical environment. This remarkable versatility usually means that one mode of metabolism is utilized at a time so as to prevent unnecessary biosynthesis of alternative energy systems. Thus photosynthetic metabolism and hence pigment biosynthesis only occurs under a limited set of conditions. In regard to their ability to produce bacteriochlorophyll there appear to be two main groups within the purple bacteria. The first group includes Rhodobacter (Rba.) sphaeroides, Rba. capsulatus, and Rhodospirillum rubrum, which photosynthesize and thus make bacteriochlorophyll anaerobically in the light; they can also synthesize significant amounts of pigment in the dark, but only under low aeration conditions. This indicates an oxygen and light dependent control over the genes encoding the enzymes involved in the synthesis of these pigments. The second group, which includes Rhodovulum sulfidophilum, Roseobacter sp. and Rubrivivax (Rvi.) gelatinosus, is able to synthesize pigments and to photosynthesize under both anaerobic and aerobic conditions in the light. As we shall see in this chapter the differences between these two groups are reflected in a number of the enzymes in the bacteriochlorophyll biosynthetic pathway. This is the first recent review devoted to the biosynthesis of bacteriochlorophylls in purple bacteria, although several recent reviews have been published on broader or more narrow aspects (Armstrong, 1998; Willows, 2003; Bollivar, 2006). As such we have attempted to include all references to the primary literature which span several decades. For some of the enzymatic steps in purple bacteria there has been limited study since the 1950s and 1960s, while a few enzymes have received considerable recent attention. In this chapter we will limit our discussion to the biosynthesis of bacteriochlorophylls which are found in the purple bacteria and the properties of the purple bacterial enzymes. We will not discuss the regulation of bacteriochlorophyll biosynthesis Abbreviations: ALA – δ-aminolevulinate; ALAD – δ-aminolevulinic acid dehydratase; ALAS – δ-aminolevulinic acid synthase; CPO – coproporphyrinogen III oxidase; DPOR – dark allowed protochlorophyllide reductase; E. – Escherichia; HMB – hydroxymethylbilane; HMBS – hydroxymethylbilane synthase; MPE – magnesium protoporphyrin IX monomethyl ester; PPO – protoporphyrinogen oxidase; Rba. – Rhodobacter; Rvi. – Rubrivivax; SAM – S-adenosylmethionine
Robert D. Willows and Alison M. Kriegel in detail as this is covered elsewhere in this volume (Chapter 39, Zeilstra-Ryalls). Hemes, chlorophylls and bacteriochlorophylls are all tetrapyrroles and have the same basic tetrapyrrole skeleton. Apart from the differences in the substituent groups attached to the macrocycle the other key difference between these molecules is the degree of reduction of the macrocycle. Fig. 1 highlights the key differences in the position and level of reduction of the macrocycle of these molecules together with that of a porphyrinogen. Porphyrinogens, porphyrins and chlorin macrocycles are all found as intermediates in the synthesis of bacteriochlorophyll a, which has a bacteriochlorin-type macrocycle. The IUPAC numbering scheme will be used throughout this chapter as detailed in Fig. 2 but the trivial names of the bacteriochlorophylls and their biosynthetic intermediates will generally be used for both clarity and consistency with the primary literature rather than using the more complex systematic names. The structures of bacteriochlorophylls a and b, which are the primary photosynthetic pigments found in the purple bacteria, are shown in Fig. 2. The alcohol esterified to the 17-propanoate is usually phytol although Rhodospirillum rubrum has geranylgeraniol. The other major bacteriochlorophyll found in the purple bacteria is bacteriochlorophyll b which differs from bacteriochlorophyll a in that it has an 8-ethylidene instead of an 8-ethyl group. Mutagenesis studies have been a key feature in identifying the structural genes for the enzymes involved in the biosynthesis of bacteriochlorophyll. The later genes which are specific for bacteriochlorophyll biosynthesis are located in what is termed the photosynthetic gene cluster in all purple bacteria which have been studied to date. The genes involved in bacteriochlorophyll biosynthesis are given the prefix ‘bch.’ The earlier genes in the pathway have the prefix ‘hem’ as they are also involved in heme biosynthesis. As such the enzymes responsible for the synthesis of bacteriochlorophyll are shared with the heme biosynthetic pathway up to the intermediate protoporphyrin IX when the two pathways diverge with either insertion of iron by ferrochelatase to make heme or the insertion of magnesium by magnesium chelatase to make the first committed precursor of bacteriochlorophyll biosynthesis, magnesium protoporphyrin IX. The following sections are divided into the reactions common to both heme and bacteriochlorophyll biosynthesis followed by the reactions starting from protoporphyrin IX, which are unique to bacteriochlorophyll biosynthesis.
Chapter 4
Bacteriochlorophyll Biosynthesis
59
Fig. 1. Tetrapyrrole macrocycles. Shading emphasizes conjugated double bond systems within the tetrapyrroles to highlight their differences.
Fig. 2. IUPAC numbering and structure of bacteriochlorophylls a and b.
Fig. 3. The two pathways for ALA biosynthesis. Gltx, glutamyl tRNA synthetase. Gtr, glutamyl tRNA reductase. HemL, glutamate-1semialdehyde aminotransferase. HemA, δ-aminolevulinate synthase
II. δ-Aminolevulinate to Protoporphyrin IX All tetrapyrroles are synthesized from the universal 5-carbon precursor δ-aminolevulinate (ALA). ALA can be synthesized via two different routes, shown in Fig. 3, which are known as the Shemin pathway and the C-5 pathway. The C-5 pathway is utilized
by all photosynthetic eukaryotes, archeae and most bacteria with the exception of the alpha-proteobacteria. Vertebrates, non-photosynthetic eukaryotes and the alpha-proteobacterial group of bacteria, which include the purple bacteria, utilize the Shemin pathway for ALA synthesis. The biosynthetic pathway from ALA to protopor-
Robert D. Willows and Alison M. Kriegel
60
Fig. 4. Biosynthesis of protoporphyrin IX from ALA. See Table 1 for details.
phyrin IX is shown in Fig. 4. Briefly, the enzyme ALA dehydratase catalyzes the condensation of 2 ALA molecules to produce the pyrrole porphobilinogen; 4 porphobilinogen molecules are then condensed to form the linear tetrapyrrole hydroxymethylbilane (HMB) by the enzyme HMB synthase; HMB is then cyclized to uroporphyrinogen III by uroporphyrinogen III synthase; uroporphyrinogen III is decarboxylated by uroporphyrinogen III decarboxylase to coproporphyrinogen III; coproporphyrinogen III is oxidatively decarboxylated to protoporphyrinogen IX by coproporphyrinogen III oxidase; and protoporphyrinogen IX is then oxidized to protoporphyrin IX by protoporphyrinogen IX oxidase. The individual reactions and the properties of the genes and enzymes are covered in detail in the following sections; mutations in genes encoding these activities as well as enzyme names and EC numbers are shown in Table 1. A. δ-Aminolevulinic acid Synthase The first committed step in the tetrapyrrole pathway is the formation of ALA. In animals, fungi, faculta-
tive aerobic bacteria and photosynthetic bacteria this reaction is catalysed by the enzyme δ-aminolevulinic acid synthase (ALAS). The enzyme was first isolated from Rba. sphaeroides (Kikuchi et al., 1958) and has since been isolated from a number of sources including other bacteria, fungi, yeast and animals (Ferreira and Gong, 1995; Oh-hama, 1997). The reaction catalysed by ALAS involves the condensation of glycine and succinyl-CoA. ALAS is highly specific for the substrate glycine. No other amino acids can be substituted in the reaction (Ferreira and Gong, 1995) while specificity for the second substrate, succinyl-CoA, is less stringent. Other acyl-CoA derivatives including malonyl-, β-hydroxybutyryl- and acetoacetyl-CoA allow reasonable enzyme activity, while acetyl- and n-valeryl-CoA are inactive substrates (Shoolingin-Jordan et al., 1997). Kinetic analysis shows the reaction takes place with an ordered bi-bi mechanism with glycine binding to the enzyme first, followed by succinyl-CoA with CoA being released first before ALA (Fanica-Gaignier and Clement-Metral, 1973). The reaction mechanism, shown in Fig. 5, is one
Chapter 4
Bacteriochlorophyll Biosynthesis
61
Table 1. Details of enzyme names, EC numbers, gene and mutant names for the purple bacterial enzymatic steps shown in Figs. 3 and 4. Enzyme ALA synthase
EC# 2.3.1.37
Gene names hemA hemT
References Neidle and Kaplan (1993b)
Notes on mutants Rba. sphaeroides mutants in which one or both genes are inactivated, mutants with a single gene mutation grew fine, whereas double mutants required exogenous ALA for growth
hemA
Wright et al. (1987)
Rba. capsulatus unable to synthesize ALA, requires exogenous ALA for aerobic growth and bacteriochlorophyll synthesis
ALA dehydratase
4.2.1.24
hem B
Hydroxymethylbilane synthase
2.5.1.61 (formerly 4.3.1.8)
hemC
Uroporphryinogen III synthase
4.2.1.75
hemD
Uroporphryinogen III decarboxylase
4.1.1.37
hemE
Coproporphryinogen oxidase
1.3.3.3 1.3.99.22
hemF hemN hemZ
Zeilstra-Ryalls and Schornberg (2006); Coomber et al. (1992)
Rba. sphaeroides requires HemF for aerobic growth
hemN (formerly hemF)
Coomber et al. (1992)
Rba. sphaeroides incapable of photosynthetic growth; excretes coproporphyrin III
Protoporphyrinogen oxidase
1.3.3.4
hemG hemY
in which glycine forms a Schiff ’s base with the pyridoxal phosphate cofactor, which is necessary for the activation of ALAS (Burnham and Lascelles, 1963). Studies with tritiated glycine show that condensation with succinyl-CoA subsequently takes place with the removal of the pro-R hydrogen atom of glycine (Akhtar and Jordan, 1968; Zaman et al., 1973). The condensation mechanism proceeds with stereochemical inversion (Abboud et al., 1974). It is, however, unclear if decarboxylation proceeds before or after release of the product from the pyridoxal cofactor. After the initial discovery of ALAS in Rba. sphaeroides (Kikuchi et al., 1958), the enzyme was purified from a number of sources. Early studies on ALAS from Rba. sphaeroides by Tuboi et al. (1970b) revealed two isoforms of the enzyme that could be separated by ion exchange chromatography on diethylaminoethyl-cellulose. These isoforms had similar kinetic properties (Km for glycine of 5 mM and for succinyl-CoA of 5 µM) but their relative
amounts differed with the growth conditions (Tuboi et al., 1970a). The first gene to encode ALAS (hemA) cloned from a photosynthetic bacterium was from a Rba. capsulatus cosmid library (Biel et al., 1988) and this is the only ALAS gene in this bacterium. Rba. sphaeroides, on the other hand, has two separate genes, as implied by the Tuboi study. The two genes, hemA and hemT (each 1,224 nucleotides in length with 407 codons), encode two isozymes which are homologous to each other as well to other previously characterized ALAS (Neidle and Kaplan, 1993a). A single functional hemA or hemT gene is sufficient to provide approximately wild-type ALA synthase levels (Neidle and Kaplan, 1993b). The two genes are, however, located on separate chromosomes and are subject to different regulation (Neidle and Kaplan, 1993a); see Chapter 39 (Zeilstra-Ryalls) on regulation of the tetrapyrrole pathway.
62
Robert D. Willows and Alison M. Kriegel
Fig. 5. Enzymatic mechanism for ALA synthase (Burnham and Lascelles, 1963; Akhtar and Jordan, 1968; Zaman et al., 1973; Abboud et al., 1974)
B. δ-Aminolevulinic acid Dehydratase The reaction catalysed by δ-aminolevulinic acid dehydratase (ALAD) represents the first step in the pathway conserved by all organisms that biosynthesize tetrapyrroles. ALAD, also called porphobilinogen synthase, catalyses the asymmetric condensation of two molecules of ALA to form porphobilinogen (PBG) and release two molecules of water. The ALAD-catalysed reaction is necessary to ensure that two ALA molecules condense in parallel, as opposed to the antiparallel reaction that occurs nonenzymatically. First identified in duck blood (Shemin and Russell, 1953), ALAD has since been isolated from a number of sources. The protein sequences
of ALADs are highly homologous across a broad range of species including mammals, yeast, plants and bacteria (Jaffe, 1995). The conserved nature of these sequences suggests that the catalytic function is probably the same for all organisms. Early work by Shemin (1976) and Nandi et al. (1968) on ALAD isolated from Rba. sphaeroides determined that the 8-subunit enzyme required the presence of a thiol reagent to maintain reducing conditions, and a metal cation to remain functional. In addition, they provided the first evidence that a Schiff ’s base is formed between the enzyme and substrate (Nandi et al., 1968; Shemin, 1976). Pyrrole synthesis requires two distinct binding sites on the ALAD enzyme for two identical ALA
Chapter 4
Bacteriochlorophyll Biosynthesis
63
substrates, termed ‘A-site’ and ‘P-site.’ The binding sites accept the ALA molecule that will form the acetic acid side chain (A-side ALA) and the propanoic acid side chain (P-side ALA) of the resulting porphobilinogen. Studies with human ALAD show that ALA binds first to the P-site of the enzyme and establishes a Schiff ’s base with a reactive lysine (Jordan and Gibbs, 1985). The detailed mechanism of the Pseudomonas aeruginosa ALAD was determined with the aid of crystal structures (Frankenberg et al., 1999; Frere et al., 2002). The reaction begins with the formation of a Schiff’s base between the 4-keto group of P-side ALA with a conserved lysine, Lys-260, in this sequence, followed by an additional molecule of ALA that binds to the A-site of the enzyme. The reaction proceeds with an aldol condensation and Schiff’s base formation with Lys-205. The removal of H+ from C-3 of the A-site yields an enamine. An aldol condensation allows the formation of a C-C bond between the A and P-site ALA before a Schiff ’s-base exchange produces an additional C-N bond. An H+ transferred to Lys 260 allows the trans-elimination of the P-site lysine. Finally, the pro-R H is removed from C5 of the P-site ALA resulting in the aromatization and release of the porphobilinogen. ALADs are metalloenzymes and the specific metal ion necessary varies among different species; bacterial enzymes rely exclusively on Mg2+ or on both Mg2+ and Zn2+, plant enzymes contain Mg2+ and mammalian enzymes generally utilize Zn2+. Some Mg2+ bacterial enzymes such as ALAD from Rba. sphaeroides are further activated by monovalent cations such as K+ (Nandi et al., 1968; Shemin, 1976). While the exact function of the metal ions has not been elucidated, Jaffe (1995) has suggested three different types of metal-binding sites called A, B and C harboring metals MeA, MeB and MeC. MeA is believed to aid in A-side ALA binding and reactivity. MeB does not appear necessary for activity, although its cysteine ligands were proposed to be involved in proton removal during product formation. MeC is proposed to bind at the subunit interface and to act as an allosteric activator.
al., 1979; Jordan et al., 1979). HMBS, encoded by the gene hemC, was originally referred to as uroporphyrinogen I synthase, thought to catalyze the formation of uroporphyrinogen I from four molecules of porphobilinogen. It is HMB, however, that cyclizes spontaneously to uroporphyrinogen I in the absence of uroporphyrinogen III synthase, the next enzyme in the pyrrole pathway (Burton et al., 1979; Jordan et al., 1979). Initial studies with HMBS isolated from Rba. sphaeroides confirmed that a covalent interaction was established with the first porphobilinogen, which becomes ring I. Subsequent porphobilinogen units are added while ring I remains covalently bound to the enzyme (Jordan and Berry, 1981). HMBSs have been isolated from a number of sources. Protein sequences are highly conserved across species. The structure of the E. coli HMBS was elucidated using X-ray crystallographic methods (Louie et al., 1992). It is made up of three distinct domains of approximately equal size connected by linkers that allow much flexibility as the pyrrole chain elongates. The activity of HMBS relies on a unique prosthetic group, a dipyrromethane cofactor. Through a thioether linkage, the cofactor is covalently attached to a cysteine residue in the active site of the enzyme (Jordan and Warren, 1987). The cofactor then serves as the covalent docking point for the 4 pyrroles to be added sequentially leading to the intermediates; ES, ES2, ES3 and ES4 (Berry et al., 1981, Warren and Jordan, 1988). The product, HMB is released while the cofactor remains attached to the enzyme. E. coli hemC mutants show the importance of highly conserved arginine residues in the binding site for the acetic acid and propionic acid side chains of the cofactor, as well as the substrate and intermediate complexes (Jordan and Woodcock, 1991; Lander et al., 1991) . The arginine residues are thought to help neutralize the amassed negative charges that build up from 12 carboxyl side chains in the ES4 intermediate. The presence of the cofactor may also help neutralize the electropositivity from these Arg residues in the active site (Louie et al., 1992).
C Hydroxymethylbilane Synthase
D. Uroporphyrinogen III Synthase
Hydroxymethylbilane synthase (HMBS), also called porphobilinogen deaminase, catalyzes the tetrapolymerization of porphobilinogen to yield HMB, an unstable linear tetrapyrrole intermediate (Burton et
Uroporphryinogen III synthase [4.2.1.75] catalyzes the conversion of HMB to uroporphyrinogen III by rearrangement and cyclization. The enzyme catalysed reaction occurs more quickly than the non-enzymatic
64 conversion to uroporphyrinogen I (Burton et al., 1979; Jordan et al., 1979) The enzyme encoded by hemD was originally called cosynthase. In this step of the tetrapyrrole pathway, the I and IV rings are joined in an asymmetrical manner to the reverse of the IV ring so that the pattern of alternating acetic and propionic acid side chains is disrupted, resulting in the type III asymmetry. Although uroporphyrinogen III synthase genes have been isolated from a number of different sources (though from an extremely limited number of purple bacteria), the proteins show little sequence similarity. Even the hemD sequences from three different strains of Rhodopseudomonas palustris show great variation with as little as 67% identity between protein sequences (Kanehisa, 1997; Kanehisa and Goto, 2000; Kanehisa et al., 2006). The crystal structure of the human uroporphyrinogen III synthase identified considerable flexibility in the active site and mutagenesis of 10 active site residues resulted in little variation in enzymatic activity (Mathews et al. 2001). Mutagenesis of the cyanobacterial uroporphyrinogen III synthase identified a tryptophan residue that was important but not essential for activity which is possibly involved in orientating the substrate in the active site (Roessner et al. 2002). Therefore little can be said for any inferred mechanism for this enzyme. E. Uroporphyrinogen III Decarboxylase The next step in the pathway involves the conversion of uroporphyrinogen III to coproporphyrinogen III and is catalyzed by the enzyme uroporphyrinogen III decarboxylase. The genes/cDNAs have been isolated from a number of prokaryotic and eukaryotic species and the HemE proteins show strong sequence identity, indicating that the enzyme mechanism is likely to be similar across different organisms (Jones and Jordan, 1993; Whitby et al., 1998). Studies with uroporphyrinogen III decarboxylase isolated from Rba. sphaeroides (Jones and Jordan, 1993) suggested the importance of both cysteine and arginine residues for the activity of the enzyme. Uroporphyrinogen III decarboxylase is interesting in that the enzyme isolated from both human (Deverneuil et al., 1983) and bovine (Straka and Kushner, 1983) sources needed no cofactors to remain active. Early studies based on porphyrin intermediates found in the feces of rats poisoned with hexachlorobenzene, suggests the decarboxylation of uroporphyrinogen III to coproporphyrinogen III takes
Robert D. Willows and Alison M. Kriegel place in an ordered sequence starting with ring IV, followed by I, II and III, consecutively (Jackson et al., 1976). In vitro studies with uroporphyrinogen III decarboxylase purified to homogeneity from Rba. sphaeroides suggest that the decarboxylation steps occur in a random order (Jones and Jordan, 1993). However, the substrate specificity of this enzyme is reduced in the presence of high substrate concentration giving a random order of decarboxylations while at low substrate concentration the decarboxylations occur in an ordered sequence (Kawanishi et al., 1983, Lash, 1991). Comparisons between the structures of both the human enzyme (Whitby et al., 1998) and that from Nicotiana tabacum (Martins et al., 2001) show a high degree of structural conservation. However, even with these structures, the mechanism of action of uroporphyrinogen III decarboxylase is still unclear. F. Coproporphyrinogen Oxidase The enzyme coproporphyrinogen III oxidase (CPO) catalyses the oxidative decarboxylation of the propanoates on rings I and II to vinyl groups to form protoporphyrinogen. Three isoforms exist; oxygendependent CPO, encoded by hemF, is found in eukaryotic and bacterial species and oxygen-independent CPO enzymes, encoded by hemN and the orthologous hemZ, are prevalent in prokaryotes that synthesize heme. The hemF genes are not widely represented among bacteria, and bacteria with hemF tend also to have hemN (Panek and O’Brian, 2002). HemN and HemF are unique in their protein sequence and utilize different cofactors. Both aerobic and anaerobic CPO activities were originally demonstrated in Rba. sphaeroides by Tait (1969). Oxygen-dependent CPO activity was localized in the soluble fraction of a crude cellular extract. This CPO activity required the presence of oxygen as the enzymatic reaction did not occur under anaerobic conditions, even when additional cofactors (see below) were added (Seehra et al., 1983). The oxygen-independent CPO required the combination of both the insoluble and soluble cellular fractions, in addition to the cofactors ATP, Mg2+, methionine, NADH and NADP+ for enzymatic activity (Seehra et al., 1983; Tait, 1972). The oxygen-independent CPO in Rba. sphaeroides was originally designated as a putative hemF as mutations in this gene resulted in coproporphyrin into the media. The sequence of this putative HemF had
Chapter 4
Bacteriochlorophyll Biosynthesis
little homology to HemF protein sequences in other species. The activity required ATP, methionine, Mg2+ and NADPH, which indicated that it was an anaerobic coproporphyrinogen oxidase (Coomber et al., 1992) and it has subsequently been determined that it is a HemN-type protein. The correct Rba. sphaeroides hemF gene, with homology to other HemF proteins was recently confirmed to be an oxygen-dependent CPO and the hemF gene is necessary for aerobic growth and is also differentially regulated by oxygen and light (Zeilstra-Ryalls and Schornberg, 2006). Two putative oxygen-independent isoenzymes, hemN (Coomber et al., 1992) and hemZ (ZeilstraRyalls and Kaplan, 1995), have been isolated from Rba. sphaeroides while Rba. capsulatus has a single isoform, orthologous to hemZ, which appears to be essential for heme production under both aerobic and anaerobic conditions (Smart et al., 2004). The oxygenindependent HemN orthologs are free radical SAM enzymes (Sofia et al., 2001). The proposed enzymatic mechanism for HemN shown in Fig. 6 was elucidated based on the crystal structure of E. coli CPO, HemN (Layer et al., 2003; Layer et al., 2005). G. Protoporphyrinogen Oxidase The final step in the formation of protoporphyrin IX is the six-electron oxidation of protoporphyrinogen IX. The enzyme protoporphyrinogen oxidase (PPO) catalyses this reaction, however, the reaction will proceed spontaneously in presence of molecular oxygen. PPO activity was demonstrated in Rba. sphaeroides by Jacobs and Jacobs (1981). Protoporphyrinogen oxidizing activity was associated with the membrane fraction under both aerobic and anaerobic growth conditions. Membranes from aerobically grown cells exhibited about 50% of the enzyme activity of membranes from anaerobically grown cells. The PPO activities of membranes from cells grown in both conditions were inhibited by cyanide. The genes encoding PPO have been identified from eukaryotes, designated hemG, and a number of prokaryotes, designated hemG or hemY. However no homologs to either hemG or hemY have been identified in the genomes of purple bacteria (Kanehisa, 1997; Kanehisa and Goto, 2000; Kanehisa et al., 2006). This situation is not unique as pointed out by Panek and O’Brian (2002); 14 out of 28 prokaryotic genera examined had no hemG or hemY homologs. The authors suggest the possibility that either an
65 unidentified PPO exists or a known protein has additional PPO activity. III. Protoporphyrin IX to Bacteriochlorophyll a and b The enzymatic steps from protoporphyrin IX onwards are unique to bacteriochlorophyll biosynthesis unlike earlier steps which are shared with the heme and vitamin B12 biosynthetic pathways. Early work on this section of the pathway utilized inhibitors of bacteriochlorophyll biosynthesis and mutants of Rba. sphaeroides to define some of the intermediates in the pathway. Some of the intermediates which accumulated in these mutants were the same as the intermediates which accumulated in Chlorella mutants characterized by Granick (Granick, 1948, 1949, 1950). This indicated that the pathways to chlorophyll and bacteriochlorophyll were likely to be similar. The use of mutants and inhibitors and the characterization of pigments which accumulated was one of the more useful methods in characterization of the pathway from protoporphyrin IX onwards (Jones, 1967). This methodology was continued using molecular genetic studies by Marrs and coworkers utilizing Rba. capsulatus to identify the genetic loci required for individual steps in bacteriochlorophyll biosynthesis (Yen and Marrs, 1976; Marrs, 1981; Zsebo and Hearst, 1984; Young et al., 1989; Bauer et al., 1991). Similar studies were carried out on Rba. sphaeroides using complementation and transposon mutagenesis approaches (Hunter and Coomber 1988; Coomber and Hunter 1989; Coomber et al., 1990b) All of these studies pointed to a clustering of genes for pigment biosynthesis and photosystem proteins in both Rba. capsulatus and Rba. sphaeroides, an arrangement that has become known as the photosynthesis gene cluster. It has since been confirmed that all of the purple bacteria examined thus far exhibit some degree of clustering of photosynthesis-related genes, with the genes arranged in subclusters encoding bacteriochlorophyll biosynthesis, carotenoid biosynthesis and photosystem components although there is overlap between these subclusters. There is also evidence of ‘superoperonal’ arrangement of genes with overlapping operons within the cluster (Bauer et al., 1991; Young et al., 1989). Initially it was believed that the order of genes within these clusters is conserved among the purple bacteria (Yildiz et al., 1992) although a comprehensive survey
66
Robert D. Willows and Alison M. Kriegel
Fig. 6. CPO free radical SAM mechanism (Layer et al., 2003; Layer et al., 2005). Scheme I shows the generation of the 5´deoxyadenosyl radical from SAM with release of methionine by HemN. Scheme II shows that reaction catalysed by HemN utilizing the 5´deoxyadenosyl radical to oxidatively decarboxylate the propionate on ring I to a vinyl group.
has recently shown that the gene arrangement is not conserved across the proteobacteriaceae although the genes are still clustered into functional groups (Béjà et al., 2002). The functions of the genes within these clusters has been probed by using transposon and insertional based mutagenesis procedures coupled with complementation and marker rescue techniques. The identity of the genes responsible for bacteriochlorophyll biosynthesis was based on analysis of the pigments which accumulated and in many cases has since been confirmed by heterologous expression of the genes in E. coli and demonstration of enzymatic activity which will be discussed in detail below for the individual enzymes. These mutational studies and the corresponding references are summarized in Table 2 and the structures of the biosynthetic intermediates are shown in Fig. 7. A. Magnesium Chelatase Magnesium chelatase catalyses the ATP hydrolysis dependent insertion of magnesium into protoporphyrin IX and is a complex enzyme consisting of three distinct types of subunits. It is the first committed
step in chlorophyll biosynthesis as the preceding steps in the pathway are shared by the heme biosynthetic pathway (Willows and Hansson, 2003). Early studies by Gorchein (1972, 1973) indicated that the magnesium chelatase is a complex enzyme. Experiments with spheroplasts of Rba. sphaeroides suggested that ATP was necessary for enzymatic function and that activity was lost when the spheroplasts were lysed (Gorchein, 1972, 1973). Gorchein subsequently developed a continuous assay using spheroplasts (Gorchein, 1994) and later developed an in vitro magnesium chelatase assay using lysed Rba. capsulatus and Rba. sphaeroides cells (Gorchein, 1997). However, the first report of in vitro magnesium chelatase activity from a bacterial source was with the Rba. sphaeroides BchI, BchD and BchH proteins which were heterologously expressed in E.coli and the extracts mixed to yield magnesium chelatase activity (Gibson et al., 1995). Most of the subsequent details of the catalytic mechanism for magnesium chelatase have been gleaned from studies of heterologously expressed Rhodobacter and Synechocystis enzymes. Although the enzymes from different sources have slightly different properties the overall mechanism is likely to be
Chapter 4
Bacteriochlorophyll Biosynthesis
67
Table 2. Details of enzyme names, EC numbers, gene names, references and properties of mutants for the purple bacterial enzymatic steps detailed in Fig. 7. Enzyme Magnesium chelatase
EC# 6.6.1.1
Gene names bchI bchD bchH
Magnesium protoporphyrin IX: SAM O-methyltransferase
2.1.1.11
bchM
Magnesium protoporphyrin IX monomethyl ester oxidative cyclase/s
1.14.13.81
bchE acsF
Taylor et al. (1983); Hunter and Coomber (1988); Pinta et al. (2002)
Accumulate MPE
Protochlorophyllide reductase
1.18.-.-
bchLNB
Biel and Marrs (1983); Bollivar et al. (1994b); Yen and Marrs (1976); Marrs (1981); Coomber et al. (1990a)
Accumulate protochlorophyllide
8-Vinyl reductase
1.3.1.75
bchJ
Suzuki and Bauer (1995); Bollivar et al. (1994b)
Excretes divinyl protochlorophyllide
Chlorin reductase
1.18.-.-
bchX, bchY and bchZ
Young et al. (1989); Bollivar et al. (1994b) McGlynn and Hunter (1993)
Accumulates 3-hydroxyethyl chlorophyllide a.
3-Vinyl bacteriochlorophyllide a hydroxylase
bchF
Bollivar et al. (1994b)
Appears to accumulate 3vinyl bacteriochlorophyll a
3-Hydroxyethyl bacteriochlorophyllide a dehydrogenase
bchC
Wellington and Beatty (1990); McGlynn and Hunter (1993)
Accumulates 3-hydroxyethyl bacteriochlorophyllide a
bchG
Bollivar et al. (1994b); Coomber et al. (1990a); Addlesee et al. (2000)
Accumulates bacteriochlorophyllide a
bchP
Bollivar et al. (1994b); Addlesee and Hunter (1999)
Accumulates bacteriochlorophyll with geranyl geraniol esterified
Bacteriochlorophyll synthase
Geranylgeranyl reductase
2.5.1.62
similar but we will concentrate on the properties of the purple bacterial enzymes. Magnesium chelatase is now known to consist of three separate subunits, with the proteins called BchI, BchD and BchH in bacteriochlorophyll biosynthesizing organisms with orthologous proteins ChlI, ChlD and ChlH in organisms which synthesize chlorophyll (Jensen et al., 1996; Willows and Hansson, 2003). The first magnesium chelatase assay using purified components used the BchH, BchI and BchD proteins of Rba. sphaeroides that were heterologously expressed in E. coli (Gibson et al., 1995). The BchI and BchD proteins were expressed together as the genes for these proteins are part of the same operon. The BchI protein was highly expressed but no ap-
References Bollivar et al. (1994b); Coomber et al. (1990a); Gorchein et al. (1993); Marrs (1981); Taylor et al. (1983); Zsebo and Hearst (1984)
Notes on mutants Accumulate protoporphyrin IX
Reported mutations not definitively in bchM gene
parent expression of the BchD protein was observed on SDS-PAGE gels. The BchH protein was highly expressed and the cells and cell lysate were visibly red from accumulation of protoporphyrin IX, which was thought to be bound to the soluble BchH protein. Magnesium chelatase activity was only reconstituted when a cell extract from the BchI plus BchD coexpression strain was mixed with an extract from cells expressing the BchH protein in a buffer containing an ATP regenerating system. This indicated that all three proteins were required for activity and that the protoporphyrin bound to the BchH subunit was used as a substrate (Gibson et al., 1995). The first magnesium chelatase proteins to be purified to apparent homogeneity were the BchI and BchH
68
Robert D. Willows and Alison M. Kriegel
Fig. 7. Biosynthetic pathway for bacteriochlorophyll biosynthesis from protoporphyrin IX. The gene products required for catalysis of each step are indicated. Further details of enzyme names are given in Table 2.
Chapter 4
Bacteriochlorophyll Biosynthesis
proteins from Rba. sphaeroides (Willows et al., 1996). The BchD protein was not purified to homogeneity but copurified with the E. coli chaperonin protein GroEL. When the BchH protein was purified to near homogeneity from the cell lysate, protoporphyrin IX was still found to be bound to the protein. Magnesium chelatase activity required all three proteins, ATP and magnesium but exogenous protoporphyrin IX was not absolutely required as it was bound to the BchH protein. The addition of protoporphyrin IX yielded higher activity and deuteroporphyin could be substituted for protoporphyrin as a substrate (Willows et al., 1996). The BchI, BchD and BchH proteins of Rba. sphaeroides and Rba. capsulatus were heterologously expressed in E.coli both with and without an N-terminal 6xHis-Tag (Willows and Beale, 1998; Gibson et al., 1999). The purified BchH proteins had protoporphyrin IX bound (Willows et al., 1996; Willows and Beale, 1998) in a pigment to protein ratio of 1.3:1 (Willows et al., 1996) indicating that the protein has at least one high affinity porphyrin binding site. The BchH protein of Rba. capsulatus was slowly inactivated by light, presumably because of the bound protoporphyrin IX (Willows and Beale, 1998). While both the BchH and BchI proteins were soluble the BchD proteins formed insoluble inclusion bodies and required refolding from 6M urea. Effective refolding of the BchD subunits required ATP, MgCl2 and BchI protein. Thus the BchI appeared to be acting as a molecular chaperone to allow BchD to fold correctly or to stabilize the BchD once it was folded (Willows and Beale, 1998; Gibson et al., 1999). Magnesium chelatase activity was only reconstituted when all three subunits were present in the assay together with protoporphyrin IX, Mg2+ and ATP. Curiously although both the 6xHis-Tagged and non-His-Tagged Rba. sphaeroides proteins were able to reconstitute magnesium chelatase activity, the His-Tagged BchI protein of Rba. capsulatus could not reconstitute magnesium chelatase activity and only the non-His-tagged BchI worked in magnesium chelatase assays (Willows and Beale, 1998). In continuous assays a lag period was observed for both the Rba. sphaeroides (Willows et al., 1996) and Rba. capsulatus (Willows and Beale, 1998) enzymes before magnesium protoporphyrin IX was produced at a constant rate. This was similar to the lag period observed for the pea enzyme (Walker and Weinstein, 1994). This lag period could be overcome by preincubation of the BchI and BchD proteins with Mg-ATP
69 prior to addition of BchH and protoporphyrin IX to initiate the assay. This lag period was suggested to be due to the formation of an ATP dependent activation complex between the BchI and BchD proteins which was dependent on both the relative concentrations of subunits and the presence of ATP (Willows et al., 1996; Willows and Beale, 1998). This activation complex catalyses magnesium insertion into protoporphyrin only when combined with the BchH protein, Mg-ATP, protoporphyrin IX and Mg2+. In addition to the activation complex the BchH protein behaves as a substrate in the magnesium chelatase reaction and the purple bacterial enzyme has a Km of 1.34 µM (Willows et al., 1996; Willows and Beale, 1998; Gibson et al., 1999). The reported Km for protoporphyrin was 1.23 µM for the Rba. capsulatus enzyme (Willows and Beale, 1998) and 0.15 µM to 0.36 µM for the Rba. sphaeroides enzyme (Willows et al., 1996; Gibson et al., 1999). The kinetic constants for ATP and Mg2+ were also reported for the Rba. sphaeroides enzyme with Km values for Mg2+ of 0.12 ± 0.02 mM (Gibson et al., 1999) to 0.166 mM (Willows et al., 1996) and Km values for ATP of 1.7 mM (Willows et al., 1996) to 3.3 mM (Gibson et al., 1999). Binding affinity of deuteroporphyrin IX to the BchH subunit of Rba. sphaeroides was estimated using tryptophan fluorescence quenching, with the Kd estimated at 0.53 µM (Karger et al., 2001). The structure of the BchI protein from Rba. capsulatus was determined by X-ray crystallography and is shown as a ribbon diagram in Fig. 8 (Fodje et al., 2001). The BchI protein has an ATPase domain, shown as sections A and B in Fig. 8, and a four helix bundle domain, section C in Fig. 8. Both the structure and primary sequence of BchI show that it belongs to the extended ATPases Associated with a variety of cellular Activities (AAA+) class of proteins. Most structures of AAA+ proteins have two domains, an ATPase domain and a second domain that is positioned close to the ATP binding site. The position of this second domain in BchI, the four helix bundle domain, differs markedly from other AAA+ proteins as it is not close to the ATP binding site. In addition the ATPase domain has a number of extensions, see section A in Fig 8, which are not present in other AAA proteins (Fodje et al., 2001). The BchD protein also has an AAA+ domain at its N-terminus. AAA+ proteins are one of the largest and most diverse classes of proteins known and are represented in all organisms from all kingdoms (Confalonieri and Duguet, 1995; Glockner et al., 2000). The AAA+ proteins are
70
Fig. 8. Structure of BchI from Rba. capsulatus (Fodje et al., 2001) as a ribbon diagram. The ATPase domain is shown as segments labeled A and B. Section C is a four helix bundle domain. See text for further details.
known to form nucleotide-dependent ring structures, which are usually hexameric, and many types can form double hexameric rings. The BchI subunit was shown to form ATP dependent oligomers (Hansson et al., 2002) and these oligomers were found to be hexameric rings using single particle analysis of negatively stained electron micrograph images (Willows et al., 2004). The overall shape of the BchI subunit is wedge shaped suggesting that it fits into a hexameric complex. Site directed mutagenesis suggested that the ATPase activity of BchI was dependent on amino acid residues predicted to lie in adjacent positions in neighboring subunits within a predicted hexameric ring structure. Specifically, Arg289 was required for ATPase activity as when mutated to a Lys, ATPase activity was lost (Hansson et al., 2002). Arg289 is a so-called arginine finger found in other AAA+ proteins and is thought to mediate the ATPase activity and this residue is only adjacent to the ATP binding site when a neighboring subunit is present in a hexameric structure (Hansson et al., 2002). In AAA+ modules which form double ring structures the second ring often has an inactive ATPase and this ring presumably has a structural role. AAA+ proteins have also been called mechanoenzymes due to observed large conformational changes that
Robert D. Willows and Alison M. Kriegel occur upon ATP hydrolysis and also the mechanical nature of the processes in which many of these proteins are involved (Glockner et al., 2000). Both BchI and BchD have AAA+ domains which led to the suggestion that these subunits also form these double hexameric ring structures (Fodje et al., 2001; Willows and Hansson, 2003). Such a double ring structure would also be consistent with the ATP dependent activation phase required between BchI and BchD (Willows and Beale, 1998; Gibson et al., 1999) and the demonstration that an I-D complex can be formed in the presence of Mg2+ and ATP. The double ring structure would presumably catalyze an ATP-dependent conformational change in BchH to effect magnesium insertion into protoporphyrin IX which is bound to BchH (Karger et al., 2001; Hansson et al., 2002; Willows and Hansson, 2003). ATPase and phosphate exchange activities have been measured for BchI of Rba. sphaeroides and the BchH and BchD subunits had ATPase activity, although the BchD subunit was not pure (Hansson and Kannangara, 1997). The BchH ortholog of Synechocystis was also reported to have ATP hydrolysis activity, although the specific activity was 155-fold lower than for ChlI (Jensen et al., 1999) indicating that there was no significant ATPase activity. However, Mg-ATP increased the KD for porphyrin binding to ChlH (Karger et al., 2001) and also protected the ChlH from inactivation by a thiol modifying reagent (Jensen et al., 2000) suggesting that ChlH has a Mg-ATP binding site possibly involved in regulation (Karger et al., 2001). In contrast Mg-ATP had no effect on porphyrin binding to BchH from Rba. sphaeroides (Karger et al., 2001). A re-examination of the ATPase activity of the purple bacterial enzyme from Rba. capsulatus has shown that no ATP hydrolysis activity is associated with BchH of Rba. capsulatus and it is not required for protoporphyrin IX binding to BchH (Sirijovski et al., 2006). The purpose of the phosphate exchange activity in BchI is unclear but it is possible that the complex of six BchI subunits hydrolyses ATP sequentially and the exchange activity may be a measure of the reverse reaction. A number of inhibitors of the purple bacterial magnesium chelatases have been reported. The most unusual inhibitors were reported for the Rba. sphaeroides magnesium chelatase which was inhibited by both chloramphenicol and p-aminosalicylic acid in a study which showed that the enzyme activity was associated with ribosomes (Kannangara et al., 1997). Light has been shown to inhibit the magnesium
Chapter 4
Bacteriochlorophyll Biosynthesis
chelatase of Rba. capsulatus (Willows and Beale, 1998) and this mode of inhibition probably occurs via photooxidative damage of the BchH subunit (Willows and Beale, 1998; Willows et al., 2003). Sulfhydryl modifying agents were reported to inhibit the purple bacterial enzyme (Gorchein, 1997). These inhibitors have been used to analyze the Synechocystis enzyme. It was discovered that the cysteines in the ChlD subunit are not essential while those in ChlI and ChlH are essential. As mentioned above, inhibition of ChlH by thiol modifying reagents was partially alleviated by preincubation of ChlH with Mg-ATP and protoporphyrin IX before the treatment (Jensen et al., 2000). B. S-Adenosyl-L-methionine:Magnesium Protoporphyrin IX-O-methyltransferase S-adenosylmethionine:magnesium protoporphyrin IX-O-methyltransferase catalyses the S-adenosylmethionine (SAM) dependent methylation of the carboxyl group of the 13-propionate on magnesium protoporphyrin IX to form magnesium protoporphyrin IX monomethyl ester (MPE). This enzyme is membrane associated and the activity has been characterized for two Rhodobacter species, reviewed in Bollivar (2003). Characterization of this activity first occurred with the Rba. sphaeroides enzyme in the 1960s. The enzyme was found to be associated with the chromatophores, which are small vesicles on which the photosystems are assembled in these bacteria. The activity of cell-free extracts was characterized in terms of the substrate specificity and inhibition by substrate analogs. The methyl donor was S-adenosylmethionine and effective porphyrin substrates were magnesium protoporphyrin IX, zinc protoporphyrin IX and calcium protoporphyrin IX. However, non metal containing porphyrins as well as ferric protoporphyrin IX, ferrous protoporphyrin IX, manganic protoporphyrin IX and manganous protoporphyrin IX were not substrates for the monomethyl ester transferase (Gibson et al., 1963). The enzyme was partially purified and kinetically characterized from chromatophores of Rba. sphaeroides. The membrane bound activity was solubilized from the chromatophores with 1% w/v sodium cholate. The solubilized enzyme was then purified using S-adenosyl homocysteine coupled agarose. The kinetic mechanism was found to be sequential equilibrium ordered and the Km for SAM was given as 106 µM but no Km was given for magnesium pro-
71 toporphyrin IX (Hinchigeri et al., 1984). It was confirmed that the bchM gene encoded the methyltransferase in both Rba. capsulatus and Rba. sphaeroides when both enzymes were heterologously expressed in E. coli and the extracts shown to have S-adenosylmethionine:magnesium protoporphyrin IX-O-methyltransferase activity (Bollivar et al., 1994a; Gibson and Hunter, 1994). The activity of S-adenosyl-L- methionine:magnesium protoporphyrin IX methyltransferase, present in the membranes of an expression strain of E. coli into which the bchM gene of Rba. capsulatus had been cloned, was enhanced when mixed with recombinant BchH protein or with membranes of a bchH mutant of Rba. capsulatus, which did not have detectable methyltransferase activity (Hinchigeri et al., 1997). The authors took this to mean that the BchH protein stimulates the BchM activity however considering that the bchH mutant also stimulated activity it could also be taken to mean that the BchH and the membranes carried a stimulatory factor for the methyltransferase. More recent work using quenched flow analysis on purified proteins from the cyanobacterium Synechocystis suggests shows that the ChlH subunit affects methyltransferase activity during the pre-steady-state phase of the reaction and hence is involved in the reaction chemistry (Shepherd et al., 2005). C. Magnesium-protoporphyrin IX Monomethylester Oxidative Cyclase An oxidative cyclization is required to create the fifth ring of bacteriochlorophyll; this reaction is catalysed by magnesium protoporphyrin IX monomethyl ester oxidative cyclase. The 131-132 acrylate, the 131-hydroxy and the 131-keto derivatives of MPE were suggested as intermediates in the conversion of MPE to divinyl-protochlorophyllide by Granick (1949). However, although the hydroxy and keto intermediates have been detected in plants there is no evidence for the acrylate intermediate; for a review see Bollivar (2003). The origin of the oxo-group in ring V has been studied in the purple bacteria using 18O2 and/or H218O. It was found in Rba. sphaeroides that the oxo group is derived from water rather than molecular oxygen indicating that the reaction is catalysed by a hydratase (Porra et al., 1995). In contrast, the purple bacteria Roseobacter denitrificans and Rhodovulum sulphidophilum appear to incorporate 18O from both oxygen and water into the oxo-group on the
72 fifth ring of bacteriochlorophyll (Porra et al., 1996, 1998). Unlike Rba. sphaeroides, these two species make bacteriochlorophyll under both anaerobic and aerobic conditions which suggested that they have two different enzymes for oxidative cyclization. Molecular oxygen is preferentially incorporated into the oxo group on ring V in bacteriochlorophyll a when cultures of these bacteria are grown under aerobic conditions and oxygen from water is preferentially incorporated when the bacteria are cultured anaerobically (Porra et al., 1996, 1998). Mutation of the bchE gene in Rba. sphaeroides (Hunter and Coomber, 1988) and Rba. capsulatus (Yen and Marrs, 1976; Bollivar et al., 1994b) resulted in accumulation of MPE although it was not confirmed until later that it was MPE which accumulated (reviewed in Bollivar, 2003). The BchE protein sequences have a predicted 4Fe-4S cluster which may be involved in the reaction mechanism (Ouchane et al., 2004). Sequence analysis of BchE by Gough et al. (2000), found a putative cobalamin binding site within the BchE sequence. This binding site was suggested to have a functional role in a cobalamin mediated free radical mechanism in which an adenosyl radical from adenosyl cobalamin was used at each of five steps in the suggested seven step reaction mechanism to generate the isocyclic ring. An in vivo oxidative cyclase assay was developed using two different vitamin B12 requiring mutants of Rba. capsulatus grown in vitamin B12 deficient media. Addition of vitamin B12 under anaerobic conditions to this assay system restored oxidative cyclase activity with cyano cobalamin being most effective followed by adenosyl cobalamin. In addition the vitamin B12 requiring mutants of Rba. capsulatus were found to accumulate MPE when vitamin B12 was limiting, even when the cells were supplemented with methionine which requires vitamin B12 for its synthesis (Gough et al., 2000). Further sequence analysis of the BchE protein suggests that it is a free radical SAM enzyme like HemN as it contains the conserved CXXXCXXC motif (Sofia et al., 2001). This motif forms the previously mentioned 4Fe-4S cluster and is used as an electron donor to generate a 5´-deoxyadenosyl radical which acts as a proton radical acceptor to generate a radical on the propanoate side chain of the porphyrin (see Fig. 6). The mechanism shown in Fig. 9 is the one suggested by Gough et al. (2000) which involves adenosyl cobalamin as the free radical donor. In this mechanism the 4Fe-4S cluster acts as the electron do-
Robert D. Willows and Alison M. Kriegel nor to generate the radical from adenosyl cobalamin. This radical may be a deoxyadenosyl radical similar to the HemN mechanism or some other radical. In contrast to HemN the radical donor is used to create a 131-radical. However, unlike coproporphyrinogen III which has a free carboxyl group, decarboxylation cannot occur on MPE as the carboxyl group is methylated. The 131-hydroxy intermediate is then generated by loss of an electron to create a carbocation which reacts with a water molecule. Subsequent rounds of radical generation would result in generation of the 131-keto- intermediate and subsequently the final product. The gene encoding a protein required for the oxidative cyclase which incorporates molecular oxygen into the oxo-group of the isocyclic ring has been identified by mutagenesis of Rvi. gelatinosus. This organism is able to synthesize bacteriochlorophyll a under both aerobic and anaerobic conditions. Disruption of the acsF gene of Rvi. gelatinosus prevents bacteriochlorophyll a synthesis and causes accumulation of MPE under aerobic conditions but not under conditions of low aeration. The designation acsF stands for aerobic cyclization system Fe-containing subunit, as AcsF and its homologs have a conserved putative binuclear-iron-cluster motif (Pinta et al., 2002). The AcsF protein is homologous to previously identified genes in Chlamydomonas reinhardtii called Crd1 (Moseley et al., 2000) and Cth1 (Moseley et al., 2002) and homologs of AcsF were also identified in Arabidopsis thaliana and Synechocystis (Pinta et al., 2002). D. Modification of Vinyl Groups on Rings I and II Virtually all photosynthetic organisms require reduction of the 8-vinyl group of chlorophyll to an ethyl group. In the purple bacteria it was discovered that disruption of the bchJ gene results in excretion of divinyl protochlorophyllide but bacteriochlorophyll a was still produced (Bollivar et al., 1994b) When protochlorophyllide reductase mutants are generated in a disrupted bchJ strain of Rba. capsulatus a mixture of mono and divinyl protochlorophyllide accumulated but with the ratio skewed in favor of the divinyl form (Suzuki and Bauer, 1995). These reports suggested that the bchJ encoded an 8-vinyl reductase and also that the protochlorophyllide reductase favored the mono-vinyl substrate over the divinyl substrate. However, recent work has shown that bchJ does not
Chapter 4
Bacteriochlorophyll Biosynthesis
73
Fig. 9. Postulated free radical mechanism for the anaerobic cyclase reaction. X and •X are either 5´ deoxy adenosine and 5´ deoxyadenosine radicals or an alternative radical and product supplied from adenosyl cobalamin respectively.
encode an 8-vinyl reductase and the 8-vinyl reductase is encoded by the bciA gene in Chlorobium tepidum. These authors suggest that BchJ may be important in substrate channeling within the pathway (Chew and Bryant, 2006). The stereochemistry of the reduction of the 8-vinyl group in the purple bacteria was analyzed in the early 1980s by Battersby and Akhtar. The saturation of the double bond occurs with proton addition to the 81 and 82 carbons occurring in trans. The addition of the proton on the 82-position occurs from the re-face (Battersby et al., 1981) of the divinyl structure as is drawn in Fig. 6, while the proton addition to the 81position occurred on the si-face (Emery and Akhtar, 1985a). The modification of the 3-vinyl group of bacteriochlorophyll has been analyzed at the genetic level. The genes responsible for conversion of the vinyl into an acetoxy group have been identified by mutagenesis. The bchF gene product BchF appears to be responsible for hydration of the vinyl group to generate a hydroxyl intermediate. The bchC gene product BchC further oxidizes the hydroxyl group to a ketone. The enzyme which catalyses the first step appears to be able to utilize both protochlorophyllide or chlorophyllide as substrate based on the accumulation both hydroxy chlorophyllide and hydroxy proto-
chlorophyllide when bchC and bchXYZ are disrupted (Wellington et al., 1991; Bollivar et al., 1994b). E. Protochlorophyllide Oxidoreductase and Chlorin Reductase Two types of enzymes have been identified that reduce the D pyrrole ring of protochlorophyllide to form chlorophyllide in both the chlorophyll and bacteriochlorophyll biosynthetic pathways. Of these two enzymes the NADPH-protochlorophyllide oxidoreductase (EC 1.3.1.33 or EC 1.6.99.1, abbreviated POR) has been the subject of a large number of reviews (Hendrich and Bereza, 1993; Fujita, 1996; Reinbothe and Reinbothe, 1996; Reinbothe et al., 1996a,b; Adamson et al., 1997; Lebedev and Timko, 1998; Schoefs, 1999, 2001; Rüdiger, 2003; Heyes and Hunter, 2005). However this light dependent enzyme is not found in the purple bacteria. The purple bacterial enzyme is also known as a protochlorophyllide reductase and it is able to convert protochlorophyllide to chlorophyllide in the absence of light and hence has been given the acronym DPOR for dark allowed protochlorophyllide reductase to distinguish it from the single subunit light dependent enzyme (Armstrong, 1998; Fujita and Bauer, 2003). The genes encoding DPOR were identified from
74 tagged mutants which accumulated protochlorophyllide. Three tagged genetic loci, known as bchB, bchL and bchN, were identified in Rba. capsulatus (Zsebo and Hearst, 1984; Yang and Bauer, 1990; Burke et al., 1993b; Bollivar et al., 1994b) and Rba. sphaeroides (Coomber et al., 1990a) that resulted in protochlorophyllide accumulation. These genes have sequence similarity to polypeptides of the multisubunit eubacterial nitrogenase enzyme. Thus BchL is similar to NifH (Hearst et al., 1985; Burke et al., 1993a,b), BchN is similar to NifD and BchB is similar to NifK(Fujita and Bauer, 2003). In addition three other genes called bchX, bchY, and bchZ encode proteins which are similar to the to the BchL, BchN and BchB respectively (Burke et al., 1993c) and when these genes are disrupted they result in accumulation of chlorophyllide and chlorophyllide derivatives (Young et al., 1989; Bollivar et al., 1994b). Functional expression and purification of both the DPOR and the chlorin reductase from Rba. capsulatus have been achieved (Fujita and Bauer, 2000; Nomata et al., 2005; Nomata et al., 2006). For the DPOR expression in Rba. capsulatus, BchN and BchL proteins were produced with an N-terminal Strep-tag (S-tag) for purification. The proteins were purified using S-agarose under anaerobic conditions and the BchB protein copurified as a complex with the S-tagged BchN in what appeared to be a 1:1 stoichiometric ratio by SDS-PAGE. Reconstitution of protochlorophyllide reductase activity required the S-tagged BchN:BchB complex, S-tagged BchL, 1 mM ATP, an ATP regeneration system, 5 mM MgCl2, 10 mM sodium dithionite (an electron donor), and 2 µM protochlorophyllide (Fujita and Bauer, 2000). Optimization of the expression system resulted in higher yields of these proteins and protein complexes required for further biochemical analysis (Nomata et al., 2005). Kinetic analysis of the DPOR revealed an apparent Km of 10.6 µM for protochlorophyllide. Dithionite could be replaced as the electron donor by ferredoxin from maize together with NADPH and ferredoxin reductase indicating that ferredoxin is the likely electron donor in vivo. Gel filtration analysis of the BchN:BchB complex gave an estimated molecular weight of 200 kDa indicative of a tetramer consisting of 2 BchN subunits and 2 BchB subunits similar to the heterotetramer of nitrogenase. BchL chromatographed as a dimer which is also similar to nitrogenase where the orthologous NifH is also a dimer (Nomata et al., 2005).
Robert D. Willows and Alison M. Kriegel The chlorin reductase has also been expressed in Rba. capsulatus using the same strategy as for DPOR subunit expression. Similar to the DPOR expression the non-S-tagged BchZ copurified with S-tagged BchY on S-agarose in what appeared to be a 1:1 ratio and S-tagged BchX was also purified using S-agarose. Reconstitution of chlorin reductase activity required the S-tagged BchY : BchZ complex, S-tagged BchX, 2 mM ATP, an ATP regeneration system, 5 mM MgCl2, 0.7 mM sodium dithionite (an electron donor), and 2 µM chlorophyllide a. The product generated in this reaction was 3-vinyl bacteriochlorophyllide a (Nomata et al., 2006). Protochlorophyllide was not a substrate for the chlorin reductase. Both of these expression systems will allow further biochemical analysis of the proteins and protein complexes and will facilitate comparative studies of the two enzymes to understand the catalytic mechanism and substrate specificity. F. Esterification to the Propionate on Ring IV Bacteriochlorophyll a synthesis is completed with the esterification of bacteriochlorophyllide a with a geranylgeraniol onto the 17 propanoate moiety. This is then followed by reduction of this esterified geranylgeraniol to phytol. The esterification reaction is catalysed by bacteriochlorophyll synthase. The bacteriochlorophyll synthase gene, bchG, has been cloned and heterologously expressed in E. coli from both Rba. capsulatus (Oster et al., 1997) and Rba. sphaeroides (Addlesee et al., 2000). BchG is orthologous to the plant chlorophyll synthase ChlG but the recombinant bacteriochlorophyll synthase could not utilize chlorophyllide as substrate. Both phytyl-pyrophosphate and geranylgeranyl-pyrophosphate are substrates for bacteriochlorophyll synthases of Rba. capsulatus with phytyl-pyrophosphate being the preferred substrate (Oster et al., 1997). The mechanism of esterification was shown by Emery and Akhtar (1985b) to be one of nucleophilic attack as the oxygens of the carboxyl group are retained in the esterification process. It appears that reduction of geranylgeraniol to phytol may occur either before or after esterification to bacteriochlorophyllide a. BchP was heterologously expressed in E. coli and the expressed product was demonstrated to convert bacteriochlorophyll a esterified with geranylgeraniol into bacteriochlorophyll a esterified with phytol. It is not clear if the enzyme could convert geranylgeranyl-pyrophosphate
Chapter 4
Bacteriochlorophyll Biosynthesis
into phytyl-pyrophosphate (Addlesee and Hunter, 1999). IV. Concluding Remarks From this chapter it is obvious that there are significant gaps in our knowledge on bacteriochlorophyll biosynthesis in purple bacteria. To correct this shortfall in our knowledge of this pathway genes encoding the protoporphyrinogen oxidase and the genes responsible for bacteriochlorophyll b synthesis need to be identified. This should be possible given that the complete genome sequences for a number of purple bacteria are known. In addition a number of biosynthetic steps have not been the subject of any significant biochemical analysis or have been recalcitrant to further analysis. These steps include those catalyzed by magnesium-protoporphyrin IX monomethylester oxidative cyclase, 3-vinyl bacteriochlorophyllide a hydroxylase and 3-hydroxyethyl bacteriochlorophyllide a dehydrogenase. It is hoped that these shortfalls in our knowledge and understanding of this pathway will be addressed in the near future. Acknowledgments This work was made possible thanks to a Macquarie University Research Development Grant to R.D.W. Thanks also to D. W. Bollivar for supplying a preprint of his review. References Abboud MM, Jordan PM and Akhtar M (1974) Biosynthesis of 5-aminolevulinic acid: Involvement of a retention-inversion mechanism. J Chem Soc Chem Comm: 643–644 Adamson HY, Hiller RG and Walmsley J (1997) Protochlorophyllide reduction and greening in angiosperms: An evolutionary perspective. J Photochem Photobiol B 41: 201–221 Addlesee HA and Hunter CN (1999) Physical mapping and functional assignment of the geranylgeranyl-bacteriochlorophyll reductase gene, bchP, of Rhodobacter sphaeroides. J Bacteriol 181: 7248–7255 Addlesee HA, Fiedor L and Hunter CN (2000) Physical mapping of bchG, orf427, and orf177 in the photosynthesis gene cluster of Rhodobacter sphaeroides: Functional assignment of the bacteriochlorophyll synthetase gene. J Bacteriol 182: 3175–3182 Akhtar M and Jordan PM (1968) Mechanism of action of δaminolaevulate synthetase and synthesis of stereospecifically tritiated glycine. Chem Comm: 1691–1692
75 Armstrong GA (1998) Greening in the dark: Light-independent chlorophyll biosynthesis from anoxygenic photosynthetic bacteria to gymnosperms. J Photochem Photobiol 43: 87–100 Battersby AR, Gutman AL, Fookes CJR, Gunther H and Simon H (1981) Stereochemistry of formation of methyl and ethyl groups in bacteriochlorophyll-a. J Chem Soc Chem Comm: 645–647 Bauer CE, Buggy JJ, Yang ZM and Marrs BL (1991) The superoperonal organization of genes for pigment biosynthesis and reaction center proteins is a conserved feature in Rhodobacter capsulatus: Analysis of overlapping bchB and puhA transcripts. Mol Gen Genet 228: 433–444 Béjà O, Suzuki MT, Heidelberg JF, Nelson WC, Preston CM, Hamada T, Eisen JA, Fraser CM and DeLong EF (2002) Unsuspected diversity among marine aerobic anoxygenic phototrophs. Nature 415: 630–633 Berry A, Jordan PM and Seehra JS (1981) The isolation and characterization of catalytically competent porphobilinogen deaminase-intermediate complexes. FEBS Lett 129: 220–224 Biel AJ and Marrs BL (1983) Transcriptional regulation of several genes for bacteriochlorophyll biosynthesis in Rhodopseudomonas capsulata in response to oxygen. J Bacteriol 156: 686–694 Biel SW, Wright MS and Biel AJ (1988) Cloning of the Rhodobacter capsulatus hemA gene. J Bacteriol 170: 4382–4384 Bollivar DW (2003) Intermediate steps in chlorophyll biosynthesis: Methylation and cyclization. In: Kadish KM, Smith K and Guilard R (eds) The Porphyrin Handbook II, Vol 13 pp 49–69. Academic Press, New York Bollivar DW (2006) Recent advances in chlorophyll biosynthesis. Photosynth Res 89: 1–22 Bollivar DW, Jiang ZY, Bauer CE and Beale SI (1994a) Heterologous expression of the bchM gene product from Rhodobacter capsulatus and demonstration that it encodes S-adenosylL-methionine:Mg-protoporphyrin IX methyltransferase. J Bacteriol 176: 5290–5296 Bollivar DW, Suzuki JY, Beatty JT, Dobrowolski JM and Bauer CE (1994b) Directed mutational analysis of bacteriochlorophyll a biosynthesis in Rhodobacter capsulatus. J Mol Biol 237: 622–640 Burke DH, Hearst JE and Sidow A (1993a) Early evolution of photosynthesis: Clues from nitrogenase and chlorophyll iron proteins. Proc Natl Acad Sci USA 90: 7134–7138 Burke DH, Alberti M and Hearst JE (1993b) bchFNBH bacteriochlorophyll synthesis genes of Rhodobacter capsulatus and identification of the third subunit of light-independent protochlorophyllide reductase in bacteria and plants. J Bacteriol 175: 2414–2422 Burke DH, Alberti M and Hearst JE (1993c) The Rhodobacter capsulatus chlorin reductase-encoding locus, bchA, consists of three genes, bchX, bchY, and bchZ. J Bacteriol 175: 2407–2413 Burnham BF and Lascelles J (1963) Control of porphyrin biosynthesis through a negative-feedback mechanism—studies with preparations of δ-aminolaevulate synthetase and δ-aminolaevulate dehydratase from Rhodopseudomonas spheroides. Biochem J 87: 462–472 Burton G, Fagerness PE, Hosozawa S, Jordan PM and Scott AI (1979) C-13 NMR evidence for a new intermediate, pre-uroporphyrinogen, in the enzymic transformation of porphobilinogen into uroporphyrinogen-I and -uroporphyrinogen-III. J Chem Soc Chem Comm: 202–204
76 Chew AGM and Bryant DA (2006) Characterization of a plantlike protochlorophyllide a divinyl reductase in green sulfur bacteria J Biol Chem 282: 2967–2975 Confalonieri F and Duguet M (1995) A 200-amino acid ATPase module in search of a basic function. Bioessays 17: 639–650 Coomber SA and Hunter CN (1989) Construction of a physical map of the 45 kb photosynthetic gene cluster of Rhodobacter sphaeroides. Archives of Microbiology 151: 454–458 Coomber SA, Chaudri M and Hunter CN (1990a) The cloning and organization of genes for bacteriochlorophyll and carotenoid biosynthesis in Rhodobacter sphaeroides. FEMS Symposium 53: 49–56 Coomber SA, Chaudhri M, Connor A, Britton G and Hunter CN (1990b) Localized transposon Tn5 mutagenesis of the photosynthetic gene cluster of Rhodobacter sphaeroides. Mol Microbiol 4: 977–989 Coomber SA, Jones RM, Jordan PM and Hunter CN (1992) A putative anaerobic coproporphyrinogen III oxidase in Rhodobacter sphaeroides. I. Molecular cloning, transposon mutagenesis and sequence analysis of the gene. Mol Microbiol 6: 3159–3169 Deverneuil H, Sassa S and Kappas A (1983) Purification and properties of uroporphyrinogen decarboxylase from humanerythrocytes. A single enzyme catalyzing the 4 sequential decarboxylations of uroporphyrinogen-I and uroporphyrinogen-III. J Biol Chem 258: 2454–2460 Emery VC and Akhtar M (1985a) Stereochemistry of the generation of the ethyl group in bacteriochlorophyll a biosynthesis. J Chem Soc Chem Comm: 1646–1647. Emery VC and Akhtar M (1985b) Mechanistic studies on the phytylation step in bacteriochlorophyll a biosynthesis: An application of the oxygen-18 induced isotope effect in carbon-13 NMR spectroscopy. J Chem Soc Chem Comm: 600–601 Fanica-Gaignier M and Clement-Metral J (1973) 5-Aminolevulinic-acid synthetase of Rhodopseudomonas spheroides Y. Kinetic mechanism and inhibition by ATP. Eur J Biochem 40: 19–24 Ferreira GC and Gong J (1995) 5-Aminolevulinate synthase and the first step of heme-biosynthesis. J Bioenerg Biomem 27: 151–159 Fodje MN, Hansson A, Hansson M, Olsen JG, Gough S, Willows RD and Al-Karadaghi S (2001) Interplay between an AAA module and an integrin I domain may regulate the function of magnesium chelatase. J Mol Biol 311: 111–122 Frankenberg N, Erskine PT, Cooper JB, Shoolingin-Jordan PM, Jahn D and Heinz DW (1999) High resolution crystal structure of a Mg2+-dependent porphobilinogen synthase. J Mol Biol 289: 591–602 Frere F, Schubert WD, Stauffer F, Frankenberg N, Neier R, Jahn D and Heinz DW (2002) Structure of porphobilinogen synthase from Pseudomonas aeruginosa in complex with 5fluorolevulinic acid suggests a double Schiff base mechanism. J Mol Biol 320: 237–247 Fujita Y (1996) Protochlorophyllide reduction: A key step in the greening of plants. Plant Cell Physiol 37: 411–421 Fujita Y and Bauer CE (2000) Reconstitution of light-independent protochlorophyllide reductase from purified BChl and BchN-BchB subunits. In vitro confirmation of nitrogenase-like features of a bacteriochlorophyll biosynthesis enzyme. J Biol Chem 275: 23583–23588 Fujita Y and Bauer C (2003) The light-independent protochlorophyllide reductase: A nitrogenase-like enzyme catalyzing a
Robert D. Willows and Alison M. Kriegel key reaction for greening in the dark. In: Kadish KM, Smith K and Guilard R (eds) The Porphyrin Handbook II, Vol 13, pp. 109–156, Academic Press, New York Gibson KD, Neuberger A and Tait GH (1963) Biosynthesis of porphyrin and bacteriochlorophyll by Rhodopseudomonas spheroides. IV. S-Adenosylmethionine-magnesium protoporphyrin methyltransferase. Biochem J 88: 325–334 Gibson LCD and Hunter CN (1994) The bacteriochlorophyll biosynthesis gene, bchM, of Rhodobacter sphaeroides encodes S-adenosyl-L-methionine: Mg protoporphyrin IX methyltransferase. FEBS Lett 352: 127–130 Gibson LCD, Willows RD, Kannanagara CG, von Wettstein D and Hunter CN (1995) Magnesium-protoporphyrin chelatase of Rhodobacter sphaeroides: Reconstitution of activity by combining the products of the bchH, -I, and -D genes expressed in Escherichia coli. Proc Natl Acad Sci USA. 92: 1941–1944 Gibson LCD, Jensen PE and Hunter CN (1999) Magnesium chelatase from Rhodobacter sphaeroides: initial characterization of the enzyme using purified subunits and evidence for a BchI-BchD complex. Biochem J 337: 243–251 Glockner G, Rosenthal A and Valentin K (2000) The structure and gene repertoire of an ancient red algal plastid genome. J Mol Evol 51: 382–390 Gorchein A (1972) Magnesium protoporphyrin chelatase activity in Rhodopseudomonas spheroides. Studies with whole cells. Biochem J 127: 97–106 Gorchein A (1973) Control of magnesium-protoporphyrin chelatase activity in Rhodopseudomonas spheroides. Role of light, oxygen, and electron and energy transfer. Biochem J 134: 833–845 Gorchein A (1994) Uptake of protoporphyrin and continuous spectrophotometric assay for magnesium chelatase in Rhodobacter spheroides. Biochem J 299: 869–874 Gorchein A (1997) Cell-free activity of magnesium chelatase in Rhodobacter sphaeroides and Rhodobacter capsulatus. Biochem Soc Trans 25: 82S Gorchein A, Gibson LCD and Hunter CN (1993) Gene expression and control of enzymes for synthesis of magnesium protoporphyrin monomethyl ester in Rhodobacter sphaeroides. Biochem Soc Trans 21: 201S Gough SP, Petersen BO and Duus JO (2000) Anaerobic chlorophyll isocyclic ring formation in Rhodobacter capsulatus requires a cobalamin cofactor. Proc Natl Acad Sci USA 97: 6908–6913 Granick S (1948) Protoporphyrin IX as a precursor of chlorophyll. J Biol Chem 172: 717–727 Granick S (1949) The structural and functional relations between heme and chlorophyll. Harvey Lectures Ser 44: 220–245 Granick S (1950) Magnesium vinyl pheoporphyrin a5, another intermediate in the biological synthesis of chlorophyll. J Biol Chem 183: 713–730 Hansson A, Willows RD, Roberts TH and Hansson M (2002) Three semidominant barley mutants with single amino acid substitutions in the smallest magnesium chelatase subunit form defective AAA+ hexamers. Proc Natl Acad Sci USA 99: 13944–13949 Hansson M and Kannangara CG (1997) ATPases and phosphate exchange activities in magnesium chelatase subunits of Rhodobacter sphaeroides. Proc Natl Acad Sci USA 94: 13351–13356 Hearst JE, Alberti M and Doolittle RF (1985) A putative nitrogenase reductase gene found in the nucleotide sequences
Chapter 4
Bacteriochlorophyll Biosynthesis
from the photosynthetic gene cluster of R. capsulata. Cell 40: 219–220 Hendrich W and Bereza B (1993) Spectroscopic characterization of protochlorophyllide and its transformation. Photosynthetica 28: 1–16 Heyes DJ and Hunter CN (2005) Making light work of enzyme catalysis: Protochlorophyllide oxidoreductase. Trends Biochem Sci 30: 642–649 Hinchigeri SB, Nelson DW and Richards WR (1984) The purification and reaction mechanism of S-adenosyl-L-methionine: magnesium protoporphyrin methyltransferase from Rhodopseudomonas sphaeroides. Photosynthetica 18: 168–178 Hinchigeri SB, Hundle B and Richards WR (1997) Demonstration that the BchH protein of Rhodobacter capsulatus activates S-adenosyl-L-methionine:magnesium protoporphyrin IX methyltransferase. FEBS Lett 407: 337–342 Hunter CN and Coomber SA (1988) Cloning and oxygen-regulated expression of the bacteriochlorophyll biosynthesis genes bch E, B, A and C of Rhodobacter sphaeroides. J Gen Microbiol 134: 1491–1497 Jackson AH, Sancovich HA, Ferramola AM, Evans N, Games DE, Matlin SA, Elder GH and Smith SG (1976) Macrocyclic Intermediates in biosynthesis of porphyrins. Phil Trans Royal Soc London B-Biol Sci 273: 191–206 Jacobs NJ and Jacobs JM (1981) Protoporphyrinogen oxidation in Rhodopseudomonas-spheroides, a step in heme and bacteriochlorophyll synthesis. Arch Biochem Biophys 211: 305–311 Jaffe EK (1995) Porphobilinogen synthase, the first source of heme symmetry. J Bioenerg Biomem 27: 169–179 Jensen PE, Gibson LCD, Henningsen KW and Hunter CN (1996) Expression of the chlI, chlD, and chlH genes from the cyanobacterium Synechocystis PCC6803 in Escherichia coli and demonstration that the three cognate proteins are required for magnesium-protoporphyrin chelatase activity. J Biol Chem 271: 16662–16667 Jensen PE, Gibson LCD and Hunter CN (1999) ATPase activity associated with the magnesium-protoporphyrin IX chelatase enzyme of Synechocystis PCC6803: evidence for ATP hydrolysis during Mg2+ insertion, and the MgATP-dependent interaction of the ChlI and ChlD subunits. Biochem J 339: 127–134 Jensen PE, Reid JD and Hunter CN (2000) Modification of cysteine residues in the ChlI and ChlH subunits of magnesium chelatase results in enzyme inactivation. Biochem J 352: 435–441 Jones OTG (1967) Intermediates in chlorophyll biosynthesis in Rhodopseudomonas spheroides. Effects of substrates and inhibitors. Phytochemistry 6: 1355–1362 Jones RM and Jordan PM (1993) Purification and properties of the uroporphyrinogen decarboxylase from Rhodobacter sphaeroides. Biochem J 293: 703–712 Jordan PM and Berry A (1981) Mechanism of action of porphobilinogen deaminase. The participation of stable enzyme substrate covalent intermediates between porphobilinogen and the porphobilinogen deaminase from Rhodopseudomonasspheroides. Biochem J 195: 177–181 Jordan PM and Gibbs PNB (1985) Mechanism of action of 5-aminolaevulinate dehydratase from human-erythrocytes. Biochem J 227: 1015–1020 Jordan PM and Warren MJ (1987) Evidence for a dipyrromethane cofactor at the catalytic site of Escherichia coli porphobilinogen deaminase. FEBS Lett 225: 87–92 Jordan PM and Woodcock SC (1991) Mutagenesis of arginine
77 residues in the catalytic cleft of Escherichia coli porphobilinogen deaminase that affects dipyrromethane cofactor assembly and tetrapyrrole chain initiation and elongation. Biochemical J 280: 445–449 Jordan PM, Burton G, Nordlov H, Schneider MM, Pryde L and Scott AI (1979) Pre-uroporphyrinogen: Substrate for uroporphyrinogen-III cosynthetase. J Chem Soc Chem Comm: 204–205 Kanehisa M (1997) A database for post-genome analysis. Trends Gen. 13: 375–376 Kanehisa M and Goto S (2000) KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acid Res 28: 27–30 Kanehisa M, Goto S, Hattori M, Aoki-Kinoshita KF, Itoh M, Kawashima S, Katayama T, Araki M and Hirakawa M (2006) From genomics to chemical genomics: New developments in KEGG. Nucleic Acid Res 34: D354–D357 Kannangara CG, Vothknecht UC, Hansson M and von Wettstein D (1997) Magnesium chelatase: Association with ribosomes and mutant complementation studies identify barley subunit Xantha-G as a functional counterpart of Rhodobacter subunit BchD. Mol Gen Genet 254: 85–92 Karger GA, Reid JD and Hunter CN (2001) Characterization of the binding of deuteroporphyrin IX to the magnesium chelatase H subunit and spectroscopic properties of the complex. Biochem 40: 9291–9299 Kawanishi S, Seki Y and Sano S (1983) Uroporphyrinogen decarboxylase. Purification, properties, and inhibition by polychlorinated biphenyl isomers. J Biol Chem 258: 4285–4292 Kikuchi G, Kumar A, Talmage P and Shemin D (1958) The enzymatic synthesis of δ-aminolevulinic acid. J Biol Chem 233: 1214–1219 Lander M, Pitt AR, Alefounder PR, Bardy D, Abell C and Battersby AR (1991) Studies on the mechanism of hydroxymethylbilane synthase concerning the role of arginine residues in substrate binding. Biochem J 275: 447–452 Lash TD (1991) Action of uroporphyrinogen decarboxylase on uroporphyrinogen-III: A reassessment of the clockwise decarboxylation hypothesis. Biochemical J 278: 901–903 Layer G, Moser J, Heinz DW, Jahn D and Schubert WD (2003) Crystal structure of coproporphyrinogen III oxidase reveals cofactor geometry of radical SAM enzymes. EMBO J 22: 6214–6224 Layer G, Grage K, Teschner T, Schunemann V, Breckau D, Masoumi A, Jahn M, Heathcote P, Trautwein AX and Jahn D (2005) Radical S-adenosylmethionine enzyme coproporphyrinogen III oxidase HemN: Functional features of the [4Fe-4S] cluster and the two bound S-adenosyl-L-methionines. J Biol Chem 280: 29038–29046 Lebedev N and Timko MP (1998) Protochlorophyllide photoreduction. Photosynth Res 58: 5–23 Louie GV, Brownlie PD, Lambert R, Cooper JB, Blundell TL, Wood SP, Warren MJ, Woodcock SC and Jordan PM (1992) Structure of porphobilinogen deaminase reveals a flexible multidomain polymerase with a single catalytic site. Nature 359: 33–39 Marrs B (1981) Mobilization of the genes for photosynthesis from Rhodopseudomonas capsulata by a promiscuous plasmid. J Bacteriol 146: 1003–1012. Mathews MA, Schubert HL, Whitby FG, Alexander KJ, Schadick K, Bergonia HA, Phillips JD and Hill CP (2001) Crystal structure of human uroporphyrinogen III synthase. EMBO J
78 20: 5832–5839 Martins BM, Grimm B, Mock HP, Huber R and Messerschmidt A (2001) Crystal structure and substrate binding modeling of the uroporphyrinogen-III decarboxylase from Nicotiana tabacum. Implications for the catalytic mechanism. J Biol Chem 276: 44108–44116 McGlynn P and Hunter CN (1993) Genetic analysis of the bchC and bchA genes of Rhodobacter sphaeroides. Mol Gen Genet 236: 227–234 Moseley J, Quinn J, Eriksson M and Merchant S (2000) The Crd1 gene encodes a putative di-iron enzyme required for photosystem I accumulation in copper deficiency and hypoxia in Chlamydomonas reinhardtii. EMBO J 19: 2139–2151 Moseley JL, Page MD, Alder NP, Eriksson M, Quinn J, Soto F, Theg SM, Hippler M and Merchant S (2002) Reciprocal expression of two candidate di-iron enzymes affecting photosystem I and light-harvesting complex accumulation. Plant Cell 14: 673–688 Nandi DL, Baker-Cohen K and Shemin D (1968) δ-aminolevulinic acid dehydratase of Rhodopseudomonas spheroides. I. Isolation and properties. J Biol Chem 243: 1224–1230. Neidle EL and Kaplan S (1993a) Expression of the Rhodobacter sphaeroides hemA and hemT genes, encoding two 5-aminolevulinic acid synthase isozymes. J Bacteriol 175: 2292–2303 Neidle EL and Kaplan S (1993b) 5-Aminolevulinic acid availability and control of spectral complex formation of HemA and HemT mutants of Rhodobacter sphaeroides. J Bacteriol 175: 2304–2313 Nomata J, Swem LR, Bauer CE and Fujita Y (2005) Overexpression and characterization of dark-operative protochlorophyllide reductase from Rhodobacter capsulatus. Biochim Biophys Acta 1708: 229–237 Nomata J, Mizoguchi T, Tamiaki H and Fujita Y (2006) A second nitrogenase-like enzyme for bacteriochlorophyll biosynthesis: Reconstitution of chlorophyllide a reductase with purified Xprotein (BchX) and YZ-protein (BchY-BchZ) from Rhodobacter capsulatus. J Biol Chem 281: 15021–15028 Oh-hama T (1997) Evolutionary consideration on 5-aminolevulinate synthase in nature. Orig Life Evol Biosph 27: 405–412 Oster U, Bauer CE and Rüdiger W (1997) Characterization of chlorophyll a and bacteriochlorophyll a synthases by heterologous expression in Escherichia coli. J Biol Chem 272: 9671–9676 Ouchane S, Steunou A-S, Picaud M and Astier C (2004) Aerobic and anaerobic Mg-protoporphyrin monomethyl ester cyclases in purple bacteria: A strategy adopted to bypass the repressive oxygen control system. J Biol Chem 279: 6385–6394 Panek H and O’Brian MR (2002) A whole genome view of prokaryotic haem biosynthesis. Microbiology 148: 2273–2282 Pinta V, Picaud M, Reiss-Husson F and Astier C (2002) Rubrivivax gelatinosus acsF (previously orf358) codes for a conserved, putative binuclear-iron-cluster-containing protein involved in aerobic oxidative cyclization of Mg-protoporphyrin IX monomethylester. J Bacteriol 184: 746–753 Porra RJ, Schaefer W, Katheder I, Scheer H, Gad’on N and Drews G (1995) Evidence for two different pathways for the formation of isocyclic ring E of bacteriochlorophyll a in Rhodobacter sphaeroides and Roseobacter denitrificans using 18 O-labeling and mass spectrometry. In: Mathis P (ed) Photosynthesis: From Light to Biosphere: Proceedings of the Xth International Photosynthesis Congress, Vol 3, pp 881–885
Robert D. Willows and Alison M. Kriegel Porra RJ, Schaefer W, Gad’on N, Katheder I, Drews G and Scheer H (1996) Origin of the two carbonyl oxygens of bacteriochlorophyll a. Demonstration of two different pathways for the formation of ring E in Rhodobacter sphaeroides and Roseobacter denitrificans, and a common hydratase mechanism for 3-acetyl group formation. Eur J Biochem 239: 85–92 Porra RJ, Urzinger M, Winkler J, Bubenzer C and Scheer H (1998) Biosynthesis of the 3-acetyl and 131-oxo groups of bacteriochlorophyll a in the facultative aerobic bacterium, Rhodovulum sulphidophilum. The presence of both oxygenase and hydratase pathways for isocyclic ring formation. Eur J Biochem 257: 185–191 Reinbothe S and Reinbothe C (1996) The regulation of enzymes involved in chlorophyll biosynthesis. Eur J Biochem 237: 323–343 Reinbothe S, Reinbothe C, Lebedev N and Apel K (1996a) PORA and PORB, two light-dependent protochlorophyllide-reducing enzymes of angiosperm chlorophyll biosynthesis. Plant Cell 8: 763–769 Reinbothe S, Reinbothe C, Apel K and Lebedev N (1996b) Evolution of chlorophyll biosynthesis—The challenge to survive photooxidation. Cell 86: 703–705 Roessner CA, Ponnamperuma K and Scott AI (2002) Mutagenesis identifies a conserved tyrosine residue important for the activity of uroporphyrinogen III synthase from Anacystis nidulans. FEBS Lett 525: 25–28 Rüdiger W (2003) The last steps of chlorophyll biosynthesis. In: Kadish KM, Smith K and Guilard R (eds), The Porphyrin Handbook II, Vol 13, pp 71–108. Academic Press, New York Schoefs B (1999) The light-dependent and light-independent reduction of protochlorophyllide a to chlorophyllide a. Photosynthetica 36: 481–496 Schoefs B (2001) The light-dependent protochlorophyllide reduction: From a photoprotecting mechanism to a metabolic reaction. Rec Res Develop Plant Physiol 2: 241–258 Seehra JS, Jordan PM and Akhtar M (1983) Anaerobic and aerobic coproporphyrinogen-III oxidases of Rhodopseudomonas spheroide. Mechanism and stereochemistry of vinyl group formation. Biochem J 209: 709–718 Shemin D (1976) 5-Aminolaevulinic acid dehydratase: Structure, function, and mechanism. Phil Trans Royal Soc London B-Biol Sci 273: 109–115 Shemin D and Russell CS (1953) δ-aminolevulinic acid, its role in the biosynthesis of porphyrins and purines. J Am Chem Soc 75: 4873–4874 Shepherd M, McLean S and Hunter CN (2005) Kinetic basis for linking the first two enzymes of chlorophyll biosynthesis. FEBS J 272: 4532–4539 Shoolingin-Jordan PM, LeLean JE and Lloyd AJ (1997) Continuous coupled assay for 5-aminolevulinate synthase. Methods Enzymol 281: 309–316 Sirijovski N, Olsson U, Lundqvist J, Al-Karadaghi S, Willows RD and Hansson M (2006) ATPase activity associated with the magnesium chelatase H-subunit of the chlorophyll biosynthetic pathway is an artefact. Biochem J 400: 477–484 Smart JL, Willett JW and Bauer CE (2004) Regulation of hem gene expression in Rhodobacter capsulatus by redox and photosystem regulators RegA, CrtJ, FnrL, and AerR. J Mol Biol 342: 1171–1186 Sofia HJ, Chen G, Hetzler BG, Reyes-Spindola JF and Miller NE (2001) Radical SAM, a novel protein superfamily linking
Chapter 4
Bacteriochlorophyll Biosynthesis
unresolved steps in familiar biosynthetic pathways with radical mechanisms: Functional characterization using new analysis and information visualization methods. Nucleic Acids Res 29: 1097–1106 Straka JG and Kushner JP (1983) Purification and characterization of bovine hepatic uroporphyrinogen decarboxylase. Biochemistry 22: 4664–4672 Suzuki JY and Bauer CE (1995) Altered monovinyl and divinyl protochlorophyllide pools in bchJ mutants of Rhodobacter capsulatus. Possible monovinyl substrate discrimination of light-independent protochlorophyllide reductase. J Biol Chem 270: 3732–3740 Tait GH (1969) Coproporphyrinogenase activity in extracts from Rhodopseudomonas spheroides. Biochem Biophys Res Comm 37: 116–122 Tait GH (1972) Coproporphyrinogenase activities in extracts of Rhodopseudomonas spheroides and Chromatium strain D. Biochem J 128: 1159–1169 Taylor DP, Cohen SN, Clark WG and Marrs BL (1983) Alignment of genetic and restriction maps of the photosynthesis region of the Rhodopseudomonas capsulatus chromosome by a conjugation-mediated marker rescue technique. J Bacteriol 154: 580–590 Tuboi S, Kim HJ and Kikuchi G (1970a) Differential induction of fraction-I and fraction-II of δ-aminolevulinate synthetase in Rhodopseudomonas spheroides under various incubation conditions. Arch Biochem Biophys 138: 155–159 Tuboi S, Kim HJ and Kikuchi G (1970b) Occurrence and properties of 2 types of δ-aminolevulinate synthetase in Rhodopseudomonas spheroides. Arch Biochem Biophys 138: 147–154 Walker CJ and Weinstein JD (1994) The magnesium-insertion step of chlorophyll biosynthesis is a two-stage reaction. Biochem J 299: 277–284 Warren M J, Jordan PM (1988) Investigation into the nature of substrate binding to the dipyrromethane cofactor of Escherichia coli porphobilinogen deaminase. Biochemistry 27: 9020–9030 Wellington CL Taggart, AKP, Beatty, JT (1991) Functional significance of overlapping transcripts of crtEF, bchCA, and puf photosynthesis gene operons in Rhodobacter capsulatus J Bact 173: 2954–2961 Whitby FG, Phillips JD, Kushner JP and Hill CP (1998) Crystal structure of human uroporphyrinogen decarboxylase. EMBO J 17: 2463–2471 Willows RD (2003) Biosynthesis of chlorophylls from protoporphyrin IX. Nat Prod Rep 20: 327–341 Willows RD and Beale SI (1998) Heterologous expression of the
79 Rhodobacter capsulatus BchI, -D, and -H genes that encode magnesium chelatase subunits and characterization of the reconstituted enzyme. J Biol Chem 273: 34206–34213 Willows RD and Hansson M (2003) Mechanism, structure and regulation of magnesium chelatase, In: Kadish KM, Smith K and Guilard R (eds) The Porphyrin Handbook II, Vol 13 pp. 1–48 Academic Press, New York Willows RD, Gibson LCD, Kanangara CG, Hunter CN and von Wettstein D (1996) Three separate proteins constitute the magnesium chelatase of Rhodobacter sphaeroides. Eur J Biochem 235: 438–443 Willows RD, Lake V, Roberts TH and Beale SI (2003) Inactivation of Mg chelatase during transition from anaerobic to aerobic growth in Rhodobacter capsulatus. J Bacteriol 185: 3249–3258 Willows RD, Hansson A, Birch D, Al-Karadaghi S and Hansson M (2004) EM single particle analysis of the ATP-dependent BchI complex of magnesium chelatase: An AAA+ hexamer. J Struct Biol 146: 227–233 Wright MS, Cardin RD and Biel AJ (1987) Isolation and characterization of an aminolevulinate-requiring Rhodobacter capsulatus mutant. J Bacteriol 169: 961–966 Yang ZM and Bauer CE (1990) Rhodobacter capsulatus genes involved in early steps of the bacteriochlorophyll biosynthetic pathway. J Bacteriol 172: 5001–5010 Yen HC and Marrs B (1976) Map of genes for carotenoid and bacteriochlorophyll biosynthesis in Rhodopseudomonas capsulata. J Bacteriol 126: 619–629 Yildiz FH, Gest H and Bauer CE (1992) Conservation of the photosynthesis gene cluster in Rhodospirillum centenum. Mol Microbiol 6: 2683–2691 Young DA, Bauer CE, Williams JC and Marrs BL (1989) Genetic evidence for superoperonal organization of genes for photosynthetic pigments and pigment-binding proteins in Rhodobacter capsulatus. Mol Gen Genet 218: 1–12 Zaman Z, Jordan PM and Akhtar M (1973) Mechanism and stereochemistry of the 5-aminolaevulinate synthetase reaction. Biochem J 135: 257–263 Zeilstra-Ryalls JH and Kaplan S (1995) Aerobic and anaerobic regulation in Rhodobacter sphaeroides 2.4.1: the role of the fnrL gene. J Bacteriol 177: 6422–6431 Zeilstra-Ryalls JH and Schornberg KL (2006) Analysis of hemF gene function and expression in Rhodobacter sphaeroides 2.4.1. J Bacteriol 188: 801–804 Zsebo KM and Hearst JE (1984) Genetic-physical mapping of a photosynthetic gene cluster from R. capsulata. Cell 37: 937–947
Chapter 5 Vitamin B12 (Cobalamin) Biosynthesis in the Purple Bacteria Martin J. Warren* and Evelyne Deery Protein Science Group, Department of Biosciences, University of Kent, Canterbury CT2 7NJ, U.K.
Summary ................................................................................................................................................................. 81 I. Background ...................................................................................................................................................... 81 II. Function of Cobalamin ..................................................................................................................................... 82 III. B12 Biosynthesis in Rhodobacter capsulatus and Rhodobacter sphaeroides ................................................ 83 IV. Summary of Events Required for Cobalamin Biosynthesis .............................................................................. 84 V. Biosynthesis of Precorrin-2 and Its Onward Route Towards Siroheme and Cobalamin .................................. 86 VI. Control and Regulation of Cobalamin Biosynthesis ......................................................................................... 92 Acknowledgments ................................................................................................................................................... 92 References .............................................................................................................................................................. 92
Summary Vitamin B12 (cobalamin), the antipernicious anemia factor, is unique among the vitamins in that its synthesis is restricted to the prokaryotic world. In fact, cobalamin is made along one of two related biosynthetic pathways, which differ in their timing of cobalt insertion and requirement for molecular oxygen. The purple bacteria appear to utilize, exclusively, the pathway requiring oxygen (aerobic route). In this chapter, the aerobic pathway for cobalamin biosynthesis is described. The biosynthetic pathway for cobalamin is shared, in part, with that for other modified tetrapyrroles such as heme, bacteriochlorophyll and siroheme. Cobalamin biosynthesis diverges from heme and chorophyll at uroporphyrinogen III, the first macrocyclic intermediate of the pathway. The transformation of uroporphyrinogen III into cobalamin requires the addition of eight S-adenosyl-L-methionine derived methyl groups, ring contraction, decarboxylation, amidation, cobalt insertion, adenosylation and the attachment of a lower nucleotide loop that houses an unusual base in the form of dimethylbenzimidazole. In total, around thirty enzymes are required for the complete de novo biosynthesis of the coenzyme form of cobalamin, a coenzyme that appears to be required for a number of cellular activities in the purple bacteria including the formation of the active photosynthetic apparatus.
I. Background Vitamin B12 (cobalamin) was first identified as a nutrient in the 1920s when Whipple and Robscheit-Robbins (1925) and Minot and Murphy (1926) demonstrated that crude liver extracts could be used to cure pernicious anemia. The active factor within the liver was purified, crystallized (Rickes et al., 1948; Smith, 1948) and the structure of the vitamin determined by the pioneering efforts of Dorothy Hodgkin (Brink et al., 1954; *Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 81–95. © 2009 Springer Science + Business Media B.V.
82
Martin J. Warren and Evelyne Deery
Fig. 1. The structure of vitamin B12. The diagram outlines the numbering and lettering system used to describe the corrin ring component of the molecule. The various components of the vitamin are described.
Hodgkin et al., 1956), exposing it as a molecule of much greater complexity than had been predicted. The structure revealed that vitamin B12 (see Fig. 1) is a modified tetrapyrrole that belongs to the same family of compounds as heme, bacteriochlorophyll and siroheme (Warren et al., 2002), prosthetic groups that are also made in the purple bacteria. Although it shares the same basic architecture as these compounds, it differs from them in a number of important ways including the identity of the central metal ion (cobalt) and the fact that the molecule contains both upper and lower axial ligands for the metal. Moreover, the tetrapyrrole framework within vitamin B12 has lost one of the bridging carbons, thereby allowing the direct connection of the pyrrole-derived rings D and A (Fig. 1). This cobalt containing, macrocyclic contracted system is called a corrin ring, whereas the corrin ring with its lower axial ligand attached through a linking aminopropanol spacer is called cobalamin. In vivo, cobalamin is generally found in one of two key functional biological forms, methylcobalamin and adenosylcobalamin, which differ in the nature of their upper axial ligand (Fig. 1) (Warren et al., 2002; Banerjee and Ragsdale, 2003). Commercial B12 is generally cyanocobalamin, where the upper Abbreviations: DMB – 5,6-dimethylbenzimidazole; HBA – hydrogenobyrinic acid; P. – Pseudomonas; Rba. – Rhodobacter; SAM – S-adenosyl-L-methionine
axial ligand has been displaced by cyanide, which is done to facilitate the purification of the vitamin. The complexity of the molecule means that a chemical synthesis is not practical and hence cobalamin is produced from bacterial cultures (Martens et al., 2002). However, although cyanocobalamin is not itself functional in biological systems, it is rapidly converted into adenosylcobalamin. Uniquely among the vitamins, B12 is made de novo only by prokaryotes; its synthesis does not seem to have survived the prokaryotic to eukaryotic transition (Roth et al., 1996; Croft et al., 2005). II. Function of Cobalamin Currently, three classes of B12-dependent enzymes are recognized and are termed the isomerases, the methyltransferases and the reductive dehalogenases (Banerjee and Ragsdale, 2003). The latter are generally found in anaerobic microbial systems and are associated with the detoxification of chlorinated carbon compounds. There are no examples of such enzymes in the purple bacteria. However, the purple bacteria do house cobalamin-dependent isomerases and methyltransferases. The isomerases catalyze the 1,2 interchange between a variable substituent and a hydrogen atom on a neighboring carbon (Banerjee and Ragsdale, 2003). The commonest B12 isomerase,
Chapter 5
B12 Biosynthesis
and perhaps the best studied, is methylmalonyl CoA mutase, which uses adenosylcobalamin as a coenzyme and promotes catalysis using radical chemistry (Marsh and Drennan, 2001). The Rhodobacter (Rba.) sphaeroides sequencing project (Mackenzie et al., 2001) has shown that this enzyme is encoded within the genome and is likely to be involved in the degradation of several amino acids and odd-chain fatty acids via propionyl-CoA to the tricarboxylic acid cycle. The B12-dependent methyltransferases are involved in the biosynthesis of the amino acid methionine and, especially in the archaea, in C1 metabolism and CO2 fixation (Banerjee and Ragsdale, 2003). In the purple bacteria, the majority of organisms appear to have a B12-dependent methionine synthase. Here the enzyme catalyses the transfer of a methyl group from a donor (methyl tetrahydrofolate) to homocysteine to give methionine and tetrahydrofolate. With reference to cobalamin in purple bacteria, there has been comparatively little work reported. There has been no study on the B12-requiring enzymes although in Rba. sphaeroides it is known that B12 is required for methionine synthase (Cauthen et al., 1967). However, it has been shown on several occasions that there is a correlation between B12 availability and the ability to form bacteriochlorophyll (Pollich and Klug, 1995; Pollich et al., 1996; Gough et al., 2000; Chapter 4, Willows and Kriegel). The correlation between B12 and bacteriochlorophyll synthesis was first noticed in green sulfur bacteria (Fuhrmann et al., 1993), where cobalamin deficiency not only led to a decrease in the bacteriochlorophyll content but also prevented the formation of chlorosomes. In purple sulfur bacteria, a parallel observation was made with B12-auxotrophs in that a lack of the vitamin led to reduced bacteriochlorophyll synthesis (Pfennig and Lippert 1966). Finally, disruption of the B12 biosynthetic pathway in Rba. capsulatus was found to cause a reduction in the formation of the photosynthetic apparatus under semiaerobic growth conditions (Pollich and Klug, 1995; Pollich et al., 1996). An explanation for the apparent relationship between B12 availability and bacteriochlorophyll synthesis was forwarded when it was suggested that B12 may function as a coenzyme in the isocyclic ring formation of bacteriochlorophyll (Gough et al., 2000). This conclusion was based on several compelling pieces of evidence, including the observation that B12 auxotrophic mutants of Rba. capsulatus, grown without B12, accumulated Mg2+-porphyrins, and that in vitro activity of the cyclase enzyme was dependent
83 upon the addition of B12 (Gough et al., 2000). However, no direct evidence for the presence of B12 in the cyclase has yet been presented. The cyclase belongs to the radical S-adenosyl-L-methionine (SAM) class of enzymes, and although the presence of a Fe4-S4 centre was reported on a recombinant form of this enzyme, there was no report of a bound cobalamin (Ouchane et al., 2004). III. B12 Biosynthesis in Rhodobacter capsulatus and Rhodobacter sphaeroides Research mainly carried out during the 1990s had revealed two pathways for cobalamin biosynthesis which are referred to as the aerobic and the anaerobic routes (Blanche, 1993; Scott, 1994; Raux et al., 1999). The aerobic route is so-called because it has been shown that the pathway requires molecular oxygen as a substrate, whereas the anaerobic pathway has no such requirement and can operate in the presence or absence of oxygen. Although the two pathways are clearly related and share a number of orthologous enzymes, they are nonetheless quite distinct at both a biochemical and a genetic level (Blanche, 1993). The two pathways differ in the timing of cobalt insertion and the method that is employed to contract the ring (Scott 1994; Raux et al., 1999). In the aerobic pathway cobalt is inserted at a comparatively late stage, after the tetrapyrrole-derived ring has undergone its ring contraction. An outline of the aerobic pathway can be viewed in Table 1. Significantly, when bacteria that operate this pathway are grown under cobalt-limiting conditions, cobalt-free corrinoids are found to accumulate (Toohey, 1965; Toohey, 1966; Dresow et al., 1980). In fact, this was first observed in a range of purple bacteria. This, together, with the initial publication of the gene sequences for cobalamin biosynthesis in Rba. capsulatus (Vlcek et al., 1997), which bore most of the genetic hallmarks of the aerobic pathway, supported the idea that the purple bacteria made their cobalamin along the aerobic pathway (Raux et al., 1999). The organization of the cobalamin biosynthetic genes in Rba. capsulatus is presented in Fig. 2. The aerobic pathway was largely elucidated from research undertaken on Pseudomonas (P.) denitrificans (Blanche, 1993; Blanche et al., 1995). In analyzing the biosynthetic pathway that is found in organisms such as Rba. capsulatus and Rba. sphaeroides, much will be drawn and inferred from
84 the earlier work on the aerobic pathway elucidated in P. denitrificans. The genetic similarities of the cobalamin biosynthetic genes/proteins between these three organisms are highlighted in Table 1. The major reason for believing that Rba. capsulatus operates an aerobic cobalamin biosynthetic pathway comes from the presence of cobN, S and T, which encode the subunits of the cobaltochelatase required for insertion of cobalt into hydrogenobyrinic acid (HBA) (Raux et al., 1999). The genome also contains cobF, which encodes a methyltransferase that also only appears to segregate with the aerobic pathway (Rodionov et al., 2003). However, somewhat worryingly, there was no sign of the monooxygenase required for the oxygen-dependent ring contraction process. The absence of this protein from the repertoire of pathway enzymes led to the suggestion that Rba. capsulatus may operate a different way of catalyzing this step (Raux et al., 1999). Indeed, as will be highlighted later, this was found to be the case (McGoldrick et al., 2005). In terms of genetic layout, in Rba. capsulatus 28 cobalamin biosynthetic genes are found at a single locus, making it one of the most complete cobalamin biosynthetic operons found in nature (Vlcek et al., 1997). Indeed, only three genes are located elsewhere in the chromosome and thus the vast majority of the enzymes required for the transformation of uroporphyrinogen III into adenosylcobalamin are sequestered at this region. The genetic organization of this region is highlighted in Fig. 2. In Rba. sphaeroides,
Martin J. Warren and Evelyne Deery the cobalamin biosynthetic genes are somewhat more dispersed. Although there is again only one major operon, it contains a mere 12 genes. The remaining B12 biosynthetic genes are found dispersed throughout the genome. A comparison of the organization of the cobalamin biosynthetic genes in Rba. capsulatus and Rba. sphaeroides is shown in Fig. 2. IV. Summary of Events Required for Cobalamin Biosynthesis The biosynthesis of adenosylcobalamin represents one of the most intricate metabolic processes undertaken by living systems, involving a multifarious cocktail of enzyme activities and cofactors (Raux et al., 2000). The synthesis is initiated by the construction of the tetrapyrrole framework, which then undergoes a range of modifications that includes peripheral methylation by the addition of 8 SAMderived methyl groups, ring contraction and decarboxylation to generate a metal-free corrinoid, HBA (Warren et al., 2002). The subsequent steps see a combination of cobalt insertion, lateral amidations and the cosseting of the cobalt ion by the attachment of the upper axial ligand in the form of adenosine. In this way the synthesis of the corrin ring is completed, and sets up the synthetic finale whereby the lower nucleotide assembly is secured on to the macrocyclic framework. This involves the piecing together of the
Fig. 2. Genetic organization of B12 operons in Rba. sphaeroides and Rba. capsulatus. The organization of the known cobalamin biosynthetic genes is outlined.
Chapter 5
B12 Biosynthesis
85
Table 1. Comparison of genes and their functions associated with cobalamin biosynthesis Reaction
Pathway Intermediate
methylation at C2 and C7
⏐ ↓
methylation at C20
⏐ ↓
C20 hydroxylation γ lactone formation
⏐ ↓
methylation at C17 ring contraction
⏐ ↓
methylation at C11
↓
Enzyme encoded in P. denitrificans
Enzyme encoded in Rba. capsulatus
Enzyme encoded in Rba. sphaeroides
CobA
CobA
CobA
CobI
CobI
CobI
CobG
CobZ
?
CobJ
CobJ
CobJ
CobM
CobM
CobM
CobF
CobF
CobF
CobK
CobK
CobK
uroporphyrinogen III
precorrin-2
precorrin-3A
precorrin-3B
precorrin-4 precorrin-5 methylation at C1 acetic acid loss
⏐ ↓
C18-C19 reduction
↓
precorrin-6A precorrin-6B methylation at C5
↓
CobL
CobL
CobL
methylation at C15 and decarboxylation
⏐ ↓
CobL
CobL
CobL
methyl rearrangement C11 to C12
⏐ ↓
CobH
CobH
CobH
a, c- amidation
↓
CobB
CobB
CobB
CobN, S & T
CobN, S & T
CobN, S & T
CobR
CobR
CobR
CobO
CobO
CobO
CobQ
CobQ
CobQ
CobD, C & protein α
CobD, C
CobD, C
CobP
CobP
CobP
?
BluB
BluB
CobU
CobU
CobU
CobV
CobV
CobV
precorrin-8x
hydrogenobyrinic acid hydrogenobyrinic acid a,c-diamide cobalt insertion
↓
cob(II)yrinic a,c-diamide cobalt reduction
↓
adenosylation
↓
cob(I)yrinic a,c-diamide ado-cob(I)yrinic a,c-diamide b, d, e, g- amidation
⏐ ↓
aminopropanol phosphate synthesis and attachment
⏐ ↓
phosphorylation & GMP addition
⏐ ↓
ado-cobyric acid
ado-cobinamide
ado-GDP-cobinamide Synthesis of dimethylbenzimidazole α-ribazole 5´ phosphate synthesis & dephosphorylation α-ribazole attachmnent
⏐ ⏐ ↓ ↓
ado-cobalamin
86 components of the lower nucleotide, which is then tethered to the corrin via a carbon-based linker to yield adenosylcobalamin (Fig. 1). V. Biosynthesis of Precorrin-2 and Its Onward Route Towards Siroheme and Cobalamin As with all modified tetrapyrroles, cobalamin is synthesized from uroporphyrinogen III (Warren et al., 2002; Chapter 4, Willows and Kriegel). In fact, the branch of the pathway responsible for cobalamin biosynthesis diverges from the branches for heme and chlorophyll at this intermediate (Fig. 3). Thus, for cobalamin biosynthesis, uroporphyrinogen III is methylated by an enzyme that transfers two SAMderived methyl groups to positions 2 and 7 of the framework to generate precorrin-2 (Fig. 3). This enzyme, termed the S-adenosyl-L-methionine uroporphyrinogen III methyltransferase, is encoded by cobA and is unique in its ability as a methyltransferase to mediate two distinct methylations within a single active site (Blanche et al., 1989). The structure of CobA has recently been determined and some of the amino acids that are key to orchestrating the temporal and regiospecificity of the enzyme have been identified (Vevodova et al., 2004). The topology of the enzyme is similar to five of the other methyltransferases (CobI, J, M, F and L) (Schubert et al., 2003) that are required to decorate this molecular masterpiece. These enzymes have clearly evolved from a common
Martin J. Warren and Evelyne Deery ancestor (Roth et al., 1993), and this relationship has important implications for understanding how the pathway may have appeared since it suggests a model of retrograde evolution (Horowitz, 1945). The synthesis of the yellow precorrin-2 brings the pathway to another junction, where the pathways for siroheme and cobalamin diverge (Fig. 3). The removal of two protons and two electrons converts precorrin-2 into sirohydrochlorin and insertion of ferrous iron yields siroheme (Fig. 3) (Raux et al., 2003). There is evidence that since Rba. capsulatus seems to harbor an assimilatory NADH-nitrite reductase, which requires siroheme as a cofactor, that the organism must be able to make this prosthetic group (Olmo-Mira et al., 2006). Moreover, in the Rba. sphaeroides genome there is also a gene that encodes a protein with a high degree of similarity to SirB, the sirohydrochlorin ferrochelatase that is required in the final step of siroheme synthesis (Raux et al., 2003). However, to direct precorrin-2 towards cobalamin synthesis, the intermediate undergoes another methylation to generate a trimethylpyrrocorphin (Fig. 4). This next enzyme in the pathway methylates precorrin-2 at position C20 to produce precorrin 3A in a reaction catalysed by CobI which, like all the cobalamin biosynthetic methyltransferases, utilizes SAM as the methyl donor (Fig. 4) (Thibaut et al., 1990a). In P. denitrificans, precorrin-3A acts as the substrate for CobG, a monooxygenase that hydroxylates precorrin-3A at C20 and stimulates gamma lactone formation on ring A (Fig. 4) (Debussche et al., 1993; Min, 1993; Spencer et al., 1994). This
Fig. 3. Transformation of uroporphyrinogen III into precorrin-2 and siroheme. Uroporphyrinogen III represents the first branchpoint in the synthesis of modified tetrapyrroles, as methylation at positions 2 and 7 produce precorrin-2 and directs the intermediate away from the branch that leads to heme and chlorophyll. Oxidation of precorrin-2 to sirohydrochlorin followed by ferrochelation generates siroheme. Although the genome of Rba. sphaeroides contains a sirohydrochlorin ferrochelatase (SirB), there is no evidence of a recognized precorrin dehydrogenase.
Chapter 5
B12 Biosynthesis
spring loads the molecule in anticipation of ring contraction. However, as mentioned earlier, there is no homolog of CobG in the genome of Rba. capsulatus (Fig. 2 and Table 1) (Vlcek et al., 1997; Raux et al., 1999). Nonetheless, the main cob operon does contain another open reading frame that is not found in other cobalamin biosynthetic operons. This gene is predicted to encode a 97 kDa protein, the N-terminus of which displays some similarity to flavoproteins such as succinate dehydrogenase and TcuA, whereas the C-terminus has similarity to a protein originally called CitB but which is now known as TcuB (Lewis et al., 2004). Moreover, the C-terminal region of the protein was predicted to contain a number of transmembrane regions, suggesting that it is an integral membrane protein (McGoldrick et al., 2005). In order to investigate whether this protein is involved in cobalamin biosynthesis, an artificial operon was constructed in which all the genes encoding the enzymes for HBA synthesis except for CobG were cloned into a suitable plasmid. When cobG was added to this plasmid, the resulting construct when transformed into E. coli endowed the host bacterium with the ability to make HBA, a compound that E. coli cannot normally make. When cobG was deleted from the plasmid and replaced with the unknown open reading frame from Rba. capsulatus, E. coli strains transformed with the new plasmid were found to also make HBA. The conclusion from these experiments was that the open reading frame within the main Rba. capsulatus B12 biosynthetic operon must encode an enzyme that is isofunctional with CobG — that is it must be able to transform precorrin-3A into precorrin-3B (McGoldrick et al., 2005). As the gene encodes a new cobalamin biosynthetic enzyme it was designated cobZ. To understand how this enzyme works, CobZ was overproduced recombinantly in E. coli. Upon short induction in E. coli a protein of 97 kDa was observed in cell extracts, but with longer induction times this protein appeared to break down and a soluble peptide of approximately 50 kDa was observed. This peptide was purified and was shown to contain the N-terminal region of CobZ. Moreover, the purified protein was yellow indicating that it contained a flavin, which was characterized as FAD. The full-length CobZ (97 kDa protein) was also purified after detergent extraction. The appearance of the purified protein here was somewhat surprising as, rather than yellow, it had an orange coloration and presented with a complex spectrum that included a large Soret peak. The Soret
87 peak could be reduced with dithionite indicating that the protein must contain a heme. Moreover, the protein sequence also suggested that the protein may contain several Fe-S centers. EPR analysis of the protein revealed that the protein did indeed contain two Fe4-S4 centers. Based on the characterization of the protein, a putative mechanism for the enzyme was proposed. In this scheme, CobZ is deemed to act as a monooxygenase. The reduced flavin, housed in the N-terminal region of the protein, reacts with molecular oxygen to form a 4-hydroxyperoxide intermediate. This is then able to methylate the incoming precorrin-3A at the C20 position and generates precorrin-3B. To regenerate the flavin cofactor for the next round of catalysis it is proposed that it is reduced by the transfer of electrons via the heme and Fe-S centers, possibly by drawing on reducing equivalents from the quinol pool in the membrane. In Rhodospirillum rubrum, cobZ is found as two separate genes, suggesting that in Rba. capsulatus CobZ may have arisen through gene fusion (McGoldrick et al., 2005). It is interesting to note that with CobZ and CobG, two quite distinct enzymes have evolved that play an essential role in the ring contraction process, highlighting how the B12 pathway has also employed a patchwork model of evolution (Bartlett et al., 2003). However, to add interest to intrigue, it is noteworthy that the genome of Rba. sphaeroides does not contain a homolog to either of these genes (Table 1 and Fig. 2), so presumably Rba. sphaeroides must employ yet another enzyme system to help promote ring contraction. The ring contraction process is completed by the action of CobJ, a methyltransferase that elicits the macrocyclic ring shrinkage (Fig. 4) (Scott et al., 1993; Thibaut, 1993). The contraction process results in the extrusion of the C20 adduct and the addition of another SAM-derived methyl group to the porphin framework, this time at C17. The product of the reaction is called precorrin-4. The next two enzymes in the pathway see further SAM-derived methyl groups added at C11 and C1, by CobM and F, respectively (Fig. 4) (Thibaut et al., 1990b; Thibaut 1993). Strangely, a look at the final cobalamin molecules reveals that there is no methyl group at C11. This is because later on in the pathway the methyl group at C11 is moved to C12. The methyltransferase that methylates at C1 is also thought to be involved in the final elimination of the C20 adduct through the deacylation of the C1 position. The actions of CobM and F see the transformation of precorrin-4 into precorrin-5 and finally onto precorrin-6A (Fig. 4).
88
Martin J. Warren and Evelyne Deery
Fig. 4. The transformation of precorrin-2 into hydrogenobyrinic acid. This transformation requires the addition of a further 6 methyl groups, ring contraction, deacylation and decarboxylation.
At this stage in the reaction sequence, the pathway intermediate is reduced by the action of CobK, an enzyme that utilizes NADPH as its source of reducing equivalents, resulting in the synthesis of precorrin 6B (Fig. 4) (Blanche et al., 1992c). The final two methyl groups are then added to the emerging corrin ring by
the action of CobL, a multifunctional enzyme that not only methylates both C5 and C15, but also decarboxylates the acetate side chain attached to C12 (Fig. 4) (Blanche et al., 1992a). In fact, it would appear from sequence analysis that CobL is really a fusion between two distinct methyltransferases, which in some organ-
Chapter 5
B12 Biosynthesis
89
Fig. 5. Synthesis of adenosylcobyric acid from hydrogenobyrinic acid. These series of reactions see the amidation of six of the carboxylic acid side chains, cobalt insertion and the adenosylation of the cobalt ion.
isms are encoded by two separate genes (Roth et al., 1993). Thus, the N-terminal region of CobL houses a canonical cobalamin biosynthetic methyltransferase, which is thought to methylate C5, whereas the C-terminal region of CobL is thought to possess a different class of methyltransferase that methylates C15 and is also responsible for the decarboxylation (Roessner and Scott, 2006; Santander et al., 2006). These deductions have actually come from work on the anaerobic pathway, where the ortholog of CobL is found as two distinct proteins termed CbiE and T. On the basis of some biosynthetic studies with CbiT, it seems likely that methylation precedes decarboxylation. The net result of the actions of CobL is the synthesis of precorrin-8 (Fig. 4). Finally, the basic corrin carbon scaffold is completed by the action of CobH which catalyses the rearrangement of the methyl group attached to C11 to C12, resulting in the synthesis of HBA (Fig. 4) (Thibaut et al., 1992; Shipman et al., 2001). A crystal structure of
this enzyme with its bound product has provided an insight into how a conserved histidine residue helps to mediate this 1,5 sigmatropic methyl shift (Shipman et al., 2001). The next steps in the biosynthesis of the corrin ring component of the macrocycle see modification to the acidic side chains and the insertion of cobalt into the core of the molecule (Fig. 5). Initially, HBA is amidated by an enzyme (CobB) that amidates the carboxylic acid groups on the a and c acetate side chains (Debussche et al., 1990). This enzyme amidates the c side chain followed by the a side chain in an ATP-dependent manner, with the amido donor being glutamine. The result of this reaction is hydrogenobyrinic acid a,c-diamide (Fig. 5), which acts as the substrate for the cobaltochelatase. The cobalt-inserting step is catalysed in a reaction that requires three subunits, CobN, S and T (Debussche et al., 1992). The three subunits, Co2+ and Mg-ATP are all required for cobalt insertion into hydrogenobyrinic
90 acid a,c-diamide, although the enzyme complex will also insert cobalt into a range of derivatives. For cobaltochelatase activity, CobS (molecular mass 38 kDa) and T (molecular mass 80 kDa) are thought to form a complex with a combined mass of 450 kDa, and this complex interacts with CobN (molecular mass of 140 kDa) in the presence of ATP to generate the product cob(II)yrinic acid a,c-diamide (Fig. 5). The cobalt chelatase displays significant similarities to the Mg2+ chelatase of bacteriochlorophyll biosynthesis, where the CobN, S and T subunits are homologous to the BchH, I and D subunits (Gibson et al., 1995). After the insertion of cobalt, the next step is to provide an upper ligand for the metal ion. This is accomplished by adenosylating the complex. Initially this requires the reduction of the cobalt ion to the Co(I) form, which acts as a powerful nucleophile to attack the C5 position on the ribose of ATP, with the release of triphosphate, to give the adenosyl form. The reduction of the cobalt ion is catalysed by an enzyme called the cob(II)yrinic acid a,c-diamide reductase, an enzyme that was purified from crude cell extracts (Blanche et al., 1992b). The enzyme was shown to be a flavoprotein and was able to catalyze the reduction of cobalt (II) to cobalt (I) (Fig. 5). However,
Martin J. Warren and Evelyne Deery although the enzyme was purified to homogeneity and its N-terminus sequenced, no gene sequence encoding the enzyme was identified. Recently, this gene has been identified and termed cobR (Warren, unpublished). Once reduced, the Co(I) form is acted upon by the adenosyltransferase, encoded by cobO, in the presence of ATP to give the adenosylcobyrinic acid a,c-diamide (Fig. 5) (Debussche et al., 1991). The remaining 4 amidations are undertaken by CobQ to generate adenosylcobyric acid (Fig. 5) (Blanche et al., 1991a). As an amidase, CobQ is related to CobB and utilizes glutamine as the amide donor. The reactions are carried out in a stepwise fashion, resulting in the amidation of carboxyl groups b, d, e, and g, although the exact order was not determined. Adenosylcobyric acid is converted into adenosylcobinamide (Fig. 6) by the attachment of an aminopropanol linking group to the f side chain of the corrin (Debussche et al., 1993; Blanche et al., 1995). In fact, this side chain is probably made from threonine O-3-phosphate by the action of CobC which, on the basis of the orthologous enzyme from the anaerobic pathway, is likely to act as a pyrixodal phosphatedependent decarboxylase that generates (R)-1-aminopropan-2-ol O-phosphate (Fig. 6) (Brushaber et al.,
Fig. 6. Adenosylcobinamide phosphate synthesis. The figure describes the synthesis and attachment of the aminopropanol linker group.
Chapter 5
B12 Biosynthesis
1998). This linker group is subsequently attached to the corrin by the action of CobD and an unknown 38 kDa protein that was termed protein α (Blanche et al., 1995). The reaction is dependent upon ATP/Mg2+ and the enzyme only recognizes the adenosylated form of cobyric acid, generating adenosylcobinamide phosphate (Fig. 6). The final stages of the synthesis of the coenzyme form of cobalamin require the synthesis and attachment of the lower nucleotide loop (Fig. 7). Initially, a transfer of GMP to adenosylcobinamide phosphate is mediated by CobP, which utilizes GTP as a substrate (Blanche et al., 1991b). The product of the reaction is adenosyl-GDP-cobinamide (Fig. 7). The lower axial ligand is an unusual nucleoside, which is sometimes referred to as the α-ribazole, containing a modified base called 5,6-dimethylbenzimadazole linked by an N-α-glycosidic bond to the ribose sugar. The modified base in the aerobic pathway is made from reduced flavin mononucleotide, in a remarkable reaction catalysed by a single enzyme called BluB (Fig. 7) (Campbell et al., 2006). The gene encoding this enzyme was first identified in
91 Rba. capsulatus, where it was implicated in the later stages of cobalamin biosynthesis (Pollich and Klug, 1995). Further work on a Sinorhizobium meliloti bluB mutant revealed that mutant strains were unable to grow in minimal media and that the cobalamin deficient defects could be rescued by the addition of the modified base 5,6-dimethylbenzimidazole (DMB) (Campbell et al., 2006). Biochemical analysis demonstrated that the bluB mutant does not produce cobalamin unless DMB is supplied. The α-ribazole is made as α-ribazole 5´-phosphate from the phosphoribosyltransferase activity of CobU that utilizes DMB and β-nicotinate mononucleotide as substrates (Fig. 7) (Cameron et al., 1991; Blanche et al., 1995; Warren et al., 2002). The final step in the synthesis then sees the transfer of adenosylcobinamide phosphate of adenosyl-GDP-cobinamide to the α-ribazole 5´-phosphate in a reaction catalysed by CobV, to generate adenosylcobalamin (Fig. 7) (Cameron et al., 1991; Blanche et al., 1995; Warren et al., 2002). Within this sequence of events there are still some genes that are thought to encode cobalamin
Fig. 7. The synthesis of adenosylcobalamin. The final stages of the synthesis of adenosylcobalamin are described, outlining the construction of the lower nucleotide loop and its attachment to the corrin ring component.
92 biosynthetic proteins but for which no role has been ascribed. These include cobW and cobE, in which mutations lead to reduced B12 levels (Blanche et al., 1995). There is some evidence that CobW may be involved in cobalt delivery as the protein contains a histidine rich region that would be capable of binding metal and, moreover, the protein is often found closely associated with the gene encoding the large cobalt chelatase subunit (CobN) (see Fig. 2). CobE may have a function in helping to open the γ-lactone and eliminating the C2 unit that is extruded during the ring contraction process (Fig. 4) as the protein shares some similarity with CbiG for which equivalent functions have been assigned (Roessner and Scott, 2006). On top of these are also some proteins that have been identified as being involved in B12 synthesis in Rba. capsulatus that have no ortholog in P. denitrificans. These include BluE and BluF (Fig. 2), where mutations in the genes produce a B12 phenotype (Pollich and Klug, 1995; Pollich et al., 1996). In fact, the mutations lead to a reduction in the formation of the photosynthetic apparatus under semiaerobic growth conditions, preventing them from ‘blushing’ (hence the prefix blu) (Pollich and Klug, 1995). The phenotype could be corrected by the addition of cobalamin, presumably reflecting the apparent need for cobalamin in bacteriochlorophyll synthesis (Gough et al., 2000). However, disruption of bluE also results in reduced expression of the puf and puc operons, which encode apoproteins of the photosynthetic apparatus (Pollich et al., 1993; Pollich and Klug, 1995). In some non-phototrophic bacteria, there is evidence that a light regulated transcription regulator may bind B12, and regulate carotenogenesis (Cervantes and Murillo, 2002). There is always the possibility that something similar happens in the purple bacteria although there is no evidence that this is the case. More detailed analyses of the bluE and bluF mutations have suggested that both their gene products are involved in the transformation of cobinamide into adenosylcobalamin as both phenotypes can be corrected by the addition of cobinamide to the culture medium (Pollich and Klug 1995; Pollich et al., 1996). However, their precise roles in cobalamin biosynthesis are not known. In the cobalamin biosynthetic operon in both Rba. sphaeroides and Rba. capsulatus a ferredoxin (fer in Fig. 2) is also found, but again no function for this protein has been described.
Martin J. Warren and Evelyne Deery VI. Control and Regulation of Cobalamin Biosynthesis There are very few B12 auxotrophs known for the purple bacteria (Pollich and Klug, 1995), suggesting that most are able to make their own cobalamin. The essential nature of the vitamin, especially for methionine synthesis and its likely involvement in bacteriochlorophyll synthesis, acts as a strong selection pressure to maintain the de novo synthesis within these bacteria. Moreover, as bacteriochlorophyll, heme and cobalamin are all made along a branched biosynthetic pathway, and as cobalamin is required for bacteriochlorophyll synthesis and heme required for cobalamin synthesis (Gough et al., 2000; McGoldrick et al., 2005), there must be regulatory mechanisms in place to ensure that these modified tetrapyrroles are produced to the correct level. The syntheses of all these compounds requires SAM as a substrate. As SAM is made from methionine in a cobalamindependent process, it is likely that there will be a larger global level of regulation involving cobalamin. However, there have been no studies on the regulation of cobalamin made along the aerobic pathways as such studies have been exclusively undertaken on the anaerobic pathway. Nonetheless, bioinformatic analysis of the Rba. capsulatus and Rba. sphaeroides genomes has revealed the presence of B12 elements, which are thought to repress gene expression by direct binding of adenosylcobalamin most likely by interfering with translation (Vitreschak et al., 2003). More research is required to understand this process in detail and to obtain a clearer overall view of tetrapyrrole metabolism in these organisms. Acknowledgments We thank the Biotechnology and Biological Sciences Research Council (BBSRC) of the UK and the European Union Viteomics Research Training Network for funding. References Banerjee R and Ragsdale SW (2003) The many faces of vitamin B12: Catalysis by cobalamin-dependent enzymes. Annu Rev Biochem 72: 209–247 Bartlett GJ, Borkakoti N and Thornton, JM (2003) Catalyzing new reactions during evolution: Economy of residues and
Chapter 5
B12 Biosynthesis
mechanism. J Mol Biol 331: 829–860 Blanche F, Debussche L, Thibaut D, Crouzet J and Cameron B (1989) Purification and characterization of S-adenosyl-Lmethionine: Uroporphyrinogen III methyltransferase from Pseudomonas denitrificans. J Bacteriol 171: 4222–4231 Blanche F, Couder M, Debussche L, Thibaut D, Cameron B and Crouzet J (1991a) Biosynthesis of vitamin B12: Stepwise amidation of carboxyl groups b, d, e, and g of cobyrinic acid a,c-diamide is catalyzed by one enzyme in Pseudomonas denitrificans. J Bacteriol 173: 6046–6051 Blanche F, Debussche L, Famechon A, Thibaut D, Cameron B and Crouzet J (1991b) A bifunctional protein from Pseudomonas denitrificans carries cobinamide kinase and cobinamide phosphate guanylyltransferase activities. J Bacteriol 173: 6052–6057 Blanche F, Famechon A, Thibaut D, Debussche L, Cameron B and Crouzet, J (1992a) Biosynthesis of vitamin B12 in Pseudomonas denitrificans: The biosynthetic sequence from precorrin-6y to precorrin-8x is catalyzed by the cobL gene product. J Bacteriol 174: 1050–1052 Blanche F, Maton L, Debussche L and Thibaut, D (1992b) Purification and characterization of cob(II)yrinic acid a,c-diamide reductase from Pseudomonas denitrificans. J Bacteriol 174: 7452–7454 Blanche F, Thibaut D, Famechon A, Debussche L, Cameron B and Crouzet J (1992c) Precorrin-6x reductase from Pseudomonas denitrificans: Purification and characterization of the enzyme and identification of the structural gene. J Bacteriol 174: 1036–1042 Blanche F, Thibaut D, Debussche L, Hertle R, Zipfel F and Müller G (1993) Parallels and decisive differences in vitamin B12 biosyntheses. Angew Chem Int Ed Engl 32: 1651–1653 Blanche F, Cameron B, Crouzet J, Debussche L, Thibaut D, Vuilhorgne M, Leeper FJ and Battersby AR (1995). Vitamin B12: How the problem of its biosynthesis was solved. Angew Chem Int Ed Engl 34: 383–411 Brink C, Hodgkin DC, Lindsey J, Pickworth J, Robertson JR and White JG (1954) X-ray crystallographic evidence on the structure of vitamin B12. Nature 174: 1169–1171 Brushaber KR, O’Toole GA and Escalante-Semerena JC (1998) CobD, a novel enzyme with L-threonine-o-3-phosphate decarboxylase activity is responsible for the synthesis of (R)-1amino-propanol o-2-phosphate, a proposed new intermediate in cobalamin biosynthesis in Salmonella typhimurium LT2. J Biol Chem 273: 2684–2691 Cameron B, Blanche F, Rouyez MC, Bisch D, Famechon A, Couder M, Cauchois L, Thibaut D, Debussche L and Crouzet J (1991) Genetic analysis, nucleotide sequence, and products of two Pseudomonas denitrificans cob genes encoding nicotinate-nucleotide: Dimethylbenzimidazole phosphoribosyltransferase and cobalamin (5´-phosphate) synthase. J Bacteriol 17: 6066–6073 Campbell GR, Taga ME, Mistry K, Lloret J, Anderson PJ, Roth JR and Walker GC (2006) Sinorhizobium meliloti bluB is necessary for production of 5,6-dimethylbenzimidazole, the lower ligand of B12. Proc Natl Acad Sci U S A 103: 4634–4639 Cauthen SE, Pattison JR and Lascelles J (1967) Vitamin B12 in photosynthetic bacteria and methionine synthesis by Rhodopseudomonas spheroides. Biochem J 102: 774–781 Cervantes M and Murillo FJ (2002) Role for vitamin B12 in light induction of gene expression in the bacterium Myxococcus
93 xanthus. J Bacteriol 184: 2215–2224 Croft MT, Lawrence AD, Raux-Deery E, Warren MJ and Smith AG (2005) Algae acquire vitamin B12 through a symbiotic relationship with bacteria. Nature 438: 90–93 Debussche L, Thibaut D, Cameron B, Crouzet J and Blanche F (1990) Purification and characterization of cobyrinic acid a,cdiamide synthase from Pseudomonas denitrificans. J Bacteriol 172: 6239–6244 Debussche L, Couder M, Thibaut D, Cameron B, Crouzet J and Blanche F (1991) Purification and partial characterization of cob(I)alamin adenosyltransferase from Pseudomonas denitrificans. J Bacteriol 173: 6300–6302 Debussche L, Couder M, Thibaut D, Cameron B, Crouzet J and Blanche F (1992) Assay, purification, and characterization of cobaltochelatase, a unique complex enzyme catalyzing cobalt insertion in hydrogenobyrinic acid a,c-diamide during coenzyme B12 biosynthesis in Pseudomonas denitrificans. J Bacteriol 174: 7445–7451 Debussche L, Thibaut D, Cameron B, Crouzet J and Blanche F (1993) Biosynthesis of the corrin macrocycle of coenzyme B12 in Pseudomonas denitrificans. J Bacteriol 175: 7430–7440 Dresow B, Schlingmann G, Ernst L and Koppenhagen VB (1980) Extracellular metal-free corrinoids from Rhodopseudomonas spheroides. J Biol Chem 255: 7637–7644 Fuhrmann S, Overmann J, Pfennig N and Fischer U (1993) Influence of vitamin B12 and light on the formation of chlorosomes in green- and brown-coloured chlorobium species. Arch Microbiol 160: 193–198 Gibson LCD, Willows RD, Kannangara CG, von Wettstein D and Hunter CN (1995) Magnesium-protoporphyrin chelatase of Rhodobacter sphaeroides: Reconstitution of activity by combining the products of the bchH, -I, and -D genes expressed in Escherichia coli. Proc Natl Acad Sci U S A 92: 1941–1944 Gough SP, Petersen BO and Duus JO (2000) Anaerobic chlorophyll isocyclic ring formation in Rhodobacter capsulatus requires a cobalamin cofactor. Proc Natl Acad Sci USA 97: 6908–6913 Hodgkin DC, Kamper J, Mackay M, Pickworth J, Trueblood KN and White JG (1956) Structure of vitamin B12. Nature 178: 64–66 Horowitz NH (1945) On the evolution of biochemical syntheses. Proc Natl Acad Sci USA 31: 153–157 Lewis JA, Horswill AR, Schwem BE and Escalante-Semerena JC (2004) The tricarballylate utilization (tcuRABC) genes of Salmonella enterica serovar Typhimurium LT2. J Bacteriol 186: 1629–1637 Mackenzie C, Choudary M, Larimer FW, Predke PF, Stilwagen S, Armitage JP, Barber RD, Donohue TJ, Hosler JP, Newman JE, Shapleigh JP, Sockett RE, Zeilstra-Ryalls J and Kaplan S (2001) The home stretch, a first analysis of the nearly completed genome of Rhodobacter sphaeroides 2.4.1. Photosynth Res 70:19–41 Marsh EN and Drennan CL (2001) Adenosylcobalamin-dependent isomerases: New insights into structure and mechanism. Curr Opin Chem Biol 5: 499–505 Martens JH, Barg H, Warren MJ and Jahn D (2002) Microbial production of vitamin B12. Appl Microbiol Biotechnol 58: 275–285 McGoldrick HM, Roessner CA, Raux E, Lawrence AD, McLean KJ, Munro AW, Santabarbara S, Rigby SE, Heathcote P, Scott AI and Warren MJ (2005) Identification and characterization of
94 a novel vitamin B12 (cobalamin) biosynthetic enzyme (CobZ) from Rhodobacter capsulatus, containing flavin, heme, and Fe-S cofactors. J Biol Chem 280: 1086–1094 Min C, Atshaves BA, Roessner CA, Stolowich NJ, Spencer JB and Scott AI (1993) Isolation, structure and genetically engineered synthesis of precorrin-5. J Am Chem Soc 115: 10380–10381 Minot GR and Murphy WP (1926) Treatment of pernicious anemia by a special diet. Journal of the American Medical Association 87: 470–476 Olmo-Mira MF, Cabello P, Pino C, Martinez-Luque M, Richardson DJ, Castillo F, Roldan MD and Moreno-Vivian C (2006) Expression and characterization of the assimilatory NADHnitrite reductase from the phototrophic bacterium Rhodobacter capsulatus E1F1. Arch Microbiol 186: 339–344 Ouchane S, Steunou AS, Picaud M and Astier C (2004) Aerobic and anaerobic Mg-protoporphyrin monomethyl ester cyclases in purple bacteria: A strategy adopted to bypass the repressive oxygen control system. J Biol Chem 279: 6385–6394 Pfennig N and Lippert KD (1966) Uber das vitamin B12-bedurfnis phototropher schwefelbakterien. Arch Mikrobiol 55: 245–256 Pollich M and Klug G (1995) Identification and sequence analysis of genes involved in late steps in cobalamin (vitamin B12) synthesis in Rhodobacter capsulatus. J Bacteriol 177: 4481–4487 Pollich M, Jock S and Klug G (1993) Identification of a gene required for the oxygen-regulated formation of the photosynthetic apparatus of Rhodobacter capsulatus. Mol Microbiol 10: 749–757 Pollich M, Wersig C and Klug G (1996) The bluF gene of Rhodobacter capsulatus is involved in conversion of cobinamide to cobalamin (vitamin B12). J Bacteriol 178: 7308–7310 Raux E, Schubert HL, Roper JM, Wilson KS and Warren MJ (1999) Vitamin B12; insights into biosynthesis’s mount improbable. Bioorganic Chem 27: 100–118 Raux E, Schubert HL and Warren MJ (2000). Biosynthesis of cobalamin (vitamin B12): A bacterial conundrum. Cell Mol Life Sci 57: 1880–1893 Raux E, Leech HK, Beck R, Schubert HL, Santander PJ, Roessner CA, Scott AI, Martens JH, Jahn D, Thermes C, Rambach A and Warren MJ (2003) Identification and functional analysis of enzymes required for precorrin-2 dehydrogenation and metal ion insertion in the biosynthesis of sirohaem and cobalamin in Bacillus megaterium. Biochem J 370: 505–516 Rickes EL, BN, Koniuszy FR, Wood TR, Folkers K (1948) Crystalline vitamin B12. Science 107: 396–397 Rodionov DA, Vitreschak AG, Mironov AA and Gelfand MS (2003) Comparative genomics of the vitamin B12 metabolism and regulation in prokaryotes. J Biol Chem 278: 41148–41159 Roessner CA and Scott AI (2006) Fine-tuning our knowledge of the anaerobic route to cobalamin (vitamin B12). J Bacteriol 188: 7331–7334 Roth JR, Lawrence JG, Rubenfield M, Kieffer-Higgins S and Church GM (1993) Characterization of the cobalamin (vitamin B12) biosynthetic genes of Salmonella typhimurium. J Bacteriol 175: 3303–3316 Roth JR, Lawrence JG and Bobik TA (1996) Cobalamin (coenzyme B12): Synthesis and biological significance. Annu Rev Microbiol 50: 137–181 Santander PJ, Kajiwara Y, Williams HJ and Scott AI (2006) Struc-
Martin J. Warren and Evelyne Deery tural characterization of novel cobalt corrinoids synthesized by enzymes of the vitamin B12 anaerobic pathway. Bioorg Med Chem 14: 724–731 Schubert HL, Blumenthal RM and Cheng X (2003) Many paths to methyltransfer: A chronicle of convergence. Trends Biochem Sci 28: 329–335 Scott AI (1994) On the duality of mechanism of ring contraction in vitamin B12 biosynthesis. Heterocycles 39: 471–476 Scott AI, Roessner CA, Stolowich NJ, Spencer JB, Min C and Ozaki SI (1993) Biosynthesis of vitamin B12. Discovery of the enzymes for oxidative ring contraction and insertion of the fourth methyl group. FEBS Lett 331: 105–108 Shipman LW, Li D, Roessner CA, Scott AI and Sacchettini JC (2001) Crystal structure of precorrin-8X methyl mutase. Structure 9: 587–596 Smith EL (1948) Purification of anti-pernicious anemia factors from liver. Nature 161: 638–639 Spencer JB, Stolowich NJ, Santander PJ, Pichon C, Kajiwara M, Tokiwa S, Takatori K and Scott AI (1994) Mechanism of the ring contraction step in vitamin B12 biosynthesis—the origin and subsequent fate of the oxygen functionalities in precorrin3X. J Am Chem Soc 116: 4991–4992. Thibaut D, Couder M, Crouzet J, Debussche L, Cameron B and Blanche F (1990a) Assay and purification of S-adenosyl-Lmethionine:Precorrin-2 methyltransferase from Pseudomonas denitrificans. J Bacteriol 172: 6245–6251 Thibaut D, Debussche L and Blanche F (1990b) Biosynthesis of vitamin B12: Isolation of precorrin-6X, a metal-free precursor of the corrin macrocycle retaining five S-adenosylmethioninederived peripheral methyl groups. Proc Natl Acad Sci U S A 87: 8795–8799 Thibaut, D Couder M, Famechon A, Debussche L, Cameron B, Crouzet J and Blanche F (1992) The final step in the biosynthesis of hydrogenobyrinic acid is catalyzed by the cobH gene product with precorrin-8X as the substrate. J Bacteriol 174: 1043–1049 Thibaut D, Debussche L, Frechet D, Herman F, Vuilhorgne M and Blanche F (1993). Biosynthesis of vitamin B12 - the structure of factor-IV, the oxidized form of precorrin-4. J Chem Soc, Chem Commun: 513 Toohey JI (1965) A vitamin B12 compound containing no cobalt. Proc Natl Acad Sci U S A 54: 934–942 Toohey JI (1966) Cobalt-free corrinoid compounds from photosynthetic bacteria. Fed Proc 25: 1628–1632 Vévodová J, Graham RM, Raux E, Schubert HL, Roper DI, Brindley AA, Scott AI, Roessner CA, Stamford NPJ, Stroupe EM, Getzoff ED, Warren MJ and Wilson KS (2004) Structure/function studies on a S-adenosyl-L-methionine-dependent uroporphyrinogen III C methyltransferase (SUMT), a key regulatory enzyme of tetrapyrrole biosynthesis. J Mol Biol 344: 419–433 Vitreschak AG, Rodionov DA, Mironov AA and Gelfand MS (2003) Regulation of the vitamin B12 metabolism and transport in bacteria by a conserved RNA structural element. RNA 9: 1084–1097 Vlcek C, Paces V, Maltsev N, Paces J, Haselkorn R and Fonstein M (1997) Sequence of a 189-kb segment of the chromosome of Rhodobacter capsulatus SB1003. Proc Natl Acad Sci USA 94: 9384–9388 Warren MJ, Raux E, Schubert HL and Escalante-Semerena JC
Chapter 5
B12 Biosynthesis
(2002) The biosynthesis of adenosylcobalamin (vitamin B12). Nat Prod Rep 19: 390–412 Whipple GH and Robscheit-Robbins FS (1925) Favourable
95 influence of liver, heart and skeletal muscle in diet on blood regeneration in anemia. Am J Physiol 72: 408–418
Chapter 6 Distribution and Biosynthesis of Carotenoids Shinichi Takaichi* Department of Biology, Nippon Medical School, Kosugi-cho 2, Nakahara, Kawasaki 211-0063, Japan
Summary ................................................................................................................................................................. 97 I. Introduction....................................................................................................................................................... 98 II. Carotenogenesis ............................................................................................................................................ 101 A. Classification of Carotenogenesis .................................................................................................... 101 B. Carotenogenesis Genes and Gene Clusters.................................................................................... 103 C. Isopentenylpyrophosphate (IPP) and Phytoene Synthesis .............................................................. 103 D. Desaturation of Phytoene to Lycopene ............................................................................................ 103 E. Biosynthesis of Spirilloxanthin .......................................................................................................... 104 1. The Normal Spirilloxanthin Pathway ....................................................................................... 104 2. The Unusual Spirilloxanthin Pathway...................................................................................... 106 3. The Spheroidene Pathway ...................................................................................................... 107 4. The Carotenal Pathway........................................................................................................... 108 F. Biosynthesis of Okenone .................................................................................................................. 108 1. The Okenone and the R.g.-keto Carotenoid Pathways .......................................................... 108 G. Biosynthesis of Carotenoid Glucosides and their Fatty Acid Esters ................................................ 109 III. Carotenoids in Purple Bacteria....................................................................................................................... 111 A. Anaerobic Purple Bacteria................................................................................................................ 111 1. Alphaproteobacteria and Betaproteobacteria ....................................................................... 111 2. Gammaproteobacteria ........................................................................................................... 112 B. Aerobic Photosynthetic Bacteria ...................................................................................................... 112 Acknowledgments ................................................................................................................................................. 114 References ............................................................................................................................................................ 114
Summary Purple bacteria including aerobic photosynthetic bacteria belong to the Proteobacteria, and 75 genera including around 160 species have been described. These bacteria produce around 100 different carotenoids, which are essential for photoprotection and light-harvesting. This chapter summarizes the distribution and biosynthesis of carotenoids in all of the purple bacteria described so far. All of the carotenogenesis genes from Rhodobacter capsulatus, Rhodobacter sphaeroides and Rubrivivax gelatinosus, and some genes from other purple bacteria have been functionally confirmed, and the characteristics of their products have been investigated. When one enzyme of the typical spirilloxanthin pathway is lacking or is present with reduced activity, the carotenoid composition of the bacterium will be expected to change; indeed, the variation of the spirilloxanthin pathways can be explained by this idea. Based on these new findings, two main pathways within purple bacteria have been proposed; the spirilloxanthin pathway (normal spirilloxanthin, unusual spirilloxanthin, spheroidene, and carotenal pathways) and the okenone pathway (okenone and R.g.-keto carotenoid pathways). In addi*Email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 97–117. © 2009 Springer Science + Business Media B.V.
98
Shinichi Takaichi
tion, carotenoid glucosides and carotenoid glucoside fatty acid esters have also been found in some species. Purple bacteria classified as Alphaproteobacteria and Betaproteobacteria have the spirilloxanthin pathway, while those in the Gammaproteobacteria have either the spirilloxanthin or the okenone pathway depending on genus or species. The aerobic photosynthetic bacteria described so far are classified as Alphaproteobacteria and Betaproteobacteria, and most species have the spirilloxanthin pathway. Furthermore most of these species also have unusual carotenoids including ‘non-photosynthetic’ carotenoids, such as carotenoid sulfates and carotenoic acids, which seem to have no photosynthetic functions. I. Introduction Anoxygenic photosynthetic bacteria include purple bacteria, green sulfur bacteria, green filamentous bacteria and heliobacteria, and they necessarily synthesize not only bacteriochlorophyll (BChl) but also carotenoids. About three decades ago, Schmidt (1978) summarized the carotenoids in 19 genera including 38 species of photosynthetic bacteria. Subsequently, Takaichi (1999) summarized carotenoids and carotenogenesis pathways in 55 genera including 133 species. Since then, many novel genera and species have been recognized, and some species have been re-classified based on the results of new classification techniques (Imhoff et al., 1998a, 1998b). Now, around 90 genera including around 200 species of photosynthetic bacteria have been described. Purple bacteria, including the aerobic photosynthetic bacteria, belong to Phylum Proteobacteria and are distributed between 3 Classes (Alphaproteobacteria, Betaproteobacteria and Gammaproteobacteria); 7 Orders, 14 Families and 75 Genera, and 164 species have been described (Table 1) (Brenner et al., 2005). Around 50 different carotenoids have been found in the anaerobic purple bacteria, and most of their chemical structures are distinct from those found in algae, fungi, higher plants and non-photosynthetic bacteria. These carotenoids have been characterized in the following ways by Takaichi (1999) and some newly found characteristics are also added: (1) They most often occur as acyclic compounds. (2) They often contain the tertiary hydroxy and the methoxy groups at C-1. (3) There are frequently double bonds in the C-3,4 position. (4) Keto groups in conjugation with the polyene chain are found attached to C-2 or C-4 of the ψ end group. (5) Aldehyde groups in conjugation Abbreviations: BChl – bacteriochlorophyll; Erb. – Erythrobacter; Erb-type – Erythrobacter-type; IPP – isopentenylpyrophosphate; LH – light-harvesting; LH1 – light-harvesting 1 complex; LH2 – light-harvesting 2 complex; Rba. – Rhodobacter; RC – reaction center; Rps. – Rhodopseudomonas; Rsp. – Rhodospirillum; Rvi. – Rubrivivax
with the polyene chain are sometimes present at the C20 position. (6) Cyclic carotenoids, such as okenone, have aromatic rings of the χ end group, synthesized from the β end group. (7) Some species have carotenoid glucosides and/or their fatty acid esters. (8) In many cases, there are small amounts of intermediates occurring later than neurosporene or lycopene. (9) Further, most aerobic photosynthetic bacteria have spirilloxanthin, and some have additional unusual carotenoids including ‘non-photosynthetic’ carotenoids, such as carotenoid sulfates, carotenoic acids and hydroxy derivatives of β-carotene. Most carotenoids have trivial names, and moreover all carotenoids have been named semisystematically based on IUPAC-IUB nomenclature (IUPAC Commission on Nomenclature of Organic Chemistry and the IUPAC-IUB Commission on Biochemical Nomenclature, 1975). A list of both the names and the structures of all known naturally occurring carotenoids, and references giving data for each compound are presented in the Carotenoids Handbook (Britton et al., 2004), and in the Web of Bioactive Lipid Database (Carotenoids) (http://lipidbank.jp/). Two main pathways for carotenogenesis within the purple bacteria are summarized based on the carotenoid compositions (Schmidt, 1978; Takaichi, 1999). Note that the compositions are rather variable depending on the culture conditions, such as light intensity (Gardiner et al., 1993), oxygen concentration (Goodwin, 1956), and growth phases (Schwerzmann and Bachofen, 1989), and also depending on the strain in some species, such as Rbl. acidophilus (previously Rps. acidophila) (Gardiner et al., 1993) and Rhodocyclus tenuis (Schmidt, 1978). Identification of carotenoids is based on the description given in each reference (Table 1), although it should also be noted that some were doubtful or were identified without sufficient data. Classically, carotenogenesis pathways in organisms were determined based on the carotenoid composition of the wild type, and comparison with the carotenoids in spontaneous mutants and the mutants induced by
Chapter 6
Carotenoid Biosynthesis
99
Table 1. Classification and carotenogenesis pathways of the purple bacteria Classification a,b,c Class I. Alphaproteobacteria Order I. Rhodospirillales (alpha-1) Family I. Rhodospirillaceae Genus I. Rhodospirillum (Rsp.) Genus V. Phaeospirillum (Phs.) Genus VI. Rhodocista (Rcs.) Genus VII. Rhodospira (Rsa.) Genus VIII. Rhodovibrio (Rhv.) Genus IX. Roseospira (Ros.) Genus. Rhodothalassium (Rts.) Family II. Acetobacteraceae * Genus II. Acidiphilium (Acp.) * Genus III. Acidisphaera (Acs.) * Genus VII. Craurococcus (Crc.) * Genus XII. Paracraurococcus (Pcr.) Genus XIII. Rhodopila (Rpi.) * Genus XIV. Roseococcus (Rsc.) * Genus XV. Rubritepida (Rut.) Order III. Rhodobacterales (alpha-3) Family I. Rhodobacteraceae Genus I. Rhodobacter (Rba.) Genus XVIII. Rhodobaca (Rca.) Genus XX. Rhodovulum (Rdv.) * Genus XXI. Roseibium (Rib.) * Genus XXII. Roseinatronobacter (Rna.) * Genus XXIII. Roseivivax (Rsv.) * Genus XXIV. Roseobacter (Rsb.) * Genus XXV. Roseovarius (Rva.) * Genus XXVI. Rubrimonas (Rum.) * Genus XXIX. Staleya (Stl.) * Genus. Dinoroseobacter * Genus. Roseibacterium * Genus. Roseicyclus * Genus. Roseisalinus Order IV. Sphingomonadales (alpha-4) Family I. Sphingomonadaceae * Genus II. Blastomonas * Genus III. Erythrobacter (Erb.) * Genus IV. Erythromicrobium (Erm.) * Genus VII. Porphyrobacter (Por.) * Genus IX. Sandaracinobacter (San.) * Genus. Citromicrobium (Cmi.) * Genus. Sandarakinorhabdus
Number of species
Carotenogenesis pathways found in the genus d,e
2 2 2 1 2 4 1
Normal, Unusual Unusual, G Normal Unusual Normal Unusual Normal
6 1 1 1 1 1 1
Normal Normal, Acid Normal, Acid Normal, Acid R.g.-keto C30-Acid, G-FA Normal
5 1 8 2 2 2 2 2 1 1 1 1 1 1
Spheroidene Spheroidene Spheroidene (no data) Spheroidene (no data) Spheroidene Normal, Spheroidene (no data) (no data) Spheroidene (no data) (no data) (no data)
2 2 3 6 1 1 1
Spheroidene Erb-type Erb-type Erb-type (no data) (no data) Erb-type
Shinichi Takaichi
100 Table 1. Continued Classification a,b,c Order VI. Rhizobiales (alpha-2) Family I. Rhizobiaceae * Genus I. Rhizobium Family V. Phyllobacteriaceae * Genus. Hoeflea Family VIII. Bradyrhizobiaceae * Genus I. Bradyrhizobium Genus VIII. Rhodoblastus (Rbl.) Genus IX. Rhodopseudomonas (Rps.) Family IX. Hyphomicrobiaceae Genus VII. Blastochloris (Blc.) Genus XVI. Rhodomicrobium (Rmi.) Genus XVII. Rhodoplanes (Rpl.) Family X. Methylobacteriaceae * Genus I. Methylobacterium (Mtb.) Family XI. Rhodobiaceae Genus I. Rhodobium (Rbi.) Genus II. Roseospirillum (Rss.) Class II. Betaproteobacteria Order I. Burkholderiales Family IV. Comamonadaceae Genus XIII. Rhodoferax (Rfx.) Genera incertae sedis * Genus IV. Roseateles (Rst.) Genus V. Rubrivivax (Rvi.) Order VI. Rhodocyclales Family I. Rhodocyclaceae Genus I. Rhodocyclus (Rcy.) Class III. Gammaproteobacteria Order I. Chromatiales Family I. Chromatiaceae Genus I. Chromatium (Chr.) Genus II. Allochromatium (Alc.) Genus IV. Halochromatium (Hch.) Genus V. Isochromatium (Isc.) Genus VI. Lamprobacter (Lpb.) Genus VII. Lamprocystis (Lpc.) Genus VIII. Marichromatium (Mch.) Genus XI. Rhabdochromatium (Rbc.) Genus XIII. Thermochromatium (Tch.) Genus XIV. Thioalkalicoccus (Tac.) Genus XV. Thiobaca (Tba.) Genus XVI. Thiocapsa (Tca.) Genus XVII. Thiococcus (Tco.)
Number of species
Carotenogenesis pathways found in the genus d,e
1
Normal
1
Spheroidene
1 2 6
Normal and Canthaxanthin Normal, Carotenal, G Normal, Unusual
2 1 2
Normal, Unusual Unusual, β-Carotene Normal
18
Normal, Acid, C-30 Acid
2 1
Normal Normal
2
Normal and Spheroidene
1 2
Normal Normal and Spheroidene
2
Carotenal
2 3 2 1 1 2 3 1 1 1 1 5 1
Okenone Normal, Unusual, Carotenal Normal, Unusual Carotenal Okenone Carotenal, Okenone Normal, Unusual, Okenone Unusual Unusual Unusual Unusual Normal, Okenone, R.g.-keto Unusual, Carotenal
Chapter 6
Carotenoid Biosynthesis
101
Table 1. Continued Classification a,b,c Genus XVIII. Thiocystis (Tcs.) Genus XIX. Thiodictyon (Tdc.) Genus XX. Thioflavicoccus (Tfc.) Genus XXI. Thiohalocapsa (Thc.) Genus XXII. Thiolamprovum (Tlp.) Genus XXIII. Thiopedia (Tpd.) Genus XXIV. Thiorhodococcus (Trc.) Genus XXV. Thiorhodovibrio (Trv.) Genus XXVI. Thiospirillum (Tsp.) Family II. Ectothiorhodospiraceae Genus I. Ectothiorhodospira (Ect.) Genus V. Halorhodospira (Hlr.) Genus IX. Thiorhodospira (Trs.)
Number of species
Carotenogenesis pathways found in the genus d,e
4 2 1 1 1 1 3 1 1
Normal, Carotenal, Okenone, R.g.-keto Carotenal Unusual Normal, Okenone, R.g.-keto Normal Okenone Unusual, Carotenal Unusual Unusual
4 4 1
Normal, Unusual Normal, Unusual, G, G-FA Unusual
a
Bremner et al. (2005), and new genera and species are added. b Genus with asterisk (*) contains aerobic photosynthetic bacteria. c Recommended 3-letter abbreviations for genera of the purple bacteria proposed by International Committee on Systematics of Prokaryotes Subcommittee on the Taxonomy of Phototrophic Bacteria (Imhoff and Madigan, 2004) are indicated in parentheses. d Carotenogenesis pathways and unusual carotenoids found in the genus are indicated. e Acid – carotenoic acid; Carotenal – carotenal pathway; Erb-type – Erythrobacter-type carotenoids; G – carotenoid glucoside; G-FA – carotenoid glucoside fatty acid ester; Normal – normal spirilloxanthin pathway; Okenone – okenone pathway; R.g.-keto – R.g.-keto carotenoid pathway; Spheroidene – spheroidene pathway; Unusual – unusual spirilloxanthin pathway.
mutagenesis. In the 1980s some carotenogenesis genes were cloned and functionally confirmed, and characteristics of enzymes were investigated. In the early 1990’s carotenogenesis gene clusters were found in some bacteria, such as Rhodobacter (Armstrong et al., 1989; Lang et al., 1995) and Pantoea (previously Erwinia) (Misawa et al., 1990), and their complete carotenogenesis pathways were determined. Furthermore, parts of the pathways were investigated in organisms lacking gene clusters, such as the green sulfur bacteria, cyanobacteria and higher plants. At present, the genome sequences of around 100 bacteria have been determined, which has led to the identification of some carotenogenesis gene clusters and some non-clustered genes, based on sequence homologies. II. Carotenogenesis A. Classification of Carotenogenesis On the basis of the analyses of carotenoid compositions Schmidt (1978) proposed the spirilloxanthin, the spheroidene and the okenone pathways for
carotenogenesis in the purple bacteria. Since then, many new species have been described, and some carotenogenesis genes have been cloned and the characteristics of their products have been investigated (Table 2) (Armstrong, 1995, 1997; Sandmann, 1994, 1997). Based on these results and the changes in the classification of the purple bacteria, two main pathways within the purple bacteria were suggested by Takaichi (1999): 1. The spirilloxanthin pathway (normal spirilloxanthin, unusual spirilloxanthin, spheroidene, and carotenal pathways: see Fig. 6). 2. The okenone pathway (okenone, and R.g.-keto carotenoid pathways: see Fig. 7). The normal spirilloxanthin pathway is fundamental and can account for the variation of carotenoids in the purple bacteria, which would arise when one enzyme of this normal spirilloxanthin pathway is lacking or is present with reduced activity. The aerobic photosynthetic bacteria described so far are classified in the purple bacteria, and although most have the spirilloxanthin pathway, some have
102
Table 2. Carotenogenesis genes in the purple bacteria Carotenogenesis genes a
Species
Reference
Rhodospirillum rubrum
(crtE)
(crtB)
(crtI)
(crtC)
(crtD)
(crtF)
Rhodobacter capsulatus
crtE
crtB
crtI
crtC
crtD
crtF
crtA
2
Rhodobacter sphaeroides
crtE
crtB
crtI
crtC
crtD
(crtF)
crtA
3
crtA
4
Rhodovulum sulfidophilum
crtC
Rhodopseudomonas palustris
(crtE)
(crtB)
crtI
crtC
(crtD)
(crtF)
Rubrivivax gelatinosus
(crtE)
crtB
crtI
crtC
crtD
crtF
Thiocapsa roseopersicina
(crtE)
(crtI)
(crtC)
crtD
(crtF)
Roseobacter denitrificans
(crtE)
(crtI)
(crtC)
(crtD)
(crtF)
(crtB)
Erythrobacter longus Bradyrhizobium ORS278 b
crtI (crtE.s)
(crtB.s)
(crtI.s)
crtE.c
crtB.c
crtI.c
1
5 crtA
6–9 10
(crtA)
11 crtY
crtC
(crtD)
12
(crtF)
13 crtY
crtW
14
Methylobacterium extorquens (crtE) (crtB) crtI (crtC) (crtD) (crtF) 15 Parentheses indicate that the functions of the genes are not confirmed, but the genes are suggested from DNA base sequence homology and their presence in the carotenogenesis gene cluster (See Table 3 and Fig. 2). b The upper row is a gene cluster for spirilloxanthin synthesis, and the lower one is that for canthaxanthin synthesis. 1. Reslewic et al. (2005); 2. Armstrong et al. (1989); 3. Lang et al. (1995); 4. Maeda et al. (2005); and Maeda, personal communication; 5. Shimada, personal communication; 6. Ouchane et al. (1997a); 7. Ouchane et al. (1997b); 8. Harada et al. (2001); 9. Pinta et al. (2003); 10. Kovács et al. (2003); 11. Swingley et al. (2007); 12. Matsumura et al. (1997); 13. Hannibal et al. (2000); 14. Giraud et al. (2004); 15. van Dien et al. (2003). a
Shinichi Takaichi
Chapter 6
Carotenoid Biosynthesis
additional unusual carotenoids as described below. The relationship between the classification of the purple bacteria, including the aerobic photosynthetic bacteria, and the carotenogenesis pathways is summarized in Table 1. B. Carotenogenesis Genes and Gene Clusters The first carotenogenesis genes to be cloned and functionally confirmed were found in a gene cluster in Rba. capsulatus (Armstrong et al., 1989) and in Pantoea ananatis (previously Erwinia uredovora) (Misawa et al., 1990). Some bacteria, including some purple bacteria (Table 2) and two species of Pantoea, have a carotenogenesis gene cluster, while others, such as Chlorobaculum (previously Chlorobium) tepidum (green sulfur bacteria), cyanobacteria, algae and higher plants, have no such clusters. The carotenogenesis gene clusters and some genes functionally confirmed from the purple bacteria are summarized in Table 2, although some genes are identified only from homology searches. C. Isopentenylpyrophosphate (IPP) and Phytoene Synthesis IPP, a C5-compound, is the source of isoprenoids, carotenoids, and phytol of (bacterio)chlorophylls. The biosynthesis of IPP via the mevalonate pathway was studied in the 1950s. More recently, in the 1990s, a new pathway via 1-deoxy-D-xylulose-5-phosphate was found, termed the DOXP pathway. In plastids of higher plants, IPP is synthesized by the DOXP pathway, whereas the mevalonate pathway is used in the cytoplasm. Rhodopseudomonas (Rps.) palustris, Rhodospirillum (Rsp.) rubrum, Rbl. acidophilus (previously Rps. acidophila), Allochromatium (previously Chromatium) vinosum and cyanobacteria produce IPP by the DOXP pathway, while Chlorobium limicola (green sulfur bacteria) and Chloroflexus aurantiacus (green filamentous bacteria) employ the mevalonate pathway. These two pathways are widely distributed among prokaryotes, and their distributions seem not to follow any pattern of taxonomic classification (Lichenthaler, 1999; Rohmer, 1999). Most carotenoids are tetraterpenoids consisting of eight IPP units. Farnesyl pyrophosphate (C15) is synthesized from three IPPs. Then, one IPP is added to farnesyl pyrophosphate by geranylgeranyl pyrophosphate synthase (CrtE) to yield geranylgeranyl
103 pyrophosphate (C20). In a condensation of the two C20-compounds, phytoene (C40) is formed by phytoene synthase (CrtB) using ATP (Sandmann, 1997). This pathway has been confirmed by the cloning of the genes from Rhodobacter and Pantoea (Sandmann, 1994, 1997; Armstrong, 1995, 1997). D. Desaturation of Phytoene to Lycopene Phytoene, the starting compound in carotenogenesis, is desaturated to neurosporene or lycopene in the purple bacteria. Subsequent reactions produce carotenes by desaturation, saturation, cyclization and aromatization, or xanthophylls by introduction of oxy-groups, such as hydroxy, methoxy, keto and aldehyde groups. These reactions found in the purple bacteria are summarized in Table 3 and Fig. 1. Four desaturation steps are known in the conversion from phytoene to lycopene by phytoene desaturase (CrtI) (Fig. 2). Usually the final product of CrtI is lycopene. This was confirmed by the functional expression of crtI genes of Pantoea ananatis (Misawa et al., 1990; Linden et al., 1991) and Rps. palustris (K. Shimada, personal communication) in Escherichia coli in a phytoene background. In the case of crtI genes of Rba. capsulatus (Linden et al., 1991) and Rba. sphaeroides (Lang et al., 1994, 1995), the final product is three-steps desaturated neurosporene. The crtI mutant of Rba. sphaeroides produces phytoene, and by the functional expression of crt genes or crtI gene from Pantoea stewartii (previously Erwinia herbicola) in the mutant, it produces lycopene (Hunter et al., 1994; Garcia-Asua et al., 2002). In the case of Rubrivivax (Rvi.) gelatinosus, the final products are around 90% neurosporene and 10% lycopene (Harada et al., 2001). The deduced amino acid sequences of the bacterial crtI genes including these species show significant similarity, but the final products are different (Sandmann, 1994; Armstrong, 1995; Harada et al., 2001). FAD is a cofactor for CrtI from Rba. capsulatus (Raisig et al., 1996). Diphenylamine is an inhibitor of CrtI (Sandmann, 1994). The mechanisms for determination of the final product, and for recognition of the desaturating position on carotenoids by CrtI, are still unknown. Phytoene takes the 15-cis form in most cases. Isomerization from the 15-cis form to the all-trans form is reported to occur non-enzymatically during the desaturation reactions, since there are no isomerase genes in the carotenogenesis gene clusters of bacteria described above, and gradual conversion of 15-cis
104
Shinichi Takaichi
Fig. 1. Types of reactions and enzymes seen in carotenogenesis within the purple bacteria. Parentheses indicate abbreviations of the reactions and the enzymes (see Table 3).
phytoene to all-trans lycopene is found when the crtI gene is expressed in Escherichia coli in a phytoene background (Misawa et al., 1990). On the other hand, poly-cis-carotenes are produced during desaturation to lycopene, and are changed to the all-trans form by light or by carotene isomerase, which is CrtH in green sulfur bacteria and cyanobacteria and CrtISO in higher plants (Giuliano et al., 2002). In the phytoene desaturation reaction, two pathways are known, via ζ-carotene or asymmetrical ζ-carotene (Fig. 2). The ζ-carotene pathway is found in Rba. capsulatus, Rba. sphaeroides (S. Takaichi, unpublished), Rvi. gelatinosus (Harada et al., 2001) and Roseobacter denitrificans (Harashima and Nakada, 1983), while the asymmetrical ζ-carotene pathway is found in Rsp. rubrum (Davies, 1970), Rhodopila globiformis (Schmidt and Liaaen-Jensen, 1973), Rhodomicrobium vannielii (Britton et al., 1975), Blastochloris (previously Rps.) viridis (S. Takaichi, unpublished) and Erythrobacter longus (Takaichi et al., 1990). These two pathways seem not to be related
to the final products of CrtI, neurosporene and lycopene, and to the classification of purple bacteria. E. Biosynthesis of Spirilloxanthin 1. The Normal Spirilloxanthin Pathway The biosynthesis of spirilloxanthin, a final product of the normal spirilloxanthin pathway, is found in a large number of species of purple bacteria (Table 1). Spirilloxanthin is a symmetrical compound containing the methoxy groups at C-1 and C-1´ and additional double bonds in the C-3,4 and C-3´,4´ positions. It has 13 conjugated double bonds (Fig. 3). A sequence of the reactions leading from lycopene to spirilloxanthin has been confirmed (Fig. 3). This sequence includes successive reactions of (1) hydration at C-1,2 by a hydratase, CrtC, (2) desaturation at C-3,4 by a desaturase, CrtD, and (3) methylation of the tertiary hydroxy group at C-1 by a methyltransferase, CrtF. These reactions occur initially on one
Chapter 6
Carotenoid Biosynthesis
105
Table 3. Types of reactions seen in carotenogenesis within the purple bacteria, and the corresponding genes (see Fig. 1) for the functions a Reactions Geranylgeranyl pyrophosphate synthesis Phytoene synthesis Desaturation of phytoene C-3,4 Desaturation C-1,2 Saturation Lycopene cyclization Aromatization from β to χ end group C-1,2 Hydration Methylation Ketolation at C-2 of ψ end group Ketolation at C-4 of ψ end group Ketolation at C-4 of β end group Hydroxylation at C-20 Aldehyde at C-20 Glucosylation Fatty acid esterification
Genes crtE crtB crtI crtD
Abbreviations b
– 2H + 2H
crtY crtC crtF crtA
+ H2O + Me 2-Keto 4-Keto
crtW 20-OH 20-Aldehyde cruC c, crtX d cruD c
a Since aerobic photosynthetic bacteria contain many unusual carotenoids, only known genes are listed. b These abbreviations are used in Fig. 1–8. c Chlorobaculum (previously Chlorobium) tepidum (Maresca and Bryant, 2006). d Pantoea ananatis (previously Erwinia uredovora) (Misawa et al., 1990).
Fig. 2. Desaturation from phytoene to lycopene by phytoene desaturase (CrtI).
half of the molecule to yield anhydrorhodovibrin, and then on the other half. The major product is
spirilloxanthin, and usually small amounts of all or a few of five intermediates are also found (Fig. 3).
Shinichi Takaichi
106
Fig. 3. The normal spirilloxanthin pathway. Enzymes are indicated in parentheses (see Table 3).
Other postulated intermediates are, however, rarely found as described below. These genes are found in the carotenogenesis gene clusters from Rps. palustris and Rsp. rubrum (Table 2). At present, the way in which the successive reactions are controlled, with reactions first on one half of the carotenoid and then on the other, has not been elucidated. 2. The Unusual Spirilloxanthin Pathway When one enzyme of the normal spirilloxanthin pathway is lacking or is present with reduced activity, the carotenoid composition of the bacterium will be expected to change (Takaichi, 1999). Indeed, some species have been reported to have such unusual compositions (Table 1). Lycopene is accumulated in Rhabdochromatium marinum (Dilling et al., 1995) and Thiobaca trueperi (Rees et al., 2002), and this may be due to low activity of C-1,2 hydratase (CrtC) (Fig. 3). Rhodopin is the major carotenoid in some species: Rsp. photometricum and Phaeospirillum (previously Rsp.) molischianum (Matsuura and
Shimada, 1993), Rhodomicrobium vannielii, and Allochromatium (previously Chromatium) vinosum (Schmidt, 1978), Thiorhodovibrio winogradskii (Overmann et al., 1992), Thiorhodococcus minor (Guyoneaud et al., 1997), and Rps. palustris, Thermochromatium (previously Chromatium) tepidum and Ectothiorhodospira marismortui (S. Takaichi, unpublished). This may be due to low activity of C-3,4 desaturase (CrtD), and the rhodopin may not be a suitable substrate for the methylation enzyme (CrtF) due to the single bond at C-3,4 (Fig. 3). On the other hand, 3,4,3´,4´-tetrahydrospirilloxanthin (Fig. 6) is the major component in Rhodospira trueperi (Pfennig et al., 1997), Thioalkalicoccus limnaeus (Bryantseva et al., 2000), Thiocapsa pfennigii (Eimhjellen et al., 1967) and Thioflavicoccus mobilis (Imhoff and Pfennig, 2001) probably due to lack of CrtD, but in these cases rhodopin can be methylated by CrtF in spite of the single bond at C-3,4. The same carotenoid is also found in a carotenogenesis mutant (crtD) of Rsp. rubrum (Komori et al., 1998).
Chapter 6
Carotenoid Biosynthesis
107
Fig. 4. The spheroidene pathway. Enzymes are indicated in parentheses (see Table 3). The pathway from chloroxanthin to 3,4-dihydrospheroidene might not be involved in vivo, and 3,4-dihydrospheroidene might not be changed to spheroidene.
3. The Spheroidene Pathway Three genera, Rhodobacter, Rhodobaca and Rhodovulum, and some genera of aerobic photosynthetic bacteria (Table 1) produce spheroidene and its derivatives. Spheroidene is an asymmetrical compound containing the same end group as spirilloxanthin on one side and the 7,8-dihydro-ψ end group on the other (Fig. 4). All of the seven carotenogenesis genes in Rba. capsulatus (Armstrong et al., 1989) and Rba. sphaeroides (Lang et al., 1995) have been found to form a gene cluster. The characteristics of their products have been investigated (Fig. 4). Phytoene desaturase (CrtI) produces neurosporene from phytoene in three desaturation steps. Then, successive reactions include hydration at C-1,2 by hydroxyneurosporene synthase (CrtC) to yield chloroxanthin (hydroxyneurosporene), desaturation at C-3,4 by methoxyneurosporene dehydrogenase (CrtD) to yield demethylspheroidene, and methylation at the C-1 hydroxy group by hydroxyneurosporene-O-methyltransferase (CrtF) to yield spheroidene. Further, CrtC can more or less hydrate at the 7´,8´-dihydro-ψ end group to yield OH-spheroidene depending on the species. Under semi-aerobic conditions, spheroidene monooxygenase (CrtA) introduces the keto group at C-2 to yield
spheroidenone. The 1´-hydroxy-7´,8´-dihydro-ψ end group cannot be modified further due to the single bond at C-7´,8´. The keto group of spheroidenone from Roseobacter denitrificans is in the single bond cis-conformation around the conjugated double bond (Fig. 4) (Takaichi et al., 1991b). Water is a substrate for CrtC in Rba. sphaeroides (Yeliseev and Kaplan, 1997). The hydrogen acceptor of CrtD in Rba. sphaeroides is molecular oxygen (Albrecht et al., 1997). The methyl residue in the methoxy groups arises from S-adenosylmethionine in Rba. sphaeroides (Singh et al., 1973) and Rba. capsulatus (Scolnik et al., 1980). The oxygen of the keto group at C-2 is derived directly from the atmosphere in Rba. sphaeroides (Shneour, 1962) and Rhodovulum sulfidophilum (Maeda et al., 2006). 3,4-Dihydrospheroidene (methoxyneurosporene), which is rarely found in the wild type strains, is found in two strains of the carotenoid mutants, Rba. capsulatus MT1131 (Frank et al., 1986) and Rba. sphaeroides Ga (S. Takaichi, unpublished), accompanied by neurosporene and chloroxanthin (Fig. 4). These strains may be crtD gene mutants. When the crtD gene of Rba. sphaeroides is disrupted, the three carotenoids described above are accumulated (Lang et al., 1995). Further, in in vitro experiments, 3,4dihydrospheroidene is not able to be the substrate
Shinichi Takaichi
108 for CrtD from Rba. sphaeroides (Albrecht et al., 1997). These results indicate that CrtC, CrtD and CrtF work sequentially in this order, and a part of the chloroxanthin can be the substrate for CrtF only when chloroxanthin accumulates significantly. Consequently, 3,4-dihydrospheroidene may not be involved in carotenogenesis in the spheroidene pathway (Fig. 4). Similarly, in the normal spirilloxanthin pathway, 3,4-dihydroanhydrorhodovibrin (Fig. 6) and 3,4-dihydrospirilloxanthin (Fig. 8), which are produced by methylation prior to desaturation, are rarely found in the wild type strains. This may be also due to similar properties of CrtD and CrtF. Since Rhodobaca bogoriensis accumulates demethylspheroidene and demethylspheroidenone, activity of CrtF of this bacterium may be low (Fig. 4) (Takaichi et al., 2001a). Rvi. gelatinosus produces carotenoids of both the spheroidene and the spirilloxanthin pathways (Table 1; see Fig. 6). When the Rvi. gelatinosus crtI gene is expressed in Escherichia coli in a phytoene background, the products are around 90% neurosporene and 10% lycopene (Harada et al., 2001). The crtC mutant accumulates both neurosporene and lycopene (Ouchane et al., 1997a). These results indicate that phytoene desaturase (CrtI) of Rvi. gelatinosus can produce both neurosporene and lycopene. When the crtD gene is disrupted, neurosporene, chloroxanthin, 3,4-dihydrospheroidene, lycopene and rhodopin are accumulated (Ouchane et al., 1997a). Thus, Rvi. gelatinosus produces spheroidene and spirilloxanthin from neurosporene and lycopene, respectively. Although the reaction of ketolation at C-2 of the ψ end group (CrtA) is usually involved in the spheroidene pathway and not in the normal spirilloxanthin pathway (Fig. 3), this bacterium can also oxidize spirilloxanthin to 2,2´-diketospirilloxanthin (Fig. 8) by CrtA. Further, Rhodoferax antarcticus (Jung et al., 2004) and Roseobacter denitrificans (Harashima and Nakada, 1983) are also reported to have both of these keto-carotenoids. Thus, the spheroidene pathway is a variant of the unusual spirilloxanthin pathway as a result of the different properties of the phytoene desaturases (CrtI) and the additional enzyme CrtA.
13-cis form of rhodopin. Usually, small amounts of rhodopinol, lycopenol and lycopenal can also be detected. Methoxylycopenal is found as a major component in Lamprocystis roseopersicina (Pfennig et al., 1968; Francis and Liaaen-Jensen, 1970), and tetrahydrospirilloxanthinal is found in Thiococcus (previously Thiocapsa) pfennigii (Eimhjellen et al., 1967). Rbl. acidophilus (previously Rps. acidophila) strain 7050 contains rhodopinal glucoside (Fig. 6) as a major component (Gardiner et al., 1993) and small amount of methoxylycopenal (Schmidt, 1971). Spirilloxanthin and its precursors are often found in Rbl. acidophilus (Schmidt, 1971; Gardiner et al, 1993), Thiocystis (previously Chromatium) violascens and Thiocystis violacea (Schmidt et al., 1965), while in other species, spirilloxanthin intermediates occurring later than rhodopin are not found. From the structure of these carotenals, Francis and Liaaen-Jensen (1970) have postulated a branched path for each of the derivatives of cross-conjugated carotenals. Up to now, further support for this has not been reported, and the partially modified carotenogenesis pathway is shown in Fig. 5. Rhodopin and lycopene are hydroxylated at C-20 to yield rhodopinol and lycopenol, and then the hydroxy groups are oxidized to the aldehyde groups to yield rhodopinal and lycopenal, respectively. Whether lycopenol and lycopenal are precursors of rhodopinol and rhodopinal, respectively, has not been confirmed. Since all of these carotenals have a C-3,4 single bond, the C-3,4 desaturase (CrtD) may be inactive. Further, the activity of methylation to the hydroxy group at C-1 (CrtF) is none or low. Additional enzymes for hydroxylation at C-20 and for oxidation to the aldehyde group may be involved in this pathway, whereas nothing is known about how the aldehyde group is introduced under anaerobic conditions. The position is necessarily at C-20, rather than C-20´ or C-19, and the 13-cis form, is found since it is more stable than the all-trans form, because of the hydroxy or the aldehyde groups at C-20. The spirilloxanthin pathways (normal spirilloxanthin, unusual spirilloxanthin, spheroidene and carotenal pathways) are summarized in Fig. 6.
4. The Carotenal Pathway
F. Biosynthesis of Okenone
Four kinds of cross-conjugated carotenals (Fig. 5) are found in 8 genera including 12 species (Table 1). Most species contain rhodopinal as a major component, which has an aldehyde group at C-20 with the
1. The Okenone and the R.g.-keto Carotenoid Pathways R.g.-keto carotenoids are found in Rhodopila globi-
Chapter 6
Carotenoid Biosynthesis
109
Fig. 5. The carotenal pathway. The predicted pathway for the carotenogenesis of cross-conjugated carotenals and their structures are indicated.
formis, and okenone, together with small amounts of R.g.-keto carotenoids, is found in 8 genera and 10 species of Gammaproteobacteria, such as Chromatium okenii and Thiocystis gelatinosa (Table 1). Okenone has one aromatic χ end group (Fig. 7), and one aliphatic ψ end group substituted with the methoxy group at C-1 and the keto group at C-4. The keto group of okenone from Marichromatium (previously Chromatium) purpuratum is in the single bond trans-conformation around the conjugated double bond (Fig. 7) (Fujii et al., 1998). R.g.-keto III is a symmetrical carotenoid having end groups the same as one end of okenone. A small amount of lycopene is found in Rhodopila globiformis (S. Takaichi, unpublished) and in the diphenylamine-inhibited culture of Chromatium okenii (Schmidt et al., 1963). The carotenogenesis pathways of okenone and R.g.-keto carotenoids could start from lycopene rather than from neurosporene (Fig. 7).
The biosynthesis of both of these carotenoids may have a close relationship (Schmidt, 1978), and these pathways may be distinguished at the level of the cyclase, γ-carotene synthase; it could be present (the okenone pathway) or absent (the R.g.-keto carotenoid pathway). Nothing is known about how the keto group is introduced under anaerobic conditions. G. Biosynthesis of Carotenoid Glucosides and their Fatty Acid Esters The occurrence of carotenoid glucosides and their fatty acid esters in the purple bacteria is regarded as unusual; however, more recently, the wide distribution of such carotenoid derivatives among various species has become clear. They have also been found in some non-photosynthetic bacteria, green sulfur bacteria, green filamentous bacteria, heliobacteria (Takaichi, 1999) and cyanobacteria. Rbl. acidophilus (previously
110
Shinichi Takaichi
Fig. 6. Summary of the spirilloxanthin pathways. The normal spirilloxanthin, unusual spirilloxanthin, spheroidene and carotenal pathways are summarized.
Rps. acidophila) has three carotenoid glucosides; rhodopin, rhodopinol and rhodopinal glucosides (Fig. 6, Fig. 8) (Schmidt et al., 1971), and the total glucoside contents are 20 to 70% of total carotenoids depending on the culture conditions and the strains (Gardiner et al., 1993). Rhodopin glycoside has also been found in Phaeospirillum (previously Rsp.) fulvum (Table 1) (S. Takaichi, unpublished). Furthermore, two species of Halorhodospira have dihydroxylycopene diglucoside and its fatty acid diesters (Fig. 8) (Takaichi et al., 2001b). Roseococcus thisulfatophilus (Yurkov et al., 1993) and Methylobacterium rhodinum (Kleinig et al., 1979) have di(acyl-glucosyl)-diapocarotenedioate (Fig. 8). In all known cases, the hydroxy or the carboxylic groups of carotenoids form a glycosidic linkage with β-D-glucoside, and often a fatty acid is esterified at the C-6 hydroxy group of glucoside (Fig. 8). In vitro experiments with Pantoea stewartii (previously Erwinia herbicola) demonstrate that UDP-glucose is the substrate for the formation of zeaxanthin diglucoside (Hundle et al., 1992). Recently, chlorobactene glucosyltransferase (CruC) and chlorobactene lauroyltransferase (CruD) have been functionally confirmed in the green sulfur bacterium Chlorobaculum (previously Chlorobium) tepidum (Maresca and Bryant, 2006).
Fig. 7. The predicted okenone and R.g.-keto carotenoid pathways and their structures.
Although the localization of rhodopin glucoside in the light-harvesting 2 (LH2) complex of Rbl. acidophilus (previously Rps. acidophila) has been established
Chapter 6
Carotenoid Biosynthesis
111
Fig. 8. Structures of some unusual carotenoids found in the purple bacteria.
by X-ray crystallography (McDermott et al., 1995; Chapter 8, Gabrielsen et al.), the functions of the glucoside and ester moieties in photosynthesis are still unknown. III. Carotenoids in Purple Bacteria A. Anaerobic Purple Bacteria 1. Alphaproteobacteria and Betaproteobacteria Purple non-sulfur bacteria belong to Alphaproteo-
bacteria and Betaproteobacteria (Table 1) (Brenner et al., 2005). Their major BChl is BChl a or b. The functions of carotenoids in photosynthetic bacteria have been investigated in most detail in the purple non-sulfur bacteria. Carotenoids necessarily exist in reaction center (RC), RC-light-harvesting 1 (LH1) and LH2 antenna complexes, and have functions of light-harvesting and/or photoprotection, as well as playing possible structural roles. These aspects of carotenoids are covered in other chapters in this book (Chapter 8, Gabrielsen et al.; Chapter 12, Frank and Polívka; Chapter 46, Braun and Fiedor). The spirilloxanthin pathway is found in all these bacteria except for one genus, Rhodopila, which
112 has the okenone pathway (Table 1). Most genera have the normal spirilloxanthin or the unusual spirilloxanthin pathways (Fig. 6). The spheroidene pathway is found only in three genera from Rhodobacterales (alpha-3) of Rhodobacter, Rhodobaca and Rhodovulum. Rhodoferax and Rubrivivax, which belong to Betaproteobacteria, have both the normal spirilloxanthin and the spheroidene pathways. All of the carotenogenesis genes have been functionally confirmed in Rba. capsulatus, Rba. sphaeroides and Rvi. gelatinosus (Table 2). The carotenal pathway is found in Rbl. acidophilus (previously Rps. acidophila) (Alphaproteobacteria) and Rhodocyclus (Betaproteobacteria). The okenone pathway (R.g.-keto carotenoid) is found only in Rhodopila globiformis (Schmidt and Liaaen-Jensen, 1973). β-Carotene is found only in Rhodomicrobium vannielii (Ryvarden and Liaaen-Jensen, 1964). Carotenoid glucoside is found in Phaeospirillum (previously Rsp.) fulvum (S. Takaichi, unpublished), Rbl. acidophilus (Gardiner et al., 1993) and Rbl. sphagricola. Carotenes with the 1,2-dihydro-ψ end group, such as 1,2-dihydroneurosporene and 1,2-dihydro-3,4-dehydrolycopene (Fig. 8) are found in Blastochloris (previously Rps.) viridis (Malhotra et al., 1970). 2. Gammaproteobacteria Purple sulfur bacteria belong to Gammaproteobacteria (Table 1) (Brenner et al., 2005). Their major BChl is BChl a or b. These bacteria have either the spirilloxanthin pathway (normal spirilloxanthin, unusual spirilloxanthin and carotenal pathways) or the okenone pathway (Table 1). In four genera, Lamprocystis, Marichromatium, Thiocapsa and Thiocystis, some species have only the spirilloxanthin pathway (Fig. 6) and others have only the okenone pathway (Fig. 7). Exceptionally, two species Thiohalocapsa (previously Thiocapsa) halophila (Caumette et al., 1991) and Thiocapsa marina (Caumette et al., 2004) have both pathways. Halorhodospira abdelmalekii and Halorhodospira halochloris (Ectothiorhodospiraceae) contain dihydroxylycopene diglucoside, its fatty acid esters and methoxy-hydroxylycopene glucoside (Fig. 8) as major components and no spirilloxanthin (Takaichi et al., 2001b). Therefore, the carotenogenesis pathways are not consistent with the classification of these bacteria in Gammaproteobacteria. BChl b is found only eight species of anaerobic
Shinichi Takaichi purple bacteria, three species of Alphaproteobacteria and five species of Gammaproteobacteria. Concerning carotenoids of these bacteria, four species have 3,4,3´,4´-tetrahydrospirilloxanthin (Fig. 6), two species of Blastochloris have the derivatives of lycopene (Fig. 8), and two species of Halorhodospira have the derivatives of dihydroxylycopene (Fig. 8) as described above. Thus, these species lack CrtD activity. Therefore, there is a relationship between the presence of BChl b and lack of CrtD activity, but the distribution of types of carotenoids and BChls is inconsistent with the classification of these purple sulfur bacteria. B. Aerobic Photosynthetic Bacteria Recently, many obligately aerobic bacteria have been found that have a photosynthetic apparatus and BChl a, and all of them belong to Alphaproteobacteria and Betaproteobacteria (Table 1) (Brenner et al., 2005; Chapter 3, Yurkov and Csotonyi). Some genera also contain non-photosynthetic species. They are distinguished from anaerobic photosynthetic bacteria in that they synthesize BChl only under aerobic conditions and cannot grow without oxygen even in the light. Photosynthetic activity has been demonstrated in some species (Harashima et al., 1982). The low content of BChl, unique composition of carotenoids, and the presence of carotenoids that appear to have no photosynthetic activities are also marked characteristics. RC complexes purified from Roseobacter denitrificans (Shimada et al., 1985; Takamiya et al., 1987) and also from Acidiphilium rubrum (Shimada et al., 1998) are similar to those of the anaerobic purple bacteria. RC-LH1 complexes have been found in most species, while many species also contain LH2 complexes. The properties of these antenna complexes are also similar to those of the anaerobic purple bacteria (Shimada, 1995). In some species, the pigments have been clearly identified, but in others no identification has been reported (Table 1). The only BChl found was BChl a with a phytol ester. Exceptionally, the acidophilic genus Acidiphilium has Zn-BChl a, where the central metal is zinc instead of the usual magnesium, although a small amount of the usual Mg-BChl a is also present (Wakao et al., 1996; Hiraishi et al., 1998). The carotenoid compositions of most of these bacteria are different from those of the anaerobic purple bacteria (Table 1) (Takaichi, 1999). Most species so
Chapter 6
Carotenoid Biosynthesis
far investigated contain spirilloxanthin, the content of which varies from low to high depending on species. These bacteria can be classified into five groups based upon their carotenoid composition. The first group has spirilloxanthin and its precursors, and spirilloxanthin is dominant. Acidiphilium (alpha-1), Rhizobium (alpha-2) and Roseoateles (beta) belong to this group. The second has spheroidene and its derivatives, and spheroidenone is dominant. Roseobacter (alpha-3) and Erythromonas (alpha-4) belong to this group. The third has a small amount of spirilloxanthin and large amounts of other carotenoids. Craurococcus (alpha-1), Paracraurococcus (alpha-1) and Methylobacterium (alpha-2) contain unidentified carotenoic acids, and Bradyrhizobium (alpha-2) contains canthaxanthin (Fig. 8). In the fourth group, Roseococcus thiosulfatophilus (alpha-1) and Methylobacterium rhodinum (alpha-2) have the diapocarotene derivative di(acyl-glucosyl)-diapocarotene-dioate (Fig. 8). Unidentified carotenoic acids in the third group seem to be different from this diapocarotene derivative (S. Takaichi, unpublished). The fifth group has unique Erythrobacter-type (Erb-type) carotenoids including spirilloxanthin and its precursors, γ-carotene and its cross-conjugated aldehyde derivative, β-carotene and its poly-hydroxy derivatives, and carotenoid sulfates. This group is found only in Sphingomonadales (alpha-4): Erythrobacter, Porphyrobacter and Erythromicrobium. Bradyrizobium sp. ORS278 produces both spirilloxanthin and canthaxanthin (Fig. 8), and two complete carotenogenesis gene clusters have been found (Table 2) (Hannibal et al., 2000; Giraud et al., 2004). Therefore, it has two each of crtE, crtB and crtI, and also lycopene cyclase (CrtY) and β-carotene ketolase (CrtW) for canthaxanthin synthesis. The functions of CrtI and CrtY from Erythrobacter (Erb.) longus (Matsumura et al., 1997) and CrtI from Methylobacterium extorquens (van Dien et al., 2003) have been confirmed. Erb. longus produces typical Erb-type carotenoids. More than 20 different carotenoids have been identified including some novel ones (Fig. 9) (Takaichi and Shimada, 1992; Takaichi et al., 1988, 1990, 1991a). About 70% of the total are the carotenoid sulfates, erythroxanthin sulfate and caloxanthin sulfate. The presence of such carotenoid sulfates is very rare in nature (Britton et al., 2004). These carotenoid sulfates do not function as light-harvesting pigments (Noguchi et al., 1992), and indeed they cannot bind to either
113 the RC or the LH1 complexes (Shimada et al., 1985). The carotenoids bound to the RC-LH1 complex are less polar ones. Among them, bacteriorubixanthinal is a unique carotenoid (Fig. 9). It has a cross-conjugated aldehyde group at C-19´, the ψ end group is methoxylated and 3,4-desaturated, and the other end group is a 3-hydroxy-β end group, which is a typical end group of plant carotenoids. β-Carotene and its poly-hydroxy derivatives, caloxanthin and nostoxanthin, are also found in the RC-LH1 complex. It should be noted that spirilloxanthin, which is a typical carotenoid in anaerobic purple bacteria, is also found in this RC-LH1 complex, although the amount is small. Bacteriorubixanthinal, and zeaxanthin or nostoxanthin are the major components in the RCLH1 complex of the Erb-type bacteria. Interestingly, these carotenoids seem to share the same binding site in the complex despite their structural differences (K. Shimada, personal communication). In Methylobacterium radiotolerans (previously Pseudomonas radiora), spirilloxanthin is the dominant carotenoid in the RC-LH1 complex (Saitoh et al., 1995), while carotenoic acids are found in the outer membranes accompanied by no BChls and have no photosynthetic functions (S. Saitoh, personal communication). Two species of Methylobacterium also have carotenoic acids similar to those of Methylobacterium radiotolerans, and two species of the fourth group described above have polar carotenoids, which are diapocarotenoic acid derivatives. These highly polar carotenoids in the third, the fourth and the fifth groups may be ‘non-photosynthetic’ carotenoids, which are not bound to the photosynthetic pigmentprotein complexes (Shimada, 1995). Although their functions are not known, there is the possibility that they protect the photosynthetic apparatus from the outside aerobic conditions. The carotenoid composition of the first and the second groups is somewhat different from the other groups, since they lack ‘non-photosynthetic’ carotenoids. The RC-LH1 complex of Acidiphilium rubrum (Wakao et al., 1996; Shimada et al., 1998), and the RC-LH1 and LH2 complexes of Roseobacter denitrificans (Shimada et al., 1985) contain all the cellular carotenoids. In conclusion, most aerobic photosynthetic bacteria have the purple bacteria-like photosynthetic apparatus including BChl a, and many species have spirilloxanthin, as well as additional polar ‘non-photosynthetic’ carotenoids.
Shinichi Takaichi
114
β-Carotene β-Cryptoxanthin
Zeaxanthin Caloxanthin Nostoxanthin Caloxanthin sulfate Erythroxanthin sulfate
R2 H H H OH OH OH H
R3 H OH OH OH OH OH O-SO2OH
R4 H2 H2 H2 H2 H2 H2 O
R2´ H H H H OH H OH
R3´ H H OH OH OH O-SO2OH OH
Fig. 9. Structures of the major Erb-type carotenoids.
Acknowledgments The author wishes to thank K. Shimada for reading the manuscript. References Albrecht M, Ruther A and Sandmann G (1997) Purification and biochemical characterization of a hydroxyneurosporene desaturase involved in the biosynthetic pathway of the carotenoid spheroidene in Rhodobacter sphaeroides. J Bacteriol 179: 7462–7467 Armstrong GA (1995) Genetic analysis and regulation of carotenoid biosynthesis: Structure and function of the crt genes and gene products. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 1135–1157. Kluwer Academic Publishers, Dordrecht Armstrong GA (1997) Genetics of eubacterial carotenoid biosynthesis: A colorful tale. Annu Rev Microbiol 51: 629–659 Armstrong GA, Alberti M, Leach F and Hearst JE (1989) Nucleotide sequence, organization, and nature of the protein products of the carotenoid biosynthesis gene cluster of Rhodobacter capsulatus. Mol Gen Genet 216: 254–268 Brenner DJ, Krieg NR and Staley JT (2005) Bergey’s Manual of Systematic Bacteriology, 2nd edition, Vol 2, The Proteobacteria. Springer, New York
Britton G, Singh RK, Goodwin TW and Ben-Aziz A (1975) The carotenoids of Rhodomicrobium vannielii (Rhodospirillaceae) and the effect of diphenylamine on the carotenoid composition. Phytochemistry 14: 2427–2433 Britton G, Liaaen-Jensen S and Pfander H (eds) (2004) Carotenoids Handbook. Birkhaüser, Basel Bryantseva IA, Gorlenko VM, Kompantseva EI and Imhoff JF (2000) Thioalkalicoccus limnaeus gen. nov., sp. nov., a new alkaliphilic purple sulfur bacterium with bacteriochlorophyll b. Int J Syst Evol Microbiol 50: 2157–2163 Caumette P, Baulaigue R and Matheron R (1991) Thiocapsa halophila sp. nov., a new halophilic phototrophic purple sulfur bacterium. Arch Microbiol 155: 170–176 Caumette P, Guyoneaud R, Imhoff JF, Süling J and Gorlenko V (2004) Thiocapsa marina sp. nov., a novel, okenone-containing, purple sulfur bacterium isolated from brackish coastal and marine environments. Int J Syst Evol Microbiol 54: 1031–1036 Davies BH (1970) A novel sequence for phytoene dehydrogenation in Rhodospirillum rubrum. Biochem J 116: 93–99 Dilling W, Liesack W and Pfennig N (1995) Rhabdochromatium marinum gen. nom. rev., sp. nov., a purple sulfur bacterium from a salt marsh microbial mat. Arch Microbiol 164: 125–131 Eimhjellen KE, Steensland H and Traetteberg J (1967) A Thiococcus sp. nov. gen., its pigments and internal membrane system. Arch Mikrobiol 59: 82–92 Francis GW and Liaaen-Jensen S (1970) Bacterial carotenoids: XXXIII. Carotenoids of Thiorhodaceae: 9. The structures of the carotenoids of the rhodopinal series. Acta Chem Scand 24: 2705–2712 Frank HA, Chadwick BW, Taremi S, Kolaczkowski S and Bowman
Chapter 6
Carotenoid Biosynthesis
MK (1986) Singlet and triplet absorption spectra of carotenoids bound in the reaction centers of Rhodopseudomonas sphaeroides R26. FEBS Lett 203: 157–163 Fujii R, Chen CH, Mizoguchi T and Koyama Y (1998) 1H NMR, electronic-absorption and resonance Raman spectra of isomeric okenone as compared with those of isomeric β-carotene, canthaxanthin, β-apo-8´-carotenal and spheroidene. Spectrochim Acta A54: 727–743 Garcia-Asua G, Cogdell RJ and Hunter CN (2002) Functional assembly of the foreign carotenoid lycopene into the photosynthetic apparatus of Rhodobacter sphaeroides achieved by replacement of the native 3-step phytoene desaturase with its 4-step counterpart from Erwinia herbicola. Mol Microbiol 44: 233–244 Gardiner AT, Cogdell RJ and Takaichi S (1993) The effect of growth conditions on the light-harvesting apparatus in Rhodopseudomonas acidophila. Photosynth Res 38: 159–167 Giraud E, Hannibal L, Fardoux J, Jaubert M, Jourand P, Dreyfus B, Sturgis JN and Verméglio A (2004) Two distinct crt gene clusters for two different functional classes of carotenoid in Bradyrhizobium. J Biol Chem 279: 15076–15083 Giuliano G, Giliberto L and Rosati C (2002) Carotenoid isomerase: A tale of light and isomers. Trends Plant Sci 7: 427–429 Goodwin TW (1956) The carotenoids of photosynthetic bacteria: II. The carotenoids of a number of non-sulphur purple photosynthetic bacteria (Athiorhodaceae). Arch Mikrobiol 24: 313–322 Guyoneaud R, Matheron R, Liesack W, Imhoff JF and Caumette P (1997) Thiorhodococcus minus, gen. nov., sp. nov., a new purple sulfur bacterium isolated from coastal lagoon sediments. Arch Microbiol 168: 16–23 Hannibal L, Lorquin J, D’Ortoli NA, Garcia N, Chaintreuil C, Masson-Boivin C, Dreyfus B and Giraud E (2000) Isolation and characterization of canthaxanthin biosynthesis genes from the photosynthetic bacterium Bradyrhzobium sp. strain ORS278. J Bacteriol 182: 3850–3853 Harada J, Nagashima KVP, Takaichi S, Misawa N, Matsuura K and Shimada K (2001) Phytoene desaturase, CrtI, of the purple photosynthetic bacterium, Rubrivivax gelatinosus, produces both neurosporene and lycopene. Plant Cell Physiol 42: 1112–1118 Harashima K and Nakada H (1983) Carotenoids and ubiquinone in aerobically grown cells of an aerobic photosynthetic bacterium, Erythrobacter species OCh 114. Agric Biol Chem 47: 1057–1063 Harashima K, Nakagawa M and Murata N (1982) Photochemical activities of bacteriochlorophyll in aerobically grown cells of aerobic heterotrophs, Erythrobacter species (OCh 114) and Erythrobacter longus (OCh 101). Plant Cell Physiol 23: 185–193 Hiraishi A, Nagashima KVP, Matsuura K, Shimada K, Takaichi S, Wakao N and Katayama Y (1998) Phylogeny and photosynthetic features of Thiobacillus acidophilus and related acidophilic bacteria: Its transfer to the genus Acidiphilium as Acidiphilium acidophium comb. nov. Int J Syst Bacteriol 48: 1389–1398 Hundle BS, O’Brien DA, Alberti M, Beyer P and Hearst JE (1992) Functional expression of zeaxanthin glucosyltransferase from Erwinia herbicola and a proposed uridine diphosphate binding site. Proc Natl Acad Sci USA 89: 9321–9325 Hunter CN, Hundle BS, Hearst JE, Gardiner AT, Takaichi S and Cogdell RJ (1994) Introduction of new carotenoids into the
115 bacterial photosynthetic apparatus by combining the carotenoid biosynthetic pathways of Erwinia herbicola and Rhodobacter sphaeroides. J Bacteriol 176: 3692–3697 Imhoff JF and Madigan MT (2004) International Committee on Systematics of Prokaryotes Subcommittee on the taxonomy of phototrophic bacteria. Minutes of the meetings, 27 august 2003, Tokyo, Japan. Int J Syst Evol Microbiol 54: 1001–1003, and Erratum 1907 Imhoff JF and Pfennig N (2001) Thioflavicoccus mobilis gen. nov., sp. nov., a novel purple sulfur bacterium with bacteriochlorophyll b. Int J Syst Evol Microbiol 51: 105–110 Imhoff JF, Petri R and Süling J (1998a) Reclassification of species of the spiral-shaped phototrophic purple non-sulfur bacteria of the α-Proteobacteria: description of the new genera Phaeospirillum gen. nov., Rhodovibrio gen. nov., Rhodothalassium gen. nov. and Roseospira gen. nov. as well as transfer of Rhodospirillum fulvum to Phaeospirillum fulvum comb. nov., of Rhodospirillum molischianum to Phaeospirillum molischianum comb. nov., of Rhodospirillum salinarum to Rhodovibrio salinarum comb. nov., of Rhodospirillum sodomense to Rhodovibrio sodomensis comb. nov., of Rhodospirillum salexigens to Rhodothalassium salexigens comb. nov. and of Rhodospirillum mediosalinum to Roseospira mediosalina comb. nov. Int J Syst Bacteriol 48: 793–798 Imhoff JF, Süling J and Petri R (1998b) Phylogenetic relationships among the Chromatiaceae, their taxonomic reclassification and description of the new genera Allochromatium, Halochromatium, Isochromatium, Marichromatium, Thiococcus, Thiohalocapsa and Thermochromatium. Int J Syst Bacteriol 48: 1129–1143 IUPAC Commission on Nomenclature of Organic Chemistry and the IUPAC-IUB Commission on Biochemical Nomenclature (1975) Nomenclature of carotenoids. Pure Appl Chem 41: 407–431 Jung DO, Achenbach LA, Karr EA, Takaichi S and Madigan MT (2004) A gas vesiculate planktonic strain of the purple non-sulfur bacterium Rhodoferax antarcticus isolated from Lake Fryxell, Dry Valleys, Antarctica. Arch Microbiol 182: 236–243 Kleinig H, Schmitt R, Meister W, Englert G and Thommen H (1979) New C30-carotenoic acid glucosyl esters from Pseudomonas rhodos. Z Naturforsch 34 c: 181–185 Komori M, Ghosh R, Takaichi S, Hu Y, Mizoguchi T, Koyama Y and Kuki M (1998) A null lesion in the rhodopin 3,4-desaturase of Rhodospirillum rubrum unmasks a cryptic branch of the carotenoid biosynthetic pathway. Biochemistry 37: 8987–8994 Kovács ÁT, Rákhely G and Kovács KL (2003) Genes involved in the biosynthesis of photosynthetic pigments in the purple sulfur photosynthetic bacterium Thiocapsa roseopersicina. Appl Environ Microbiol 69: 3093–3102 Lang HP, Cogdell RJ and Hunter CN (1994) Early steps in carotenoid biosynthesis: Sequence and transcriptional analysis of the crtI and crtB genes of Rhodobacter sphaeroides. Overexpression and reactivation of crtI in Escherichia coli and Rhodobacter sphaeroides. J Bacteriol 176: 3859–3869 Lang HP, Cogdell RJ, Takaichi S and Hunter CN (1995) Complete DNA sequence, specific Tn5 insertion map, and gene assignment of the carotenoid biosynthesis pathway of Rhodobacter sphaeroides. J Bacteriol 177: 2064–2073 Lichtenthaler HK (1999) The 1-deoxy-D-xylulose-5-phosphate pathway of isoprenoid biosynthesis in plants. Annu Rev Plant
116 Physiol Plant Mol Biol 50: 47–65 Linden H, Misawa N, Chamovitz, D, Pecker I, Hirscherg and Sandmann G (1991) Functional complementation in Escherichia coli of different phytoene desaturase genes and analysis of accumulated carotenes. Z Naturforsch 46 c: 1045–1051 Maeda I, Yamashiro H, Yoshioka D, Onodera M, Ueda S, Miyasaka H, Umeda F, Kawase M, Takaichi S and Yagi K (2005) Unusual accumulation of demethylspheroidene in anaerobicphototrophic growth of crtA-deleted mutants of Rhodovulum sulfidophilum. Curr Microbiol 51: 193–197 Maeda I, Yamashiro H, Yoshioka D, Onodera M, Ueda S, Kawase M, Miyasaka H and Yagi K (2006) Colorimetric dimethyl sulfide sensor using Rhodovulum sulfidophilum cells based on intrinsic pigment conversion by CrtA. Appl Microbiol Biotechnol 70: 397–402 Malhotra HC, Britton G and Goodwin TW (1970) A novel series of 1,2-dihydro carotenoids. Int J Vit Res 40: 315–322 Maresca JA and Bryant DA (2006) Two new encoding new carotenoid-modifying enzymes in the green sulfur bacterium Chlorobium tepidum. J Bacteriol 188: 6217–6223 Matsuura K and Shimada K (1993) Electrochromic spectral band shift of carotenoids in the photosynthetic membranes of Rhodospirillum molischianum and Rhodospirillum photometricum. Biochim Biophys Acta 1140: 293–296 Matsumura H, Takeyama H, Kusakabe E, Burgess JG and Matsunaga T (1997) Cloning, sequencing and expressing the carotenoid biosynthesis genes, lycopene cyclase and phytoene desaturase, from the aerobic photosynthetic bacterium Erythrobacter longus sp. strain OCh 101 in Escherichia coli. Gene 189: 169–174 McDermott G, Prince SM, Freer AA, Hawthornthwaite-Lawless AM, Papiz MZ, Cogdell RJ and Isaacs NW (1995) Crystal structure of an integral membrane light-harvesting complex from photosynthetic bacteria. Nature 374: 517–521 Misawa N, Nakagawa M, Kobayashi K, Yamano S, Izawa Y, Nakamura K and Harashima K (1990) Elucidation of the Erwinia uredovora carotenoid biosynthetic pathway by functional analysis of gene products expressed in Escherichia coli. J Bacteriol 172: 6704–6712 Noguchi T, Hayashi H, Shimada K, Takaichi S and Tasumi M (1992) In vivo states and functions of carotenoids in an aerobic photosynthetic bacterium, Erythrobacter longus. Photosynth Res 31: 21–30 Ouchane S, Picaud M, Vernotte C, Reiss-Husson F and Astier C (1997a) Pleiotropic effects of puf interposon mutagenesis on carotenoid biosynthesis in Rubrivivax gelatinosus. J Biol Chem 272: 1670–1676 Ouchane S, Picaud M, Vernotte C and Astier C (1997b) Photooxidative stress stimulates illegitimate recombination and mutability in carotenoid-less mutants of Rubrivivax gelatinosus. EMBO J 16: 4777–4787 Overmann J, Fischer U and Pfennig N (1992) A new purple sulfur bacterium from saline littoral sediments, Thiorhodovibrio winogradskyi gen. nov. and sp. nov. Arch Microbiol 157: 329–335 Pfennig N, Markham MC and Liaaen-Jensen S (1968) Carotenoids of Thiorhodaceae: 8. Isolation and characterization of a Thiothece, Lamprocystis and Thiodictyon strain and their carotenoid pigments. Arch Mikrobiol 62: 178–191 Pfennig N, Lünsdorf H, Süling J and Imhoff JF (1997) Rhodospira trueeperi gen. nov., spec. nov., a new phototrophic Proteobacterium of the alpha group. Arch Microbiol 168: 39–45
Shinichi Takaichi Pinta V, Ouchane S, Picaud M, Takaichi S, Astier C and ReissHusson F (2003) Characterization of unusual hydroxy- and ketocarotenoids in Rubrivivax gelatinosus: Involvement of enzyme CrtF or CrtA. Arch Microbiol 179: 354–362 Raisig A, Bartley G, Scolnik P and Sandmann G (1996) Purification in an active state and properties of the 3-step phytoene desaturase from Rhodobacter capsulatus overexpressed in Escherichia coli. J Biochem 119: 559–564 Rees GN, Harfoot CG, Janssen PH, Schoenborn L, Kuever J and Lünsdorf H (2002) Thiobaca trueperi gen. nov., sp. nov., a phototrophic purple sulfur bacterium isolated from freshwater lake sediment. Int J Syst Evol Microbiol 52: 671–678 Reslewic S, Zhou S, Place M, Zhang Y, Briska A, Gokdstein S, Churas C, Runnheim R, Forrest D, Lim A, Lapidus A, Han CS, Roberts GP and Schwartz DC (2005) Whole-genome shotgun optical mapping of Rhodospirillum rubrum. Appl Environ Microbiol 71: 5511–5522 Rohmer M (1999) The discovery of a mevalonate-independent pathway for isoprenoid biosynthesis in bacteria, algae and higher plants. Nat Prod Rep 16: 565–574 Ryvarden L and Liaaen-Jensen S (1964) Bacterial carotenoids: XIV. The carotenoids of Rhodomicrobium vannielii. Acta Chem Scand 18: 643–654 Saitoh S, Takaichi S, Shimada K and Nishimura Y (1995) Identification and subcellular distribution of carotenoids in the aerobic photosynthetic bacterium, Pseudomonas radiora strain MD-1. Plant Cell Physiol 36: 819–823 Sandmann G (1994) Carotenoid biosynthesis in microorganisms and plants. Eur J Biochem 223: 7–24 Sandmann G (1997) High level expression of carotenogenic genes for enzyme purification and biochemical characterization. Pure Appl Chem 69: 2163–2168 Schmidt K (1971) Carotenoids of purple nonsulfur bacteria: Composition and biosynthesis of the carotenoids of some strains of Rhodopseudomonas acidophila, Rhodospirillum tenue, and Rhodocyclus purpureus. Arch Mikrobiol 77: 231–238 Schmidt K (1978) Biosynthesis of carotenoids. In: Clayton RK and Sistrom WR (eds) The Photosynthetic Bacteria, pp 729–750. Plenum Press, New York Schmidt K and Liaaen-Jensen S (1973) Bacterial carotenoids: XLII. New keto-carotenoids from Rhodopseudomonas globiformis (Rhodospirillaceae). Acta Chem Scand 27: 3040–3052 Schmidt K, Liaaen-Jensen S and Schlegel HG (1963) Die Carotinoide der Thiorhodaceae: I. Okenon als Hauptcarotinoid von Chromatium okenii Perty. Arch Mikrobiol 46: 117–126 Schmidt K, Pfennig N and Liaaen-Jensen S (1965) Carotenoids of Thiorhodaceae: IV. The carotenoid composition of 25 pure isolates. Arch Mikrobiol 52: 132–146 Schmidt K, Francis GW and Liaaen-Jensen S (1971) Bacterial carotenoids: XXXVI. Remarkable C43-carotenoid artefacts of cross-conjugated carotenals and new carotenoid glucosides from Athiorhodaceae spp. Acta Chem Scand 25: 2476–2486 Schwerzmann RU and Bachofen R (1989) Carotenoid profiles in pigment-protein complexes of Rhodospirillum rubrum. Plant Cell Physiol 30: 497–504 Scolnik PA, Walker MA and Marrs BL (1980) Biosynthesis of carotenoids derived from neurosporene in Rhodopseudomonas capsulata. J Biol Chem 255: 2427–2432 Shimada K (1995) Aerobic anoxygenic phototrophs. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and
Chapter 6
Carotenoid Biosynthesis
Respiration, Vol 2), pp 105–122. Kluwer Academic Publishers, Dordrecht Shimada K, Hayashi H and Tasumi M (1985) Bacteriochlorophyll-protein complexes of aerobic bacteria, Erythrobacter longus and Erythrobacter species OCh 114. Arch Microbiol 143: 244–247 Shimada K, Itoh S, Iwaki M, Nagashima KVP, Matuura K, Kobayashi M and Wakao N (1998) Reaction center complex based on Zn-bacteriochlorophyll from Acidiphilium rubrum. In: Garab G (ed) Photosynthesis: Mechanism and Effects, Vol II, pp 909–912. Kluwer Academic Publishers, Dordrecht Shneour EA (1962) The source of oxygen in Rhodopseudomonas spheroides carotenoid pigment conversion. Biochim Biophys Acta 65: 510–511 Singh RK, Britton G and Goodwin TW (1973) Carotenoid biosynthesis in Rhodopseudomonas spheroides: S-adenosylmethionine as the methylating agent in the biosynthesis of spheroidene and spheroidenone. Biochem J 136: 413–419 Swingley WD, Sadekar S, Mastrian SD, Matthies HJ, Hao J, Ramos H, Acharya CR, Conrad AL, Taylor HL, Dejesa LC, Shah MK, O’Huallachain ME, Lince MT, Blankenship RE, Beatty JT and Touchman JW (2007) The complete genome sequence of Roseobacter denitrificans reveals a mixotrophic rather than photosynthetic metabolism. J Bacteriol 189: 683–690 Takaichi S (1999) Carotenoids and carotenogenesis in anoxygenic photosynthetic bacteria. In: Frank HA, Young AJ, Britton G and Cogdell RJ (eds) The Photochemistry of Carotenoids (Advances in Photosynthesis and Respiration, Vol 8), pp 39–69. Kluwer Academic Publishers, Dordrecht Takaichi S and Shimada K (1992) Characterization of carotenoids in photosynthetic bacteria. Methods Enzymol 213: 374–385 Takaichi S, Shimada K and Ishidsu J (1988) Monocyclic crossconjugated carotenal from an aerobic photosynthetic bacterium, Erythrobacter longus. Phytochemistry 27: 3605–3609 Takaichi S, Shimada K and Ishidsu J (1990) Carotenoids from the aerobic photosynthetic bacterium, Erythrobacter longus: β-carotene and its hydroxyl derivatives. Arch Microbiol 153: 118–122 Takaichi S, Furihata K, Ishidsu J and Shimada K (1991a) Carot-
117 enoid sulphates from the aerobic photosynthetic bacterium, Erythrobacter longus. Phytochemistry 30: 3411–3415 Takaichi S, Furihata K and Harashima K (1991b) Light-induced changes of carotenoid pigments in anaerobic cells of the aerobic photosynthetic bacterium, Roseobacter denitrificans (Erythrobacter species OCh 114): Reduction of spheroidenone to 3,4-dihydrospheroidenone. Arch Microbiol 155: 473–476 Takaichi S, Jung DO and Madigan MT (2001a) Accumulation of unusual carotenoids in the spheroidene pathway, demethylspheroidene and demethylsphroidenone, in an alkaliphilic purple nonsulfur bacterium Rhodobaca bogoriensis. Photosynth Res 67: 207–214 Takaichi S, Maoka T, Hanada S and Imhoff JF (2001b) Dihydroxylycopene diglucoside diesters: a novel class of carotenoids from the phototrophic purple sulfur bacteria Halorhodospira abdelmalekii and Halorhodospira halochloris. Arch Microbiol 175: 161–167 Takamiya K, Iba K and Okamura K (1987) Reaction center complex from an aerobic photosynthetic bacterium, Erythrobacter species OCh 114. Biochim Biophys Acta 890: 127–133 van Dien, SJ, Marx CJ, O’Brien BN and Lidstrom ME (2003) Genetic characterization of the carotenoid biosynthetic pathway in Methylobacterium extorquens AM1 and isolation of a colorless mutant. App Environ Microbiol 69: 7563–7566 Wakao N, Yokoi N, Isoyama N, Hiraishi A, Shimada K, Kobayashi M, Kise H, Iwaki M, Itoh S, Takaichi S and Sakurai Y (1996) Discovery of natural photosynthesis using Zn-containing bacteriochlorophyll in an aerobic bacterium Acidiphilium rubrum. Plant Cell Physiol 37: 889–893 Yeliseev AA and Kaplan S (1997) Anaerobic carotenoid biosynthesis in Rhodobacter sphaeroides 2.4.1: H2O is a source of oxygen for the 1-methoxy group of spheroidene but not for the 2-oxo group of spheroidenone. FEBS Lett 403: 10–14 Yurkov V, Gad’on N and Drews G (1993) The major part of polar carotenoids of the aerobic bacteria Roseococcus thiosulfatophilus RB3 and Erythromicrobium ramosum E5 is not bound to the bacteriochlorophyll a-complexes of the photosynthetic apparatus. Arch Microbiol 160: 372–376
Chapter 7 Membrane Lipid Biosynthesis in Purple Bacteria Banita Tamot and Christoph Benning* Department of Biochemistry and Molecular Biology, Michigan State University, East Lansing, MI 48824-1319
Summary ............................................................................................................................................................... 119 I. Introduction..................................................................................................................................................... 120 II. Fatty Acids...................................................................................................................................................... 122 III. Phosphoglycerolipids ..................................................................................................................................... 123 A. Phosphatidic Acid and Cytidine Diphosphate Diacylglycerol ........................................................... 123 B. Phosphatidylethanolamine ............................................................................................................... 124 C. Phosphatidylglycerol and Cardiolipin ............................................................................................... 124 D. Phosphatidylcholine ......................................................................................................................... 125 IV. Glycoglycerolipids .......................................................................................................................................... 126 A. Glycerolipids Containing Glucose or Galactose ............................................................................... 126 B. Sulfoquinovosyldiacylglycerol........................................................................................................... 127 V. Betaine Lipid................................................................................................................................................... 128 VI. Ornithine Lipid ................................................................................................................................................ 129 VII. Lipid Function ................................................................................................................................................. 130 A. Changes in Lipid Composition in Response to the Environment ..................................................... 130 B. Interaction of Lipids with Membrane Proteins .................................................................................. 131 VIII. Perspectives ................................................................................................................................................... 131 Acknowledgments ................................................................................................................................................. 131 References ............................................................................................................................................................ 132
Summary Membranes are essential to all living cells. They provide the boundary to the surrounding environment, allow the controlled exchange of compounds through membrane transporters, and serve as a matrix for membrane associated enzymes and protein complexes involved in the generation of energy or communication with the environment. Biomembranes are built from amphipathic, polar lipids that either on their own or in mixtures with other lipids spontaneously form a bilayer in aqueous solutions. Proteins are embedded into this lipid matrix serving many different functions. Photosynthetic purple bacteria have a very rich complement of membrane lipids including phospholipids not commonly found in bacteria such as phosphatidylcholine, glycolipids typical for plant chloroplasts such as sulfoquinovosyldiacylglycerol, and the betaine and ornithine lipids. These latter lipids lack phosphorus presumably allowing purple bacteria to outcompete other organisms in a phosphorusdepleted environment. Advances in the genetic analysis of lipid metabolism of purple bacteria and related bacteria of the α-proteobacteria group have provided us with many genes encoding enzymes for the biosynthesis of polar membrane lipids beyond those described for Escherichia (E.) coli. Lipid genes discovered first in *Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 119–134. © 2009 Springer Science + Business Media B.V.
120
Banita Tamot and Christoph Benning
Rhodobacter (Rba.) sphaeroides include the sqd genes required for sulfoquinovosyldiacylglycerol biosynthesis and the bta genes required for betaine lipid biosynthesis. Similar approaches in the closely related bacterium Sinorhizobium (Sr.) meliloti in combination with genome comparison provided the genes of purple bacteria encoding proteins of ornithine lipid biosynthesis and a new pathway for phosphatidylcholine biosynthesis. Advances in the analysis of membrane associated protein complexes gave new insights into specific interactions of lipids with these complexes. I. Introduction Much of what we know today about membrane lipid biosynthesis and the basic pathways for the biosynthesis of phospholipids in Gram-negative bacteria is based on work with E. coli as recently reviewed (Cronan, 2003). However, E. coli membranes are limited in the number of polar lipids with the bulk provided by phosphatidylethanolamine (PtdEtn), phosphatidylglycerol (PtdGro) and cardiolipin (CL). Beyond these lipids, membranes of purple bacteria such as Rba. sphaeroides and Rba. capsulatus contain phosphatidylcholine (PtdCho), glycolipids such as sulfoquinovosyldiacylglycerol (SQDG), the betaine lipid diacylglycerol-N,N,N,-trimethylhomoserine (DGTS) and the ornithine lipid (OL). Moreover, some strains of Rba. sphaeroides are known to produce considerable amounts of the xenobiotic lipid phosphatidyl Tris when grown in tris-hydroxymethylaminomethane (Tris) buffer (Schmid et al., 1991). The polar lipid and fatty acid composition of different purple bacteria and other anoxygenic photosynthetic bacteria have been extensively studied and reviewed (Imhoff et al., 1982; Imhoff, 1991; Imhoff and Bias-Imhoff, 1995; Benning, 1998b). Whenever fatty acid and polar lipid profiles are considered, it is important to note that the lipid composition of purple bacteria and other bacteria is dependent on the growth conditions. In particular phosphate-limited growth conditions can drastically increase the relative proportions of non-phosphorous lipids in purple bacteria (Benning et al., 1995), restricting the usefulness of lipid profiles for taxonomic classification of purple bacteria. However, because the completed genome sequences Abbreviations: ACP – acyl carrier protein; AdoHcy – S-adenosylhomocysteine; AdoMet – S-adenosylmethionine; CDP – cytidine diphosphate; CL – cardiolipin; CMP – cytidine monophosphate; CTP – cytidine triphosphate; DAG – diacylglycerol; DGTS – dicaylglycerol-N,N,N,-trimethylhomoserine; E. – Escherichia; OL – ornithine lipid; PtdCho – phosphatidylcholine; PtdEtn – phosphatidylethanolamine; PtdGro – phosphatidylglycerol; Rba. – Rhodobacter; SQDG – sulfoquinovosyldiacylglycerol; Sr. – Sinorhizobium; Tris – tris-hydroxymethylaminomethane
of many bacteria are now publicly available, direct DNA sequence comparison has become the primary measure for evolutionary relationships and the basis for taxonomic classifications of bacteria. As of September, 2006, the US National Center for Biological Information (NCBI) genome web site (http://www.ncbi.nlm.nih.gov/genomes/lproks.cgi) lists 48 completed genomes of α-proteobacteria including 9 genomes of purple bacteria. A large number of additional α-proteobacterial genome projects are under way. Here we take advantage of this information and provide the accession numbers for membrane lipid genes of Rba. sphaeroides 2.4.1 as an example representative for the purple bacteria (Table 1). We chose to focus on this particular organism because sequences for its two chromosomes and five plasmids (Mackenzie et al., 2001) are publicly available at NCBI and because several genes encoding enzymes involved in the biosynthesis of lipids specific to purple bacteria such as SQDG and DGTS were originally discovered in this bacterium. For the description of enzymes of basic phospholipid metabolism we rely on the annotations extrapolated from the experimentally confirmed orthologs of E. coli. In the case of PtdCho and OL biosynthesis the original work has been performed in Sr. meliloti or Rba. capsulatus and we refer to these organisms. In general, the focus will be on the biosynthesis of the major lipids found in the membranes of purple bacteria and the currently known biosynthetic pathways for these lipids. The availability of lipid-deficient genetic mutants and the presence of specific lipids in high-resolution crystal structures of several membrane associated complexes of purple bacteria led to new clues towards the function of membrane lipids. These findings are now providing a promising framework to better understand the complex lipid composition of purple bacterial membranes, as well as the remodeling of membranes under adverse conditions such as phosphate deficiency. These findings will be briefly summarized.
Chapter 7
Membrane Lipids
121
Table 1. Assignments of lipid genes in the Rba. sphaeroides 2.4.1 reference genome1 Enzyme
Gene
Chr
GenBank Protein ID
References2
Phospholipids Glycerol 3-phosphate 1-O- acyltransferase
plsB
unidentified
1-O-acylglycerol 3-phosphate 2-O-acyltransferase Phosphatidyl cytidyltransferase
plsC
1
YP_353812
cdsA
1
YP_352763
Phosphatidylserine synthase
pss
1
YP_353797
Phosphatidylserine decarboxylase
psd
1
YP_353796
Phosphatidylglycerol 3-phosphate synthase Phosphatidylglycerol 3-phosphate phosphatase
pgsA pgpA
1 1
YP_354158 YP_352878
Cardiolipin synthase
cls
1
YP_353546
Phosphatidylethanolamine N-methyltransferase Phosphatidylcholine synthase
pmtA pcs
1 1
YP_353798 YP_353639
Glycolipids Diacylglycerol glycosyltransferase Monoglycosyldiacylglycerol glycosyltransferase UDP-sulfoquinovose synthase
sqdB
1
YP_352627
Required for SQDG biosynthesis Required for SQDG biosynthesis
sqdA sqdC
1 1
YP_352034 YP_352625
UDP-sulfoquinovose sulfoquinovosyltransferase
sqdD
1
YP_352626
Betaine lipid Diacylglycerolhomoserine synthase
btaA
1
YP_353939
Diacylglycerolhomoserine N-methyltransferase
btaB
1
YP_353940
Ornithine lipid Lyso-ornithine beta-hydroxy acyltransferase
olsA
predicted (ubiquitous in Gram-negative bacteria) predicted (ubiquitous in Gram-negative bacteria) predicted (ubiquitous in Gram-negative bacteria) predicted (ubiquitous in Gram-negative bacteria) confirmed (Dryden and Dowhan, 1996) predicted (ubiquitous in Gram-negative bacteria) confirmed (Tamot and Benning, unpublished) confirmed (Arondel et al., 1993) predicted from confirmed Sinorhizobium meliloti ortholog (Sohlenkamp et al., 2000; Martinez-Morales et al., 2003)
unidentified unidentified confirmed (Benning and Somerville, 1992a; Benning et al., 1993) confirmed (Benning and Somerville, 1992b) confirmed (Benning and Somerville, 1992a; Rossak et al., 1997) confirmed (Benning and Somerville, 1992a; Rossak et al., 1995)
confirmed (Klug and Benning, 2001; Riekhof et al., 2005a) confirmed (Klug and Benning, 2001; Riekhof et al., 2005a)
predicted from confirmed Rba. capsulatus ortholog (Aygun-Sunar et al., 2006) Ornithine 2-N-acyltransferase olsB 2 YP_354511 predicted from confirmed Rba. capsulatus ortholog (Aygun-Sunar et al., 2006) 1 Chromosome 1, NC_007493, Chromosome 2, NC_007494, Copeland, A., Lucas, S., Lapidus, A., Barry, K., Detter, J.C., Glavina, T., Hammon, N., Israni, S., Pitluck, S., Richardson, P., Mackenzie, C., Choudhary, M., Larimer, F., Hauser, L.J., Land, M., Donohue, T.J. and Kaplan, S., The Rhodobacter sphaeroides Genome Consortium. GenBank submission by the US DOE Joint Genome Institute. 2 The notation ‘confirmed’ indicates that genetic or biochemical evidence is available for Rba. sphaeroides. In some cases, a confirmed sequence from a closely related bacterium was used to identify the putative Rba. sphaeroides 2.4.1 ortholog. 2
YP_354512
Banita Tamot and Christoph Benning
122 II. Fatty Acids With few exceptions, the biosynthesis of fatty acids and the proteins involved in this process have not been directly studied in purple bacteria. An acyl carrier protein (ACP) from Rba. sphaeroides has been purified (Cooper et al., 1987) as well as two long-chain acyl-CoA thioesterases involved in the metabolism of fatty acids taken up from the medium (Boyce and Lueking, 1984; Seay and Lueking, 1986). However, the inner workings of the type II fatty acid synthase typical for proteobacteria were elucidated primarily in E. coli. Crystal structures for most of the components are available and the complex was mechanistically explored in great detail as recently reviewed (White et al., 2005). For details, the reader is referred to this review and the references therein. Type II fatty acid synthases assemble from multiple components encoded by different genes. Their β-ketoacyl-ACP component is highly sensitive to cerulenin, which also inhibits fatty acid synthesis at low concentrations in Rba. sphaeroides (Broglie and Niederman, 1979). Furthermore, the anti-infective agent triclosan inhibits the enoyl-ACP-reductase associated with type II fatty acid synthases (McMurry et al., 1998). Based on the similarity of confirmed
E. coli proteins to translated peptide sequences from the genome of Rba. sphaeroides, the orthologous components of type II fatty acid synthase can be predicted as summarized in Table 2. Ultimately, these assignments need to be confirmed by biochemical analysis. As already mentioned, the genome of Rba. sphaeroides is comprised of two chromosomes and several plasmids (Mackenzie et al., 2001). Most of the proteins involved in fatty acid biosynthesis are encoded on chromosome 1 except for one β-ketoacyl-ACP synthase (FabA), β-hydroxydecanoyl-ACP dehydratase (FabA) and one enoyl-ACP reductase (Table 2). The fatty acid composition of membrane lipids in purple bacteria is relatively simple (Imhoff et al., 1982; Imhoff, 1991; Imhoff and Bias-Imhoff, 1995) as they lack desaturases that introduce specific double bonds into acyl chains post synthesis. Fatty acids typically found in membrane lipids of purple bacteria include palmitate, stearate and vaccenate (Fig. 1A). In addition, the ornithine lipid contains 3-hydroxy stearate. In glycerolipids, these acyl chains are esterified to the sn-1 and sn-2 position of the glycerol giving rise to the diacylglycerol moiety of the lipid molecule (Fig. 1B). An added layer of complexity arises from the fact that different molecular species
Table 2. Assignments of fatty acid biosynthetic genes in the Rba. sphaeroides 2.4.1 reference genome1
1
Gene
Enzyme
Chromosome
GenBank Protein ID2
acpP
Acyl carrier protein (ACP)
1
YP_352523
acpS
ACP synthase
1
YP_351722
accA
Carboxyltransferase subunit (CT-α) of AcetylCoA carboxylase (ACC)
1
YP_351821
accB
Biotin carboxyl carrier protein subunit of ACC (BCCP)
1
YP_353262 YP_353263
accC
Biotin carboxylase subunit of ACC (BC)
1
accD
Carboxyltransferase subunit of ACC (CT-β)
1
YP_354012
fabD
Malonyl-CoA:ACP-trasnacylase
1
YP_352738
fabH
β-ketoacyl-ACP synthase III
1
YP_352670
fabB
β-ketoacyl-ACP synthase I
2
YP_354693
fabF
β-ketoacyl-ACP synthase II
1
YP_352524
fabG
β-ketoacyl-ACP reductase
1
YP_352522
fabA
β-hydroxydecanoyl-ACP dehydratase
2
YP_354694
fabZ
β-hydroxyl-ACP dehydratase
1
YP_352769
fabI
Enoyl-ACP reductase I
1
YP_352401
Enoyl-ACP reductase similar to FabI
2
YP_354692
Enoyl-ACP reductase similar to FabI
1
YP_354336
Chromosome 1, NC_007493, Chromosome 2, NC_007494, Copeland, A., Lucas, S., Lapidus, A., Barry, K., Detter, J.C., Glavina, T., Hammon, N., Israni, S., Pitluck, S., Richardson, P., Mackenzie, C., Choudhary, M., Larimer, F., Hauser, L.J., Land, M., Donohue, T.J. and Kaplan, S., The Rhodobacter sphaeroides Genome Consortium. GenBank submission by the US DOE Joint Genome Institute. 2 All assignments are predictions based on sequence comparison with the E. coli orthologs.
Chapter 7
Membrane Lipids
123
Fig. 1. Common acyl groups found in membrane lipids of purple bacteria (A) and structure of divaccenoylglycerol (B). The full carbon numbering is given for stearate (A) and the glycerol moiety (B).
in each lipid class contain different combinations of acyl groups in the diacylglycerol moiety. III. Phosphoglycerolipids In phosphoglycerolipids a phosphate diester links the diacylglycerol sn-3 hydroxyl group with the terminal hydroxyl of the lipid head group. Phosphoglycerolipids are the prevalent lipids in cell membranes of animals and model organisms such as E. coli. Accordingly, phosphoglycerolipid biosynthesis has been well characterized in E. coli, as recently reviewed (Cronan, 2003), providing the paradigm for proteobacteria. Table 1 shows predicted and in some cases directly confirmed orthologs of enzymes involved in glycerolipid biosynthesis in Rba. sphaeroides. An overview of the pathways leading to the biosynthesis of the different phosphoglycerolipids is shown in Fig. 2. All phosphoglycerolipids described, up to cardiolipin, are also present in E. coli. The other lipids described below reflect the complex nature of membrane lipids and their biosynthesis in purple bacteria.
A. Phosphatidic Acid and Cytidine Diphosphate Diacylglycerol Phosphoglycerolipid biosynthesis starts with the formation of phosphatidic acid by sequential acylation of glycerol 3-phosphate at the sn-1 and sn-2 positions (Fig. 2). Different acyl-ACPs serve as acyl donors in these reactions. In E. coli and presumably in purple bacteria two acyltransferases PlsB and PlsC are involved, but sequence comparison identified only a possible ortholog for PlsB in Rba. sphaeroides (Table 1). There are other open reading frames encoding proteins with acyltransferase motifs, but the specific gene encoding the glycerol 3-phosphate acyltransferase remains to be identified by using genetic or biochemical methods. The assembly of some of the non-phosphorous glycerolipids present in purple bacteria starts with the formation of diacylglycerol. A phosphatidic acid phosphatase could be involved, but has not yet been implicated in purple bacteria. For the biosynthesis of phosphoglycerolipids, phosphatidic acid is converted to CDP-diacylglycerol. A predicted ortholog of the
124
Banita Tamot and Christoph Benning
Fig. 2. Biosynthesis of phosphoglycerolipids in purple bacteria. Enzyme nomenclature as shown in Table 1. DMPE, N,N-dimethyl phosphatidylethanolamine; MMPE, N-monomethyl phosphatidylethanolamine.
E. coli phosphatidate cytidylyltransferase, CdsA, is encoded in the genome of Rba. sphaeroides (Table 1). B. Phosphatidylethanolamine Phosphatidylethanolamine (PtdEtn) is at 70 – 80 mol% the predominant membrane lipid in E. coli. Phosphatidylethanolamine is required for growth unless large concentrations of divalent cations are present in the medium (DeChavigny et al., 1991). Furthermore, it acts as a molecular chaperone for the assembly of polytopic integral membrane proteins (Bogdanov et al., 1996). In purple bacteria such as Rba. sphaeroides PtdEtn is present at lower concentrations of 35–45 mol% (Russell and Harwood, 1979; Benning and Somerville, 1992b), but it is still the most abundant membrane lipid. The biosynthesis of PtdEtn from CDP-diacylglyc-
erol involves the transfer of serine and the release of CMP (Fig. 2). This reaction is catalyzed in E. coli by phosphatidylserine synthase (Pss) and a predicted ortholog is present in Rba. sphaeroides (Table 1). Biochemical analysis of Pss in Rba. sphaeroides suggests that unlike the E. coli ortholog the Rba. sphaeroides enzyme is membrane–bound and magnesium dependent (Radcliffe et al., 1989). Phosphatidylserine is an intermediate that does not accumulate, as it is rapidly converted to PtdEtn by phosphatidylserine decarboxylase (Psd). A predicted ortholog is present in Rba. sphaeroides suggesting that the biosynthesis in purple bacteria proceeds as determined for E. coli. A direct analysis of these enzymes in purple bacteria is still lacking. C. Phosphatidylglycerol and Cardiolipin Like PtdEtn, phosphatidylglycerol (PtdGro) is
Chapter 7
Membrane Lipids
synthesized from CDP-diacylglycerol. Glycerol 3phosphate serves as the head group donor leading to the formation of CMP and phosphatidylglycerol 3-phosphate (Fig. 2). The enzyme responsible, phosphatidylglycerol 3-phosphate synthase, PgsA (Table 1), was biochemically characterized in Rba. sphaeroides (Radcliffe et al., 1989), and the respective gene was identified (Dryden and Dowhan, 1996). Expression of the Rba. sphaeroides gene in an E. coli pgsA mutant could functionally replace the E. coli enzyme and overexpression of the gene in Rba. sphaeroides led to increased phosphatidylglycerol 3-phosphate synthase activity (Dryden and Dowhan, 1996). However, the phospholipid composition of the resulting strain was not significantly affected suggesting that the enzyme is regulated post-transcriptionally. Similarly, overexpression in E. coli did not affect the lipid composition suggesting that the Rba. sphaeroides and E. coli orthologs are similarly regulated. Phosphatidylglycerol is formed following the dephosphorylation of phosphatidylglycerol 3-phosphate. The gene encoding the respective phosphatase, PgpA, has been predicted for Rba. sphaeroides (Table 1). In bacteria the combination of two PtdGro molecules releases glycerol and leads to the formation of bisphosphatidylglycerol (cardiolipin, CL; Fig. 2). The CL synthase from Rba. sphaeroides, Cls, has been recently identified (Table 1) and a CL-deficient Rba. sphaeroides cls disruption mutant has been generated (Tamot and Benning, unpublished). Expression of the open reading frame in trans led to the restoration of CL biosynthesis. As observed for the respective E. coli mutant (Pluschke et al., 1978; Nishijima et al., 1988), small amounts of cardiolipin are still formed in the Rba. sphaeroides mutant. This mutant also produces more PtdGro, the precursor of CL biosynthesis. Results from a cls, pss-1 double mutant of E. coli suggest that phosphatidylserine synthase might be responsible for the formation of small amounts of cardiolipin in the cls mutant (Shibuya et al., 1985). The presence of an alternative Cls in Rba. sphaeroides has not yet been explored. The relative amount of CL is strongly increased in Rba. sphaeroides following osmotic stress (Catucci et al., 2004). Whether this is due to the upregulation of Cls activity or the activation of an alternative pathway is not yet clear. D. Phosphatidylcholine Phosphatidylcholine (PtdCho) is found in most eu-
125 karyotic cells with few exceptions, e.g. the green alga Chlamydomonas reinhardtii (Giroud et al., 1988), but it is not as prevalent in bacteria (Sohlenkamp et al., 2003). Based on estimates derived from sequenced bacterial genomes, approximately 10–20% of bacteria contain genes predicted to encode enzymes involved in PtdCho biosynthesis. The first gene directly required for bacterial PtdCho biosynthesis, pmtA, was discovered in Rba. sphaeroides (Arondel et al., 1993) and encodes a PtdEtn N-methyltransferase. Expression of the pmtA gene in E. coli led to the accumulation of PtdCho, which is normally absent from E. coli. The PmtA enzyme catalyzes the transfer of three methyl groups from AdoMet onto the terminal amino group of PtdEtn (Fig. 2). Interestingly, Sr. meliloti, an α-proteobacterium closely related to Rba. sphaeroides, uses a different form of PmtA (de Rudder et al., 1997). In general, bacteria capable of synthesizing PtdCho use either one or the other type of PmtA (Sohlenkamp et al., 2003). Introduction of PmtA from Sr. meliloti into E. coli led to the formation of PtdCho, thus confirming its potential for PtdCho biosynthesis. However, inactivation of the pmtA gene in Sr. meliloti did not cause a loss of PtdCho. Geiger and colleagues subsequently discovered a second pathway of PtdCho biosynthesis in Sr. meliloti (de Rudder et al., 1999) involving the transfer of choline onto CDP-DAG (Fig. 2). The enzyme responsible is the unique bacterial PtdCho synthase, Pcs, (Sohlenkamp et al., 2000). A survey of a number of bacteria suggests that most, but not all bacterial species with PmtA activity also have Pcs activity in their cell extracts (Martinez-Morales et al., 2003). One notable exception was Rba. sphaeroides that did not incorporate labeled choline into PtdCho, consistent with previous observations that the pmtA mutant of Rba. sphaeroides lacked PtdCho (Arondel et al., 1993). As the Rba. sphaeroides genome harbors a gene potentially coding for a protein similar to Pcs of Sr. meliloti (Martinez-Morales et al., 2003; Sohlenkamp et al., 2003), it is not clear whether Pcs activity is truly absent, or whether Pcs is only active under certain growth conditions. In this context, it is important to note that the predicted Pcs listed in Table 1 still needs to be verified. One can only speculate why some bacteria have two pathways for PtdCho biosynthesis. The PmtA pathway can be considered a de novo synthesis pathway for PtdCho required for bacteria to grow outside possible host cells. The Pcs pathway is a salvage pathway that could provide an energy advantage to
126 the bacterium when choline is present, e.g., when a bacterium has entered a symbiotic relationship with a host as is the case for plant nodule-forming bacteria such as Sr. meliloti. Similarly, pathogenic bacteria inside host cells would benefit from choline provided by the host. Depending on the environment to which a bacterium has adapted, one or the other pathway, or both, might have been lost during evolution. IV. Glycoglycerolipids Glycoglycerolipids are non-phosphorous because their head groups are not connected to the diacylglycerol moiety by phosphodiester bonds. Instead, a hexose is found in a glycosidic linkage with the sn-3 hydroxyl group of the diacylglycerol moiety (Fig. 3). The most abundant glycoglycerolipids found in nature are the plant galactoglycerolipids (Dörmann and Benning, 2002). It is generally assumed that the prevalence of these lipids in plants with their extensive photosynthetic membrane system provides a strategy for conserving phosphorus, because inorganic phosphate is one of the limiting factors for biomass production in land-based and marine environments
Banita Tamot and Christoph Benning (Vance et al., 2003; Van Mooy et al., 2006). Bacteria can remodel their membrane lipid composition in response to phosphate deprivation by degrading phosphoglycerolipids and synthesizing glycoglycerolipids (Minnikin et al., 1974; Benning et al., 1993, 1995). Recently, support for this hypothesis was provided in a natural environment. Apparently, substitution of the phospholipid PtdGro by the sulfoglycolipid sulfoquinovosyldiacylglycerol (SQDG; Fig. 3B) enables marine picocyanobacteria to outcompete heterotrophic microorganisms (Van Mooy et al., 2006). Likewise, purple bacteria synthesize neutral glycoglycerolipids and the acidic sulfolipid SQDG (Fig. 3), particularly during phosphate limitation. A. Glycerolipids Containing Glucose or Galactose Beyond plants, glycoglycerolipids are present in many bacteria, but their structures differ in subtle ways. Accordingly, the responsible lipid glycosyltransferases are not necessarily directly related in different organisms. For example, the biosynthesis of the major plant galactolipid involves the transfer of a galactosyl residue from UDP-Gal onto diacylglycerol giving rise
Fig. 3. Biosynthesis of glycoglycerolipids in Rba. sphaeroides. (A) Proposed formation of β-D-galactosyldiacylglycerol (MGDG) and α-D-glucosyl-(1→4)-O-β-D-galactosyldiacylglycerol (GGDG), and (B) of sulfoquinovosyldiacylglycerol (SQDG). LGT1 and LGT2 designate proposed lipid glycosyltransferases that still need to be identified. Loci genetically implicated in different steps of SQDG biosynthesis are indicated (sqdA, B, C, D). DAG, diacylglycerol; NDP-Gal, nucleoside diphosphogalactose; NDP-Glu, nucleoside diphosphoglucose; UDP-Glc, UDP-glucose; UDP-SQ, UDP-sulfoquinovose.
Chapter 7
Membrane Lipids
to β-D-galactosyldiacylglycerol. A second transfer of a galactosyl residue from UDP-Gal leads to the formation of α-D-galactosyl-(1→6)-O-β-D-galactosyldiacylglycerol. The two enzymes involved in Arabidopsis are non-processive galactosyltransferases, MGD1 and DGD1 (Benning and Ohta, 2005). Two galactoglycerolipids identical in head group structure to the plant galactoglycerolipids exist in cyanobacteria, but the first enzyme is a glucosyltransferase transferring a glucosyl residue onto diacylglycerol followed by an epimerase converting the glucosyl residue into the galactosyl head group (Awai et al., 2006). In Bacillus subtilis and in Staphylococcus aureus processive glucosyltransferases produce βD-glucosyldiacylglycerol and β-D-glucosyl-(1→6)O-β-D-glucosyldiacylglycerols (Jorasch et al., 1998, 2000). In the mycoplasma Acholeplasma laidlawii the formation of α-D-glucosyldiacylglycerol, and αD-glucosyl-(1→2)-O-β-D-glucosyldiacylglycerol is catalyzed by a distinct set of retaining, non-processive UDP-glucose dependent glucosyltransferases (Berg et al., 2001; Edman et al., 2003). Purple bacteria, as analyzed thus far, produce a set of glycoglycerolipids different from those found in plants and other bacteria. Membranes of Blastochloris viridis contain β-D-galactosyldiacylglycerol and β-D-galactosyl-(1→6)-O-β-D-galactosyldiacylglycerol (Linscheid et al., 1997). Following phosphate-deprivation, Rba. sphaeroides produces β-D-galactosyldiacylglycerol and α-D-glucosyl-(1→ 4)-O-β-D-galactosyldiacylglycerol (Benning et al., 1995). In both cases the identity of the glycosyltransferases is not known. Given the diversity of the lipid glycosyltransferases in different organisms, we could not predict Rba. sphaeroides glycosyltransferases involved in the biosynthesis of this lipid. However, based on analogy to other bacterial enzyme systems, we propose that the biosynthesis of α-D-glucosyl(1→4)-O-β-D-galactosyldiacylglycerol involves two different glycosyltransferases (LGT1 and LGT2; Fig. 3). One can only marvel at the diversity of glycoglycerolipids in bacteria and speculate about their function and origin. In general, the respective monohexosyldiacylglycerol represents a non-bilayer forming lipid, while the respective dihexosyldiacylglycerol usually forms bilayers in mixtures with water (Webb and Green, 1991). Mixtures of both lipid types may provide organisms with a lipid environment suitable for proper function of the membrane. Indeed, expression of the gene for monoglucosyldiacylglycerol
127 synthase from A. laidlawii in the PtdEtn deficient pss mutant of E. coli alleviated the growth defects (Wikstrom et al., 2004; Xie et al., 2006), indicating that compositionally very different non-bilayer-forming glycerolipids can substitute for each other. The great diversity of glycoglycerolipids in bacteria due to the recruitment of different glycosyltransferases is presumably facilitated by the fact that different lipids can be readily substituted as long as their basic biophysical properties are similar. B. Sulfoquinovosyldiacylglycerol The sulfolipid, sulfoquinovosyldiacylglycerol (SQDG), was discovered by Benson and coworkers in plants, algae, and the purple bacterium Rhodospirillum rubrum (Benson et al., 1959). It is a glycoglycerolipid with a 6-deoxy-6-sulfo-α-D-glucosyl head group bound to the sn-3 hydroxyl of the diacylglycerol backbone (Fig. 3B). It should be noted that the head group is a sulfonate and not a sulfate ester. As such its pKa is low and it is resistant to loss of the sulfur group by hydrolysis. In essence, SQDG is an anionic glycoglycerolipid at biological pH values. This particular lipid is widespread in bacteria and plants and its presence in different species of purple bacteria is well documented (Wood et al., 1965; Imhoff, 1991; Gage et al., 1992; Imhoff and Bias-Imhoff, 1995). However, in some purple bacteria, including Rba. capsulatus, no SQDG was detected (Imhoff, 1991). Searching protein sequences encoded by the Rba. capsulatus genome with the highly conserved protein sequence for SqdB (see below) from Rba. sphaeroides at the Integrated Genomics web site (http://www. ergo-light.com/ERGO/) did not provide a positive result suggesting that Rba. capsulatus might indeed lack the capability to synthesize SQDG. On the other hand, sulfolipid is present in the related, but nonphotosynthetic nodule-forming bacterium Sr. meliloti (Cedergren and Hollingsworth, 1994). Genetic analysis of SQDG metabolism in Rba. sphaeroides was instrumental in identifying the bacterial, and subsequently the plant genes involved in SQDG biosynthesis (Benning, 1998a; Frentzen, 2004). Mutants of Rba. sphaeroides deficient in SQDG biosynthesis were identified by thin-layer chromatography of lipid extracts from chemically mutagenized cells. Restoration of SQDG biosynthesis following the introduction of wild-type DNA led to the isolation of the sqdA gene (Benning and Somerville, 1992b) and an operon encoding the sqd-
128 BCD open reading frames (Benning and Somerville, 1992a). Subsequently, the sqdBCD open reading frames were individually inactivated and shown to be required for SQDG biosynthesis (Benning et al., 1993; Rossak et al., 1995, 1997). The sqdB gene of Rba. sphaeroides served in the identification of the cyanobacterial ortholog (Güler et al., 1996) and later the plant SQD1 protein (Essigmann et al., 1998). Of the four proteins encoded by the sqd genes, SqdB is most conserved and best studied. The SqdB protein is similar to sugar nucleotide-modifying proteins (Benning and Somerville, 1992a). While the Rba. sphaeroides SqdB enzyme has not been further analyzed, the recombinant plant ortholog SQD1 catalyzes the incorporation of sulfite into UDP-Glc producing UDP-sulfoquinovose (UDP-SQ, Fig. 3B) (Sanda et al., 2001). The structure of the SQD1 protein was modeled and determined by X-ray crystallography providing insights into the reaction mechanism of this enzyme (Essigmann et al., 1999; Mulichak et al., 1999). The SQD1 enzyme is a dimer and contains a tightly bound NAD+ nucleotide which participates in the reaction cycle and is intermittently reduced. Much less is known about the biochemical function of the other three Sqd proteins from Rba. sphaeroides. The predicted SqdA protein is similar to bacterial acyltransferases, SqdC to a small molecule reductase, and SqdD to glycogenin, an autoglycosylating glycosyltransferase. Specific inactivation of SqdD led to the accumulation of UDP-SQ in Rba. sphaeroides (Rossak et al., 1995). In view of the similarity of SqdD to glycosyltransferases this result suggests that SqdD transfers sulfoquinovose from UDP-SQ onto an acceptor. Unlike the SqdB protein, the SqdD protein is not conserved in cyanobacteria and plants. Both contain similar SQDG synthases, but both are different from SqdD of Rba. sphaeroides (Güler et al., 2000; Yu et al., 2002). Biochemical analysis of the plant enzyme suggest that it uses diacylglycerol as an acceptor (Heinz et al., 1989; Seifert and Heinz, 1992), but this may not be the case for SqdD from Rba. sphaeroides. Indeed, when the sqdC gene was specifically inactivated, sulfoquinovosyl-1-O-dihydroxyacetone accumulated in the mutant (Rossak et al., 1997). This result suggests that a compound such as dihydroxyacetone or dihydroxyacetone phosphate serves as an acceptor for the UDP-SQ:sulfoquinovosyl transferase presumably encoded by sqdD. The SqdC amino acid sequence is similar to that of a small molecule reductase which potentially could reduce the β-ketogroup of the dihydroxyacetone resi-
Banita Tamot and Christoph Benning due to the glycerol. As the SqdA amino acid sequence is similar to that of an acyltransferase, the SqdA protein could be involved in acylation of one or both hydroxyl groups of a sulfoquinovosyl 1-O-glycerol or a sulfoquinovosyl-1-O-dihydroxyacetone precursor. At this time, the proposed scenario is speculative, but it is very likely that SQDG is assembled in Rba. sphaeroides by a route different from that in plants or cyanobacteria. A detailed biochemical analysis of the Sqd proteins should provide the answer. V. Betaine Lipid The betaine lipid diacylglycerol-N,N,N,-trimethylhomoserine (DGTS) is a non-phosphorous glycerolipid with its head group ether-linked to the sn-3 hydroxyl of the diacylglycerol moiety (Fig. 4). This interesting lipid is similar in structure and biophysical properties to PtdCho (Sato and Murata, 1991). It is present in many algae and non-seed plants, as well as fungi as summarized by (Klug and Benning, 2001). DGTS has been discovered in bacteria such as Rba. sphaeroides (Benning et al., 1995) and Sr. meliloti (Geiger et al., 1999). In both bacteria this lipid accumulates following phosphate deprivation. Labeling studies with Rba. sphaeroides (Hofmann and Eichenberger, 1996) suggested that DGTS is produced in this bacterium by transfer of the 3-amino-3-carboxypropyl group of AdoMet onto diacylglycerol (Fig. 4), giving rise to diacylglycerolhomoserine. Subsequent transfer of three methyl groups from AdoMet would lead to the end product, DGTS. This reaction sequence was similar to the proposed biosynthesis pathway in eukaryotes based on labeling studies as summarized by Klug and Benning (2001). The discovery of DGTS in Rba. sphaeroides provided an ideal opportunity to identify the first genes involved in DGTS biosynthesis. A similar genetic approach as described for the sqd genes above was used to isolate an operon consisting of two open reading frames, btaA and btaB, involved in DGTS biosynthesis in Rba. sphaeroides (Klug and Benning, 2001). Mutation of the btaA gene resulted in the loss of DGTS accumulation during phosphate starvation. When the btaB gene was disrupted, a new lipid, diacylglycerylhomoserine, accumulated (Klug and Benning, 2001). Based on these results it was proposed that btaA encodes an AdoMet:diacylglycerol 3-amino-3-carboxypropyl transferase. This hypothesis was recently confirmed by expression of
Chapter 7
Membrane Lipids
129
Fig. 4. Biosynthesis of diacylglycerol-N,N,N,-trimethylhomoserine and the role of the two enzymes BtaA and BtaB. AdoHcy, S-adenosylhomocysteine; AdoMet, S-adenosylmethionine; DAG, diacylglycerol; DGHS, Diacylglycerolhomoserine; DGTS, diacylglycerolN,N,N,-trimethylhomoserine; 5´MTA, 5´-methylthioadenosine.
the btaA gene in E. coli and the direct biochemical analysis of the BtaA protein (Riekhof et al., 2005a). Co-expression of the btaA and btaB genes in E. coli led to DGTS biosynthesis in this bacterium which normally lacks DGTS. This result also confirmed that btaB encodes the N-methyltransferase (Riekhof et al., 2005a). The discovery of the two genes involved in DGTS biosynthesis in Rba. sphaeroides led to the identification of an orthologous protein, BTA1, in the eukaryotic alga Chlamydomonas reinhardtii (Riekhof et al., 2005b). Interestingly the BTA1 amino acid sequence contains two domains, each similar to the bacterial BtaA and BtaB sequences, respectively.
VI. Ornithine Lipid All of the membrane lipids described above are glycerolipids. In contrast, in ornithine lipids of purple bacteria and related α-proteobacteria, the α-amino group of ornithine is amide-linked to a 3-hydroxy fatty acid (Fig. 1 and Fig. 5). A second fatty acid, e.g. vaccenic acid, is esterified to the 3-hydroxy group (Brooks and Benson, 1972; Gorchein, 1973). This lipid is also present in Sr. meliloti and an operon with two open reading frames, olsA and olsB, was isolated using a genetic complementation approach as described above for the isolation of the sqd and bta genes from
Fig. 5. Proposed biosynthesis of ornithine lipid (OL). The two acyltransferases OlsA and OlsB proposed to be involved are indicated. ACP, acyl carrier protein.
130 Rba. sphaeroides (Weissenmayer et al., 2002; Gao et al., 2004). Acylation of lyso-ornithine lipid in protein extracts of wild type cells and reduced acylation activity in extracts of an olsA mutant of Sr. meliloti (Weissenmayer et al., 2002) indicated that OlsA is an acyltransferase catalyzing the second reaction of ornithine lipid biosynthesis (Fig. 5). Expression of olsB in E. coli led to the formation of lyso-ornithine lipid, consistent with a role for OlsB in the first acylation reaction (Gao et al., 2004). The olsAB genes of Rba. capsulatus were recently identified during a screen targeted at the identification of cytochrome c deficient respiratory mutants (Aygun-Sunar et al., 2006). Several c-type cytrochromes were found to be absent in mutants, which were primarily deficient in the biosynthesis of ornithine lipid. We used these olsAB genes from Rba. capsulatus to identify putative orthologs in the genome of Rba. sphaeroides as shown in Table 1. VII. Lipid Function The extensive repertoire of membrane lipids in purple bacteria raises questions regarding the biological role of this complexity, and the mechanisms that regulate membrane lipid composition. The lipid composition under defined growth conditions is generally stable (homeostatic) as bacterial cells grow and divide, but is known to alter in bacteria in response to changing environmental factors such as nutrient supply. Understanding the regulation of membrane lipid homeostasis and of changes in lipid composition in response to environmental cues remains one of the frontiers in cell biology and biochemistry. Two recent advances have been most helpful in assigning specific functions to individual membrane lipids beyond their role as bulk membrane building blocks. These are the increasing availability of mutants with disrupted lipid biosynthetic genes facilitated by the availability of genome sequences of different purple bacteria, and an increasing number of crystal structures for integral membrane protein complexes. A. Changes in Lipid Composition in Response to the Environment Phosphate deprivation is the most effective known environmental factor that leads to membrane lipid composition changes, as first recognized for Pseudomonas sp. (Minnikin and Abdolrahimzadeh, 1974;
Banita Tamot and Christoph Benning Minnikin et al., 1974). Purple bacteria (Benning et al., 1993, 1995) and related α-proteobacteria such as Sr. meliloti (Geiger et al., 1999) also change their lipid composition in response to decreases in phosphate concentration in the growth medium. In fact, some of the lipids discussed above, e.g. the betaine lipid, are not even detectable unless the cells are phosphate depleted (Benning et al., 1995; Geiger et al., 1999). It has been hypothesized that SQDG can substitute for the anionic phospholipid PtdGro in Rba. sphaeroides, based on the fact that an SQDG deficient mutant contains increased levels of PtdGro and ceases to grow sooner than the wild type strain under phosphate-limited conditions (Benning et al., 1993). Similar conclusions with regard to the role of SQDG based on mutant analysis are drawn for cyanobacteria (Güler et al., 1996) and plants (Yu et al., 2002). This glycolipid vs. phospholipid hypothesis has recently gained further recognition due to the discovery that SQDG is prevalent in marine picocyanobacteria. The ability to synthesize SQDG presumably provides these photosynthetic bacteria with a competitive advantage over heterotrophic microorganisms lacking the ability to synthesize SQDG (Van Mooy et al., 2006). Similar studies of DGTS-deficient mutants are not yet available. However, it is reasonable to assume that, because the structural and biophysical properties of DGTS resemble within limitations those of PtdCho (Sato and Murata, 1991), the betaine lipid DGTS can substitute for PtdCho in purple bacteria and α-proteobacteria under phosphate-limited conditions. The expression of the bta genes of Sr. meliloti responsible for DGTS biosynthesis was found to be under the control of the PhoB regulator, which governs the expression of genes responsive to phosphate limitation (Geiger et al., 1999). Most intriguing is the observation that addition of Tris to the growth medium leads to the accumulation of phosphatidyl Tris in membranes of Rba. sphaeroides (Schmid et al., 1991). This is a xenobiotic compound which apparently provides the cell with a suitable bulk membrane lipid. How this lipid is synthesized is not known, but it seems possible that Tris enters phospholipid biosynthesis at the level of one of the CDP-diacylglcyerol utilizing enzymes (Fig. 2). Certain strains of Rba. sphaeroides, including strain 2.4.1, discriminate against Tris and cannot form phosphatidyl Tris (Donohue et al., 1982), which was initially misidentified as N-acylphosphatidylserine. It is interesting to note in this context that Rba. sphaer-
Chapter 7
Membrane Lipids
oides 2.4.1 also apparently lacks PtdCho synthase (Pcs) activity (Arondel et al., 1993; Sohlenkamp et al., 2003), and it seems possible that Pcs (Fig. 2) might be responsible for the formation of phosphatidyl Tris in some strains of Rba. sphaeroides. However, this hypothesis remains to be tested. Environmental factors more subtly affecting membrane lipid composition in purple bacteria are light illumination (quality and intensity) and oxygen tension (Russell and Harwood, 1979; Onishi and Niederman, 1982; Campbell and Lueking, 1983). Furthermore, increasing osmolarity in the medium led to an increased conversion of PtdGro to CL (Catucci et al., 2004), presumably stabilizing the cell membrane bilayer under these conditions. The availability of a new CL-deficient mutant (Tamot and Benning, unpublished) will allow this hypothesis to be tested.
131 Paracoccus denitrificans (Ostermeier et al., 1997) and bovine heart (Tsukihara et al., 2003) showed that most alkyl chains resolved in the structures were conserved. Several of these belong to a previously identified CL associated with the complex. A recently isolated CL-deficient mutant of Rba. sphaeroides (Tamot and Benning, unpublished) should provide a tool to test the function of this CL associated with CcO. Already, work on an ornithine lipid-deficient mutant of Rba. capsulatus has shown that OL is required to maintain steady state amounts of c-type cytochromes (AygunSunar et al., 2006). Given that CL is found tightly associated with the photosynthetic reaction center and the CcO integral membrane complexes, one might expect to observe drastic effects on the growth and physiology of a CL-deficient mutant. VIII. Perspectives
B. Interaction of Lipids with Membrane Proteins The recent progress in the structural elucidation of integral membrane complexes from different organisms including purple bacteria has led to the discovery of specific lipid protein interactions at a level previously not possible (Palsdottir and Hunte, 2004; Hunte, 2005). Several crystal structures are available for the Photosystem II-like reaction center of Rba. sphaeroides and one emerging theme is the presence of one CL molecule in what appears to be a specific binding site at the protein lipid interface (McAuley et al., 1999; Wakeham et al., 2001; Jones et al., 2002; Katona et al., 2003). In addition, PtdCho and glucosylgalactosyldiacylglycerol were found in the structure of the Rba. sphaeroides reaction center complex, and were confirmed by mass spectrometry (Camara-Artigas et al., 2002). The functionality of the lipid interaction with the reaction center was tested in vitro and it was shown that CL, PtdGro and PtdCho affect the free energy levels of the quinones bound in the reaction center complex (Rinyu et al., 2004; Nagy et al., 2004). A second integral membrane complex that received much attention is a terminal complex of the respiratory chain, cytochrome c oxidase (CcO). Six PtdEtn molecules were resolved in the crystal structure of the four subunit CcO complex of Rba. sphaeroides (Svensson-Ek et al., 2002). Overlaying a more recent subunit I-II structure of CcO of Rba. sphaeroides (Quin et al., 2006) with orthologous structures from
Due to their lipid-rich membrane composition, purple bacteria provide ideal models to study the biosynthesis and function of some of the more unusual lipids found in nature. In addition, their ability to adapt to different environments by changing their membrane lipid composition provides fertile ground for studies addressing the regulation of lipid homeostasis. Furthermore, the flexibility in membrane lipid composition of purple bacteria and their ability to incorporate even xenobiotic compounds into their membranes challenge our current hypotheses about the roles of specific lipids, in particular phospholipids. The increasing repertoire of lipid biosynthetic mutants in Rba. sphaeroides provides the tools to test the function of individual lipid classes as predicted by their presence in high-resolution structures of integral membrane proteins. Moreover, as Rba. sphaeroides is proposed to serve as a suitable host for high-yield and possibly high-throughput production of integral membrane proteins (Laible et al., 2004), lipid mutants could be developed into a tool kit to adjust the lipid composition of the host cell, optimizing the lipid environment for membrane protein production. Acknowledgments We are grateful to Carrie Hiser and Wayne Riekhof for carefully reading the manuscript and helpful discussions. Work on membrane lipids in the Benning Lab is supported in part by grants from the US
132 National Science Foundation, the US Department of Energy, and the Michigan State University Research of Excellence Center for the Structural Biology of Membrane Proteins. References Arondel V, Benning C and Somerville CR (1993) Isolation and functional expression in Escherichia coli of a gene encoding phosphatidylethanolamine methyltransferase (EC 2.1.1.17) from Rhodobacter sphaeroides. J Biol Chem 268: 16002–16008 Awai K, Kakimoto T, Awai C, Kaneko T, Nakamura Y, Takamiya K, Wada H and Ohta H (2006) Comparative genomic analysis revealed a gene for monoglucosyldiacylglycerol synthase, an enzyme for photosynthetic membrane lipid synthesis in cyanobacteria. Plant Physiol 141: 1120–1127 Aygun-Sunar S, Mandaci S, Koch HG, Murray IV, Goldfine H and Daldal F (2006) Ornithine lipid is required for optimal steady-state amounts of c-type cytochromes in Rhodobacter capsulatus. Mol Microbiol 61: 418–435 Benning C (1998a) Biosynthesis and function of the sulfolipid sulfoquinovosyl diacylglycerol. Annu Rev Plant Physiol Plant Mol Biol 49: 53–75 Benning C (1998b) Membrane lipids in anoxygenic photosynthetic bacteria. In P-A Siegenthaler, N Murata, eds, Lipids in Photosynthesis: Structure, Function and Genetics, pp 83–101. Kluwer Academic Publishers, Dordrecht Benning C and Ohta H (2005) Three enzyme systems for galactoglycerolipid biosynthesis are coordinately regulated in plants. J Biol Chem 280: 2397–2400 Benning C and Somerville CR (1992a) Identification of an operon involved in sulfolipid biosynthesis in Rhodobacter sphaeroides. J Bacteriol 174: 6479–6487 Benning C and Somerville CR (1992b) Isolation and genetic complementation of a sulfolipid-deficient mutant of Rhodobacter sphaeroides. J Bacteriol 174: 2352–2360 Benning C, Beatty JT, Prince RC and Somerville CR (1993) The sulfolipid sulfoquinovosyldiacylglycerol is not required for photosynthetic electron transport in Rhodobacter sphaeroides but enhances growth under phosphate limitation. Proc Natl Acad Sci USA 90: 1561–1565 Benning C, Huang ZH and Gage DA (1995) Accumulation of a novel glycolipid and a betaine lipid in cells of Rhodobacter sphaeroides grown under phosphate limitation. Arch Biochem Biophys 317: 103–111 Benson AA, Daniel H and Wiser R (1959) A sulfolipid in plants. Proc Natl Acad Sci USA 45: 1582–1587 Berg S, Edman M, Li L, Wikstrom M and Wieslander A (2001) Sequence properties of the 1,2-diacylglycerol 3-glucosyltransferase from Acholeplasma laidlawii membranes. Recognition of a large group of lipid glycosyltransferases in eubacteria and archaea. J Biol Chem 276: 22056–22063 Bogdanov M, Sun J, Kaback HR and Dowhan W (1996) A phospholipid acts as a chaperone in assembly of a membrane transport protein. J Biol Chem 271: 11615–11618 Boyce SG and Lueking DR (1984) Purification and characterization of a long-chain acyl coenzyme A thioesterase from Rho-
Banita Tamot and Christoph Benning dopseudomonas sphaeroides. Biochemistry 23: 141–147 Broglie RM and Niederman RA (1979) Membranes of Rhodopseudomonas sphaeroides: Effect of cerulenin on assembly of chromatophore membrane. J Bacteriol 138: 788–798 Brooks JL and Benson AA (1972) Studies on the structure of an ornithine-containing lipid from Rhodospirillum rubrum. Arch Biochem Biophys 152: 347–355 Camara-Artigas A, Brune D and Allen JP (2002) Interactions between lipids and bacterial reaction centers determined by protein crystallography. Proc Natl Acad Sci USA 99: 11055–11060 Campbell TB and Lueking DR (1983) Light-mediated regulation of phospholipid synthesis in Rhodopseudomonas sphaeroides. J Bacteriol 155: 806–816 Catucci L, Depalo N, Lattanzio VM, Agostiano A and Corcelli A (2004) Neosynthesis of cardiolipin in Rhodobacter sphaeroides under osmotic stress. Biochemistry 43: 15066–15072 Cedergren RA and Hollingsworth RI (1994) Occurrence of sulfoquinovosyl diacylglycerol in some members of the family Rhizobiaceae. J Lipid Res 35: 1452–1461 Cooper CL, Boyce SG and Lueking DR (1987) Purification and characterization of Rhodobacter sphaeroides acyl carrier protein. Biochemistry 26: 2740–2746 Cronan JE (2003) Bacterial membrane lipids: Where do we stand? Annu Rev Microbiol 57: 203–224 de Rudder KE, Thomas-Oates JE and Geiger O (1997) Rhizobium meliloti mutants deficient in phospholipid N-methyltransferase still contain phosphatidylcholine. J Bacteriol 179: 6921–6928 de Rudder KE, Sohlenkamp C and Geiger O (1999) Plant-exuded choline is used for rhizobial membrane lipid biosynthesis by phosphatidylcholine synthase. J Biol Chem 274: 20011–20016 DeChavigny A, Heacock PN and Dowhan W (1991) Sequence and inactivation of the pss gene of Escherichia coli. Phosphatidylethanolamine may not be essential for cell viability. J Biol Chem 266: 5323–5332 Donohue TJ, Cain BD and Kaplan S (1982) Alterations in the phospholipid composition of Rhodopseudomonas sphaeroides and other bacteria induced by Tris. J Bacteriol 152: 595–606 Dörmann P and Benning C (2002) Galactolipids rule in seed plants. Trends Plant Sci 7: 112–118 Dryden SC and Dowhan W (1996) Isolation and expression of the Rhodobacter sphaeroides gene (pgsA) encoding phosphatidylglycerophosphate synthase. J Bacteriol 178: 1030–1038 Edman M, Berg S, Storm P, Wikstrom M, Vikstrom S, Ohman A and Wieslander A (2003) Structural features of glycosyltransferases synthesizing major bilayer and nonbilayer-prone membrane lipids in Acholeplasma laidlawii and Streptococcus pneumoniae. J Biol Chem 278: 8420–8428 Essigmann B, Güler S, Narang RA, Linke D and Benning C (1998) Phosphate availability affects the thylakoid lipid composition and the expression of SQD1, a gene required for sulfolipid biosynthesis in Arabidopsis thaliana. Proc Natl Acad Sci USA 95: 1950–1955 Essigmann B, Hespenheide BM, Kuhn LA and Benning C (1999) Prediction of the active-site structure and NAD+ binding in SQD1, a protein essential for sulfolipid biosynthesis in Arabidopsis. Arch Biochem Biophys 369: 30–41 Frentzen M (2004) Phosphatidylglycerol and sulfoquinovosyldiacylglycerol: Anionic membrane lipids and phosphate regulation. Curr Opin Plant Biol 7: 270–276
Chapter 7
Membrane Lipids
Gage DA, Huang ZH and Benning C (1992) Comparison of sulfoquinovosyl diacylglycerol from spinach and the purple bacterium Rhodobacter sphaeroides by fast atom bombardment tandem mass spectrometry. Lipids 27: 632–636 Gao JL, Weissenmayer B, Taylor AM, Thomas-Oates J, LopezLara IM and Geiger O (2004) Identification of a gene required for the formation of lyso-ornithine lipid, an intermediate in the biosynthesis of ornithine-containing lipids. Mol Microbiol 53: 1757–1770 Geiger O, Röhrs V, Weissenmayer B, Finan TM and ThomasOates JE (1999) The regulator gene phoB mediates phosphate stress-controlled synthesis of the membrane lipid diacylglyceryl-N,N,N-trimethylhomoserine in Rhizobium (Sinorhizobium) meliloti. Mol Microbiol 32: 63–73 Giroud C, Gerber A and Eichenberger W (1988) Lipids of Chlamydomonas reinhardtii. Analysis of molecular species and intracellular site(s) of biosynthesis. Plant Cell Physiol 29: 587–595 Gorchein A (1973) Structure of the ornithine-containing lipid from Rhodopseudomonas spheroides. Biochim Biophys Acta 306: 137–141 Güler S, Seeliger A, Härtel H, Renger G and Benning C (1996) A null mutant of Synechococcus sp. PCC7942 deficient in the sulfolipid sulfoquinovosyl diacylglycerol. J Biol Chem 271: 7501–7507 Güler S, Essigmann B and Benning C (2000) A cyanobacterial gene, sqdX, required for biosynthesis of the sulfolipid sulfoquinovosyldiacylglycerol. J Bacteriol 182: 543–545 Heinz E, Schmidt H, Hoch M, Jung KH, Binder H and Schmidt RR (1989) Synthesis of different nucleoside 5´-diphospho-sulfoquinovoses and their use for studies on sulfolipid biosynthesis in chloroplasts. Eur J Biochem 184: 445–453 Hofmann M and Eichenberger W (1996) Biosynthesis of diacylglyceryl-N,N,N-trimethylhomoserine in Rhodobacter sphaeroides and evidence for lipid-linked N methylation. J Bacteriol 178: 6140–6144 Hunte C (2005) Specific protein-lipid interactions in membrane proteins. Biochem Soc Trans 33: 938–942 Imhoff JF (1991) Polar lipids and fatty acids in the genus Rhodobacter. System Appl Microbiol 14: 228–234 Imhoff JF and Bias-Imhoff U (1995) Lipids, quinones and fatty acids of anoxygenic phototrophic bacteria. In: RE Blankenship, MT Madigan, CE Bauer, eds, Anoxygenic Photosynthetic Bacteria, pp 179–205. Kluwer Academic Publishers, Dordrecht Imhoff JF, Kushner DJ, Kushwaha SC and Kates M (1982) Polar lipids in phototrophic bacteria of the Rhodospirillaceae and Chromatiaceae families. J Bacteriol 150: 1192–1201 Jones MR, Fyfe PK, Roszak AW, Isaacs NW and Cogdell RJ (2002) Protein-lipid interactions in the purple bacterial reaction centre. Biochim Biophys Acta 1565: 206–214 Jorasch P, Wolter FP, Zahringer U and Heinz E (1998) A UDP glucosyltransferase from Bacillus subtilis successively transfers up to four glucose residues to 1,2-diacylglycerol: Expression of ypfP in Escherichia coli and structural analysis of its reaction products. Mol Microbiol 29: 419–430 Jorasch P, Warnecke DC, Lindner B, Zahringer U and Heinz E (2000) Novel processive and nonprocessive glycosyltransferases from Staphylococcus aureus and Arabidopsis thaliana synthesize glycoglycerolipids, glycophospholipids, glycosphingolipids and glycosylsterols. Eur J Biochem 267: 3770–3783 Katona G, Andreasson U, Landau EM, Andreasson LE and
133 Neutze R (2003) Lipidic cubic phase crystal structure of the photosynthetic reaction centre from Rhodobacter sphaeroides at 2.35Å resolution. J Mol Biol 331: 681–692 Klug RM and Benning C (2001) Two enzymes of diacylglycerylO-4´-(N,N,N,-trimethyl)homoserine biosynthesis are encoded by btaA and btaB in the purple bacterium Rhodobacter sphaeroides. Proc Natl Acad Sci USA 98: 5910–5915 Laible PD, Scott HN, Henry L and Hanson DK (2004) Towards higher-throughput membrane protein production for structural genomics initiatives. J Struct Funct Genomics 5: 167–172 Linscheid M, Diehl BW, Övermöhle M, Riedl I and Heinz E (1997) Membrane lipids of Rhodopseudomonas viridis. Biochim Biophys Acta 1347: 151–163 Mackenzie C, Choudhary M, Larimer FW, Predki PF, Stilwagen S, Armitage JP, Barber RD, Donohue TJ, Hosler JP, Newman JE, Shapleigh JP, Sockett RE, Zeilstra-Ryalls J and Kaplan S (2001) The home stretch, a first analysis of the nearly completed genome of Rhodobacter sphaeroides 2.4.1. Photosynth Res 70: 19–41 Martinez-Morales F, Schobert M, Lopez-Lara IM and Geiger O (2003) Pathways for phosphatidylcholine biosynthesis in bacteria. Microbiology 149: 3461–3471 McAuley KE, Fyfe PK, Ridge JP, Isaacs NW, Cogdell RJ and Jones MR (1999) Structural details of an interaction between cardiolipin and an integral membrane protein. Proc Natl Acad Sci USA 96: 14706–14711 McMurry LM, Oethinger M and Levy SB (1998) Triclosan targets lipid synthesis. Nature 394: 531–532 Minnikin DE and Abdolrahimzadeh H (1974) The replacement of phosphatidylethanolamine and acidic phospholipids by an ornithine-amide lipid and a minor phosphorus-free lipid in Pseudomonas fluorescens NCMB 129. FEBS Lett 43: 257–260 Minnikin DE, Abdolrahimzadeh H and Baddiley J (1974) Replacement of acidic phosphates by acidic glycolipids in Pseudomonas diminuta. Nature 249: 268–269 Mulichak AM, Theisen MJ, Essigmann B, Benning C and Garavito RM (1999) Crystal structure of SQD1, an enzyme involved in the biosynthesis of the plant sulfolipid headgroup donor UDPsulfoquinovose. Proc Natl Acad Sci USA 96: 13097–13102 Nagy L, Milano F, Dorogi M, Agostiano A, Laczko G, Szebenyi K, Varo G, Trotta M and Maroti P (2004) Protein/lipid interaction in the bacterial photosynthetic reaction center: Phosphatidylcholine and phosphatidylglycerol modify the free energy levels of the quinones. Biochemistry 43: 12913–12923 Nishijima S, Asami Y, Uetake N, Yamagoe S, Ohta A and Shibuya I (1988) Disruption of the Escherichia coli cls gene responsible for cardiolipin synthesis. J Bacteriol 170: 775–780 Onishi JC and Niederman RA (1982) Rhodopseudomonas sphaeroides membranes: Alterations in phospholipid composition in aerobically and phototrophically grown cells. J Bacteriol 149: 831–839 Ostermeier C, Harrenga A, Ermler U and Michel H (1997) Structure at 2.7 Å resolution of the Paracoccus denitrificans two-subunit cytochrome c oxidase complexed with an antibody FV fragment. Proc Natl Acad Sci USA 94: 10547–10553 Palsdottir H and Hunte C (2004) Lipids in membrane protein structures. Biochim Biophys Acta 1666: 2–18 Pluschke G, Hirota Y and Overath P (1978) Function of phospholipids in Escherichia coli. Characterization of a mutant deficient in cardiolipin synthesis. J Biol Chem 253: 5048–5055 Quin L, Hiser C, Mulichak A, Garavito RM and Ferguson-Miller
134 S (2006) Identification of conserved lipid/detergent binding sites in a high resolution structure of the membrane protein cytochrome c oxidase. Proc Natl Acad Sci USA in press: Radcliffe CW, Steiner FX, Carman GM and Niederman RA (1989) Characterization and localization of phosphatidylglycerophosphate and phosphatidylserine synthases in Rhodobacter sphaeroides. Arch Microbiol 152: 132–137 Riekhof WR, Andre C and Benning C (2005a) Two enzymes, BtaA and BtaB, are sufficient for betaine lipid biosynthesis in bacteria. Arch Biochem Biophys 441: 96–105 Riekhof WR, Sears BB and Benning C (2005b) Annotation of genes involved in glycerolipid biosynthesis in Chlamydomonas reinhardtii: Discovery of the betaine lipid synthase BTA1Cr. Eukaryot Cell 4: 242–252 Rinyu L, Martin EW, Takahashi E, Maroti P and Wraight CA (2004) Modulation of the free energy of the primary quinone acceptor (QA) in reaction centers from Rhodobacter sphaeroides: contributions from the protein and protein-lipid(cardiolipin) interactions. Biochim Biophys Acta 1655: 93–101 Rossak M, Tietje C, Heinz E and Benning C (1995) Accumulation of UDP-sulfoquinovose in a sulfolipid-deficient mutant of Rhodobacter sphaeroides. J Biol Chem 270: 25792–25797 Rossak M, Schäfer A, Xu N, Gage DA and Benning C (1997) Accumulation of sulfoquinovosyl–1-O-dihydroxyacetone in a sulfolipid-deficient mutant of Rhodobacter sphaeroides inactivated in sqdC. Arch Biochem Biophys 340: 219–230 Russell NJ and Harwood JL (1979) Changes in acyl lipid composition of photosynthetic bacteria grown under photosynthetic and non-photosynthetic conditions. Biochem J 181: 339–345 Sanda S, Leustek T, Theisen M, Garavito M and Benning C (2001) Recombinant Arabidopsis SQD1 converts UDP-glucose and sulfite to the sulfolipid head precursor UDP-sulfoquinovose in vitro. J Biol Chem 276: 3941–3946 Sato N and Murata N (1991) Transition of lipid phase in aqueous dispersions of diacylglyceryltrimethylhomoserine. Biochim Biophys Acta 1082: 108–111 Schmid PC, Kumar VV, Weis BK and Schmid HH (1991) Phosphatidyl-Tris rather than N-acylphosphatidylserine is synthesized by Rhodopseudomonas sphaeroides grown in Tris-containing media. Biochemistry 30: 1746–1751 Seay T and Lueking DR (1986) Purification and properties of acyl coenzyme A thioesterase II from Rhodopseudomonas sphaeroides. Biochemistry 25: 2480–2745 Seifert U and Heinz E (1992) Enzymatic characteristics of UDPsulfoquinovose:diacylglycerol sulfoquinovosyltranferase from chloroplast envelopes. Bot Acta 105: 197–205 Shibuya I, Miyazaki C and Ohta A (1985) Alteration of phospholipid composition by combined defects in phosphatidylserine and cardiolipin synthases and physiological consequences in Escherichia coli. J Bacteriol 161: 1086–1092 Sohlenkamp C, de Rudder KE, Rohrs V, Lopez-Lara IM and Gei-
Banita Tamot and Christoph Benning ger O (2000) Cloning and characterization of the gene for phosphatidylcholine synthase. J Biol Chem 275: 18919–18925 Sohlenkamp C, Lopez-Lara IM and Geiger O (2003) Biosynthesis of phosphatidylcholine in bacteria. Prog Lipid Res 42: 115–162 Svensson-Ek M, Abramson J, Larsson G, Tornroth S, Brzezinski P and Iwata S (2002) The X-ray crystal structures of wild-type and EQ(I–286) mutant cytochrome c oxidases from Rhodobacter sphaeroides. J Mol Biol 321: 329–339 Tsukihara T, Shimokata K, Katayama Y, Shimada H, Muramoto K, Aoyama H, Mochizuki M, Shinzawa-Itoh K, Yamashita E, Yao M, Ishimura Y and Yoshikawa S (2003) The low-spin heme of cytochrome c oxidase as the driving element of the proton-pumping process. Proc Natl Acad Sci USA 100: 15304–15309 Van Mooy BA, Rocap G, Fredricks HF, Evans CT and Devol AH (2006) Sulfolipids dramatically decrease phosphorus demand by picocyanobacteria in oligotrophic marine environments. Proc Natl Acad Sci USA 103: 8607–8612 Vance CP, Uhde-Stone C and Allan DL (2003) Phosphorus acquisition and use: critical adaptations by plants for securing a nonrenewable resource. New Phytol 157: 423–447 Wakeham MC, Sessions RB, Jones MR and Fyfe PK (2001) Is there a conserved interaction between cardiolipin and the type II bacterial reaction center? Biophys J 80: 1395–1405 Webb MS and Green BR (1991) Biochemical and biophysical properties of thylakoid acyl lipids. Biochim Biophys Acta 1060: 133–158 Weissenmayer B, Gao JL, Lopez-Lara IM and Geiger O (2002) Identification of a gene required for the biosynthesis of ornithine-derived lipids. Mol Microbiol 45: 721–733 White SW, Zheng J, Zhang YM and Rock (2005) The structural biology of type II fatty acid biosynthesis. Annu Rev Biochem 74: 791–831 Wikstrom M, Xie J, Bogdanov M, Mileykovskaya E, Heacock P, Wieslander A and Dowhan W (2004) Monoglucosyldiacylglycerol, a foreign lipid, can substitute for phosphatidylethanolamine in essential membrane-associated functions in Escherichia coli. J Biol Chem 279: 10484–10493 Wood BJB, Nichols BW and James AT (1965) The lipids and fatty acid metabolism of photosynthetic bacteria. Biochim Biophys Acta 106: 261–273 Xie J, Bogdanov M, Heacock P and Dowhan W (2006) Phosphatidylethanolamine and monoglucosyldiacylglycerol are interchangeable in supporting topogenesis and function of the polytopic membrane protein lactose permease. J Biol Chem 281: 19172–19178 Yu B, Xu C and Benning C (2002) Arabidopsis disrupted in SQD2 encoding sulfolipid synthase is impaired in phosphate-limited growth. Proc Natl Acad Sci USA 99: 5732–5737
Chapter 8 Peripheral Complexes of Purple Bacteria Mads Gabrielsen, Alastair T. Gardiner and Richard J. Cogdell* Microbial Photosynthesis Laboratory, Glasgow Biomedical Research Centre, 120 University Place, University of Glasgow, G12 8TA, U.K.
Summary ............................................................................................................................................................... 135 I. Introduction..................................................................................................................................................... 136 II. Structure ......................................................................................................................................................... 136 A. The α/β Heterodimer ........................................................................................................................ 137 B. Pigments .......................................................................................................................................... 139 C. Low-light Conditions ......................................................................................................................... 143 D. Organization of Detergents in Crystals of the Light-Harvesting 2 Complex .................................... 145 III. The Biology of Purple Bacterial Antenna Complexes..................................................................................... 146 A. The Role of Peripheral Antenna Within the Photosynthetic Unit ...................................................... 146 B. Synthesis of Peripheral Antenna Complexes ................................................................................... 147 1. The Upper Pigmented Band ................................................................................................... 147 2. The PucC Protein ................................................................................................................... 147 C. Variations on a Theme ..................................................................................................................... 148 D. Have We Only Scratched the Surface of Natural Variation?............................................................ 149 1. Genes of Purple Peripheral Antenna Complexes .................................................................. 149 2. Too Many Peripheral Antenna Polypeptides Make Life Difficult ............................................. 150 3. Nomenclature .......................................................................................................................... 150 IV. Final Remarks ................................................................................................................................................ 151 Acknowledgments ................................................................................................................................................. 151 References ............................................................................................................................................................ 151
Summary This chapter uses the LH2 complex from Rhodopseudomonas acidophila strain 10050 as an example to describe the current understanding of the structure of purple bacterial peripheral antenna complexes. It summarizes both what is and what is not understood. So far the structures of ‘standard’ LH2 complexes, such as those from Rhodopseudomonas acidophila, are rather well characterized. In contrast, there is a dearth of structural information on complexes with more unusual spectroscopic properties. There is also very little information available on how these antenna complexes are assembled.
*Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 135–153. © 2009 Springer Science + Business Media B.V.
136
Mads Gabrielsen, Alastair T. Gardiner and Richard J. Cogdell
I. Introduction The first committed, irreversible step in the photosynthetic light-reactions is the separation of charge across the membrane. This charge separation event occurs in the reaction center (RC). However, if photosynthesis relied solely on photons that can be harvested by the pigments of the RC then the majority of the RCs would spend most of their time waiting for the arrival of an incoming photon. To get around this problem, sets of pigment-protein complexes have evolved that are able to harvest a far greater proportion of photons striking the membrane and then to pass on this energy to the RC. These pigment-protein complexes, which effectively increase the cross-sectional area available for photon capture, are called antenna complexes. In purple bacteria if the cell is growing photosynthetically, invaginations from the inner membrane, termed the intracytoplasmic membrane (ICM), contain mature photosynthetic complexes. These pigment-protein complexes consist of integral membrane proteins that bind, non-covalently, the light absorbing pigments, bacteriochlorophyll (BChl) and carotenoids. Typically there are two types of antenna complexes. The LH1 antenna complex is intimately associated with the RC and is commonly referred to as the ‘core’ complex. All species of purple bacteria have this core complex. Most species then also have a second type of antenna complex called LH2. The LH2 complexes are arranged more peripherally and funnel excitation energy to the RC via LH1. This funnel exists because the Qy absorption bands of the BChls present within LH2 are at shorter wavelengths (at higher energy) than the corresponding absorption band in LH1. For example, the well known LH2 complex from Rhodopseudomonas (Rps.) acidophila strain 10050 has near infra-red (NIR) absorption maxima at 800 and 850 nm (hence this complex is also called B(bulk)800-850), whereas the LH1 complex absorbs at 880nm (B880). To illustrate this gradient, Abbreviations: (B)Chl – (bacterio)chlorophyll a; AFM – atomic force microscopy; B – bulk; Chr. – Chromatium; DPG – diphosphatidylglycerol; HL – high light; HPTO – heptane-1,2,3-triol; ICM – intracytoplasmic membrane; LDAO – lauryl dimethyl amine-N-oxide; LH – light-harvesting; LL – low light; NIR – near infra-red; PC – phosphatidylcholine; PE – phosphatidylethanolamine; Phs. – Phaeospirillum; PSU – photosynthetic unit; Rba. – Rhodobacter; RC – reaction center; Rdv. – Rhodovulum; Rps. – Rhodopseudomonas; Rsp. – Rhodospirillum; Rvi. – Rubrivivax; UDAO – N,N-dimethyundecylamine-N-oxide; UPB – upper pigmented band
Fig.1. Absorbance spectra of the peripheral, LH2 antenna (solid line) and the RC-LH1 ‘core’ complex from Rps. acidophila strain 10050 (dashed line). The Qy absorption bands of the BChl molecules present in LH2 absorb at 800 nm and 850 nm, hence this complex is also called B800-850. Similarly the Qy absorption band of the BChls in LH1 absorbs at approximately 890 nm (approximately 875 nm in some species). The smaller absorbance bands in the RC-LH1 complex at approximately 800 and 760 nm arise from the RC.
absorbance spectra of purified Rps. acidophila LH2 B800-850 and RC-LH1 ‘core’ complexes have been overlaid in Fig. 1. This chapter will examine peripheral, LH2 antenna complexes from purple bacteria from structural, molecular biological and physiological viewpoints. The biophysical processes involved in light-harvesting will be mentioned only as they relate to an understanding of the structure. The reader is directed to further chapters in Part 3 of this volume for complementary perspectives on peripheral antennas and/or the RC-LH1 ‘core’ complexes and for details of the energy-transfer reactions that take place in the LH2 complexes. II. Structure The structure of the peripheral antenna complexes from both Rps. acidophila and Phaeospirillum (Phs.) molischianum have been solved using Xray crystallography at resolutions between 2.0 and 2.5 Å (McDermott et al., 1995; Koepke et al., 1996; McLuskey et al., 2001; Papiz et al., 2003; Cherezov et al., 2006). The structure described in detail below is the one from Rps. acidophila, strain 10050, solved by McDermott et al. (1995) and Prince et al. (1997) to 2.5 Å and further improved by Papiz et al. (2003) to 2.0 Å [PDB accession codes 1KZU and 1NKZ
Chapter 8
Peripheral Complexes of Purple Bacteria
respectively]. PDB nomenclature has been adopted for designating atoms in the polypeptide chain. A. The α/β Heterodimer The antenna complex is composed of protomers, formed by heterodimers of polypeptide chains, named α and β, and their associated pigments (Figs. 2a–f), which aggregate to form the intact oligomeric structure. The short polypeptide chains (53 and 41 amino acids, respectively) span the membrane, with the N-termini on the cytoplasmic side (Papiz et al., 2003). The α-chain starts with a modified methionine with a carboxyl group attached N-terminally that is ligated to the Mg2+ ion of the B800 BChl (Papiz et al., 2003), followed by a three-residue loop involved in coordination of a BChl a molecule (Prince et al., 1997), giving way to a 310 helix that penetrates the membrane surface. This 310 helix is linked to a seventurn transmembrane helix via a three-residue section. The transmembrane helix is approximately parallel to the membrane normal. The C-terminal part of the α-chain is made up of a short linking section, a twoturn α-helix and an extended stretch (Prince et al., 1997; Papiz et al., 2003). The β-chain N-terminal residues begin in an extended section, followed by a single transmembrane helix, which is curved slightly with an inclination angle of about 15° towards the membrane normal (Prince et al., 1997). Apart from Asp17 and Arg20, which form a salt bridge, the transmembrane helical part of β is made up of hydrophobic residues (McDermott et al., 1995). The C-terminal polypeptides terminate in a short loop. The turns at both ends are exposed to the aqueous environment, and are characterized by side-chain to main-chain hydrogen bonds involving hydrophilic residues (Prince et al., 1997). The α-chains form a tightly-bound ring in the membrane, maintained by helix-helix interactions. The β-chains form a more open outer ring where the distances between the polypeptides do not allow any direct β- to β- interactions (Fig. 2b, 3a). No α/β apoprotein helix-helix interactions occur within the transmembrane part of the complex (Prince et al., 1997). The inner ring is stabilized by hydrogen bonds between the C-terminal part of one polypeptide and residues in adjacent chains, advanced or retarded by one unit. The α/β pairs interact via hydrogen bonds between residues in the N-terminal loops of both chains, but largely through protein-pigment
137
interactions (McDermott et al., 1995). Details of the hydrogen bonding between the polypeptides are presented in Table 1. The B800-850, LH2, complex from Rps. acidophila strain 10050 is made up of nine α/β pairs, forming a double ring with diameters of 36 Å and 68 Å for the inner and outer circles, respectively (Fig. 3a, 4). At the time of writing, three high-resolution crystal structures of LH2s are available; the B800-850 complex from Rps. acidophila strain 10050 (McDermott et al., 1995), the B800-820 from Rps. acidophila strain 7050 (McLuskey et al., 2001) and the B800850 from Phs. molischianum (Koepke et al., 1996). The Rps. acidophila structures are nonamers of the basic α/β-heterodimer repeating unit, whereas the Phs. molischianum structure is an octamer. Electron crystallography and atomic force microscopy (AFM) experiments performed on Rba. sphaeroides (Walz et al., 1998; Scheuring et al., 2003), Rhodovulum sulfidophilum (Montoya et al., 1995) and Rubrivivax (Rvi.) gelatinosus (Scheuring et al., 2001) all confirm that the LH2 complexes from these species are also nonameric, whereas the low-resolution crystal structure of low-light B800 LH2 from Rps. palustris (Hartigan et al., 2002), and the crystal structure of Phs. molischianum (Koepke et al., 1996) (Fig. 3b) both show octameric assemblies. For the first time, one report (Scheuring et al., 2004) has asserted that the ICM of Rhodospirillum (Rsp.) photometricum contains LH2 molecules of variable stoichiometry, with decamers and octamers being present in addition to the majority nonamers (see also Chapter 14, Sturgis and Niederman; Chapter 47, Scheuring). If this work can be confirmed by biochemical purification and spectroscopic analysis of the putative 8-, 9- and 10-mer ring sizes, they will prove an invaluable aid towards understanding the factors controlling ring size. The B800-850, LH2 complex from Phs. molischianum was solved to a resolution of 2.4 Å by Koepke et al. (1996) [PDB accession code 1LGH]. The α- and β-apoproteins from Phs. molischianum have a sequence identity of 26 % and 31 %, with respect to the corresponding apoproteins from Rps. acidophila. The Phs. molischianum structure is built up by protomer units (Fig. 2e) similar to those found in LH2 from Rps. acidophila strain 10050 (Fig. 2d), however, the overall complex is an octamer rather than a nonamer, leaving the diameters of the α-inner and β-outer helical rings as 31 Å and 62 Å respectively (Koepke et al., 1996) (Fig. 3b). The polypeptides of
138
Mads Gabrielsen, Alastair T. Gardiner and Richard J. Cogdell
Fig. 2. Representations of the β/α-protomer building blocks that comprise LH2. The β-chains are shown in black and the α-chains in pale gray. The BChl and the carotenoid are shown in wireframe. The C-terminal regions are uppermost. a) A view of the β/α protomer with the pigments removed. The main BChl liganding amino acid side chains are shown. b) A view of how two adjacent α/β protomers interact with each other, again with the pigments removed. The chains, from left to right, are in the order α, β, α, β. c) Similar to b) but looking down on to the two adjacent α/β-protomers from the C-terminal side of the complex. The β chains are darker than the α chains. d) The same view as a) but with the pigments included. e) A view of the β/α protomer from Phs. molischianum. The main differences between d) and e) are the altered orientation of the B800 bacteriochlorin ring and the different organization of the C-terminal region of the α-apoprotein. f) A view of the β/α protomer from Rps. acidophila strain 7050. Note the different amino acids shown at the C-terminal end of the α-apoprotein that are responsible for the altered absorption properties of this B800-820 complex. The structures shown in a–d) are representations of the structure of B800-850, the LH2 complex from Rps. acidophila strain 10050. See also Color Plate 3, Fig. 3.
the nonameric and the octameric structures are very similar in the transmembrane region, but differ at their ends. The α-polypeptide from Phs. molischianum has a longer C-terminal helix, and this helix is at a greater angle to the membrane-spanning helix compared to the α-chain from Rps. acidophila (140°
and 100°, respectively) (Koepke et al., 1996). The Nterminal 310 helices of the α-apoproteins are similar in structure, despite an important sequence difference in the N-terminal residue of the Phs. molischianum α-chain, which is a serine, as opposed to a methionine in Rps. acidophila.
Chapter 8
Peripheral Complexes of Purple Bacteria
139
Fig. 3. A comparison of the LH2 complexes from Rps. acidophila strain 10050 and Phs. molischianum. a) The Rps. acidophila 10050 nonameric structure viewed perpendicular to the membrane normal with b) the Phs. molischianum octameric structure. The doughnut shape of the whole molecular complex is apparent, with the two polypeptide rings (β- outside, α-inside) darkly shaded and the pigments shown with lighter tones.
Recent computational studies suggest that the difference in oligomerization is based on differences in the interaction angle between two subunits. In theory these are 45° for a ring consisting of eight protomers, and 40° for a ring consisting of nine protomers (Janosi et al., 2006). Modeling how two protomers associate reveals that the free energy profile minimum is located at about 45.2° for Phs. molischianum and at 38.5° for Rps. acidophila. Further, the free energy profiles exhibit a clear angular separation, requiring 5.5 kcal/ mol to force the subunits from Phs. molischianum into the Rps. acidophila angle, and 8.5 kcal/mol vice versa (Janosi et al., 2006). The data presented by these authors suggests that the preferred angle between the subunits plays an important role in the assembly of the complex and at least partly determines the ring size and oligomerization state. The modeling data suggest that the preferred angles between the two subunits are determined by the surface interactions in the transmembrane regions (Janosi et al., 2006). The authors speculate that there is a motif in the Cterminal ends of the Rps. acidophila polypeptides that plays a role in constraining the angle between the subunits in addition to the hydrophobic surface contacts. This double constraint on the angle may explain why nonamers are found more often than other ring sizes (Janosi et al., 2006).
B. Pigments The double ring of α- and β-polypeptides functions as a scaffold for the pigments. There are three BChls a per LH2 protomer, two B850 molecules sandwiched between the polypeptides, and one B800 parallel to the membrane surface (Fig. 2d, 5). The two B850 molecules form closely coupled dimers perpendicular to the membrane surface, so that they are edge on when observed from above (Freer et al., 1996). The intact holocomplex LH2 from Rps. acidophila then consists of a ring of 18 overlapping B850 pigments (McDermott et al., 1995). The primary contact between the B850 molecules and the peptide chains is co-ordination of the central magnesium ion through histidine residues (αHis 31 and the Mg2+ of αB850, and βHis30 and the Mg2+ of βB850) (Prince et al., 1997). The B850 α/β pair has an Mg-Mg separation of 9.6 Å, whereas the adjacent protomer Mg is separated by 8.9 Å (Freer et al., 1996). The closest ring–ring distances are 3.54 Å within the B850 α/β pair and 3.63 Å between adjacent pairs. A hydrogen bond is formed between the C3-acetyl group on ring A of BChl and αTrp45 NE1 for αB850 and (+)αTyr44 OH for βB850 (a bracketed plus or minus sign indicates that the residue belongs to an adjacent protomer). Further, both B850 molecules interact with their respective apoproteins so the phytyl chains are
140
Mads Gabrielsen, Alastair T. Gardiner and Richard J. Cogdell
Fig. 4. The LH2 complex from Rhodopseudomonas (Rps.) acidophila strain 10050. a) A view looking down on the top of the complex from the presumed periplasmic surface of the membrane. (b) A side view looking from within the presumed photosynthetic membrane with the periplasmic side of the complex uppermost. In both panels the LH2α polypeptide is in gray, LH2β is black. See also Color Plate 3, Fig. 4.
directed from the bacteriochlorin plane towards the transmembrane helices, forming hydrophobic interactions with the polypeptides (Prince et al., 1997). The remaining protein-pigment interactions are made up of van der Waals interactions and are detailed in Table 2. The phytyl chain of αB850 makes contact with a number of residues on two α-chains, before passing the edge of the B800 bacteriochlorin ring, making contacts with O1A, O1D and CBD (Papiz et al., 2003). The βB850 phytyl chain forms contacts
with residues on the β-chain, passing over the B800 bacteriochlorin ring of this BChl in the same protomer, forming contacts between the two pigments (Papiz et al., 2003). There are nine B800 BChl molecules present in the antenna complex located approximately 16 Å into the membrane and inserted between the β-polypeptides. These BChl molecules lie in a plane parallel to the membrane surface (Figs. 2b, 5). The primary contact between the polypeptides and the B800 molecule is the co-ordination of the magnesium central ion and the carboxylated αMet1 (Papiz et al., 2003). This modification is supported by several hydrogen bonds donated from the surrounding residues N-αAsn2, N-αGln3 and NE2-βHis12 (Papiz et al., 2003). The C3-acetyl group on ring A forms a hydrogen bond to βArg20 (Prince et al., 1997). The last pigment making up the protomer is the carotenoid, which, in Rps acidophila, strain 10050, is rhodopin glucoside. The glucoside head group is the only part capable of forming strong interactions such as hydrogen bonds, it is, therefore, interesting that the head group seems able to adopt two distinct conformations (Prince et al., 1997). The carotenoid runs in the space between the α and β pairs, in the same region as the phytyl chains of the BChls and is important for the structural stability of the transmembrane units (Lang and Hunter, 1994). The protein-carotenoid and carotenoid-BChl contacts are detailed in Table 3. The carotenoid takes up a helical confirmation, which imparts chirality to the conjugated double-bond system (Prince et al., 1997). This may affect the photochemical properties of the carotenoid (Nagae et al., 1993). It has been suggested that evidence in favor of a second carotenoid can be found in the uninterpreted non-protein parts of the electron density map of LH2, and this has been modeled in the structure solution by Papiz et al (2003). However, a recent structure of LH2 from Rps. acidophila strain 10050, obtained following crystallization using lipidic cubic phases, shows that this density is better explained by a superposition of mixtures of β-octylglucoside and lauryl dimethyl amine-N-oxide (LDAO) detergent molecules (Cherezov et al., 2006) rather than a second carotenoid. The density described in the Papiz structure is most likely, therefore, to be caused by a mixture of detergents binding throughout the crystal, rather than a pigment. This explanation is strongly supported by recent spectroscopic studies, which have shown that there is no second carotenoid present, either in solu-
Chapter 8
Peripheral Complexes of Purple Bacteria
141
Table 1. Hydrogen bonds between polypeptides Residue COO-Met α1
Atom O1
Residue His β12
Atom NE2
Distance (Å) 2.96
COO-Met α1
SD
Arg β20
NH2
3.31
Gly α4
O
Ser β8
OG
2.65
Gly α4
O
Ser β8
O
3.30
Trp α7
N
Ser β8
OG
3.07
Trp α7
NE1
His β12
ND1
3.01
Trp α7
O
Leu β3
N
2.96
Asn α11
OD1
Ala β1
N
2.67
Thr α39
N
Gln (+)α46
OE1
2.85
Thr α39
OG1
Gln (+)α46
NE2
2.83
Trp α40
NE1
Trp (+)α45
O
2.89
Tyr α44
OH
Trp (-)β39
NE1
3.31
Trp α45
O
Trp (-)α40
NE1
2.87
Gln α46
OE1
Thr (-)α39
N
2.90
Gln α46
NE2
Thr (-)α39
OG1
2.88
Lys α50 NZ His (-)β41 ND1 2.86 The (+) represents an adjacent polypeptide in a counterclockwise direction (as seen from the N-terminal side of the complex). The (-) represents an adjacent polypeptide in a clockwise direction (as seen from the N-terminal side of the complex). Modified from Prince et al. (1997).
tion or in crystals of LH2 (Gall et al., 2006). Lycopene is the major carotenoid present in the Phs. molischianum B800-850 complex (Germeroth et al., 1993), with rhodopin as a minor species (Koepke et al., 1996). Although different from the carotenoid found in the Rps. acidophila structure, they adopt a similar orientation, with a minor deviation close to the B850 molecules (Koepke et al., 1996). Some density has been found in the region where the ‘second carotenoid’ has been suggested in the structure by Papiz et al. (2003), but has been modeled as a detergent molecule in Koepke et al. (1996). The BChl binding in the crystal structure of the Phs. molischianum LH2 (Figs. 2e, 3b) is similar to that in the Rps. acidophila complex. The overall complex comprises a ring of 8 B850 dimers, in a circle perpendicular to the membrane surface, and another ring made up of 8 B800 BChls parallel to
the membrane surface. The two apoprotein rings are stabilized by hydrogen bonds detailed in Table 4. The finding that these two hydrogen bonds are confined within an α/β protomer in Phs. molischianum may well be why this complex is more like LH1. The B850 BChls in Phs. molischianum are co-ordinated with the polypeptide chains via the central magnesium ions and conserved histidines (α34 and β35, respectively), with minimum distances between the αB850 and βB850 rings within a dimer of 3.66 Å, and of 3.74 Å to the next protomer. Two tryptophans (αTrp45 and βTrp44) are involved in hydrogen bonds between the C3-acetyl group on ring A of BChl a αB850 and βB850, respectively (Koepke et al., 1996). The binding environment for the B800 bacteriochlorin ring is hydrophilic, in contrast to the B850 binding site. The B800 bacteriochlorin ring in
Mads Gabrielsen, Alastair T. Gardiner and Richard J. Cogdell
142
Table 2. Contacts formed between the polypeptides and the B850 bacteriochlorophylls of Rps. acidophila 10050 LH2 B850 atom αMg
Residue His α31
Atom NE2
Distance (Å) 2.34
αO3
Trp α45
NE1
2.97
αC5
Phe α41
CZ
3.73
αC32
Val (–)α30
CG1
αC71
Tyr α44
αC81
Table 3. Rhodopin glucoside contacts to the polypeptides from Rps. acidophila 10050 LH2. The carotenoid labeling scheme used is that outlined by Freer et al. (1996). Residue
Atom
Ile (–)α6
CA
3.81
4.36
Lys (–)α5
O
3.78
CE2
4.00
Val (–)α9
CG2
3.70
Trp α40
CH2
3.73
CM3
Tyr β14
CD2
3.43
αC88
Ile α34
CD1
3.98
C6
Leu β11
CD2
4.10
αC121
Ala β33
CB
4.14
CM4
Gln (–)α3
OE1
3.54
αC13
Ala β29
CB
3.43
C12
Val β15
O
3.87
αO131
His β30
CE1
3.19
C13
Leu α20
CD2
3.82
Ala β26
CA
3.43
CM3
Gly β18
O
3.66
αC134
Leu β25
CB
3.56
C14
Thr β19
OG1
3.37
αO173
Ala α27
CB
3.35
C15
Phe β22
CB
4.26
αOP
Phe β22
CE1
3.34
C18
Val α23
CG2
3.64
βMg
His β30
NE2
2.34
C25
Ile α26
CG2
3.73
βO31
Tyr (+)α44
OH
2.64
CM8
Ala (+)α27
CB
4.06
βC10
Ile α34
CD1
3.63
Ile (+)α28
CG1
4.02
βC7
Trp β39
CE2
3.43
βC71
Trp (+)α45
CD1
3.54
βC82
Trp α40
CH2
3.98
Thr β37
CG2
4.09
Ala β33
CB
3.88
βC12
Val α30
CG1
3.64
βC121
His α31
CE1
3.41
βO131
Ala α27
CB
3.51
βC134
Val α23
O
3.72
Ile α26
CG2
4.12
Leu β23
CA
3.86
Ala β26
CB
3.28
1
βO173
βOP Phe β22 CD2 3.48 The (+) represents an adjacent polypeptide in a counterclockwise direction (as seen from the N-terminal side of the complex). The (–) represents an adjacent polypeptide in a clockwise direction (as seen from the N-terminal side of the complex). Modified from Prince et al. (1997).
Rhodopin glucoside atom C4
Distance (Å)
C27 His (+)α31 NE2 4.11 The (+) represents an adjacent polypeptide in a counterclockwise direction (as seen from the N-terminal side of the complex). The (–) represents an adjacent polypeptide in a clockwise direction (as seen from the N-terminal side of the complex). Modified from Prince et al. (1997).
Phs. molischianum is tilted 20° with respect to the corresponding one in Rps. acidophila (Koepke et al., 1996). Furthermore, in Phs. molischianum the B800 bacteriochlorin ring is rotated by 90° relative to that in Rps. acidophila. The carboxylated methionine in the N-terminal sequence of Rps. acidophila is replaced by a serine in the sequence in Phs. molischianum. This sequence difference leads to the central Mg2+ of the B800 BChl molecule being co-ordinated via the OD1 of αAsp6 (2.45 Å) so that the second carboxyl oxygen of αAsp6 in close proximity to the amide ND of αAsn2, creates a strong hydrogen bond between these residues (Koepke et al., 1996). The binding of the B800 molecule is further secured via a water molecule and a contact between NE2 of βHis17 and a carbonyl oxygen on the BChl ring (Koepke et al., 1996).
Chapter 8
Peripheral Complexes of Purple Bacteria
143
Fig. 5. An overall picture of the arrangement of the BChls and carotenoids in the LH2 complex from Rps. acidophila strain 10050. At the top the ring of 18 tightly coupled BChl molecules that give rise to the absorbance band at 850nm can be clearly (dark wireframe). The well separated B800 BChl molecules (gray wireframe) can be seen with their bacteriochlorin rings lying perpendicular to those of the B850 ring. The carotenoid molecules can be seen in an all-trans configuration with a pronounced twist along their long axes.
Table 4. Hydrogen bonds involved in the complexation of Phs. molischianum LH2 Residue Trp α10
Atom O
Residue Ser β6
Atom OG
Distance (Å) 2.50
Leu α11
O
Ser β6
N
2.75
Leu α11
O
Leu β5
N
3.08
Tyr α7
ON
Glu β10
OE2
2.83
Trp α10
NE1
His β17
ND1
3.12
Asn α2
ND2
Gln (+)β19
OD1
3.16
Asn α2
OD1
Gln (+)β19
NE2
3.12
Ser α53 OG Pro (-)β43 O 3.53 The (+) represents an adjacent polypeptide in a counterclockwise direction (as seen from the N-terminal side of the complex). The (-) represents an adjacent polypeptide in a clockwise direction (as seen from the N-terminal side of the complex). Modified from Prince et al. (1997).
C. Low-light Conditions The differences in the spectroscopic properties between the different LH2 complexes must be explained
through differences in their structures. Comparing the B800-850 LH2 complex from Rps. acidophila strain 10050 with the structure of the low-light B800-820 complex from Rps. acidophila strain 7050 (also
Mads Gabrielsen, Alastair T. Gardiner and Richard J. Cogdell
144
known as the LH3 complex), solved by McLuskey et al (2001) [PDB accession code 1IJD], reveals that the overall structures are similar with nine protomers forming a circle. The protomers are composed of an α- and a β-polypeptide as well as three BChl a molecules and when superposed onto α- and β- from B800-850 gives a root mean square deviation of 0.34 Å for all main chain atoms (McLuskey et al., 2001) (Fig. 2f). The B800 molecule is bound to what has been assigned as a formylated methionine residue (McLuskey et al., 2001), but it may be that a higher resolution structure will reveal that this methionine is carboxylated like B800-850 from Rps. acidophila (Papiz et al., 2003). Due to the relatively low resolution of this structure (3.0 Å), the identity of the carotenoid could not be assigned unambiguously (McLuskey et al., 2001). However, previous biochemical studies have shown that rhodopinal glucoside is the major carotenoid present in B800-820 (Gardiner et al., 1993). The residues involved in the contacts to the pigments are detailed in Table 5. When aligning the amino acid sequence of B800850 and B800-820 α- and β-polypeptides are highly similar, and only a few residues differ (Fig. 6). These
residues include αTyr44 and αTrp45 in B800-850, which form hydrogen bonds with the C3-acetyl group on ring A on B850. The equivalent residues in the B800-820 complex are Phe and Leu, both unable to form hydrogen bonds. This leaves the βB820 molecule co-ordinated through its central Mg2+ to a His residue, but leaves the C3-acetyl group and the C131 keto group without any hydrogen bonds. The αB820 is co-ordinated through its central Mg2+ to a His residue, and by a hydrogen bond formed between the C3-acetyl group and αTyr41, leaving the C131 keto group with no H-bond (McLuskey et al., 2001). This lack of a H-bond results in the acetyl group being free to rotate relative to the plane of the bacteriochlorin ring and giving B820 C3-acetyl group torsion angles of 132° and 135° (for α- and β-, respectively,) compared with the equivalent values from B800-850 of –166° and 156° (McLuskey et al., 2001). It is currently believed that this rotation is the major cause of the difference in absorption maxima between the B800-850 and B800-820 complexes. This theory is supported by the results described by Fowler et al. (1992), who performed mutational studies on two C-terminal residues on the α chain of the LH2 complex from Rhodobacter (Rba.)
Table 5. Contacts between the pigments and the residues of Rps. acidophila strain 7070 B800-820 Pigment atom B800 Mg
Residue f-Met α1
Atom OF
B800 O31
Arg β21
NE
3.08
αB820 Mg
His α31
NE2
2.42
αB820 O3
Tyr α41
OH
2.74
βB820 Mg
His α31
NE2
2.43
αB820 C2
(+)βB820
C2
3.95
αB820 C12
βB820
C12
3.52
Carotenoid C26
(+)αB820
C20
1
Distance (Å) 2.49
3.93 ´
Carotenoid C11
(-)B800
O13
3.47
B800 Mg
(+)αB820
Mg
17.62
B800 Mg
βB820
Mg
18.35
αB820 Mg
βB820
Mg
9.51
βB820 Mg (+)αB820 Mg 8.97 The (+) represents an adjacent polypeptide in a counterclockwise direction (as seen from the N-terminal side of the complex). The (-) represents an adjacent polypeptide in a clockwise direction (as seen from the N-terminal side of the complex). Modified from Prince et al. (1997).
Chapter 8
Peripheral Complexes of Purple Bacteria
145
Fig. 6. A sequence comparison of a selection of LH2 α- and β-apoproteins. Pal- Rps. palustris, Ac- Rps. acidophila, sph- Rba. sphaeroides, mol- Phs. molischianum. The numbers beside Ac represent the different strains of Rps. acidophila. The Rps. palustris sequences are deduced from gene sequences. Those from Rps. acidophila are all protein sequences, apart from one example labeled (gene). Both the Rba. sphaeroides and the Phs. molischianum sequences are from the proteins. The degree of identity is represented by the intensity of the greyscale. It is noticeable that the gene sequences tend to show C-terminal extensions that are processed away in the mature polypeptides. At present there is no information about how this is accomplished or for what reason. The sequences Pal-PucBb and PalPucAb appear to be anomalous. Whether in fact these sequences do represent a LH2 complex is not yet clear. The sequence differences that result in the absorption shift from B50 to B820 in Rps. acidophila have been highlighted in the box. References for these sequences are given in the text.
sphaeroides, Tyr44 and Tyr45 (in Rps. acidophila, these residues correspond to αTyr44 and αTrp45, respectively). When a single mutant, αTyr44 → Phe, was introduced, the absorbance maximum of the 850 nm band is shifted 11 nm to 839 nm (Fowler et al., 1992). A double mutant, αTyr44 → Phe, αTyr45 → Leu, led to the absorbance maximum shifting 24 nm, to 826 nm, resembling the B800-820 complex from Rps. acidophila. The differences in the Qy absorption bands in LH2 and LH1 create an energetic funnel as described above. These differences are dependent on the environment and the organization of the BChl molecules, where the Qy absorption bands can be shifted from 800 nm up to 875 nm. The arrangement of the bacteriochlorin rings in LH2 is such that the Qy transition dipoles of both the B800 and B850 BChls are approximately parallel to each other, even though their bacteriochlorin rings are perpendicular to each other. The C(3) acetyl group discussed elsewhere in this chapter differs between the B800 and the B850
molecules in that the torsion angle is opposite that of the B850 molecules for B800 (180° compared with 0° for B850). The B800 carbonyl points toward the Qy axis in B800, and the reverse occurs for B850, leading to a slight rotation of the Qy transition dipole in B800 with respect to B850 (Freer et al., 1996). D. Organization of Detergents in Crystals of the Light-Harvesting 2 Complex Crystallization of membrane proteins is dependent on the presence of detergents as these interact with the hydrophobic segments of membrane proteins to maintain their stability and solubility. The packing of detergents around the LH2 complex from Rps. acidophila strain 10050 has been studied in detail using neutron diffraction (Prince et al., 2003). The outer surface of the complex (β-polypeptides) is surrounded by a 15 Å thick band (at maximum) of detergent tails, with the height of the external detergent tails varying between 26 Å at maximum height to less than 10 Å
146
Mads Gabrielsen, Alastair T. Gardiner and Richard J. Cogdell
(Prince et al., 2003). On the N-terminal side of LH2, the detergent tails reach the level of the array of B800 BChl molecules, and the charge distribution from the Arg/Asp ion pair in this region seems to delimit the band, whereas on the C-terminal side the limit is more variable (Prince et al., 2003). Within the LH2 complex a continuous detergent cylinder was observed corresponding to the hydrophobic segments of the α-polypeptides. The height of this detergent cylinder is 40 Å corresponding to full thickness of the LH2 complex. The detergent tails seem to associate with the protein parts in an individual manner and do not appear to be involved in the crystalline lattice, confirming these crystals as Type II (Prince et al., 2003). It is possible that residual phospholipid or the detergent LDAO remains in the micelle. However, these would only be observed if they form distinct volumes within the crystals. Displacement of the observed detergent tails’ match-points for the hydrogenated and deuterated detergent samples suggests that a degree of mixing of detergent/lipid tails, with the major factor in the variation not caused by the presence of residual LDAO or phospholipid (Prince et al., 2003). The central detergent cylinder is consistent with it being filled by two β-octylglucoside micelles fused with the inner surface of the central hole. The interface may contain a significant number of benzamidine molecules (the amphiphile used in the crystal conditions) intercalating with the glucoside groups of the detergent moieties (Prince et al., 2003). In the ICM, this central cylinder would be filled by naturally present phospholipids. In Phs. molischianum density was observed that was assigned to the detergent N,N-dimethyundecylamine-N-oxide (UDAO). UDAO was fitted close to the B850 BChl, in the position where the putative ‘second carotenoid’ was fitted in the structure from Rps. acidophila by Papiz (2003). The amphiphile heptane-1,2,3-triol (HPTO) was fitted in another pocket on the outside of the octamer. Interestingly, within the central cylinder several pieces of electron density were observed, which have been assigned to three HPTO molecules and one detergent molecule. The amphiphile is in direct contact with the protein, whereas the detergent forms a second layer beyond the HPTO molecules (Koepke et al., 1996). It is possible that there is actually a mix of residual phospholipids with the detergents Work has been done on the presence of lipids in Rps. acidophila, strain 10050, from the whole cell membranes and in the purified B 800-850 complex
(Russell et al., 2002). Compared with whole cell membranes, the antenna complex only retained approximately equal proportions of phosphatidylethanolamine (PE) and phosphatidylcholine (PC), together with 15% diphosphatidylglycerol (DPG). However, there were no traces of phosphatidylglycerol (PG) (Russell et al., 2002). Further, the proportions of some of the unsaturated lipids are altered between the whole cell membranes, and the purified B850-800 sample, e.g., 2-OH 16:1 and 18:1∆11 were increased whilst 16:1∆9 was decreased (Russell et al., 2002). There is no information on whether specific residual lipids are structurally or functionally important for LH2 complexes. III. The Biology of Purple Bacterial Antenna Complexes A. The Role of Peripheral Antenna Within the Photosynthetic Unit Through classic experiments with the green alga Chlorella, Emerson and Arnold (1932) were the first to advance the concept of the ‘photosynthetic unit’ (PSU). The PSU has in subsequent years become established to mean the number of interconnected pigments per RC. Over the years a detailed, kinetic, but by no means complete, understanding of the energy transfer events within the bacterial PSU has been achieved; Fleming and van Grondelle (1997), Sundström et al. (1999) and van Grondelle and Novoderezhkin (2001) are useful starting references and interested readers can also refer to the other chapters in Part 3 of this volume (Chapter 11, Robert; Chapter 13, van Grondelle and Novoderezkhin; Chapter 14, Sturgis and Niederman; Chapter 15, Şener and Schulten; Chapter 47, Scheuring). X-ray crystallography has made it possible to obtain high-resolution structural data on the individual components of the PSU. However, what is now needed is a detailed description of the in vivo organization of the PSU. AFM has emerged in the past few years as an extremely useful method for measuring the topology of membranes and surfaces (refer to Chapter 47, Scheuring), enabling the direct visualization of reconstituted complexes (Stamouli et al., 2003; Gonçalves et al., 2005b) in lipid bilayers or the native ICM and its constituent components (Bahatyrova et al., 2004; Scheuring et al., 2006). AFM images, though, must be treated with great caution to avoid over interpretation. Neverthe-
Chapter 8
Peripheral Complexes of Purple Bacteria
less, images now being produced from the laboratories of Scheuring and Hunter are revealing new insights into the long-range organization of the PSU, i.e., the in vivo relationship between LH2 and the RC-LH1 complexes. This point is illustrated by recent AFM experiments on Rba. sphaeroides (Bahatyrova et al., 2004), which revealed for the first time two distinct domains of LH2 present in the ICM. The first forming regions of LH2 (‘energy conduits’) interspersed closely within the rows of dimer RC-LH1-PufX complexes, resembling closely a model proposed in 1976 on the basis of fluorescence quenching experiments (Monger and Parson, 1977), and a second clustered pool of non-associated LH2s (for more details readers are asked to refer to Chapter 14, Sturgis and Niederman). As an aside, interestingly from a historical perspective, these images support strongly the idea that energy transfer within the PSU is a combination of both the ‘lake’ and ‘puddle’ models of exciton migration (Vredenberg and Duysens, 1963). It can be envisioned that the LH2 molecules associated with the energy conduits are ‘puddle’ molecules, those free in the membrane are ‘lake’ molecules but with both functioning in concert to supply excitons to the RC-LH1-PufX arrays. Furthermore, the proposal that there are two distinct domains of LH2 molecules, one relatively invariant in response to changes in light intensity (the rows of LH2 ‘energy conduits’), and the other a variable LH2 pool without the LH1-RCPufX rows. This ground-breaking work has enabled a great stride to be taken towards understanding the native molecular organization of the ICM within chromatophores from Rba. sphaeroides. It is now seems clear from the various AFM studies published from a wide range of purple bacteria (Scheuring et al., 2001, 2003, 2004, 2006; Bahatyrova et al., 2004; Gonçalves et al., 2005a) that the previous assumption of LH2 complexes uniformly distributed within the ICM lipid bilayer is no longer valid. Rather, although the precise details may vary, there are clearly different domains of LH2 complexes and these modulate the energy transfer processes within the PSU and the photosynthetic membrane. B. Synthesis of Peripheral Antenna Complexes It has long been a challenge for biologists to understand the cellular processes whereby a translated polypeptide becomes part of a large, membrane bound, multimeric complex. In the case of LH2
147
complexes, although the apoproteins are very small and traverse the membrane only once, they act as a scaffold to hold the pigments in a very precise orientation. The complex assembly machinery must bring together two different 5–7 KDa polypeptides, α- and β-, three BChl molecules and one carotenoid to form a protomer. The protomers must then be assembled into the intact oligomeric holocomplex. Although all the molecular details of LH2 assembly are far from being understood, two components have been identified as being crucial to this process. The first is the upper-pigmented band (UPB) and the second is a protein called PucC. 1. The Upper Pigmented Band It has previously been shown that when Rba. sphaeroides cells are disrupted a membrane fraction can be isolated from sucrose density gradients which is less dense than the standard chromatophores (Niederman et al., 1979). This membrane fraction has been called the upper-pigmented band (UPB). It contains high levels of BChl synthase, the terminal enzyme of the BChl pathway. Pulse-chase experiments with radio-labeled, newly-synthesized photosynthetic complexes showed that initially they passed through the UPB on their way into the mature ICM (Reilly and Niederman, 1986). Further complementary membrane fractionation experiments and spectroscopic measurements have led to the conclusion that the UPB is the link between the cytoplasmic membrane and ICM. It appears that the UPB represents sites of ICM development (Hunter et al., 2005). 2. The PucC Protein In purple bacteria the structural genes that encode the β- and α-polypeptides are called pucBA and are part of the larger puc operon. Downstream of pucBA other genes have been found whose products have a role in LH2 biosynthesis. Original experiments with Rba. capsulatus showed that in this species the puc operon consists of pucBACDE. In particular it was suggested that the PucC protein is intimately associated with assembly of the LH2 complex (Tichy et al., 1991). PucC genes have also been described downstream from a puc multigene cluster in Rps. acidophila and downstream from pucBA in Rubrivivax (Rvi.) gelatinosus, although in this latter case it does not form an operon with pucBA as the pucC gene is transcribed off the opposite strand (Simmons et al.,
148
Mads Gabrielsen, Alastair T. Gardiner and Richard J. Cogdell
2000). Genes encoding PucC or related proteins have been identified from Rhodovulum sulfidophilum and Rsp. rubrum (Bérard et al., 1989) as well. Deletion of pucC in either Rba. capsulatus or Rba. sphaeroides produces cells that lack LH2 (Gibson et al., 1992). However, at least in Rvi. gelatinosus, other factors are certainly also involved since in this case a PucC mutant has been shown to still be able to produce LH2 under photosynthetic conditions albeit at very low-levels (Steunou et al., 2004). Two additional Rba. capsulatus genes within the photosynthetic cluster, namely lhaA and orf428, have also been suggested to encode proteins with analogous functions to PucC, but which are involved in the assembly of LH1 and membrane located BChl synthesis respectively (Bauer et al., 1991). A sequence alignment shows high similarity between the various PucC proteins (Simmons et al., 2000) and hydropathy analysis predicts that it is probably an integral membrane protein with 12 membrane spanning helices (LeBlanc and Beatty, 1996) as, indeed, is LhaA (Young and Beatty, 1998). The precise role of PucC in LH2 assembly is still unresolved, however it is thought to function in the delivery of BChl to the assembling LH2 complex ( Young and Beatty, 1998; Young et al., 1998). Both PucC and LhaA have homology to other members of the Major Facilitator Superfamily (MFS), a loosely related group of membrane bound secondary carrier proteins that transport a large range of substrates; for general MFS reviews see Pao et al. (1998) and Chang et al. (2004). C. Variations on a Theme In response to growth at high light (HL) or low light (LL) intensities, Rps. acidophila strain 10050 adjusts the amount of the LH2 complex present in the ICM probably in a manner broadly similar to that described above for Rba. sphaeroides. Even at very HL intensities, this strain of Rps. acidophila always synthesizes some LH2 (i.e., the LH2/LH1 ratio never falls below a certain value). As yet no growth conditions have yet been discovered where Rps. acidophila strain 10050 synthesizes any LH2 complex other than B800-850 (Figs. 1 and 7a). However, other strains of Rps. acidophila have been described that are able to produce, in response to changes in growth conditions, different LH2 complexes with altered NIR absorption spectra (Gardiner et al., 1993). Rps. acidophila strain 7750 produces a different LH2 complex in
response to growth at LL and/or less than the optimum growth temperature, while strain 7050 produces its different type of LH2 complex only in response to LL growth (Fig. 7b). These additional types of LH2 complexes have NIR absorption maxima blue-shifted from 850 nm to 820 nm, hence, the complex is called B800-820 (as described above). The carotenoid composition of the 7750 peripheral complexes is rather independent of light intensity, always containing the orange-colored rhodopin glucoside. In strain 7050, however, the carotenoid composition is markedly different in the B800-820 complex. In this case rhodopin glucoside is replaced by the purple colored carotenoid rhodopinal glucoside (Gardiner et al., 1993; Chapter 6, Takaichi). Other species, in addition to Rps. acidophila, also have the ability to synthesize a range of spectroscopically different LH2 complexes in response to changes in growth conditions. Rps. palustris produces a complex at LL with a greatly reduced 850 nm NIR absorption maximum (Fig. 7c) (van Mourik et al., 1992; Gall and Robert, 1999; Tharia et al., 1999). This complex has been called both B800 and LH4 (Evans et al., 2005). Also in response to LL, Chromatium (Chr.) vinosum will produce a B800 complex (Fig. 7d) (Thornber, 1970) and Rps. cryptolactis a B800-820 complex (Halloren et al., 1995). For completeness and rather intriguingly, under extreme selective stress, i.e., deletion of the B800-850 complex, revertants of Phs. molischianum are able to express a B800-820 complex (Koepke et al., 1996). This species had not previously been observed to change the type of LH2 in response to changes in light intensity, so this novel B800-820 complex had never been reported before. Some other species such as Chr. purpuratum only produce a B800-830 complex (Cogdell et al., 1990). The question can be asked as to why some species of purple bacteria, under LL conditions, have evolved an ability to produce a completely different LH2? The answer would appear to be that this complex confers a selective advantage on the cells under these conditions. An indication of the nature of this advantage has been gained from fluorescence excitation and annihilation experiments on Rps. acidophila strain 7750 (Deinum et al., 1991). At HL the ICM contains B800-850 and RC-LH1 with a BChl absorption at 880 nm (B880). Excitons are transferred from the B850 BChl of LH2 to the B880 BChl with a rate constant of 2–4 ps (Hess et al., 1995; Sundström et al., 1999). However, due to a large spectral overlap the ‘back-reaction’, i.e., from B880 to B850, can
Chapter 8
Peripheral Complexes of Purple Bacteria
also occur. This escape of the exciton back into the peripheral antenna is not detrimental to the cell as at HL photons are plentiful. However, under LL conditions the number of photons available is limited and must be harvested efficiently to enable the cell to maintain a competitive advantage. Under LL conditions Rps. acidophila 7050 produces the B800-820 complex. Forward energy transfer is fast and efficient from B820 to B880 in spite of the smaller spectral overlap, however, the energetically uphill, back-reaction from B880 to B820 is much less favorable. The energy jump (0.11 eV) provides an efficient barrier that serves to restrict the loss of the exciton, therefore, after arriving at the RC-LH1 complex the exciton is unable to escape and is retained in close proximity to the RC and so favors trapping. Similar reasoning is thought to apply to the LL complexes present in Rps. palustris, Chr. vinosum, etc. Interestingly, the carotenoid rhodopinal glucoside, which is present in the B800-820 complex at LL from Rps. acidophila 7050 also has an increased light harvesting capability compared with rhodopin glucoside in HL B800-850 (Angerhofer et al., 1986). When studying the types of LH2 complexes that any given species of purple bacteria are able to synthesize, it is important to realize the differences between growing cells in the laboratory under well controlled, standard growth conditions and those which cells will experience in the wild. Natural growth conditions are much more variable and usually less optimal than those we all use in the laboratory. It is probable that we will only understand the reasons for the wide range of types of LH2 when we are able to explore the effects of natural growth conditions on their competitiveness. D. Have We Only Scratched the Surface of Natural Variation? Purple photosynthetic bacteria are recognized as being among the most metabolically versatile of all organisms. Purple non-sulfur bacteria are capable of growth by aerobic and anaerobic respiration, fermentation and anoxygenic photosynthesis. Purple sulfur bacteria have less metabolic versatility but are able to utilize a bewildering number of sulfur compounds as electron donors (Chapter 30, Sander and Dahl). This versatility has ensured that a wide range of ecologically distinct habitats have been colonized by purple bacteria. It follows, therefore, that the total number of species under cultivation in the laboratory are prob-
149
ably only a very small sample of the many species that exist in the wild. Current strategies for isolating purple bacteria from the environment involve ‘enrichment culture.’ This has been a very successful strategy but has the disadvantage that the choice of the enrichment media and growth conditions biases which species grow. Very often when the enrichment conditions are varied new species are found. This is how Howard Gest discovered Heliobacterium chlorum (H. Gest, personal communication) when his technician made a mistake with the standard enrichment media. Some purple bacterial species have already been characterized that require ‘non-standard’ growth conditions such as the thermotolerant Rps. cryptolactis (Stadtwald-Demchick et al., 1990), the halophilic Rhodobium marinum (Hiraishi et al., 1995) and the acidophilic Rps. acidophila (Pfennig, 1969). It is worth bearing in mind, therefore, that our current knowledge about purple bacteria in general and their photosynthetic complexes in particular, is founded on a very limited number of species that, in the main, are those that are easily cultured. We expect that many interesting variations on the ‘standard’ B800-850 LH2 complex still remain to be discovered! 1. Genes of Purple Peripheral Antenna Complexes One would expect intuitively that a LH2 complex containing one α- and one β- polypeptide would be expressed by a single pucBA gene pair. This appears to be the case in Rvi. gelatinosus where one, and so far only one, pucBA gene locus has been identified (Simmons et al., 1999; Steunou et al., 2004). Also where multiple types of LH2 complexes exist multiple puc operons might be expected. It has been known for many years that the genomes of Rps. palustris and Rps. acidophila do indeed contain multiple, additional, copies of the puc operon genes (Fig. 6) (Tadros and Waterkamp, 1989; Tadros et al., 1993; Gardiner et al., 1996). However, there are more gene copies than might be predicted based on the number of types of LH2 complex. There are five different, but highly homologous pucBA gene copies present in the Rps. palustris genome (pucAB-a through to pucAB-e), with their expression differentially regulated by light-intensity. Similarly, Rps. acidophila strains 10050 (Simmons et al., 2000) and 7050 (Gardiner et al., 1996) each have at least four highly homologous pucBA gene copies (not eight in strain 10050 as erroneously reported in Simmons et al.,
150
Mads Gabrielsen, Alastair T. Gardiner and Richard J. Cogdell
2000). The obvious question to ask, therefore, is if a single β/α pair is all that is required to form an LH2 complex then why is it necessary to have so many pucBA genes? Furthermore, if only some but not all of these genes supply polypeptides for LH2 complexes, then why in a fluid, continually recombining bacterial genome under constant selective pressure are so many, apparently superfluous gene pairs, retained by the cell? The complete genome of Rps. palustris has been sequenced (Larimer et al., 2004) so the total number of pucBA gene pairs from this organism will remain at five. However, the estimate of four pucBA gene pairs from Rps. acidophila is almost certainly a minimum number, (as none of the four predicted Rps. acidophila 7050 pucBA-encoded polypeptides has 100% sequence identity with the polypeptides isolated from either the B800-850 or B800-820 complexes) and probably the true number is much higher. Even the superficially straightforward example of pucBA in Rba. sphaeroides turns out not to be so straightforward after all. For many years it was presumed that the pucBAC operon was present as a single copy in the genome. Recently a second pair of pucBA genes has been discovered, designated puc2BA, located on Chromosome 1 at approximately 1.36Mbp clockwise from the original pucBAC operon (Zeng et al., 2003). The puc2B-encoded polypeptide is predicted to have 94% identity with the original β-polypeptide, whereas the puc2A is predicted to encode 263aa compared with the 54aa puc1A-encoded polypeptide (although the first 48aa of the putative Puc2A have 58% aa identity to the original puc1A-encoded polypeptide). The puc2BA genes are expressed and approximately 30% of the wild-type LH2 complexes contain the Puc2B polypeptide. In contrast, puc1A is the sole source of α-polypeptides for the LH2 complex in this species, as neither Puc2A nor its truncated 48aa N-terminal derivative are able to enter into LH2 formation. Furthermore, if puc1B is deleted in-frame, then the ability to form LH2 is strongly reduced. This is a surprising result bearing in mind that Puc2B can be incorporated into the native complex. Interestingly, when the gene sequences for pucA are compared with those of PucA proteins, there is usually significant C-terminal processing (Fig. 6). This means that the mature proteins are always shorter than might be expected from the gene sequences. There is no information as to the mechanism of this process. Why do some species exist successfully with a single pucBA gene pair while others require
multiple copies? At present we only have the question not the answer. 2. Too Many Peripheral Antenna Polypeptides Make Life Difficult There is also a low-resolution (7.5 Å) projection map available for a peripheral antenna complex from Rps. palustris strain 2.1.6 that is synthesized during growth at low-light intensities (Hartigan et al., 2002). This complex has a single BChl absorption at 800nm, hence the B800 (and also LH4) designation. The structure reveals an octameric ring with very similar dimensions to the high-resolution, octameric B800850 structure from Phs. molischianum (Koepke et al., 1996). AFM pictures from LL-adapted Rps. palustris show that the great majority of complexes present in these membranes are octamers (Scheuring et al., 2006; Chapter 47, Scheuring). However, the spectrum presented in Fig. 7c also originates from a complex purified from LL-adapted Rps. palustris membranes yet this spectrum has a pronounced peak at 850nm. Does, therefore, the spectrum in Fig. 7c arise from a heterogeneous population of B800 and residual B800-850 molecules? Or is the Fig. 7c spectrum a homogeneous, true B800-850 preparation, which if over-purified would produce a modified B800 only absorbing form? Perhaps this complex is in fact not a single complex, rather a product of multiple pucBA genes. 3. Nomenclature At the time of writing, there are seventeen species stored by the German Collection of Microorganisms and Cell Cultures (Deutsche Sammlung von Mikroorgnismen und Zellkulturen GmbH, Braunschweig, Germany; DSMZ) for the genus Rhodopseudomonas alone, a number that will certainly grow. There are about ten different genera of purple bacteria in common laboratory ‘use.’As more purple bacterial species are characterized, (hopefully) many more interesting and novel antenna complexes will be discovered. We would like to remind all researchers in the field that just as B875, B890 and B1020 spectral forms are all LH1 complexes that it would be best to consider B800-850, B800-820, B800 and B800-830 all to be LH2s. To us it seems sensible to avoid the confusion involved with adopting the naming sequence LH3, LH4 and so on and to persist with LH1 and LH2.
Chapter 8
Peripheral Complexes of Purple Bacteria
151
Fig. 7. Examples of variable NIR absorbance spectra found in different species of purple photosynthetic bacteria. a) the B800-850 complex from Rps. acidophila strain 10050, b) the low light B800-820 complex from Rps acidophila strain 7050, c) the low light B800-850 complex from Rps. palustris and d) the low light B800 complex from Chr. vinosum.
The spectral form can then be designated separately and everybody will then understand what type of complex is being discussed. IV. Final Remarks A lot is known about the structure of the peripheral light-harvesting complexes and the influence of the structure upon their spectroscopic properties. The protomer of each complex is very similar, regardless of multimer state (Fig. 2d-f). However, much is still unclear, e.g., what decides the quaternary structure, or ring size, of the complex? Is this an inherent property in the α/β dimer, or is it controlled during the assembly process? Can a single complex be formed from a heterologous mixture of antenna apoproteins? Are there LH2 complexes with ring sizes different from the currently available 8- and 9-mers? Future research will be needed to answer these questions. Acknowledgments The authors would like to thank the Biotechnology and Biological Sciences Research Council for financial support.
References Angerhofer A, Cogdell RJ and Hipkins MF (1986) A spectral characterization of the light-harvesting pigment-protein complexes from Rhodopseudomonas acidophila. Biochim Biophys Acta 848: 333–341 Bahatyrova S, Frese RN, Siebert CA, Olsen JD, van der Werf KO, van Grondelle R, Niederman RA, Bullough PA, Otto C and Hunter CN (2004) The native architecture of a photosynthetic membrane. Nature 430: 1058–1062 Bauer CE, Buggy JJ, Yang ZM and Marrs BL (1991) The superoperonal organization of genes for pigment biosynthesis and reaction center proteins is a conserved feature in Rhodobacter capsulatus — analysis of overlapping bchB and puhA transcripts. Mol Gen Gen 228: 433–444 Bérard J, Bélanger G and Gingras G (1989) Mapping of the puh messenger-RNAs from Rhodospirillum rubrum — evidence for tandem promoters. J Biol Chem 264: 10897–10903 Cherezov V, Clogston J, Papiz MZ and Caffrey M (2006) Room to move: Crystallizing membrane proteins in swollen lipidic mesophases. J Mol Biol 357: 1605–18 Cogdell RJ, Hawthornthwaite AM, Evans MB, Ferguson LA, Kerfeld C, Thornber JP, van Mourik F and van Grondelle R (1990) Isolation and characterization of an unusual antenna complex from the marine purple sulfur photosynthetic bacterium Chromatium purpuratum Bn5500. Biochim Biophys Acta 1019: 239–244 Deinum G, Otte SCM, Gardiner AT, Aartsma TJ, Cogdell RJ and Amesz J (1991) Antenna organization of Rhodopseudomonas acidophila — a study of the excitation migration. Biochim Biophys Acta 1060: 125–131 Evans K, Fordham-Skelton AP, Mistry H, Reynolds CD, Lawless AM and Papiz MZ (2005) A bacteriophytochrome regulates the synthesis of LH4 complexes in Rhodopseudomonas palustris. Photosynth Res 85: 169–180
152
Mads Gabrielsen, Alastair T. Gardiner and Richard J. Cogdell
Fleming GR and van Grondelle R (1997) Femtosecond spectroscopy of photosynthetic light-harvesting systems. Cur Op Struc Biol 7: 738–748 Fowler GJ, Visschers RW, Grief GG, van Grondelle R and Hunter CN (1992) Genetically modified photosynthetic antenna complexes with blueshifted absorbance bands. Nature 355: 848–850 Freer A, Prince S, Sauer K, Papiz M, Hawthornthwaite-Lawless A, McDermott G, Cogdell R and Isaacs NW (1996) Pigment-pigment interactions and energy transfer in the antenna complex of the photosynthetic bacterium Rhodopseudomonas acidophila. Structure 4: 449–462 Gall A and Robert B (1999) Characterization of the different peripheral light-harvesting complexes from high- and low-light grown cells from Rhodopseudomonas palustris. Biochemistry 38: 5185–5190 Gall A, Gardiner AT, Cogdell RJ and Robert B (2006) Carotenoid stoichiometry in the LH2 crystal: No spectral evidence for the presence of the second molecule in the α/β-apoprotein dimer. FEBS Lett 580: 3841–3844 Gardiner AT, Cogdell RJ and Takaichi S (1993) The effect of growth-conditions on the light-harvesting apparatus in Rhodopseudomonas acidophila. Photosynth Res 38: 159–167 Gardiner AT, MacKenzie RC, Barrett SJ, Kaiser K and Cogdell RJ (1996) The purple photosynthetic bacterium Rhodopseudomonas acidophila contains multiple puc peripheral antenna complex (LH2) genes: Cloning and initial characterisation of four beta/alpha pairs. Photosynth Res 49: 223–235 Germeroth L, Lottspeich F, Robert B and Michel H (1993) Unexpected similarities of the B800-850 light-harvesting complex from Rhodospirillum molischianum to the B870 light-harvesting complexes from other purple photosynthetic bacteria. Biochemistry 32: 5615–5621 Gibson LCD, McGlynn P, Chaudhri M and Hunter CN (1992) A putative anaerobic coproporphyrinogen III oxidase in Rhodobacter sphaeroides II Genetic analysis of a region of the genome encoding hemF and the puc operon. Mol Micro 6: 3171–3186 Gonçalves RP, Bernadac A, Sturgis JN and Scheuring S (2005a) Architecture of the native photosynthetic apparatus of Phaeospirillum molischianum. J Struc Biol 152: 221–228 Gonçalves RP, Busselez J, Lévy D, Seguin J and Scheuring S (2005b) Membrane insertion of Rhodopseudomonas acidophila light harvesting complex 2 investigated by high resolution AFM. J Struc Biol 149: 79–-86 Halloren E, McDermott G, Lindsay JG, Miller C, Freer AA, Isaacs NW and Cogdell RJ (1995) Studies on the light-harvesting complexes from the thermotolerant purple bacterium Rhodopseudomonas cryptolactis. Photosynthesis Research 44: 149–155 Hartigan N, Tharia HA, Sweeney F, Lawless AM and Papiz MZ (2002) The 7.5-Å electron density and spectroscopic properties of a novel low-light B800 LH2 from Rhodopseudomonas palustris. Biophys J 82: 963–977 Hess S, Chachisvilis M, Timpmann K, Jones MR, Fowler GJS, Hunter CN and Sundström V (1995) Temporally and spectrally resolved subpicosecond energy transfer within the peripheral antenna complex (LH2) and from LH2 to the core antenna complex in photosynthetic purple bacteria. PNAS 92: 12333–12337 Hiraishi A, Urata K and Satoh T (1995) A new genus of marine
budding phototrophic bacteria, Rhodobium gen-nov, which includes Rhodobium orientis Sp-Nov and Rhodobium marinum Comb-Nov. Int J Syst Bacteriol 45: 226–234 Hunter CN, Tucker JD and Niederman RA (2005) The assembly and organization of photosynthetic membranes in Rhodobacter sphaeroides. Photochem Photobiol Sci 4: 1023–1027 Janosi L, Keer H, Kosztin I and Ritz T (2006) Influence of subunit structure on the oligomerization state of light-harvesting complexes: A free energy calculation study. Chem Phys 323: 117–128 Koepke J, Hu X, Muenke C, Schulten K and Michel H (1996) The crystal structure of the light-harvesting complex II (B800-850) from Rhodospirillum molischianum. Structure 4: 581–97 Lang HP and Hunter CN (1994) The relationship between carotenoid biosynthesis and the assembly of the light-harvesting LH2 complex in Rhodobacter sphaeroides. Biochem J 298: 197–205 Larimer FW, Chain P, Hauser L, Lamerdin J, Malfatti S, Do L, Land ML, Pelletier DA, Beatty JT, Lang AS, Tabita FR, Gibson JL, Hanson TE, Bobst C, Torres J, Peres C, Harrison FH, Gibson J and Harwood CS (2004) Complete genome sequence of the metabolically versatile photosynthetic bacterium Rhodopseudomonas palustris. Nature Biotechnology 22: 55–61 LeBlanc HN and Beatty JT (1996) Topological analysis of the Rhodobacter capsulatus PucC protein and effects of C-terminal deletions on light-harvesting complex II. J Bact 178: 4801–4806 McDermott G, Prince SM, Freer AA, Hawthornthwaite-Lawless AM, Papiz MZ, Cogdell RJ and Isaacs NW (1995) Crystal structure of an integral membrane light-harvesting complex from photosynthetic bacteria. Nature 374: 517–521 McLuskey K, Prince SM, Cogdell RJ and Isaacs NW (2001) The crystallographic structure of the B800-820 LH3 light-harvesting complex from the purple bacteria Rhodopseudomonas acidophila strain 7050. Biochemistry 40: 8783–8789 Monger TG and Parson WW (1977) Singlet-triplet fusion in Rhodopseudomonas sphaeroides chromatophores — probe of organization of photosynthetic apparatus. Biochim Biophys Acta 460: 393–407 Montoya G, Cyrklaff M and Sinning I (1995) Two-dimensional crystallization and preliminary structure analysis of light harvesting II (B800-850) complex from the purple bacterium Rhodovulum sulfidophilum. J Mol Biol 250: 1–10 Nagae H, Kakitani T, Katoh T and Mimuro M (1993) Calculation of the excitation transfer-matrix elements between the S2 or S1 state of carotenoid and the S(2) or S(1) state of bacteriochlorophyll. J Chem Phys 98: 8012–8023 Niederman RA, Mallon DE and Parks LC (1979) Membranes of Rhodopseudomonas sphaeroides: Isolation of a fraction enriched in newly synthesized bacteriochlorophyll alpha-protein complexes. Biochim Biophys Acta 555: 210–220 Papiz MZ, Prince SM, Howard T, Cogdell RJ and Isaacs NW (2003) The structure and thermal motion of the B800-850 LH2 complex from Rps. acidophila at 2.0 Å resolution and 100K: New structural features and functionally relevant motions. J Mol Biol 326: 1523–38 Pfennig N (1969) Rhodopseudomonas acidophila, Sp N, a new species of budding purple nonsulfur bacteria. J Bact 99: 597–602 Prince SM, Papiz MZ, Freer AA, McDermott G, HawthornthwaiteLawless AM, Cogdell RJ and Isaacs NW (1997) Apoprotein
Chapter 8
Peripheral Complexes of Purple Bacteria
structure in the LH2 complex from Rhodopseudomonas acidophila strain 10050: modular assembly and protein pigment interactions. J Mol Biol 268: 412–423 Prince SM, Howard TD, Myles DA, Wilkinson C, Papiz MZ, Freer AA, Cogdell RJ and Isaacs NW (2003) Detergent structure in crystals of the integral membrane light-harvesting complex LH2 from Rhodopseudomonas acidophila strain 10050. J Mol Biol 326: 307–315 Reilly PA and Niederman RA (1986) Role of apparent membrane growth initiation sites during photosynthetic membrane-development in synchronously dividing Rhodopseudomonas sphaeroides. J Bact 167: 153–159 Russell NJ, Coleman JK, Howard TD, Johnston E and Cogdell RJ (2002) Rhodopseudomonas acidophila strain 10050 contains photosynthetic LH2 antenna complexes that are not enriched with phosphatidylglycerol, and the phospholipids have a fatty acyl composition that is unusual for purple non-sulfur bacteria. Biochim Biophys Acta 1556: 247–253 Scheuring S, Reiss-Husson F, Engel A, Rigaud JL and Ranck JL (2001) High-resolution AFM topographs of Rubrivivax gelatinosus light-harvesting complex LH2. EMBO J 20: 3029–3035 Scheuring S, Seguin J, Marco S, Lévy D, Breyton C, Robert B and Rigaud JL (2003) AFM characterization of tilt and intrinsic flexibility of Rhodobacter sphaeroides light harvesting complex 2 (LH2). J Mol Biol 325: 569–580 Scheuring S, Rigaud JL and Sturgis JN (2004) Variable LH2 stoichiometry and core clustering in native membranes of Rhodospirillum photometricum. EMBO J 23: 4127–4133 Scheuring S, Gonçalves RP, Prima V and Sturgis JN (2006) The photosynthetic apparatus of Rhodopseudomonas palustris: Structures and organization. J Mol Biol 358: 83–96 Simmons AE, Mackenzie RC and Cogdell RJ (1999) Cloning and sequencing of the pucBA genes from two strains of Rubrivivax gelatinosus. Photosynth Res 62: 99–106 Simmons AE, Barrett SJ, Hunter CN and Cogdell RJ (2000) Cloning, sequencing and analysis of the pucC genes from Rubrivivax gelatinosus strain 151 and Rhodopseudomonas acidophila strain 10050. Photosynth Res 65: 69–82 Stadtwald-Demchick R, Turner FR and Gest H (1990) Rhodopseudomonas cryptolactis, Sp-Nov, a new thermotolerant species of budding phototrophic purple bacteria. FEMS Microbiol Lett 71: 117–121 Stamouli A, Kafi S, Klein DC, Oosterkamp TH, Frenken JW, Cogdell RJ and Aartsma TJ (2003) The ring structure and organization of light harvesting 2 complexes in a reconstituted lipid bilayer, resolved by atomic force microscopy. Biophys J 84: 2483–2491 Steunou AS, Ouchane S, Reiss-Husson F and Astier C (2004) Involvement of the C-terminal extension of the α polypeptide and of the PucC protein in LH2 complex biosynthesis in Rubrivivax gelatinosus. J Bact 186: 3143–3152 Sundström V, Pullerits T and van Grondelle R (1999) Photosyn-
153
thetic light-harvesting: Reconciling dynamics and structure of purple bacterial LH2 reveals function of photosynthetic unit. J Phys Chem B 103: 2327–2346 Tadros MH and Waterkamp K (1989) Multiple copies of the coding regions for the light-harvesting B800-850 alpha-polypeptide and beta-polypeptide are present in the Rhodopseudomonas palustris genome. EMBO J 8: 1303–1308 Tadros MH, Katsiou E, Hoon MA, Yurkova N and Ramji DP (1993) Cloning of a new antenna gene-cluster and expression analysis of the antenna gene family of Rhodopseudomonas palustris. Eur J Biochem 217: 867–875 Tharia HA, Nightingale TD, Papiz MZ and Lawless AM (1999) Characterisation of hydrophobic peptides by RP-HPLC from different spectral forms of LH2 isolated from Rps. palustris. Photosynth Res 61: 157–167 Thornber JP and Sokoloff MK (1970) Photochemical reactions of purple bacteria as revealed by studies of 3 spectrally different carotenobacteriochlorophyll-protein complexes isolated from Chromatium, Strain-D. Biochemistry 9: 2688–2698 Tichy HV, Albien KU, Gadon N and Drews G (1991) Analysis of the Rhodobacter capsulatus puc operon — the pucC gene plays a central role in the regulation of LH2 (B800-850 complex) expression. EMBO J 10: 2949–2955 van Grondelle R and Novoderezhkin V (2001) Dynamics of excitation energy transfer in the LH1 and LH2 light-harvesting complexes of photosynthetic bacteria. Biochemistry 40: 15057–15068 van Mourik F, Hawthornthwaite AM, Vonk C, Evans MB, Cogdell RJ, Sundström V and van Grondelle R (1992) Spectroscopic characterization of the low-light B800-850 light-harvesting complex of Rhodopseudomonas palustris, Strain 216. Biochim Biophys Acta 1140: 85–93 Vredenberg WJ and Duysens LNM (1963) Transfer of energy from bacteriochlorophyll to a reaction centre during bacterial photosynthesis. Nature 197: 355–357 Walz T, Jamieson SJ, Bowers CM, Bullough PA and Hunter CN (1998) Projection structures of three photosynthetic complexes from Rhodobacter sphaeroides: LH2 at 6 Å, LH1 and RC-LH1 at 25 Å. J Mol Biol 282: 833–45 Young CS, Reyes RC and Beatty JT (1998) Genetic complementation and kinetic analyses of Rhodobacter capsulatus ORF1696 mutants indicate that the ORF1696 protein enhances assembly of the light-harvesting I complex. J Bact 180: 1759–1765 Young CY and Beatty JT (1998) Structural and functional analysis of the orf1696/PucC family of light-harvesting complex assembly proteins. In: Peschek GA, Loeffelhardt W and Schmetterer G (eds) The Phototrophic Prokaryotes. 113–126 Kluwer Academic Publishers/Plenum Press. New York Zeng XH, Choudhary M and Kaplan S (2003) A second and unusual pucBA operon of Rhodobacter sphaeroides 2.4.1: genetics and function of the encoded polypeptides. J Bact 185: 6171–6184
Chapter 9 Reaction Center-Light-Harvesting Core Complexes of Purple Bacteria Per A. Bullough, Pu Qian and C. Neil Hunter* Krebs Institute for Biomolecular Research, Department of Molecular Biology and Biotechnology, University of Sheffield, Firth Court, Western Bank, Sheffield S10 2TN, U.K.
Summary ............................................................................................................................................................... 155 I. Introduction..................................................................................................................................................... 156 II. The Building Blocks: The α and β Polypeptides of Light-Harvesting Complex 1, and PufX........................... 156 A. Nuclear Magnetic Resonance Studies of Light-Harvesting Complex 1 Polypeptides ...................... 156 B. PufX ................................................................................................................................................. 157 C. The Solution Structure of PufX......................................................................................................... 159 III. Circles, Arcs and Ellipses — The Light-Harvesting 1 Complex ...................................................................... 160 IV. Monomeric Reaction Center-Light-Harvesting 1 Complexes ......................................................................... 162 A. The Reaction Center-Light-Harvesting 1 Complex of Blastochloris viridis ..................................... 162 B. The Reaction Center-Light-Harvesting 1 Complex of Phaeospirillum molischianum ...................... 163 C. The Reaction Center-Light-Harvesting 1 Complex of Rhodospirillum rubrum ................................ 163 D. Mutant Reaction Center-Light-Harvesting 1 Complexes of Rhodobacter sphaeroides .................. 164 E. The Reaction Center-Light-Harvesting 1 Complex of Rhodospirillum photometricum ................... 165 F. Models of Monomeric Core Complexes .......................................................................................... 165 V. Monomeric Reaction Center-Light-Harvesting 1-PufX Complexes ................................................................ 166 A. The Rhodopseudomonas palustris Reaction Center-Light-Harvesting 1-W Complex..................... 166 A. The Rhodobacter veldkampii Reaction Center-Light-Harvesting 1-PufX Complex ......................... 166 VI. Dimeric Reaction Center-Light-Harvesting 1-PufX Complexes ...................................................................... 168 A. The Rhodobacter sphaeroides and Rhodobacter blasticus Core Dimers ...................................... 168 B. Structural and Functional Roles for PufX ......................................................................................... 170 1. PufX: Core Dimerization and Membrane Curvature................................................................ 170 2. PufX: Membrane Domains, Excitation Sharing and Quinone Sequestration .......................... 172 VII. The Biogenesis of Core Complexes ............................................................................................................... 174 Acknowledgments ................................................................................................................................................. 175 References ............................................................................................................................................................ 175
Summary Reaction center-light-harvesting 1 (RC-LH1) complexes, the fundamental photosynthetic units, are fascinatingly diverse in their aggregation state, with both monomers and dimers found in membranes. Some complexes also contain the PufX polypeptide in addition to the RC and LH1 complexes. Many structural techniques have been applied to RC-LH1 complexes, in order to analyze their shape, variation in conformation, the individual polypeptide structures, and the overall organization in the membrane. Thus, the unique capabilities of cryo- and negative stain electron microscopy (EM), X-ray crystallography, nuclear magnetic resonance (NMR) spectros*Author for correspondence, email: c.n.hunter@sheffield.ac.uk
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 155–179. © 2009 Springer Science + Business Media B.V.
156
Per A. Bullough , Pu Qian and C. Neil Hunter
copy and atomic force microscopy (AFM) have all complemented one another in the analysis of membranes and purified protein complexes, from both wild type and mutated strains of photosynthetic bacteria. However, many unsolved problems remain; for example, the ways in which quinones and quinols exchange at the RC QB site despite the apparent barrier arising from the encircling LH1 complex are still not resolved, and this hints at still unexplored flexibility and dynamics associated with the LH1 complex and with the photochemistry of the RC. Other controversies currently involve the structure, location and function of the PufX polypeptide, and most basic of all, there is still a need for high-resolution structures of core complexes. Finally, the assembly pathways of core complexes are not known. This chapter attempts to convey the diversity of experimental approaches used to investigate core complexes and it summarizes the current state of progress in this evolving field. I. Introduction RC-LH1 complexes are often termed ‘core’ complexes, since, as the name implies, they are central to bacterial photosynthesis. Indeed many bacteria, such as Rhodospirillum (Rsp.) rubrum, employ core complexes exclusively for both light harvesting and photochemical functions, and they dispense with peripheral complexes such as LH2, at the expense of a lowered capacity to absorb light. The complex which is solely responsible for light harvesting in the core, LH1, surrounds and interconnects one or two RCs. The RC, the site of the photochemical reactions which generate quinol for consumption by the cytochrome bc1 complex (see Chapters 22, Berry et al. and Chapter 23, Kramer et al. for reviews of the cytochrome bc1 complex), is not considered further in this chapter as the structural and functional properties of this complex merit an entire section of this book (Chapters 16 through 20). The main purpose of this chapter is to summarize the current state of structural information available for core complexes, starting with the structures of the small repeating units which comprise the LH1 antenna, the LH1 α and β polypeptides, and building up towards the whole core complex. Chapter 10 by Loach and Parkes-Loach summarizes the in vitro reconstitution of LH complexes from LH1 polypeptides. Several approaches for the analysis of core complexes (directed mutagenesis, NMR, EM, X-ray crystallography, AFM) have revealed either monomeric or dimeric structures, depending on the bacterium used, as well as the presence in some cases Abbreviations: 2-D – two-dimensional; 3-D – three-dimensional; AFM – atomic force microscopy; BChl – bacteriochlorophyll; Blc. – Blastochloris; Cryo-EM – cryo-electron microscopy; LD – linear dichroism; LH – light-harvesting; NMR – nuclear magnetic resonance; Phs. – Phaeospirillum; Rba. – Rhodobacter; RC – reaction center; Rps. – Rhodopseudomonas; Rsp. – Rhodospirillum
of additional components such as PufX. The structures represented in this chapter reflect this diversity. The consequences of building a photosynthetic unit from monomeric or dimeric cores are felt at the level of membrane organization; for a discussion of these issues see Chapter 14, Sturgis and Niederman and Chapter 47, Scheuring. However, many uncertainties remain, partly through a lack of atomic level structural data. This is reflected in a number of alternative structural models in the current literature. II. The Building Blocks: The ␣ and  Polypeptides of Light-Harvesting Complex 1, and PufX A. Nuclear Magnetic Resonance Studies of Light-Harvesting Complex 1 Polypeptides Bacterial light-harvesting complexes are composed of α and β polypeptides. The pioneering studies of Zuber and coworkers, who determined the primary structures of antenna polypeptides from a diverse selection of photosynthetic bacteria, revealed the similarities and differences between the α and β components. In general they are relatively short (between 40 and 70 amino acid residues), they possess a ‘three-domain’ structure in which polar N- and C-terminal domains lie either side of a hydrophobic transmembrane span, and they possess a conserved histidine residue which was proposed as the ligand to bacteriochlorophyll (BChl) (Zuber, 1985; Brunisholz and Zuber, 1992; Zuber and Cogdell, 1995). This residue, termed His0, was used as the reference point for numbering all LH polypeptides, so that a residue N-terminal to His0 was designated Trp–10, for example. This numbering system will be used in this chapter. Further analysis by Zuber revealed correlations between antenna type, namely LH2 and LH1 complexes, as well as suggestions for residues that
Chapter 9
Core Complexes of Purple Bacteria
determine wavelength shifts in absorption maxima. This primary sequence information was then used to produce models of the basic αβ unit (Olsen and Hunter, 1994; Zuber and Cogdell, 1995), accurately predicting turns and hydrogen bonding residues. The way in which αβ units formed beautiful circular structures was finally revealed for the LH1 and LH2 complexes by EM and X-ray crystallographic studies (Karrasch et al., 1995; McDermott et al., 1995; see below and also Chapter 8, Gabrielsen et al.). At the time of writing, the LH1 complex has resisted attempts to provide a level of structural detail comparable to that available for LH2, although much progress has been made, and this is reviewed elsewhere in this chapter. However, the small size of LH polypeptides does provide an alternative method for structural analysis, namely NMR. Since LH1 complexes are assembled from multiple copies of small repeating units, the LH1α and β polypeptides, the solution structures of the individual components provide important clues to the structure of the whole complex. LH1β was chosen for the first NMR study (Conroy et al., 2000), because the reconstitution studies of Loach and co-workers (see Chapter 10, Loach and Parkes-Loach, and also Chapter 46, Braun and Fiedor) had previously shown that although a B820 subunit complex (a building block of the whole LH1 complex) could be formed from LH1α and β polypeptides, the β polypeptide alone could also form a B820-type subunit complex (putatively ββ BChl2; Parkes-Loach et al., 1988). The solution structure of the LH1β polypeptide from Rba. sphaeroides was determined in organic solvent (Conroy et al., 2000), the rationale being that structural integrity is generally maintained in organic solvents in spite of removal of the protein from its native environment (Pervushin et al., 1994). The LH1β structure comprises two alpha-helical regions; the N-terminal helix (residues –34 to –15) is suggested to lie in the lipid head groups or in the cytoplasm, and the C-terminal helix (–11 to +6) forms the membrane-spanning region (Fig. 1c). These two helices are connected by a flexible linker consisting of four residues (MSGL) in the –14 to –10 region. A subsequent NMR study of the LH1β polypeptide from Rba. sphaeroides determined a similar structure, but in a detergent micelle, which more closely mimics a membrane environment (Sorgen et al., 2002). BChl binding to the LH1β polypeptides was possible, although the micelle environment did not permit the hydrogen bond known to exist in intact, membrane-
157 bound LH1 complexes between the β Trp+9 and the C2 acetyl carbonyl of one of the LH1 BChls (Sturgis et al., 1997). The micellar environment (Sorgen et al., 2002) accentuated the bend in LH1β, with the N-terminal helix lying along the micelle surface, in contrast to the less ordered ensemble of structures in organic solvent (Conroy et al., 2000). However, it is likely that whatever its disposition in either organic solvent or micelles this N-terminal helix has to accommodate the enclosed RC complex in vivo. This would be the case if the model proposed by Conroy et al. (2000) and refined subsequently (Fotiadis et al., 2004) is accurate (see also section IV.D). One interesting aspect of the helix-hinge-helix solution structure is a possible role for LH1β in allowing rapid quinone exchange. Indeed the projection structure of the RCLH1-PufX dimer complex of Rba. sphaeroides shows a relatively diffuse density for LH1β in the vicinity of the RC QB site (see Section VI), which has been suggested to arise from the flexible hinge region, and which could promote rapid quinol/quinone exchange (Qian et al., 2005). In light of the two NMR studies on the LH1β polypeptide from Rba. sphaeroides (Conroy et al., 2000; Sorgen et al., 2002), the solution structures of both the LH1α and β polypeptides from Rsp. rubrum were especially interesting, given the monomeric nature of the RC-LH1 core complex in this bacterium, and the lack of a PufX component of the LH1 ring (Wang et al., 2005). LH1α was shown to consist of a short Nterminal helix of four residues, then a three residue unstructured loop followed by a long, slightly curved 32-residue helix (Fig. 1a). In the middle of this helix there is the BChl ligand, His0 (corresponding to His29 in the conventional numbering system). In marked contrast to the helix-hinge-helix arrangement found in the LH1β polypeptide from Rba. sphaeroides, the LH1β from Rsp. rubrum is composed of a single membrane-spanning helix of 32 amino acid residues with the pigment-coordinating residue His0 (His38) at a position close to the C-terminal end of the helix (Fig. 1b). The availability of a structure for both LH1 polypeptides of Rsp. rubrum allowed Wang et al. (2005) to propose a model for the αβBChl2 (B820) subunit, the building block of the LH1 complex identified and characterized in reconstitution experiments (see Chapter 10, Loach and Parkes-Loach). B. PufX PufX was shown to be important for photosynthetic
158
Per A. Bullough , Pu Qian and C. Neil Hunter
Fig. 1. Ribbon representation of solution structures of (a) the LH1α polypeptide from Rsp. rubrum (PDB code 1XRD) (b) LH1β from Rsp. rubrum (1WRG) (c) the LH1β polypeptide from Rba. sphaeroides (1DX7) (d) the PufX polypeptide from Rba. sphaeroides (2NRG). The ribbons denote α-helices. Positive and negative symbols denote lysine/arginine and glutamic/aspartic acids, respectively. Glycines are represented by dark shading, and H denotes histidine. Tryptophan sidechains are shown. A suggested position for the membrane is shown, with the darker gray regions showing the lipid head group region and the lighter gray regions representing the hydrophobic core of the membrane. Figure prepared by Richard Tunnicliffe.
growth in a series of deletion and complementation experiments conducted on Rba. capsulatus and Rba. sphaeroides (Klug and Cohen, 1988; Farchaus et al., 1990; Lilburn et al., 1992). However, no such requirement was seen in a mutant lacking LH1 (McGlynn et al., 1994), and a series of papers showed that intergenic suppressor mutants could at least partially compensate for loss of PufX, by altering or destabilizing the LH1 complex so it could not form a barrier to quinol release ( Lilburn and Beatty, 1992; Lilburn et al., 1992, 1995; Barz and Oesterhelt, 1994). A detailed study of the PufX/LH1 relationship revealed that the loss of PufX is only deleterious when the size of the LH1 ring exceeds 20 BChls per RC (McGlynn et al., 1996). All of these observations were consistent with the idea that PufX acted in some way to restrain LH1 from completely encircling the RC, and in doing so, this restraint permitted quinol/quinone exchange at the QB site of the RC (Barz et al., 1995a,b). The PufX polypeptide was suggested to lie adjacent to the RC QB site and to form part of the LH1 assembly (Cogdell et al., 1996). A kinetic study showed that removal of PufX and closure of the LH1 ring created a barrier adjacent to the RC QB site, so that the time taken for quinone turnover increased from ~1.6 to
~3.5 ms, and the time for reduction of the cytochrome bc1 complex by quinol increased from ~7 to ~14 ms (Comayras et al., 2005). However, these slowed reactions are not thought to explain the inability of the PufX-minus mutant to sustain photosynthetic growth under reducing conditions. Instead, these authors conclude that the lack of PufX modifies the RC QB pocket and stabilizes the RC-bound QB– state; this in turn results in an unexpected accumulation of closed RC complexes in the QA– state. In the last ten years in vivo and in vitro analyses have helped to establish the membrane topology and functional roles of the extrinsic and membrane spanning regions of PufX. The PufX polypeptide has a predicted molecular mass of approximately 9 kDa corresponding to 82 residues in Rba. sphaeroides, but there is relatively little homology (29%) between the PufX polypeptides of Rba. sphaeroides and Rba. capsulatus, despite homologies of between 70 and 80% for LH and RC subunits. Nevertheless, it was found that the Rba capsulatus PufX could substitute for the Rba. sphaeroides protein (Fulcher et al., 1998). The N-terminus of PufX is exposed on the cytoplasmic surface of the cell membrane (Pugh et al., 1998). Mild detergent fractionation revealed that LH1 is essential
Chapter 9
Core Complexes of Purple Bacteria
for the correct binding of PufX to the core complex; however, in the growing photosynthetic membrane the PufX polypeptide is incorporated into the assembling RC-LH1-PufX core complex before the encirclement of the RC by LH1 α and β polypeptides, presumably to avoid obstruction of the RC QB site (Pugh et al., 1998). It was proposed that a RC-PufX-(LH1αβBChl2) complex is formed initially at the earliest stages of the assembly of the core, and that subsequently LH1αβBChl2 units are progressively added to this complex (Pugh et al., 1998). Evidence for a tight association between LH1 α and PufX emerged during attempts to purify PufX from Rba. sphaeroides and Rba. capsulatus (Recchia et al., 1998). Analysis of purified PufX indicated that C-terminal processing takes place, with 12 and 9 residues missing from the proteins, respectively (Parkes-Loach et al., 2001). Experiments with isolated, chemically synthesized components showed that segments of PufX containing the putative transmembrane segment inhibited the oligomerization of B820 units to form the complete LH1 ring, and that neither the N- nor the C-terminal regions of PufX had this property. Interactions of the glycine-rich transmembrane region of PufX were proposed, firstly with a similarly glycine-rich region within the first membrane-spanning region of the RC M subunit, and secondly with LH1 α (Parkes-Loach et al., 2001). Finally, it has been shown that PufX can bind BChl in vitro, possibly at the C-terminal (residues 60–68) region (Law et al., 2003), or at an alternative binding site involving two glycines at either end of the transmembrane segment (Aklujkar and Beatty, 2006). These authors also propose that PufX modulates the binding of carotenoids to LH1 (Aklujkar and Beatty, 2005). C. The Solution Structure of PufX The small size of PufX makes it suitable for structure determination by NMR. The Rba. sphaeroides PufX was overproduced in Escherichia coli, labeled with 15 N and 13C, and purified. Determination of the solution structure in CD3OH/CDCl3 (1:1) revealed a long α-helical structure of 34 residues with a significant bend of approximately 120° approximately in the middle (Tunnicliffe et al., 2006). A model of PufX shows that the most N-terminal region of PufX is almost unstructured; it contains the site of proteinase K digestion determined in studies of PufX topology in the membrane bound RC-LH1-PufX complex (Pugh et al., 1998). This is followed by a structured helical
159 domain positioned so that Trp and basic residues are at the membrane interface (Fig. 1d), allowing the basic amino acids within the N-terminal part of the helix to lie near the membrane surface, and Trp 22 to play a membrane anchoring role. There is a bent, less rigid, region in the PufX polypeptide in the region of residues 30–35 (GAGWA), and the transmembrane region spans residues 30–52. This positioning of PufX requires that the C-terminal helix, residues 30–52, is almost perpendicular to the membrane plane. The exposure of the PufX C-terminus on the periplasmic side of the membrane is consistent with the observation of posttranslational cleavage, which removes its last 12 residues (Parkes-Loach et al., 2001). It has been shown that PufX can bind BChl in vitro, with a binding site proposed in this C-terminal (residues 60–68) region (Law et al., 2003), The NMR data do not rule out BChl binding in vivo, although there is a lack of obvious candidates for BChl ligands. Tunnicliffe et al. (2006) also titrated labeled PufX with unlabelled LH1 α and β polypeptides, and found that chemical shift changes with LH1 α were much larger, indicating specific binding, in this solvent environment at least. The PufX residues interacting with LH1 α were Val 51, Met 54, Arg 53 and Ile 57, close to the periplasmic end of the C-terminal helix. A subsequent NMR paper proposed a different solution structure for PufX, with no bend in the polypeptide (Wang et al., 2007). PufX is depicted as a long, almost straight helix running from residues 19 to 51 and flanked by disordered, but straight, Nand C-terminal domains, and therefore, somewhat surprisingly, projecting out from the membrane at both ends, with the hydophilic Lys29 within the transmembrane region. A projecting N-terminus has not been observed in AFM topographs of this side of the core complex in native membranes (Bahatyrova et al., 2004). The transmembrane part of this structure is proposed to span residues 24 to 44, which places the glycine-rich sequence GAGWA in the middle of the transmembrane domain, rather than forming a bend near the membrane interface as in Tunnicliffe et al. (2006). This glycine-rich domain is proposed to form a site for quinone exchange between the RC QB site and the external quinone pool. Other consequences of a transmembrane location for glycine-rich GAGWA sequence include a potential for interaction with GXXXG and other dimerization motifs, which is precluded by the bent structure in Tunnicliffe et al. (2006). In fact, PufX possesses a ‘double’ GXXXG mo-
160 tif, and such sequences are known to promote the dimerization of transmembrane helices (Russ and Engelman, 2000). This glycine-rich region is located towards the cytoplasmic surface of the membrane bilayer in Fig. 1d, but in the transmembrane region in the model of Wang et al. (2007). The most obvious role of these glycine residues, if the bent structure for PufX is correct, is to allow a structurally and functionally necessary bend to form between the Nterminal and transmembrane helices (Fig. 1d), but it has been proposed that homodimerization of PufX transmembrane regions drives dimerization of the core, on the basis of TOXCAT analysis (Aklujkar and Beatty, 2006; Russ and Engelman, 1999) and AFM and EM data (Scheuring et al., 2004, 2005; Busselez et al., 2007). At the time of writing, many aspects of the structure and function of PufX are uncertain, and even controversial, although undeniably interesting and important. A more detailed discussion of the structural role of PufX can be found in Sections V.B and VI.B of this chapter. III. Circles, Arcs and Ellipses — The LightHarvesting 1 Complex Moving on from the structures of the individual LH1 and PufX components of the core complex we now discuss the structures of higher order molecular assemblies involving these components, firstly assemblies of LH1 subunits and then complexes of LH1 and PufX with the RC. As noted earlier, Chapter 10, Loach and Parkes-Loach and Chapter 46, Braun and Fiedor cover the assembly of LH1 assemblies using in vitro reconstitution methods. The LH1 complex was first purified from Rba. sphaeroides and Rsp. rubrum (Broglie et al., 1980; Picorel et al., 1983). Subsequent work used lithium dodecyl sulfate-polyacrylamide gel electrophoresis to fractionate the Rba. sphaeroides LH1 into a series of LH1αβBChl2 oligomers (Westerhuis et al., 2002), estimated to range in size from (αβ)4 to (αβ)13 (Westerhuis et al., 1999). The first structural work on LH1 complexes was performed on the LH1 from Rsp. rubrum, which was found to form vesicular structures spontaneously in solution, which, when examined by freeze-fracture electron microscopy, contained pseudo-hexagonally close-packed 120 Å particles (Ghosh et al., 1990). Further studies on negatively stained 2-D crystals reconstituted from a carotenoidless LH1 from Rsp. rubrum revealed
Per A. Bullough , Pu Qian and C. Neil Hunter pseudo-hexagonally packed ring-like structures with an outer diameter of 120 Å and an inner diameter of 55 Å (Ghosh et al., 1993). dA similar crystal form was observed for LH1 purified from Rhodopseudomonas (Rps.) marina (Meckenstock et al., 1992). In the absence of higher resolution information at the time, it seemed natural to propose a hexagonal ring symmetry for these purified LH1 complexes. However Karrasch et al. (1995) ultimately achieved a resolution high enough to allow the first reliable symmetry determination. 2-D crystals of LH1 from a strain of Rsp. rubrum containing wild-type carotenoids were reconstituted from detergent-solubilized protein complexes with a synthetic phospholipid, dioleoyl-9, 10-phosphatidylcholine (DOPC). Frozen hydrated samples were analyzed by cryo-EM and a projection map was calculated to 8.5 Å. The map showed 16 subunits in a 116 Å diameter ring with a 68 Å hole in the center (Fig. 2a). These dimensions are sufficient to incorporate a single RC in vivo. Within each subunit, densities for the α- and the β-polypeptide chains were clearly resolved, and the density for the BChl could be assigned. The inner face of the ring shows 16 very large densities separated by ~15 Å. These peaks have the appearance of α-helices viewed end on, and indeed a significant proportion (40%) of the two polypeptide chains had been predicted to be in an α-helical conformation with the helices lying roughly perpendicular to the membrane plane. The outer face of the ring consists of 16 peaks of density separated from each other by ~20 Å and from the inner peaks by ~15 Å. These peaks represent the projected α-helical density from the second polypeptide chain. Between the inner and outer rings of the peaks lies a third ring of densities, the smallest density of the three sets and lying 11 Å away from both the inner and outer peaks. It is likely that some of this central density represents an edge-on view of a pair of BChl a molecules bound between the inner and outer helices (Fig. 2a). The apparently closed ring structure raised questions about the mechanism of quinone transfer between the RC and the spatially distant cytochrome bc1 complex during cyclic electron transport. It was suggested that breathing motions of the LH1 ring might allow transient opening for quinone passage (Walz and Ghosh, 1997), and indeed there was clear evidence of flexibility in LH1 with Karrasch et al. (1995) observing LH1 rings with varying degrees of ellipticity. Furthermore, subsequent AFM analysis of individual LH1 rings from Rba. sphaeroides was not
Chapter 9
Core Complexes of Purple Bacteria
161
Fig. 2. Projection maps of the potential distribution within 2-D crystals of photosynthetic complexes of Rsp. rubrum. Lighter regions of density represent the protein. The circle diameter is approximately 110 Å. (a) projection of LH1 showing inner and outer rings of α-helical densities sandwiching density arising partly from the BChl (Karrasch et al., 1995) (b) RC-LH1 complex in a tetragonal crystal form, showing a central complex of LH1 subunits surrounding a 16-fold averaged density for the RC. Segments of neighboring complexes within the crystal can also be seen (Jamieson et al., 2002) (c) RC-LH1 in an orthorhombic crystal form, again showing a central complex surrounded by segments from neighboring complexes. The small deviation of the LH1 subunits from a perfect circle can be seen. The RC density is only two-fold averaged in this crystal form (Jamieson et al., 2002).
dependent on averaging of multiple images, unlike cryoEM; this explicitly demonstrated for the first time an extraordinary variety of shapes and sizes adopted by LH1 (Bahatyrova et al., 2004b), only hinted at by Karrasch et al. (1995). Although this AFM work was performed on Rba. sphaeroides LH1 complexes (Fig. 3) the particular LH1-only strain used in this study did not contain the pufX gene, and thus the complex was analogous to the LH1 complex from Rsp. rubrum or Blastochloris (Blc.) viridis. In contrast to LH1, Bahatyrova and co-workers (Bahatyrova et al., 2004b) could find no similar pliability in the LH2 complex of Rba. sphaeroides and they suggested that this reflected differences in the arrangement of H-bonds between the C-terminal regions of LH complexes and the C2 acetyl carbonyls of the bound BChl. The effect is that each αβBChl2 unit within LH1 has a certain degree of autonomy within the complex. Indeed, in a detergent environment, LH1 can be disassembled into individual αβBChl2 units (B820) (Miller et al., 1987). It can also be fractionated into a series of oligomers varying in size from 2–3 to 10–11 αβBChl2 units (Westerhuis et al., 2002). LH1 can also assemble in vitro from its constituent αβBChl2 units (Miller et al., 1987) and indeed the 16-fold LH1 ring imaged by Karrasch et al. (1995) was constructed from solubilized αβBChl2 units. The work of Bahatyrova et al. (2004b) demonstrated that the linkage between adjacent αβBChl2 units is flexible and even breakable at room temperature in a detergent-free membrane bilayer (Fig. 3). This makes the argument for the transient opening and closing of LH1 units as a means for quinone exchange additionally appealing. A recent steered molecular dynamics
Fig. 3. (a) Schematic diagram of an αβBChl2 unit, depicted as a building block of the LH1 complex in (b). The four panels show high magnification AFM topographs of individual LH1 complexes prepared from an LH1-only mutant of Rba. sphaeroides (Bahatyrova et al., 2004b).
simulation of quinone shuttling across a ‘closed’ LH1 boundary calculated that this diffusion process took ~8 × 10–3 s (Aird et al., 2007), a timescale compatible with known turnover rates determined for PufX– RCLH1 complexes (Comyras et al., 2005). To date, there is no high-resolution 3-D structure for
162 an LH1 ring. This is partly a reflection of the flexibility of the complex making it difficult to grow large, well ordered crystals for X-ray diffraction. Attempts have been made to determine the 3-D structure of LH1 from Rsp. rubrum by cryo-EM of tilted 2-D crystals. However difficulties in indexing Fourier transforms of the images from tilted crystals are likely to make this impractical. IV. Monomeric Reaction Center-LightHarvesting 1 Complexes A number of low-resolution electron microscopic studies have shown that the RC forms an intimate complex with LH1 (Miller, 1982; Stark et al., 1984; Engelhardt et al., 1986; Meckenstock et al., 1992; Boonstra et al., 1994; Walz et al., 1998; Jungas et al., 1999; Siebert et al., 2004). Based on higher resolution data from ice-embedded samples, Karrasch et al. (1995) proposed that one RC can be accommodated within the 16-fold LH1 ring of Rsp. rubrum. However, the LH1 crystals in this study were reconstituted from detergent-solubilized αβ subunits, raising the question of whether the intact LH1 antenna surrounding the RC really forms a complete, circular ring of 16 subunits in vivo. A. The Reaction Center-Light-Harvesting 1 Complex of Blastochloris viridis The earliest detailed analysis of electron micrographs was performed on photosynthetic membranes from Blc. viridis, which appeared to contain hexagonally close-packed photoreceptor units (Welte and Kreutz, 1982). Miller (1982) determined the 3-D structure of this RC-LH1 complex at low resolution by EM and Fourier analysis of images of natural 2-D crystals embedded in negative stain. A large central structure was revealed to protrude from both surfaces of the membrane apparently surrounded by six smaller centers of mass. It was proposed that this represented a RC surrounded by light-harvesting BChl-protein complexes. It should be noted, however, that the interpretation of low resolution images of close-packed arrays of RC-LH1 complexes to show six-fold or 12-fold symmetry within the LH1 ring has largely arisen from the tendency of many such complexes to adopt a pseudo-hexagonal packing arrangement. In cases where higher resolution was achievable,
Per A. Bullough , Pu Qian and C. Neil Hunter subsequent work has without exception revealed the true symmetry to be higher. Stark and co-workers extended the resolution achieved by Miller (Stark et al., 1984). They undertook an electron crystallographic and heavy metal shadowing analysis of membranes treated with the detergent Triton X-100. These workers characterized the photoreceptor unit as consisting of a central core extending through the membrane and protruding on the periplasmic side by ~40 Å. The core was proposed to be surrounded by a ring of 12 subunits giving a diameter of ~120 Å on the cytoplasmic side and ~100 Å on the periplasmic side. Scheuring et al. (2003) presented high-resolution AFM topographs of the photosynthetic core complex in native Blc. viridis membranes (For a review of the AFM of this and other core complexes, see Chapter 47, Scheuring). The topographs of Blc. viridis membranes at 10 Å lateral and ~1 Å vertical resolution revealed a single RC surrounded by a closed ellipsoidal ring of 16 LH1 subunits. The LH1 subunits rearranged into a circle after removal of the RC from the core complex using the AFM tip. Protrusions from the membrane surface of ~48 Å in height were interpreted as being the tetraheme cytochrome subunit suggesting that only the periplasmic surface was exposed to the AFM tip. Given that the samples were grown by fusion of smaller membrane fragments this implies that fusion always took place between identically oriented membranes. Removal of the putative tetraheme cytochrome subunit led to a change of height, resulting in an RC–LH1 complex protruding ~15 Å from the membrane bilayer. Removal of the tetraheme cytochrome–RC complex led to an empty LH1 ring complex protruding ~7 Å and forming a closed circle with an ~100 Å top diameter. The particles from Blc. viridis appeared to have 16fold rotational symmetry as previously shown for Rsp. rubrum (Karrasch et al., 1995; Jamieson et al., 2002). Scheuring and coworkers proposed that the RC had a ‘noncentered’ location within the LH1 (Scheuring et al., 2003). However unlike cryoEM, AFM does not reveal the internal structure of the transmembrane domain of the RC (the L and M subunits) and this finding is somewhat at odds with the conclusions of Jamieson et al. (2002) working on Rsp. rubrum using cryoEM. However, as had previously been shown the LH1 complex can adopt an ellipticity of between 5% and 10% in the presence of RC, indicating a likely flexibility in the LH1 ring (Jamieson et al., 2002).
Chapter 9
Core Complexes of Purple Bacteria
B. The Reaction Center-Light-Harvesting 1 Complex of Phaeospirillum molischianum The structures of two types of isolated light-harvesting antenna complexes from Phs. molischianum were studied by EM and image analysis (Boonstra et al., 1994). For the RC-LH1 complexes, two distinct projections were obtained, namely top and side views. The RC-LH1 complex formed an almost circular particle with a diameter in the plane of the membrane of about 110 Å. It was suggested that a complex consists of a RC surrounded by most likely 12 LH1 subunits. An asymmetrical protrusion was seen in side views, consisting of the cytochrome subunit giving the RC an apparent height of ~115 Å. The cytochrome subunit was not lost during the purification procedure, which suggests a similar mode of binding to that seen in Blc. viridis (Deisenhofer et al., 1985). The core complexes appeared to have a larger diameter on the periplasmic side and a smaller diameter on the cytosolic side, suggesting that the LH1 subunits are tilted relative to the membrane plane. Gonçalves et al. (2005) obtained AFM images of native membranes allowing the calculation of averages of the photosynthetic complexes using single particle alignment procedures (see also Chapter 47, Scheuring). The LH1 assembly appeared slightly elliptical, with long and short axes of 95 ± 3 Å and 85 ± 3 Å, although the authors do not explicitly state which side of the molecule was viewed. Nevertheless, this clearly showed the LH1 completely surrounding the RC, but the resolution was insufficient to determine the number of LH1 subunits; the authors speculated that it would be 16. C. The Reaction Center-Light-Harvesting 1 Complex of Rhodospirillum rubrum 8.5 Å resolution projection structures of the Rsp. rubrum RC-LH1 complex (Figs. 2b and c) were determined from two crystal forms by cryo-EM (Jamieson et al., 2002). Cryo-EM probed the internal structure of the complex as opposed to the essentially surface views revealed by AFM and negative stain EM. The projection maps were sufficiently detailed to clearly show, for the first time in any species, 16 LH1 subunits arranged in a closed ring around the central RC density. The rotational power spectrum of the tetragonal crystal form showed an unambiguous 16-fold symmetry; no significant difference was observed between the densities representing the crys-
163 tallographically independent LH1 subunits, leading to the conclusion that the RC is surrounded by 16 identical protein subunits (Fig. 2c). In maps of both crystal forms, the area enclosed within the LH1 ring was occupied by significant density corresponding to the RC, as proposed by earlier workers for RC-LH1 complexes from a number of species (Miller, 1982; Stark et al., 1984; Engelhardt et al., 1986; Boonstra et al., 1994; Karrasch et al., 1995; Walz and Ghosh, 1997; Stahlberg et al., 1998; Walz et al., 1998). Tetragonal crystals from a carotenoidless mutant had previously been reported by Walz and Ghosh (1997) and Stahlberg et al. (1998); by treating RC-LH1 complexes in negatively stained crystals as single particles, they showed that individual RCs adopted one of four possible orientations with respect to the crystal axes. This could account for the four-fold symmetry seen in the analysis of one of the crystal forms found by Jamieson et al. (2002). Although the resolution was higher, their strictly crystallographic approach to the image analysis necessarily resulted in the superposition of all RCs in these four possible orientations to give an apparent four-fold symmetry. However, the rotational power spectrum of the RC density at higher radius, in addition to showing a four-fold component, showed an even stronger 16fold harmonic, suggesting that there must also be a 16-fold symmetry to the RC density. This implied a propensity of the RC to adopt up to 16 discrete orientations thereby suggesting that there must be a specific interaction binding the RC within the 16fold LH1 ring. An analysis of a second, orthorhombic, crystal form by Jamieson et al. (2002) indicated only two major orientations for the RC. The structure of this orthorhombic form clearly demonstrated for the first time the ability of the LH1 ring to adopt circular and elliptical conformations (Fig. 2c). The LH1 ring could clearly distort to pack round the RC, when subjected to external packing constraints in this orthorhombic crystal lattice. Jamieson et al. modeled this two-fold symmetric RC density using atomic coordinates derived from the Rba. sphaeroides RC structure. The long axis of the RC model coincides approximately with the major axis of the elliptical LH1 ring. The pseudo two-fold axis of the RC running through the Fe and between the ‘special pair’ of BChls lies within 2−3 Å of the crystallographic two-fold axis. With the uncertainties that arise from calculating model density at low resolution in such a complex environment of lipid, protein, pigment and glucose, the uncertainty
164 in orientation must be at least ±3°. However, within this degree of uncertainty, the planes of the BChl porphyrin rings make an angle of ~43° with the major axis of the LH1 ring. Within the membrane, the only RC transmembrane helices that could interact with the putative LH1 α-helices would be LA, MA and possibly LB and MB; all other helices are much further than 10 Å from their nearest LH1 neighbors. This would be consistent with a helix−helix packing arrangement of LA and MA against nearest-neighbor LH1 α-subunits. This type of packing is also observed in the higher resolution X-ray structure of the core complex of Rps. palustris between LA and its nearest LH1 α-subunit (Roszak et al., 2003). One difference between the RC-LH1 complex of Rsp. rubrum and core complexes from other bacteria is that the BChl macrocycle is esterified with geranylgeraniol, instead of the phytol found elsewhere. Phytol or geranylgeraniol constitute ~30% of the molecular mass of the pigment, and play a significant structural role in assembly and stability of the core complex. Rsp. rubrum is now known to harbor a mutation in the bchP gene, preventing the final step of BChl biosynthesis, which converts geranylgeraniol to phytol (Addlesee and Hunter, 2002). This lost function was restored to Rsp. rubrum by transferring the bchP gene from Rba. sphaeroides, and this new RC-LH1 core complex was purified and reconstituted into wellordered 2-D crystals with a p4212 plane group (Qian et al., 2003). Although interesting differences were observed in the carotenoid content and spectroscopic properties of the complex, a projection map calculated to 9 Å showed that a 16-subunit LH1 ring was present in the mutant that surrounds the RC, as found for the wild type complex (Jamieson et al., 2002). The projection maps appeared to confirm the proposal by Karrasch et al. (1995) that the RC is indeed completely enclosed by 16 identical LH1 subunits. Thus the mechanism of quinone transfer remained unclear. In the absence of any evidence for a ‘quinone channel’ or fixed opening in the LH1 ring, the alternative possibilities of a close approach of quinone molecules either side of the LH1 or ‘breathing motions’ in LH1 must be entertained, at least for quinone transfer in Rsp. rubrum and probably for all monomeric RC-LH1 complexes, most of which are likely to contain ~16 LH1 subunits. Certainly, the LH1 is not a rigid assembly, as shown by the circular and elliptical forms found in this study and that of Scheuring et al. (2003) and in work on LH1 complexes in the absence of the RC (Bahatyrova et al., 2004b).
Per A. Bullough , Pu Qian and C. Neil Hunter As mentioned earlier, a recent molecular dynamics study demonstrated that diffusion of quinone through an LH1 assembly is feasible (Aird et al., 2007). The Rsp. rubrum core complex has so far defied all attempts at growing diffraction quality crystals for X-ray work. Moreover, attempts at imaging and analysis of 2-D crystals in three dimensions have encountered the same problems as Rsp. rubrum LH1 (see above, Qian, Hunter and Bullough, unpublished). In an alternative approach to visualizing the 3-D structure, high-resolution atomic force microscopy (AFM) was used to image 2-D crystals of the RC-LH1 complex (Fotiadis et al., 2004). The AFM topographs showed that the RC-LH1 complex is ~94 Å in height, the RC-H subunit protrudes from the cytoplasmic face of the membrane by 40 Å, and it sits 21 Å above the highest point of the surrounding LH1 ring (Fig. 4.a and g). In contrast, the RC on the periplasmic side is at a lower level than LH1, which protrudes from the membrane by 12 Å, roughly comparable with the equivalent height measured for Blc. viridis (Scheuring et al., 2003). Remarkably, in this exceptionally high resolution AFM study, two distinct rings, one for LH1α and the other for LH1β, could be seen (Fig. 4c). AFM imaging demonstrated the general ellipticity of the LH1 ring at the cytoplasmic and periplasmic sides of the membrane, in both the presence and absence of the RC. However, the RC-LH1 complex was seen to adopt an irregular shape in regions of uneven packing forces in the crystals, reflecting a likely flexibility in the natural membrane and which might indeed be functionally important for allowing the export of quinol. D. Mutant Reaction Center-Light-Harvesting 1 Complexes of Rhodobacter sphaeroides Wild-type RC-LH1 complexes from Rba. sphaeroides RC-LH1 contain an additional polypeptide, PufX, which promotes the formation of dimeric RC-LH1 complexes (see below). The present discussion will be restricted to RC-LH1 complexes from mutant strains of Rba. sphaeroides that assemble monomeric RC-LH1 complexes. The earliest negative stain EM study was carried out by Walz et al. (1998) on 2-D crystals reconstituted from RC-LH1 complexes, purified from an LH2-minus, PufX-minus strain of Rba. sphaeroides, which appeared to contain roughly circular LH1 rings completely surrounding the RC. At this resolution (~30 Å) the RC appeared to adopt random orientations with respect to the crystal lattice.
Chapter 9
Core Complexes of Purple Bacteria
165
Fig. 4. (a) High-resolution AFM topograph of a single molecule of the Rsp. rubrum RC-LH1 complex. (b) Topograph of a complex lacking the RC-H subunit, showing underlying L and M subunits, (c) topograph of a complex with no central RC complex, showing only the LH1 ring. (d) Tilted representation of the RC-LH1 complex of Rsp. rubrum. The RC is the Rba. sphaeroides homolog (1M3X) structure, surrounded by the LH1 complex, which was modeled on the LH1αβ dimer from Conroy et al. (2000) and fitted to the EM projection map of the Rsp. rubrum RC-LH1 complex in Jamieson et al. (2002). (e) Tilted view of the RC-LH1 complex with no RC-H subunit. (f) Tilted view of the LH1 complex with no RC. (g) Side view of the complex. The relative heights of the RC and LH1 complexes and their positions in the membrane were obtained from the AFM data in Fotiadis et al. (2004). HG is the head group region of the membrane, and HC represents the hydrophobic core region of the lipids. The dimensions of the complex and membrane bilayer are indicated in Å. (h) Side view, showing the positions of RC and LH1 BChl cofactors, and their location with respect to the membrane bilayer.
Native membranes isolated from such a strain were shown to contain closely-packed hexagonal arrays of monomeric RC-LH1 complexes, reminiscent of those naturally found in wild-type Blc. viridis, which also lacks a PufX analog (Siebert et al., 2004). In fact, mutants of the core complex date back to early studies of bacterial photosynthesis in the 1960s by Clayton, who isolated a carotenoid-less mutant, R26 (LH2-minus, LH1-RC), which played a pivotal role in the development of the reaction center concept, and which facilitated the subsequent purification and crystallization of the RC complex (Clayton and Smith, 1960; Clayton, 1962, Clayton and Wang, 1971; Feher, 1971; Allen et al., 1987). This is just one example, and the topic of mutant core complexes is too large to be adequately reviewed here, but from a purely structural point of view it is interesting to note that it is possible to generate a series of mutants with decreasing ratios of LH1 BChls per RC. Starting from 29 LH1 BChls/RC in the wild-type core complex, progressive deletion of between 5 and 16 residues from the C-terminus of LH1 α yielded gradually diminished core complexes, eventually with only 10–11 LH1 BChls/RC for the largest deletion. The effect was to create a permanent gap for quinone exchange, so that PufX was no longer necessary. A threshold level of LH1/RC BChls was found, so that when there were fewer than 20 LH1 BChls per RC (about 10 αβ-units) there was no requirement for PufX (McGlynn et al., 1996).
E. The Reaction Center-Light-Harvesting 1 Complex of Rhodospirillum photometricum Scheuring and Sturgis (2005) imaged the cytoplasmic membrane surface of intact chromatophores from Rsp. molischianum; large elliptical shapes were identified as core complexes. The LH1 array of the core complex was found to form a closed ellipse of 16 LH1 subunits surrounding a single RC reminiscent of that observed in Blc. viridis (Scheuring et al., 2003) and Rsp. rubrum (Jamieson et al., 2002). A recent review summarizes the AFM data on a variety of core complexes from photosynthetic bacteria (Scheuring, 2006). Additional information can be found in Chapter 14, Sturgis and Niederman, and Chapter 47, Scheuring. F. Models of Monomeric Core Complexes Early attempts to model the LH1 ring involved creation of a ‘generic’ model of LH1 using the atomic coordinates of the αβ-pairs of LH2 from Rps. acidophila as a framework and building them into a 16-fold ring (Papiz et al., 1996). Subsequently, the LH1 complex of Rba. sphaeroides was modeled computationally as a hexadecamer of αβ-heterodimers, based on a close homology of the heterodimer to that of LH2 of Rsp. molischianum (Hu and Schulten, 1998). The resulting structure yielded a projection map showing good general agreement with the experimentally determined projection map of LHI from Rsp. rubrum (Karrasch et al., 1995).
166 A 3-D model of the RC-LH1 complex is shown in Fig. 4g, which is based on a number of experimental inputs. The lateral positions of the LH1 polypeptides are based on the cryo-EM data (Jamieson et al., 2002) and the vertical positions are taken from the AFM height measurements (Fotiadis et al., 2004). The LH1β structure is taken from the NMR data on this polypeptide (Conroy et al., 2000), and the LH1α structure is based on LH2α from the structure of the Phs. molischianum LH2 complex (Koepke et al., 1996). The reason for this was the similarity of the LH polypeptides of this complex to LH1, rather than LH2 sequences (Germeroth et al., 1993). The sequence of the Rba. sphaeroides LH1α was interpolated into this structure and additional constraints for the BChl binding sites and turns in the C-terminal regions were provided from site directed mutagenesis and resonance Raman studies of the Rba. sphaeroides LH1 complex (Olsen et al., 1994, 1997; Sturgis et al., 1997). In this model there are two belts of aromatic residues for the LH1 complex, which coincide with the measured position of the membrane interface (Fig. 4g). The positive belts of surface charges on both the cytoplasmic and periplasmic sides are in a position to interact with the lipid phosphate groups (Fotiadis et al., 2004). In Fig. 4h it can also be seen that the special pair BChls are located in the hydrophobic core of the membrane. The Fe atom on the cytoplasmic side of the complex lies just outside the head group region of the membrane. The ring of LH1 BChls falls into alignment with the special pair BChls of the RC, in such a way that excitation energy could transfer along a horizontal plane from LH1 to the RC special pair BChls, as suggested by Hunter et al. (1989). However, in the absence of sufficient experimental data, the more detailed features of all such models should be viewed with some caution. V. Monomeric Reaction Center-LightHarvesting 1-PufX Complexes A. The Rhodopseudomonas palustris Reaction Center-Light-Harvesting 1-W Complex The highest resolution 3-D view of a bacterial core complex was provided by the X-ray crystallographic study of Roszak et al. (2003) who determined the structure of the RC-LH1-helix W complex of Rps.
Per A. Bullough , Pu Qian and C. Neil Hunter palustris to 4.8 Å.. The resolution was sufficient to reveal the basic features of the complex, namely an elliptical ring of 15 LH1 αβ subunits surrounding the RC (modeled on the Rba. sphaeroides complex). More recently, in a novel application of single molecule spectroscopy, modeling of the experimental data was shown to be consistent with an elliptical array of 15 BChl dimers, as well as a gap in the ellipse along the long side of the array (Richter et al., 2007a,b). Another notable feature of the structure determined by crystallography is the presence of helix W which lies near to the gap in the LH1 ring which would correspond to a missing sixteenth LH1 αβ subunit (Fig. 5a). The elliptical shape of the LH1 ring is consistent with previous cryo-EM data on the Rsp. rubrum RC-LH1 core complex (Jamieson et al., 2002; see below), as well as with AFM images of the Blc. viridis RC-LH1 complex (Scheuring et al., 2003), but the Rps. palustris LH1 is more tightly wrapped round the RC than in these other complexes, as discussed in Qian et al. (2005). The structural basis for this flexibility in LH1, which accommodates varying degrees of ellipticity, is discussed in Section III above. Helix W has not been identified at present, but in view of its location, adjacent to the hydrophobic tail of the ubiquinone molecule which projects outwards from the RC QB site, and the known role of PufX in the Rhodobacter core complex as a facilitator of quinone exchange (Barz et al., 1995a), it is likely that helix W is a functional homolog of PufX. The monomeric nature of the Rps. palustris core complex in the 3-D structure (Roszak et al., 2003) appears to reflect its native state in the membrane. AFM studies of membranes from this bacterium show the presence of monomeric RC-LH1-W complexes of a size and shape that broadly agree with the crystallographic data (Scheuring et al., 2006; Chapter 47, Scheuring). The exception to this is that the topographic feature assigned to helix W in this AFM study is not restricted to one position in relation to the enclosed RC complex, which argues against its role in quinone exchange, and which also presents difficulties in preparing homogeneous samples for 3-D crystallization. This conflict requires improved resolution for both crystallographic and AFM approaches. A. The Rhodobacter veldkampii Reaction Center-Light-Harvesting 1-PufX Complex Gubellini et al. (2006) performed a structural and
Chapter 9
Core Complexes of Purple Bacteria
167
Fig. 5. Schematic representation of (a) the monomeric Rps. palustris and (b) the dimeric Rba. sphaeroides RC-LH1-PufW/X complexes. Transmembrane helices and pigments are shown in cartoon form and stick form, respectively. Working from the outside of the complex they are: LH1β ring (unshaded), BChl (black), LH1α ring (shaded dark grey), single PufW or PufX helix (unshaded) and reaction center (highlighted with a grey surround) with black pigments and white quinone. The RC protein is composed of the three chains: L (grey boxes), M (unshaded) and H (diagonal lines). The figure was constructed initially with Pymol (DeLano Scientific) using the PDB coordinates 1PYH of the complete Rps. palustris complex. A crude model of the Rba. sphaeroides complex was constructed using EM data from Qian et al. (2005) as a basis along with the reaction center crystal structure (PDB code 1PCR), the solution structures of PufX (2NRG), LH1α from Rsp. rubrum (1XRD) and LH1β from Rba. sphaeroides (1DX7), and LH1 BChl coordinates from the aforementioned Rps. palustris complex. For clarity, the coordinates were edited to show only transmembrane helices and phytol chains were removed from pigments. For consistency, coordinates of QB from the Rba. sphaeroides reaction center were modeled to the Rps. palustris complex based on the homology of the two structures.
functional analysis of the photosynthetic apparatus of Rhodobacter veldkampii. A core complex was purified and analyzed by sedimentation, size exclusion chromatography, mass spectrometry, and EM. A PufX subunit identified by MALDI-TOF was found to be associated with the core complex. Sedimentation and electron microscopy showed the core complex to be monomeric, suggesting that in Rba. veldkampii, PufX is involved in the photosynthetic growth but is unable to induce the dimerization of the core complex. It should be noted that in Rba. veldkampii, the requirement of PufX for anaerobic, photosynthetic growth has not been proven by deletion of the pufX gene. However, analysis of Cyt b561 reduction kinetics showed that the quinol-quinone exchange at the QB site of the RC occurred in Rba. veldkampii membranes as rapidly as in wild type Rba. sphaeroides chromatophores, suggesting that the PufX protein might also facilitate cyclic electron flow in Rba. veldkampii. Gubellini et al. (2006) concluded that the native membrane has either a monomeric organization or a much less stable dimeric organization of the core complex than that prevalent in Rba. sphaeroides and Rba. blasticus. They suggested that amino acids in PufX involved in the core complex dimerization may lie within the N-terminal region, since it is known
that N-terminal deletion of PufX in Rba. sphaeroides reduces the level of core complex dimerization (Francia et al., 2002). This deletion work had suggested a role for amino acids 7–19 in the interaction between two PufX proteins; alignment of the PufX N-terminus from Rba. sphaeroides (amino acids 9–16) and Rba. capsulatus (amino acids 9–16) revealed two conserved oppositely charged amino acids in the DXXNXXXK sequence that might be involved in the stabilization of the dimer. It was suggested that the corresponding DXXXK sequence in Rba. veldkampii might be too short for such an interaction (Gubellini et al., 2006). In a recent development of this work cryo-EM was used for a 3-D reconstruction of the Rba. veldkampii core complex at 12 Å resolution (Busselez et al., 2007). Interestingly, this complex was found to be larger and less elliptical than the monomeric RC-LH1-W complex from Rps. palustris, but the two complexes are similar in overall architecture, in the sense that the RC is surrounded by an LH1 assembly of 15 subunits, and there is a gap in the structure, in the case of the Rba. veldkampii of ~40 Å. The QB site of the RC faces the gap. The preferred model of this complex employs the ‘straight’ PufX NMR structure (Wang et al., 2007), as opposed to the ‘bent’ structure determined by Tunnicliffe et al. (2006), since only
168 the former sterically allows two adjacent GXXXG (GXXXV in Rba. veldkampii) putative dimerization motifs to drive the association of two monomers to form a dimer. In the model favored by Qian et al. (2005) the much greater distance between two PufX molecules within a dimer requires that, instead, the N-termini are essential for stabilizing the core dimer. There are several problems with reconciling the Qian et al. cryo-EM data with the PufX-PufX helix-helix model (Scheuring et al., 2004, 2005; Busselez et al., 2007): (1) PufX is located near to the RC QB site, so (2) a PufX-PufX transmembrane association therefore requires that the RC is rotated within the dimer structure to a point where it no longer fits with the known EM density, nor with the linear dichroism (LD) data for the Qy transition of the RC special pair BChls determined by Frese et al. (2000), then applied to the negative stain EM of dimer-only membranes (Siebert et al., 2004). However, only better structural data can resolve these differences. VI. Dimeric Reaction Center-LightHarvesting 1-PufX Complexes A. The Rhodobacter sphaeroides and Rhodobacter blasticus Core Dimers Much of the work on RC-LH1-PufX complexes has been facilitated by mutants lacking LH2 complexes. The elongated tubular membranes from these mutants, which have diameters of between 50 and 75 nm and lengths of 50 to 2000 nm (Siebert et al., 2004), are ideal material for negative stain analysis by EM because they contain naturally crystalline helical arrays of RC-LH1-PufX complexes (Jungas et al., 1999; Siebert et al., 2004). Freeze-fracture electron microscopy of whole cells of M21, a mutant lacking LH2 complexes (Ashby et al., 1987), also showed the presence of elongated tubular membranes with a highly ordered helical array of closely packed, dimeric particles with dimensions of ~100 Å × ~200 Å lying along the longitudinal axis. It was suggested that these particles represent dimers of RC-LH1-PufX complexes (Westerhuis et al., 2002). Subsequently, membranes from mutants lacking LH2 complexes provided a convenient starting point for purification of RC-LH1-PufX dimers, and their crystallization (Qian et al., 2005). Jungas et al. (1999) analyzed negatively stained tubular membranes and obtained a low-resolution
Per A. Bullough , Pu Qian and C. Neil Hunter (~25 Å) structure of the RC-LH1-PufX complex. This first view of the Rba. sphaeroides core complex revealed a dimeric structure in projection, with two RC complexes enclosed by an S-shaped LH1 assembly. Some of the electron density had been tentatively assigned to the cytochrome bc1 complex, so a subsequent study was conducted to resolve this issue and also to investigate the effect of deleting PufX (Siebert et al., 2004). Highly purified tubular membranes were prepared, and the absence of the cytochrome bc1 complex was verified by western blotting. These membranes were composed solely of dimeric RC-LH1-PufX complexes. The absence of PufX had a remarkable effect, transforming the tubular cell membranes into flattened sheets comprised of closely packed hexagonal arrays of monomeric RC-LH1 complexes. The low-resolution structures of these monomers showed that an LH1 ring completely enclosed each RC (Siebert et al., 2004). Francia et al. (1999) had already shown that dimerization of the core was abolished in the absence of PufX, and the EM study (Siebert et al., 2004) showed the macroscopic consequences of deleting PufX. Following the initial observations of Jungas et al. (1999) independent verification of the ordered nature of dimeric RC-LH1-PufX complexes was obtained by measuring the capacity of oriented tubular membranes to absorb polarized light. This LD study clearly showed that PufX induces the long-range alignment of transition dipoles of BChls within the RC (Frese et al., 2000). Importantly, LD also established the angle between the RC special pair BChls and the longitudinal tube (orientation) axis. Superposition of tshe LD data onto the negative stain projection map places the RC QB site adjacent to the opening in the LH1 assembly (Siebert et al., 2004), an arrangement that would permit quinol access to the membrane milieu. The suggestion was made that PufX was also near this opening in the LH1 structure, consistent with the 500-fold reduction in Q/QH2 exchange observed in a mutant with no PufX (Barz et al., 1995b). This proposed location of PufX, the orientation of the RC within the complex, and the extent to which the LH1 complex was really ‘open’ were all aspects of the dimer structure in need of improved structural resolution, beyond the capacity of the naturally crystalline membrane material used, and analysis by negative stain EM. Using such RC-LH1-PufX (no LH2) membranes as a starting point, highly purified dimeric complexes were prepared, and 2-D crystals were grown with
Chapter 9
Core Complexes of Purple Bacteria
169
Fig. 6. Comparison of the 8.5 Å resolution projection map of the Rba. sphaeroides RC-LH1-PufX dimer with previous core complex structures. (a) Comparison with negatively stained complexes from natural tubular membranes of Rba. sphaeroides (Siebert et al., 2004). (b) Comparison of the X-ray crystal structure of RC-LH1-helix W complex of Rps. palustris (Roszak et al., 2003).
sufficient size and order to enable the calculation of a projection map to 8.5 Å (Qian et al., 2005). This revealed the α-helices that comprise the RC and LH1 components, as well as a possible location for PufX. A continuous LH1 assembly, consisting of 28 αβ subunits, snakes round the two RC complexes within the dimer structure (Fig. 5b). Figure 6a compares the negative stain projection map (Siebert et al., 2004) with the higher resolution structure (Qian et al., 2005). Examination of the cryo-EM projection map shows that there is no ‘gap’ as such in each half of the LH1 assembly; however, the αβ pairs are more separated in this region and the relatively weak and diffuse density could arise from mobility of LH1β, namely movement of the whole polypeptide, or selective motion of just part of LH1β (Fig. 6a). This possibility was raised by the solution structure of LH1β (Conroy et al., 2000). One crucial feature of this projection map is the orientation of the RC within the dimer, unequivocally established by fitting the known RC structure known from X-ray crystallographic data to the projection map obtained by cryo-EM (Fig. 5b). This places the RC in an orientation which not only locates the QB site adjacent to the more widely spaced and mobile LH1β polypeptide, it also places the Qy transition dipole of the special pair BChls in approximate agreement with the orientation proposed on the basis of earlier LD and negative stain data (Frese et al., 2000; Siebert et al., 2004). Finally, this orientation of the RC places the QB site next to the ‘extra’ density in the cryo-EM projection map, which lies outside both the LH1 arc and the envelope of density arising from the RC (Fig. 6a). On the basis that this extra region of density arises from a single transmembrane helix, and that the only unassigned
component in the structure is PufX, this density was assigned to PufX (Fig. 5b), close to the region proposed earlier (Siebert et al., 2004). The RC provides a valuable reference structure for aligning and comparing structures of core complexes; accordingly it was possible to compare the dimeric core complex from Rba. sphaeroides with the monomeric Rps. palustris complex (Roszak et al., 2003). The result is that helix W of the Rps. palustris structure and the PufX location proposed for the dimer structure lie within ~10 Å of each other. However, another EM analysis of 2-D crystals prepared from the Rba. sphaeroides core dimer has concluded that there are 24 LH1 αβ subunits (compared to 28 in the 8.5 Å structure), and that the two PufX polypeptides of the dimer are located at the junction between the two halves of the structure (Scheuring et al., 2004). PufX is known to drive the dimerization process (Francia et al., 1999; Siebert et al., 2004), and this arrangement allows the proximity of the transmembrane helices of PufX to drive the formation of the dimer, whereas in the 8.5 Å structure (Qian et al., 2005) an interaction between the N-termini is more likely. The proposed location of the PufX transmembrane helix adjacent to the RC QB site, a position analogous to that of helix W in the monomeric RC-LH1 complex (Roszak et al., 2003), requires an N-terminal region of at least 40 Å reaching towards the dimer interface to promote dimer formation (Qian et al., 2005). This is compatible with the solution structure of the Rba. sphaeroides PufX polypeptide, showing such an extended N-terminal region (Tunnicliffe et al. 2006; Fig. 1d). Consistent with this idea, a mutagenesis study demonstrated that deletion of N-terminal residues destabilizes dimer formation (Francia et al., 2002).
170 An AFM analysis of the core complex from Rba. blasticus (Scheuring et al., 2005) also proposes a PufX-PufX transmembrane interaction at the dimer junction (Fig. 7), as well as an orientation of the RC which differs from that proposed by Qian et al. (2005). Homodimerization of the PufX transmembrane regions (Fig. 7) is proposed to drive core dimer formation (Scheuring et al., 2004; Scheuring et al., 2005; Aklujkar and Beatty, 2006). The interaction of two such closely associated transmembrane helices implies a center-to-center helix spacing of ~10 Å (MacKenzie et al., 1997). The densities in the 8.5 Å projection map of the Rba. sphaeroides core complex dimer reveal no such spacings at the interface between the two halves of the dimer (Qian et al., 2005). However, as already pointed out in the earlier section on the monomeric Rba. veldkampii core complex, only a higher resolution structure will resolve these issues. Whatever the determinants of PufX are that drive dimerization, they can be disrupted in such a way that a homogeneous population of monomeric core complexes is produced. This is shown clearly by the work of Abresch et al. (2005), who used an LH2minus mutant, together with a 6-Histidine tag on the RC-H subunit, as a source of material. It is possible that the use of the detergent diheptanoyl-sn-glycero-3phosphocholine (DHPC) for solubilization is responsible for conversion of dimers into monomers, since this was also the experience of the authors (P. Qian, C.N. Hunter and P.A. Bullough, unpublished). The fact that a homogeneous population of monomers is produced, to the extent that it was possible to use them to grow 3-D crystals that diffracted to 12 Å, shows that core dimers split symmetrically to form a uniform population; Abresch et al. (2005) estimated that each monomer comprised 13.3 ± 0.9 LH1 subunits per RC, in good agreement with the ratio of 14 found for each half of the dimer (Qian et al. 2005). B. Structural and Functional Roles for PufX It has been suggested that there is a relationship between PufX and the N-terminal segment of the cytochrome subunit which forms an integral component of the RC complex from the aerobic bacterium Roseobacter denitrificans (Hucke et al., 2003). Although such comparisons are hampered by the inherently low similarity between even better known PufX homologs, this interesting suggestion provides another possible insight into the origin of PufX and
Per A. Bullough , Pu Qian and C. Neil Hunter
Fig. 7. The Rba. blasticus dimeric core complex, imaged by AFM. (a) Two-fold symmetrized image of the periplasmic surface of a single molecule of the complex. Scale bar: 5 nm; (b) corresponding assembly model. Two RC complexes are surrounded by two C-shaped LH1 assemblies each consisting of 13 subunits, and connected by a PufX dimer at the two-fold axis. Adapted from Scheuring et al. (2004).
its interaction with the RC and LH1 complexes. PufX and its analogs would appear to fulfill two roles in some cases, namely to facilitate quinonequinol exchange and to stabilize dimeric core complexes. Clearly dimerization is not essential for a photosynthetic phenotype, as evidenced by the monomeric core complexes discussed above. It is possible, though, that dimerization is important for membrane organization, antenna connectivity, excitation transfer and for quinone-mediated communication with cytochrome bc1 complexes by influencing distribution of the quinone pool (Frese et al., 2000; Francia et al., 2004; Siebert et al., 2004; Bahatyrova et al., 2004a; Comayras et al., 2005). See Chapter 14, Sturgis and Niederman, and Chapter 15, Şener and Schulten, for discussions of membrane architecture and the functional consequences of differing arrangements and stoichiometries of complexes. 1. PufX: Core Dimerization and Membrane Curvature A recent 3-D reconstruction of the RC-LH1-PufX core dimer from Rhodobacter sphaeroides, calculated from single particle EM analysis of negatively stained complexes (Qian et al., 2008; P. Qian, P.A. Bullough and C.N. Hunter, unpublished), shows how PufX can indirectly mediate membrane curvature. This study reveals that the two halves of the dimer molecule incline toward each other on the periplasmic side at an angle of ~146º, forming a distinct ‘V’ shape (Fig. 8a). Such a possibility was raised earlier by Geyer and Helms (2006a) who predicted an angle of 154º. The intrinsic curvature of this dimer drives the for-
Chapter 9
Core Complexes of Purple Bacteria
171
Fig. 8. (a) 3-D reconstruction of the core RC-LH1-PufX dimer of Rba. sphaeroides. Left: examples of the classes of negatively stained single particles used for the 3-D reconstruction. Middle: surface views of the 3-D reconstruction of the dimeric RC-LH1-PufX complex viewed from different angles. LH1 appears to form a continuous density, which surrounds central densities arising from the RCs. The RC-H subunits, protruding from LH1 ring, can be seen clearly on the cytoplasmic surface of the complex. The two LH1 rings are inclined towards each other on the periplasmic side forming a ‘V’ shape. Right: 3-D model of the arrangement of RC-LH1-PufX dimers in the native tubular membrane. The long pitch helix formed from associations between the long axes of the dimers is colored in dark gray. The outer, cytoplasmic face of the tubular membrane is shown. (b) The influence of PufX, and the consequent formation of core dimers and monomers, on membrane morphology in LH2-minus mutants. Elongated cells (top) as well as tubular membranes (bottom) are assembled from PufX+ dimer-containing cores (core model on left), but flattened membrane sheets result from PufX-minus monomeric cores (core model on right). The structural models of the core complexes are shown in more detail in Fig. 9.
mation of the membrane that houses it, and the angle of 146º explains the size and shape of the tubular membranes observed in negatively stained thin sections of core-only mutants that lack the LH2 complex (Hunter et al., 1988; Kiley et al., 1988; Westerhuis et al., 2002). These early EM thin section and freezefracture studies showed that, in the absence of any other interactions, core dimers of Rba. sphaeroides have a tendency to pack to form helical arrays. Many such arrays form elongated tubes, with a diameter of approximately 70 nm (Hunter et al., 1988; Jungas et al., 1999; Siebert et al., 2004). Figure 8b shows the effect of deletion of pufX; in
such mutants (which also lack LH2) membranes comprising flattened sheets of pseudo-hexagonally packed arrays of monomeric core complexes form, rather than tubes (Siebert et al., 2004). Figure 9 shows a comparison of the monomeric RC-LH1-helix W core complex from Rps. palustris (Roszak et al., 2003) and the modeled dimeric RC-LH1-PufX complex of Rba. sphaeroides, using the location of PufX assigned in Qian et al. (2005), and a space-filling representation in each case. These contrasting shapes of core complex can be expected to impose differing morphologies on the membranes in which they sit. In the wild-type, as opposed to the situation in LH2-
172
Per A. Bullough , Pu Qian and C. Neil Hunter
Fig. 9. Structural models of monomeric and dimeric core complexes. In each case a side view and a space-filling representation is used. For the RC the complete structure is present (shown in black), but for the LH1 (gray) and helix W/PufX components (white) only the transmembrane helices are shown. a. The monomeric RC-LH1-W core complex from Rps. palustris (Roszak et al., 2003). b. The dimeric RC-LH1-PufX core complex of Rba. sphaeroides. The coordinates of the LH1 αβ pairs were adopted from Roszak et al. (2003), fitted into the tilted 3-D dimer structure from single particle analysis (Chapter 9, Bullough et al.), and adjusted according to the cryo-EM projection map in Qian et al. (2005). The transmembrane domain of PufX was taken from Tunnicliffe et al. (2006). See also color plate 5, Fig. 7.
minus mutants, this tendency of core dimers to stack has to be resisted; if it were otherwise, the spherical invaginations seen in the wild-type could not assemble, and tubular structures would form instead. Thus, only a few dimers — a rough model suggests no more than 4 or 5 — can stack along their long axes before it becomes necessary to ‘cap’ the growing membrane with LH2-rich domains (Fig. 10). It is interesting that in the AFM images of wild-type membrane patches, the number of dimers in a row rarely exceeded 5 (Bahatyrova et al., 2004a). Thus, although invagination of the intracytoplasmic membrane proceeds with the preferential assembly of core complexes, it is important that LH2 incorporation prevails during the latter stages of membrane invagination. It is relevant to note here that Niederman et al. (1976) observed a stepwise sequence of photosynthetic unit assembly with LH1 complexes dominating the early stages of membrane growth, eventually giving way to LH2 in the later stages. The suitability of LH2 domains for ‘capping’ growing membrane invaginations is consistent with the known intrinsic curvature of the LH2 complex in Rba. sphaeroides, and it is known that LH2-only membranes form spherical structures (Hunter et al., 1988). It is not clear, in the absence of a high resolution structure of the Rba. sphaeroides LH2 complex, whether the intrinsic curvature arises from the protein complex itself, or from tightly bound lipids. An LD study of the organization of complexes in intact wild-type membrane vesicles concluded that rows of dimers, running ‘north-south’ were capped by LH2-rich domains (Frese et al., 2004).
From the above, it is clear that the well-documented role of PufX in driving dimerization of core complexes (Francia et al., 1999; Siebert et al., 2004) also has the effect of promoting membrane curvature in Rba. sphaeroides. Although the effect of PufX on membrane curvature is observed in Rba. sphaeroides mutants (Fig. 8), this work raises a more general possibility that bacteria with monomeric complexes (see sections IV and V) will tend to assemble lamellar structures, rather than spherical or tubular membranes, and that these alternative morphologies are related in some way to the absence of PufX or to the lack of dimerization motifs in the PufX homologs present. However, there is a notable exception; Rsp. rubrum assembles spherical intracytoplamic membranes packed with monomeric core complexes. In order to understand more about membrane curvature and domain formation, it has been necessary to invoke the application of colloid theory (see below). 2. PufX: Membrane Domains, Excitation Sharing and Quinone Sequestration AFM of Rba. sphaeroides membranes shows that the core dimers and LH2 complexes partition within the membrane, but the mechanisms for formation of domains of membrane proteins within a membrane bilayer are unclear. Why do core and LH2 complexes not distribute randomly throughout the membrane in Rba. sphaeroides? A recent modeling study used Monte Carlo simulations to demonstrate that depletion interactions, known from colloid theory (for
Chapter 9
Core Complexes of Purple Bacteria
173
Fig. 10. Model of spherical intracytoplasmic membrane vesicles of Rba. sphaeroides, showing rows of dimers interspersed with LH2 complexes, as in Bahatyrova et al. (2004), and modeled to correspond approximately to the arrangement of complexes in Sener et al. (2007). A large domain composed of LH2 complexes caps the vesicle and prevents the rows of dimers from extending the invagination further, a process which would eventually form a tubular structure.
example, Eldridge et al., 1993), could produce ordered states within the membrane (Frese et al., 2008). Domain formation was shown to be driven, at least in part, by the size mismatch between the relatively large, elongated core dimers and the smaller, circular LH2 complexes. An additional factor also came into play in this analysis, namely curvature mismatch between the interacting outer faces of core and LH2 complexes as they pack together. For example, LH2 complexes tend to associate with one another to form tightly curved spherical arrays ~50 nm in diameter, in contrast to the ~70 nm tubular membranes formed by core dimers. Thus, incompatibilities between the interacting surfaces of these complexes can arise in mixed populations of LH2 and core dimers, giving rise to the curvature mismatch factor. Thus, taking all of the work detailed in this section into account, it is possible to trace connections between PufX, core dimerization, membrane curvature and domain formation. For further analysis and a review of membrane organization in bacterial photosynthetic membranes see Chapter 14, Sturgis and Niederman. The tendency of core dimers to stack has a structural role, but is there a functional role as well? Modelling of excitation transfer in wild-type membranes predicts a significant degree of excitation sharing within a dimer array (Şener et al., 2007; Chapter 15, Şener and Schulten). It is possible that, in the event of one or more RCs being in the closed state, linear arrays of core dimers might optimize energy trapping by allowing migration of the excitation along
the dimer array to an open RC. The necessity for such an arrangement might be dictated not only by energy transfer considerations, but also by turnover at the RC QB site, and migration of quinol/quinone and cytochrome c2 to and from the cytochrome bc1 complex. The interdependence of these processes is nicely illustrated in a modeling paper by Geyer and Helms (2006b). Another interesting aspect of the 3-D reconstruction of the Rba. sphaeroides core dimer is the identification of an apparent space within each half of the molecule, situated within the LH1 ring near the LH1 αβ14 subunit (Fig. 11) (Qian et al., 2008). The favored position of PufX assigned in Qian et al. (2005), but still subject to some debate, as noted in previous sections, is indicated by a single helix. A space-filled RC QB molecule is also shown in Fig. 11. A detailed analysis of the kinetics of charge separation, together with the phospholipid and quinone content of core dimers, showed that approximately 10–15 quinone molecules and 80–90 phospholipid molecules per reaction center are retained by purified, detergent washed core dimer complexes (Dezi et al., 2007). The 3-D model of the dimer in Fig. 11 shows that there is a space between the inner face of the LH1 ring and the outer surface of the RC which could accommodate lipids. An area of approximately 250 Å2 is required to accommodate ~10 quinone molecules within each half of the core complex, and lipid molecules are also likely to be present in this region, since they form a flexible environment for quinone movement (Jones, 2007). Each of the two spaces
174
Per A. Bullough , Pu Qian and C. Neil Hunter
Fig. 11. 3-D model of the RC-LH1-PufX dimer viewed from the cytoplasmic side showing the transmembrane helices of the LH1 complex positioned according to the cryo-EM projection map in Qian et al. (2005). The LH1 αβ14 pair is colored in dark gray to differentiate it from the other LH1 αβ pairs. The transmembrane domain of PufX, represented by a ribbon and labeled ‘X’, was taken from Tunnicliffe et al. (2006). The 3-D reconstruction from single particle data (see Fig. 8) is represented by a mesh. The RC QB molecule, labeled U10, is shown using a space-filling representation. The space adjacent to LH1 αβ14, which might accommodate a quinone pool, is labeled ‘Q’.
within the dimer in Fig. 11 is roughly equivalent to a circular area of diameter ~20 Å, consistent with the estimate in Dezi et al. (2007). This space appears to have been penetrated by stain, which may reflect the more open and mobile nature of the quinone/lipid environment compared with the dense hydrophobic stain-excluding interior of the protein. This space within the complex occupies a good strategic location for a holding area or vestibule for quinones, since it is adjacent to PufX, the RC QB site and the loose region of the LH1 ring, where the flexibility of the αβ transmembrane helices could facilitate the exit and ingress of quinol/quinone molecules. Clearly, higher resolution 3-D structural data are required to examine this point. VII. The Biogenesis of Core Complexes The early work of Youvan, Marrs and Hearst discovered genes involved in assembly of the core complex, and one of these, orf1696, is required for assembly of the RC-LH1-PufX complex in Rba. capsulatus (Youvan et al., 1982; Zsebo and Hearst, 1984). Orf1696 encodes a hydrophobic protein of approximately 50 kDa. Subsequent studies verified its role in core complex assembly (Young et al.,
1998), and proposed a topological model comprising 12 transmembrane spans (Young and Beatty, 1998). The sequence similarity between the gene products of orf1696, designated as lhaA, and pucC, an analogous assembly factor for LH2, was also noted (Young and Beatty, 1998). Disruption of both the lhaA and RC-H subunit genes showed the likelihood of an LhaA assembly factor in Rba. sphaeroides (Sockett et al., 1989). Other insertional mutagenesis work had already revealed another gene for assembly of core complexes, orf214, which is located next to the puhA gene and near to lhaA (Wong et al., 1996), and yet another gene product, Orf162b, appears to play a role in enhancing the assembly or stability of RC-LH1-PufX complexes (Aklujkar et al., 2000). As more genomic sequences become available, it is clear that lhaA, orf214 and orf162b all have counterparts in other photosynthetic bacteria (Aklujkar et al., 2000; Aklujkar et al., 2005). Further analysis of orf241, which designated this gene as puhB, assigned the encoded protein as a RCspecific assembly factor with only an indirect effect on LH1; a role in dimerization of the RC-LH1-PufX complex was proposed (Aklujkar et al., 2005). One view of the RC-H subunit is that it has a role in core complex assembly, since early work on puhA mutants showed that only LH2 assembly appeared to survive
Chapter 9
Core Complexes of Purple Bacteria
inactivation of puhA. In contrast, inactivation of the puhA gene in Rsp. rubrum does not impair assembly of the LH1 complex, unless carotenoid synthesis is also prevented (Lupo and Ghosh, 2004). As a result the RC-H subunit was proposed to have a chaperonelike function in forming assembly sites prior to the arrival of the other RC subunits (Varga and Kaplan, 1993). Subsequently, a different model proposed that, instead, RC-M is a nucleus for assembly, with the arrival of RC-L followed by RC-H (Tehrani et al., 2003; Tehrani and Beatty, 2004). Niederman and co-workers investigated the timing of assembly of the core complex by initiating membrane assembly in low aeration cell suspensions, following a transfer from high aeration conditions which suppress the biogenesis of photosynthetic complexes (Niederman et al., 1976). Subsequently, it was shown that prior to the induction of membrane assembly, which took place over a 400 minutes time course, the RC-H subunit was detectable by western blotting (Pugh et al., 1998). Levels of RC-M and RC-L were not monitored. The next protein to be synthesized was PufX, and this was followed by LH1 α and β polypeptides. It was suggested that a RCPufX-LH1αβ complex is formed at the earliest stages of the assembly of the core, and also that LH1αβBChl2 complexes are progressively added to this complex to drive the encirclement of the RC. The early arrival of PufX would prevent blockage of the RC QB site by LH1αβBChl2 units. The fact that LH1 complexes from a variety of bacteria can be reversibly dissociated into αβBChl2 (B820) complexes in the presence of detergents supports this concept of LH1αβBChl2 units as building blocks of LH1, both in vitro and in vivo (see Section III above, and Chapter 10, Loach and Parkes-Loach for a thorough discussion of B820 complexes, and Chapter 46, Braun and Fiedor, for a model of LH1 assembly). Resonance Raman studies of site directed mutants of LH1 revealed that the network of H-bonds stabilizes each αβBChl2 unit (Olsen et al., 1994, 1997; Sturgis et al., 1997). Thus, the four possible H-bonds to the pair of BChls are ‘internal’ to the αβBChl2 unit (Fig. 3), providing a significant driving force to stabilize this putative assembly intermediate (Davis et al., 1997). More recently, a possible intermediate of LH1 assembly, iB873, has been demonstrated using a new in vitro reconstitution method, the important difference from earlier work being that this complex contains carotenoids. It was speculated that iB873 is a smaller complex than the mature LH1 assembly (Fiedor and Scheer, 2005).
175 Several factors that influence the insertion of LH1 α and β polypeptides into the membrane have been
investigated in Rba. capsulatus. This topic has been reviewed (Drews, 1996; Drews and Niederman, 2002), but briefly it has been shown that a positively charged N-terminal domain of LH1α is essential for assembly of the whole LH1 complex. Surprisingly, the corresponding domain of the LH1β polypeptide is negatively charged (Stiehle et al., 1990). The highly conserved residues Trp–24 and Pro–19 located within the LH1α N-terminal domain were also examined, with the former residue essential for LH1 assembly, but mutations in the Pro were tolerated to some extent (Richter et al., 1991). Phosphorylation of LH1α has also been reported, detected using both monoclonal antibodies and radio-labeling approaches. The role of this processing in assembly of the LH1 complex is still unclear, as is the proportion of LH1 polypeptides that undergo this modification, but the factors that influence phosphorylation include a light-driven proton gradient and a redox potential in excess of +200mV (Pucheu et al., 1996). Certainly, the value of using an in vitro transcription-translation system is clear, since this allows some control over the proteins and insertion factors present (Chory and Kaplan, 1982; Meryandini and Drews, 1996). Acknowledgments The authors gratefully acknowledge funding from the Biotechnology and Biological Research Council, U.K. References Abresch EC, Axelrod HLA, Beatty JT, Johnson JA, Nechustai R and Paddock ML (2005) Characterization of a highly purified, fully active, crystallizable RC-LH1-PufX core complex from Rhodobacter sphaeroides. Photosynth Res 86: 61-70 Addlesee HA and Hunter CN (2002) Rhodospirillum rubrum possesses a variant of the bchP gene, encoding geranylgeranylbacteriopheophytin reductase. J Bacteriol 184: 1578–1586 Aklujkar M and Beatty JT (2005) The PufX protein of Rhodobacter capsulatus affects the properties of bacteriochlorophyll a and carotenoid pigments of light-harvesting complex 1. Arch Biochem Biophys 443: 21–32 Aklujkar M and Beatty JT (2006) Investigation of Rhodobacter capsulatus PufX interactions in the core complex of the photosynthetic apparatus. Photosynth Res 88: 159–171 Aklujkar M, Harmer AL, Prince RC and Beatty JT (2000) The orf162b sequence of Rhodobacter capsulatus encodes a protein
176 required for optimal levels of photosynthetic pigment-protein complexes. J Bacteriol 182: 5440–5447 Aklujkar M, Prince RC and Beatty JT (2005) The PuhB protein of Rhodobacter capsulatus functions in photosynthetic reaction center assembly with a secondary effect on light-harvesting complex 1. J Bacteriol 187: 1334–1343 Allen JP, Feher G, Yeates TO, Komiya H and Rees DC (1987) Structure of the reaction center from Rhodobacter sphaeroides R26: The protein subunits. Proc Natl Acad Sci USA 84: 6162–6166 Ashby MK, Coomber SA and Hunter CN (1987) Cloning, nucleotide sequence and transfer of genes for the B800-850 light-harvesting complex of Rhodobacter sphaeroides. FEBS Lett 213: 245–248 Bahatyrova S, Frese RN, Siebert CA, Olsen JD, van der Werf KO, van Grondelle R, Niederman RA, Bullough PA, Otto C and Hunter CN (2004a) The native architecture of a photosynthetic membrane. Nature 430: 1058–1062 Bahatyrova S, Frese RN, van der Werf KO, Otto C, Hunter CN and Olsen JD (2004b) Flexibility and size heterogeneity of the LH1 light-harvesting complex revealed by atomic force microscopy: Functional significance for bacterial photosynthesis. J Biol Chem 279: 21327–21333 Barz WP and Oesterhelt D (1994) Photosynthetic deficiency of a pufX deletion mutant of Rhodobacter sphaeroides is suppressed by point mutations in the light-harvesting complex genes pufB and pufA. Biochemistry 33: 9741–9752 Barz WP, Francia F, Venturoli G, Melandri BA, Verméglio A and Oesterhelt D (1995a) Role of the PufX protein in photosynthetic growth of Rhodobacter sphaeroides. 1. PufX is required for efficient light-driven electron transfer and photophosphorylation under anaerobic conditions. Biochemistry 34: 15235–15247 Barz WP, Verméglio A, Francia F, Venturoli G, Melandri BA and Oesterhelt D (1995b) Role of the PufX protein in photosynthetic growth of the Rhodobacter sphaeroides. 2. PufX is required for efficient ubiquinone/ubiquinol exchange between the reaction centre QB site and the cytochrome bc1 complex. Biochemistry 34: 15248–15258 Boonstra AF, Germeroth L and Boekema EJ (1994) Structure of the light-harvesting antenna from Rhodospirillum molischianum studied by electron microscopy. Biochim Biophys Acta 1184: 227–234 Broglie RM, Hunter CN, Delepelaire P, Niederman RA, Chua NH and Clayton RK (1980) Isolation and characterization of the pigment-protein complexes of Rhodopseudomonas sphaeroides by lithium dodecyl sulfate/polyacrylamide gel electrophoresis. Proc Natl Acad Sci USA 77: 87–91 Brunisholz RA and Zuber H (1992) Structure, function and organization of antenna polypeptides and antenna complexes from the three families of Rhodospirillaneae. J Photochem Photobiol B-Biol 15: 113–140 Busselez J, Cottevieille M, Cuniasse P, Gubellini F, Boisset N and Lévy D (2007) Structural basis for the PufX-mediated dimerization of bacterial photosynthetic core complexes. Structure 15: 1674–1683 Chory J and Kaplan S (1982) The in vitro transcription-translation of DNA and RNA templates by extracts of Rhodopseudomonas sphaeroides. Optimization and comparison of template specificity with Escherichia coli extracts and in vivo synthesis. J Biol Chem 257: 15110–15121 Clayton RK (1962) Primary reactions in bacterial photosynthesis.
Per A. Bullough , Pu Qian and C. Neil Hunter I. The nature of light-induced absorbency changes in chromatophores; evidence for a special bacteriochlorophyll component. Photochem Photobiol 1: 201–210 Clayton RK and Smith C (1960) Rhodopseudomonas sphaeroides: High catalase and blue-green double mutants. Biochem Biophys Res Commun 3: 143–145 Clayton RK and Wang RT (1971) Photochemical reaction centers from Rhodopseudomonas sphaeroides. In: San Pietro A (ed) Methods in Enzymology Vol XXIII, pp 696–704. Academic Press, New York Cogdell RJ, Fyfe PK, Barrett SJ, Prince SM, Freer AA, Isaacs NW, McGlynn P and Hunter CN (1996) The purple bacterial photosynthetic unit. Photosynth Res 48: 55–63 Comayras R, Jungas C and Lavergne J (2005) Functional consequences of the organization of the photosynthetic apparatus in Rhodobacter sphaeroides — II. A study of PufX(-) membranes. J Biol Chem 280: 11214–11223 Conroy MJ, Westerhuis WH, Parkes-Loach PS, Loach PA, Hunter CN and Williamson MP (2000) The solution structure of Rhodobacter sphaeroides LH1β reveals two helical domains separated by a more flexible region: structural consequences for the LH1 complex. J Mol Biol 298: 83–94 Davis CM, Bustamante PL, Todd JB, Parkes-Loach PS, McGlynn P, Olsen JD, McMaster L, Hunter CN and Loach PA (1997) Evaluation of structure-function relationships in the core lightharvesting complex of photosynthetic bacteria by reconstitution with mutant polypeptides. Biochemistry 36: 3671–3679 Deisenhofer J, Epp O, Miki K, Huber R and Michel H (1985) Structure of the protein subunits in the photosynthetic reaction centre of Rhodopseudomonas viridis at 3 Å resolution. Nature 318: 618–624 Dezi M, Francia F, Mallardi A, Colafemmina G, Palazzo G and Venturoli G (2007) Stabilization of charge separation and cardiolipin confinement in antenna reaction center complexes purfied from Rhodobacter sphaeroides. Biochim Biophys Acta 1767: 1041–1056 Drews G (1996) Formation of the light-harvesting complex I (B870) of anoxygenic phototrophic purple bacteria. Arch Microbiol 166: 151–159 Drews G and Niederman RA (2002) Membrane biogenesis in anoxygenic photosynthetic prokaryotes. Photosynth Res 73: 87–94 Eldridge MD, Madden PA and Frenkel D (1993) Entropy-driven formation of a superlattice in hard-sphere binary mixture. Nature. 365: 35–37 Engelhardt H, Engel A and Baumeister W (1986) Stoichiometric model of the photosynthetic unit of Ectothiorhodospira halochloris. Proc Natl Acad Sci USA 83: 8972–8976 Farchaus JW, Gruenberg H and Oesterhelt D (1990) Complementation of a reaction center-deficient Rhodobacter sphaeroides pufLMX deletion strain in trans with pufBALM does not restore the photosynthesis-positive phenotype. J Bacteriol 172: 977–985 Feher G (1971) Some chemical and physical properties of a bacterial reaction center particle and its primary photochemical reactants. Photochem Photobiol 14: 373–387 Fiedor L and Scheer H (2005) Trapping of an assembly intermediate of photosynthetic LH1 antenna beyond B820 subunit: Significance for the assembly of photosynthetic LH1 antenna. J Biol Chem 280: 20921–20926 Fotiadis D, Qian P, Philippsen A, Bullough PA, Engel A and Hunter
Chapter 9
Core Complexes of Purple Bacteria
CN (2004) Structural analysis of the RC-LH1 photosynthetic core complex of Rhodospirillum rubrum using atomic force microscopy. J Biol Chem 279: 2063–2068 Francia F, Wang J, Venturoli G, Melandri BA, Barz WP and Oesterhelt D (1999) The reaction center-LH1 antenna complex of Rhodobacter sphaeroides contains one PufX molecule which is involved in dimerization of this complex. Biochemistry 38: 6834–6845 Francia F, Wang J, Zischka H, Venturoli G and Oesterhelt D (2002) Role of the N- and C-terminal regions of the PufX protein in the structural organization of the photosynthetic core complex of Rhodobacter sphaeroides. Eur J Biochem 269: 1877–1885 Francia F, Dezi M, Rebecchi A, Mallardi A, Palazzo P, Melandri BA and Venturoli G (2004) Light-harvesting complex 1 stabilizes P+QB– charge separation in reaction centres of Rhodobacter sphaeroides. Biochemistry 43: 14199–14210 Frese RN, Olsen JD, Branvall R, Westerhuis WHJ, Hunter CN and van Grondelle R (2000) The long-range supraorganization of the bacterial photosynthetic unit: A key role for PufX. Proc Natl Acad Sci USA 97: 5197–5202 Frese R, Siebert CA, Niederman RA, Hunter CN, Otto C and van Grondelle R (2004) The long-range organization of a native photosynthetic membrane. Proc Natl Acad Sci USA 101: 17994–17999 Frese RN, Pàmies JC, Olson JD, Bahatyrova S, van der Weij-de Wit CD, Aartsma TJ, Otto C, Hunter CN, Frenkel D and van Grondelle R (2008) Protein shape and crowding drive domain formation and curvature in biological membranes. Biophys J 19: 640–647 Fulcher TK, Beatty JT and Jones MR (1998) Demonstration of the key role played by the PufX protein in the functional and structural organization of native and hybrid bacterial photosynthetic core complexes. J Bacteriol 180: 642–646 Germeroth L, Lottspeich F, Robert B and Michel H (1993) Unexpected similarities of the B800-B850 light-harvesting complex of Rhodospirillum molischianum to the B870 light-harvesting complexes from other purple photosynthetic bacteria. Biochemistry 32: 5615–5621 Geyer T and Helms V (2006a) A spatial model of the chromatophore vesicles of Rhodobacter sphaeroides and the position of the cytochrome bc1 complex. Biophys J 91: 921–926. Geyer T and Helms V (2006b) Reconstruction of a kinetic model of the chromatophore vesicles from Rhodobacter sphaeroides. Biophys J 91: 927–937. Ghosh R, Hoenger A, Hardmeyer A, Mihailescu D, Bachofen R, Engel A and Rosenbusch JP (1993) Two-dimensional crystallization of the light-harvesting complex from Rhodospirillum rubrum. J Mol Biol 231: 501–504 Ghosh R, Kessi J, Hauser H, Wehrli E and Bachofen R (1990) Quarternary structure of the B875 light-harvesting complex from Rhodopseudomonas rubrum G9+. FEMS Microbiol Lett 53: 245–252 Gonçalves RP, Bernadac A, Sturgis JN and Scheuring S (2005) Architecture of the native photosynthetic apparatus of Phaeospirillum molischianum. J Struct Biol 152: 221–228 Gubellini F, Francia F, Busselez J, Venturoli G and Levy D (2006) Functional and structural analysis of the photosynthetic apparatus of Rhodobacter veldkampii. Biochemistry 45: 10512–10520 Hu X and Schulten K (1998) Model for the light-harvesting complex I (B875) of Rhodobacter sphaeroides. Biophys J
177 75: 683–694 Hucke O, Schiltz E, Drews G and Labahn A (2003) Sequence analysis reveals new membrane anchor of reaction centrebound cytochromes possibly related to PufX. FEBS Lett 535: 166–170 Hunter CN, Pennoyer JD, Sturgis JN, Farrelly D and Niederman RA (1988) Oligomerisation states and associations of light-harvesting pigment-protein complexes of Rhodobacter sphaeroides as analysed by lithium dodecyl sulphate-polyacrylamide gel electrophoresis. Biochemistry 27: 3459—3467 Hunter CN, van Grondelle R and Olsen JD (1989) Photosynthetic antenna proteins: 100 ps before photochemistry starts. Trends Biochem Sci 14: 72–76 Jamieson SJ, Wang P, Qian P, Kirkland JY, Conroy MJ, Hunter CN and Bullough PA (2002) Projection structure of the photosynthetic reaction centre-antenna complex of Rhodospirillum rubrum at 8.5 Å resolution. EMBO J 21: 3927–3935 Jones MR (2007) Lipies in photosynthetic reaction centres: Structural roles and functional holes. Prog Lipid Res 46: 56–87 Jungas C, Ranck JL, Rigaud JL, Joliot P and Verméglio A (1999) Supramolecular organization of the photosynthetic apparatus of Rhodobacter sphaeroides. EMBO J 18: 534–542 Karrasch S, Bullough PA and Ghosh R (1995) The 8.5 Å projection map of the light-harvesting complex I from Rhodospirillum rubrum reveals a ring composed of 16 subunits. EMBO J 14: 631–638 Kiley PJ, Varga A and Kaplan S (1988) Physiological and structural analysis of light-harvesting mutants of Rhodobacter sphaeroides. J Bacteriol 170: 1103–1115 Klug G and Cohen SN (1988) Pleiotropic effects of localized Rhodobacter capsulatus puf operon deletions on production of light-absorbing pigment-protein complexes. J Bacteriol 170: 5814–5821 Koepke J, Hu XC, Muenke C, Schulten K and Michel H (1996) The crystal structure of the light-harvesting complex II (B800-B850) from Rhodospirillum molischanum. Structure 4: 581–597 Law CJ, Chen J, Parkes-Loach PS, Loach PA (2003) Interaction of bacteriochlorophyll with the LH1 and PufX polypeptides of photosynthetic bacteria: Use of chemically synthesized analogs and covalently attached fluorescent probes. Photosynth Res 75: 193–210 Lilburn TG and Beatty JT (1992) Suppressor mutants of the photosynthetically incompetent pufX deletion mutant Rhodobacter capsulatus ∆RC6(pTL2). FEMS Microbiol Lett 100: 155–159 Lilburn TG, Haith CE, Prince RC and Beatty JT (1992) Pleiotropic effects of pufX gene deletion on the structure and function of the photosynthetic apparatus of Rhodobacter capsulatus. Biochim Biophys Acta 1100: 160–170 Lilburn TG, Prince RC and Beatty JT (1995) Mutation of the Ser2 codon of the light-harvesting B870 polypeptide of Rhodobacter capsulatus partially suppresses the pufX phenotype. J Bacteriol 177: 4593–4600 Lupo D, Ghosh R (2004) The reaction center H subunit is not required for high levels of light-harvesting complex 1 in Rhodospirillum rubrum mutants. J Bacteriol 186: 5585–5595 MacKenzie KR, Prestegard J and Engelman DM (1997) A transmembrane helix dimer: Structure and implications. Science 276: 131–133 McDermott G, Prince SM, Freer AA, Hawthornthwaite-Lawless AM, Papiz MZ, Cogdell RJ and Isaacs NW (1995) Crystal
178 structure of an integral membrane light-harvesting complex from photosynthetic bacteria. Nature 374: 517–521 McGlynn P, Hunter CN and Jones MR (1994) The Rhodobacter sphaeroides PufX protein is not required for photosynthetic competence in the absence of a light-harvesting system. FEBS Lett 349: 349–353 McGlynn P, Westerhuis WH, Jones MR and Hunter CN (1996) Consequences for the organization of reaction centre-lightharvesting antenna 1 (LH1) core complexes of Rhodobacter sphaeroides arising form deletion of amino acid residues at the C terminus of the LH1 α polypeptide. J Biol Chem 271: 3285–3292 Meckenstock RU, Brunisholz RA and Zuber H (1992) The light-harvesting core-complex and the B820-subunit from Rhodopseudomonas marina. 1. Purification and characterization. FEBS Lett 311: 128–134 Meryandini M, Drews G (1996) Import and assembly of the α and β-polypeptides of the light-harvesting complex I (B870) in the membrane system of Rhodobacter capsulatus investigated in an in vitro translation system. Photosynth Res 47: 21–31 Miller JF, Hinchigeri SB, Parkes-Loach PS, Callahan PM, Sprinkle JR, Riccobono JR and Loach PA (1987) Isolation and characterization of a subunit form of the light-harvesting complex of Rhodospirillum rubrum. Biochemistry 26: 5055–5062 Miller KR (1982) Three-dimensional structure of a photosynthetic membrane. Nature 300: 53–55 Niederman RA, Mallon DE and Langan JJ (1976) Membranes of Rhodopseudomonas sphaeroides. IV. Assembly of chromatophores in low-aeration cell suspensions. Biochim Biophys Acta 440: 429–447 Olsen JD and Hunter CN (1994) Protein structure modeling of the bacterial light-harvesting complex. Photochem Photobiol 60: 521–535 Olsen JD, Sockalingum GD, Robert B and Hunter CN (1994) Modification of a hydrogen bond to a bacteriochlorophyll a molecule in the light-harvesting 1 antenna of Rhodobacter sphaeroides. Proc Natl Acad Sci USA 91: 7124–7128 Olsen JD, Sturgis JN, Westerhuis WH, Fowler GJS, Hunter CN and Robert B (1997) Site-directed modification of the ligands to the bacteriochlorophylls of the light-harvesting LH1 and LH2 complexes of Rhodobacter sphaeroides. Biochemistry 36: 12625–12632 Papiz MZ, Prince SM, Hawthornthwaite-Lawless AM, McDermott G, Freer AA, Isaacs NW and Cogdell RJ (1996) A model for the photosynthetic apparatus of purple bacteria. Trends Plant Sci 1: 198–206 Parkes-Loach PS, Sprinkle JR and Loach PA (1988) Reconstitution of the B873 light-harvesting complex of Rhodospirillum rubrum from the separately isolated α- and β-polypeptides and bacteriochlorophyll a. Biochemistry 27: 2718–2727 Parkes-Loach PS, Law CJ, Recchia PA, Kehoe J, Nehrlich S, Chen J and Loach PA (2001) Role of the core region of the PufX protein in inhibition of reconstitution of the core light-harvesting complexes of Rhodobacter sphaeroides and Rhodobacter capsulatus. Biochemistry 40: 5593–5601 Pervushin KV, Orekhov VY, Popov AI, Musina LY and Arseniev AS (1994) Three-dimensional structure of (1-71)bacterioopsin solubilized in methanol/chloroform and SDS micelles determined by 15N-1H heteronuclear NMR spectroscopy. FEBS J 219: 571–583 Picorel R, Belanger G and Gingras G (1983) Antenna holochrome-
Per A. Bullough , Pu Qian and C. Neil Hunter B880 of Rhodospirillum rubrum S1-pigment, phospholipid, and polypeptide composition. Biochemistry 22: 2491–2497 Pucheu NL, Kerber NL, Pardo P, Brand M, Drews G and Garcia AF (1996) Bioenergetic factors controlling in vitro phosphorylation of LHIα (B870) polypeptides in membranes isolated from Rhodobacter capsulatus. Arch of Microbiol 165: 119–125 Pugh RJ, McGlynn P, Jones MR and Hunter CN (1998) The LH1RC core complex of Rhodobacter sphaeroides: Interaction between components, time-dependent assembly, and topology of the PufX protein. Biochim Biophys Acta 1366: 301–316 Qian P, Addlesee HA, Ruban AV, Wang P, Bullough PA and Hunter CN (2003) A reaction center-light-harvesting 1 complex (RC-LH1) from a Rhodospirillum rubrum mutant with altered esterifying pigments. J Biol Chem 278: 23678–23685 Qian P, Hunter CN and Bullough PA (2005) The 8.5 Å projection structure of the core RC-LH1-PufX dimer of Rhodobacter sphaeroides. J Mol Biol 349: 948–960 Qian P, Bullough PA and Hunter CN (2008) 3-D reconstruction of a membrane-bending complex: The RC-LH1-PufX core dimer of Rhodobacter sphaeroides. J Biol Chem 283: 14002-14011 Recchia PA, Davis CM, Lilburn TG, Beatty JT, Parkes-Loach PS, Hunter CN and Loach PA (1998) Isolation of the PufX protein from Rhodobacter capsulatus and Rhodobacter sphaeroides: Evidence for its interaction with the α-polypeptide of the core light-harvesting complex. Biochemistry 37: 11055–11063 Richter MF, Baier J, Prem T , Oellerich S, Francia F, Venuroli G, Oesterhelt D, Southall J, Cogdell RJ and Köhler J (2007a) Symmetry matters for the electronic structure of core complexes from Rhodopseudomonas palustris and Rhodobacter sphaeroides PufX–. Proc Natl Acad Sci USA 104: 6661–6665 Richter MF, Baier J, Southall J, Cogdell RJ, Oellerich S and Köhler J (2007b) Refinement of the X-ray structure of the RC-LH1 core complex from Rhodopseudomonas palustris by single-molecule spectroscopy. Proc Natl Acad Sci USA 104: 20280–20284 Richter P, Cortez N and Drews G (1991) Possible role of the highly conserved amino-acids Trp-8 and Pro-13 in the N-terminal segment of the pigment-binding polypeptide LH1α of Rhodobacter-capsulatus. FEBS Lett 285: 80–84 Roszak AW, Howard TD, Southall J, Gardiner AT, Law CJ, Isaacs NW and Cogdell RJ (2003) Crystal structure of the RC-LH1 core complex from Rhodopseudomonas palustris. Science 302: 1969–1972 Russ WP and Engelman DM (1999) TOXCAT: A measure of transmembrane helix association in a biological membrane. Proc Natl Acad Sci USA 96: 863–868 Russ WP and Engelman DM (2000) The GxxxG motif: A framework for transmembrane helix-helix association. J Mol Biol 296: 911–919 Scheuring S (2006) AFM studies of the supramolecular assembly of bacterial photosynthetic core-complexes. Curr Opin in Chem Biol 10: 387–393 Scheuring S and Sturgis JN (2005) Chromatic adaptation of photosynthetic membranes. Science 309: 484—487 Scheuring S, Seguin J, Marco S, Levy D, Robert B and Rigaud JL (2003) Nanodissection and high-resolution imaging of the Rhodopseudomonas viridis photosynthetic core complex in native membranes by AFM. Proc Natl Acad Sci USA 100: 1690–1693 Scheuring S, Francia F, Busselez J, Melandri BA, Rigaud JL and
Chapter 9
Core Complexes of Purple Bacteria
Levy D (2004) Structural role of PufX in the dimerization of the photosynthetic core-complex of Rhodobacter sphaeroides. J Biol Chem 279: 3620–3626 Scheuring S, Busselez J, Levy D (2005) Structure of the dimeric PufX-containing core complex of Rhodobacter blasticus by in situ atomic force microscopy. J Biol Chem 280: 1426–1431 Scheuring S, Gonçalves RP, Prima V, Sturgis JN (2006) The photosynthetic apparatus of Rhodopseudomonas palustris: Structures and organization. J Mol Biol 358: 83–96 Şener MK, Olsen JD, Hunter CN and Schulten K (2007) Atomic level structural and functional model of a bacterial photosynthetic membrane vesicle (2007) Proc Natl Acad Sci USA 104: 15273–15278. Siebert CA, Qian P, Fotiadis D, Engel A, Hunter CN and Bullough PA (2004) Molecular architecture of photosynthetic membranes in Rhodobacter sphaeroides: The role of PufX. EMBO J 23: 690–700 Sockett RE, Donohue TJ, Varga AR and Kaplan S (1989) Control of photosynthetic membrane assembly in Rhodobacter sphaeroides mediated by puhA and flanking sequences. J Bacteriol 171: 436–446 Sorgen PL, Cahill SM, Krueger-Koplin RD, Krueger-Koplin ST, Schenck CC and Girvin ME (2002) Structure of the Rhodobacter sphaeroides light-harvesting 1 beta subunit in detergent micelles. Biochemistry 41: 31–41 Stahlberg H, Dubochet J, Vogel H and Ghosh R (1998) Are the light-harvesting I complexes from Rhodospirillum rubrum arranged around the reaction centre in a square geometry? J Mol Biol 282: 819–831 Stark W, Kühlbrandt W, Wildhaber I, Wehrli E and Muhlethaler K (1984) The structure of the photoreceptor unit of Rhodopseudomonas viridis. EMBO J 3: 777–783 Stiehle H, Cortez N, Klug G and Drews G (1990) A negatively charged N-terminus in the α-polypeptide inhibits formation of light-harvesting complex-I in Rhodobacter capsulatus. J Bacteriol 172: 7131–7137 Sturgis JN, Olsen JD, Robert B and Hunter CN (1997) Functions of conserved tryptophan residues of the core light-harvesting complex of Rhodobacter sphaeroides. Biochemistry 36: 2772–2778 Tehrani A and Beatty JT (2004) Effects of precise deletions in Rhodobacter sphaeroides reaction center genes on steady-state levels of reaction center proteins: A revised model for reaction center assembly. Photosynth Res 79: 101–108 Tehrani A, Prince RC and Beatty JT (2003) Effects of photosynthetic reaction center H protein domain mutations on photosynthetic properties and reaction center assembly in Rhodobacter sphaeroides. Biochemistry 42: 8919–8928 Tunnicliffe RB, Ratcliffe EC, Hunter CN and Williamson MP (2006) The solution structure of the PufX polypeptide from Rhodobacter sphaeroides. FEBS Lett 580: 6967–6971 Varga AR and Kaplan S (1993) Synthesis and stability of reaction center polypeptides and implications for reaction center assembly in Rhodobacter sphaeroides. J Biol Chem 268: 19842–19850
179 Walz T and Ghosh R (1997) Two-dimensional crystallization of the light-harvesting I reaction centre photounit from Rhodospirillum rubrum. J Mol Biol 265: 107–111 Walz T, Jamieson SJ, Bowers CM, Bullough PA and Hunter CN (1998) Projection structures of three photosynthetic complexes from Rhodobacter sphaeroides: LH2 at 6 Å, LH1 and RC-LH1 at 25 Å. J Mol Biol 282: 833–845 Wang ZY, Gokan K, Kobayashi M and Nozawa T (2005) Solution structures of the core light-harvesting α and β polypeptides from Rhodospirillum rubrum: Implications for the pigment-protein and protein-protein interactions. J Mol Biol 347: 465–477 Welte W and Kreutz W (1982) Formation, structure and composition of a planar hexagonal lattice composed of specific proteinlipid complexes in the thylakoid membranes of Rhodopseudomonas viridis. Biochim Biophys Acta 692: 479–488 Westerhuis WHJ, Hunter CN, van Grondelle R and Niederman RA (1999) Modeling of oligomeric-state dependent spectral heterogeneity in the B875 light-harvesting complex of Rhodobacter sphaeroides by numerical simulation. J Phys Chem B 103: 7733–7742 Westerhuis WHJ, Sturgis JN, Ratcliffe EC, Hunter CN and Niederman RA (2002) Isolation, size estimates, and spectral heterogeneity of an oligomeric series of light-harvesting 1 complexes from Rhodobacter sphaeroides. Biochemistry 41: 8698–8707 Wong DH, Collins WJ, Harmer A, Lilburn TG and Beatty JT (1996) Directed mutagenesis of the Rhodobacter capsulatus puhA gene and Orf 214: Pleiotropic effects on photosynthetic reaction center and light-harvesting 1 complexes. J Bacteriol 178: 2334–2342 Young CS and Beatty JT (1998) Topological model of the Rhodobacter capsulatus light-harvesting complex I assembly protein LhaA (previously known as ORF1696). J Bacteriol 180: 4742–4745 Young CS, Reyes RC and Beatty JT (1998) Genetic complementation and kinetic analyses of Rhodobacter capsulatus ORF1696 mutants indicate that the ORF1696 protein enhances assembly of the light-harvesting I complex. J Bacteriol 180: 1759–1765 Youvan DC, Elder JT, Sandlin DE, Zsebo K, Alder DP, Panopoulos NJ, Marrs BL and Hearst JE (1982) R-prime site-directed transposon Tn7 mutagenesis of the photosynthetic apparatus in Rhodopseudomonas capsulata. J Mol Biol 162: 17–41 Zsebo KM and Hearst JE (1984) Genetic-physical mapping of a photosynthetic gene cluster from Rhodopseudomonas capsulata. Cell 37: 937–947 Zuber H (1985) Structure and function of light-harvesting complexes and their polypeptides. Photochem Photobiol 42: 821–844 Zuber H and Cogdell RJ (1995) Structure and organization of purple bacterial antenna complexes. In: Blankenship RE, Madigan, MT, Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 315–348. Kluwer Academic Publishers
Chapter 10 Structure-Function Relationships in Bacterial LightHarvesting Complexes Investigated by Reconstitution Techniques Paul A. Loach* and Pamela S. Parkes-Loach Department of Biochemistry, Molecular Biology and Cell Biology, Northwestern University, 2205 Tech Drive, Evanston, IL 60208-3500, U.S.A.
Summary ............................................................................................................................................................... 181 I. Introduction..................................................................................................................................................... 182 II. Reversible Dissociation of Core Light-Harvesting 1 Complexes to a Subunit Form (B820)........................... 182 III. Reversible Dissociation of B820 to its Fundamental Components................................................................. 183 IV. Cofactor Requirements .................................................................................................................................. 184 V. Cofactor – Protein Interactions ....................................................................................................................... 184 A. Mg Coordination of Bacteriochlorophyll............................................................................................ 184 B. Bacteriochlorophyll – Protein Hydrogen Bonds ................................................................................ 186 C. Steric Requirements for Bacteriochlorophyll Binding ....................................................................... 187 VI. Cofactor – Cofactor Interactions ..................................................................................................................... 188 VII. Protein – Protein Interactions .......................................................................................................................... 188 A. B820 ................................................................................................................................................. 188 B. Light-Harvesting 1 and 2 Complexes .............................................................................................. 190 VIII. Effect of PufX on Reconstitution of Light-Harvesting 1 Complexes ............................................................... 191 IX. In vitro versus In vivo Assembly of Complexes .............................................................................................. 192 A. Oligomerization of B820 ................................................................................................................... 192 B. Reconstitution of B820 and Reaction Center Complexes to Form a Photoreceptor Complex ......... 192 C. Role of Carotenoids ......................................................................................................................... 193 X. Reconstitution of the Reaction Center............................................................................................................ 193 Acknowledgments ................................................................................................................................................. 195 References ............................................................................................................................................................ 195
Summary Exploration of conditions to achieve the reversible dissociation of the core light-harvesting complexes (LH1) of photosynthetic bacteria led to the isolation of the fundamental subunit complex, B820. Further reversible dissociation of B820 provided conditions for reconstitution of both this subunit complex and LH1 from separately isolated polypeptides and bacteriochlorophyll (BChl). Native-like LH1 complexes have also been reconstituted using BChl, isolated polypeptides from Rhodospirillum (Rsp.) rubrum or Rhodobacter (Rba.) sphaeroides and carotenoid. The reconstitution methodology has also been applied to the peripheral light-harvesting complex (LH2) of Phaeospirillum (Phs.) molischianum in which it was demonstrated that LH2 contained the same B820-type subunit structure found in LH1. Since the interaction between B820 and its individual components could be studied under equilibrium condi*Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 181–198. © 2009 Springer Science + Business Media B.V.
182
Paul A. Loach and Pamela S. Parkes-Loach
tions, many structure-function questions could be addressed as well as thermodynamic parameters established. The subunit structure was shown to consist of α1β1•2BChl with overlap at rings III and V of the BChls. Minimal requirements for B820 formation include (1) a polypeptide of at least 26 amino acids, 18 of which constitute an α-helix with hydrophobic side chains, (2) a His residue for coordination and hydrogen bonding to BChl and (3) a Trp residue for hydrogen bonding to the C31 carbonyl of BChl. The His interaction accounted for over half of the stabilization energy of the B820 complex. Hydrogen bonds involving the Trp residues provide a second major stabilization of approximately 3.5 kcal/Trp. Finally, some stabilization is provided by specific interactions between the α and β polypeptides in their N-terminal regions, most likely exhibiting the same structural motif as found in the crystal structure of Phs. molischianum LH2. Reconstitution methodology has also been used to study formation of the reaction center-LH1 (RC-LH1) core complex, which exhibited physical properties analogous to the native complex, and to study the interaction and function of PufX. Currently, the isolation and reconstitution methodology is being applied to reconstitute the RC from its fundamental components. I. Introduction In order to thoroughly understand the structurefunction relationships of supramolecular systems, it is immensely useful if the system can be reversibly dissociated into sub-complexes. Even more powerful in probing how structure enables function is the ability to reconstitute a supramolecular system from its fundamental components. Because of limited solubility, development of methodology for reconstitution of supramolecular complexes that reside in membranes has been especially challenging. This chapter reviews the successful reversible dissociation and reconstitution of light-harvesting (LH) and photoreceptor complexes (PRC) of photosynthetic bacteria. First, the core (LH1) and secondary (LH2) light-harvesting complexes will be discussed, then reconstitution of PRC from the LH1 and RC complexes and finally progress on reconstitution of the RC from its fundamental components. An emphasis is placed on the wealth of information and insights that can be obtained using reconstitution methodology. Abbreviations: BChl – bacteriochlorophyll; BChl a is implied unless BChl b is indicated; CD – circular dichroism; EM – electron microscopy; HPLC – high performance liquid chromatography; KA – association constant for binding; KD – dissociation constant for binding; LH1 – core light-harvesting complex, also called B875, B890, etc. depending on the far-red absorption maximum; LH2 – peripheral light-harvesting complex, also called B800-850, for example; NMR – nuclear magnetic resonance; PAGE – polyacrylamide gel electrophoresis; Phs. – Phaeospirillum; PRC – photoreceptor complex which contains LH1 and RC; QB – quinone on the B-side of the reaction center; Rba. – Rhodobacter; RC – reaction center; Rps. – Rhodopseudomonas; RR – resonance Raman; Rsp. – Rhodospirillum; SDS – sodium dodecyl sulfate; UQ – ubiquinone; β-OG – n-octyl β-D-glucopyranoside; λmax – wavelength of maximum absorption
Ultimately, complete understanding of how supramolecular complexes work can only be achieved by putting the system together from its simplest parts. II. Reversible Dissociation of Core LightHarvesting 1 Complexes to a Subunit Form (B820) Since LH1 of photosynthetic bacteria is a complex containing 12 to 16 molecules each of α and β polypeptides and 24 to 32 molecules of bacteriochlorophyll (BChl) (Loach and Sekura, 1968; Francke and Amesz, 1995; Loach and Parkes-Loach, 1995; Chapter 9, Bullough et al.), it seemed likely that this structure was formed from oligomerization of a subunit complex whose structure would consist of αnβn•2nBChl where n is a small number (Loach and Sekura, 1968; Miller et al., 1987; Francke and Amesz, 1995; Loach and Parkes-Loach, 1995). Early attempts to use detergents to reversibly dissociate LH1 were largely unsuccessful until carotenoid components were first removed either by extraction or by studying mutants in which carotenoid biosynthesis was blocked (Miller et al., 1987; Loach and Parkes-Loach, 1995). In addition, the availability of new types of detergent (e.g., n-octyl β-D-glucopyranoside (β-OG), lauryl maltoside) that were more gentle than lauryl dimethyl amine oxide, Triton X-100 and sodium dodecyl sulfate (SDS), but more effective than sodium cholate and deoxycholate, enabled the preparation of subunit complexes that could be re-associated to form LH1 (Miller et al., 1987). Careful control of dissociating conditions enabled a quantitative yield of B820 from Rsp. rubrum LH1 which could also be quantitatively re-associated to re-form LH1.
Chapter 10
Light-Harvesting Reconstitution Studies
The B820 complex has been extensively studied. It was initially prepared from LH1 complexes of Rsp. rubrum (Miller et al., 1987), Rba. sphaeroides (Chang et al., 1990b) and Rba. capsulatus (Heller and Loach, 1990). Subsequently, it has also been prepared from LH1 complexes of Rhodopseudomonas (Rps.) viridis (Parkes-Loach et al., 1994), Rps. marina (Meckenstock et al., 1992), Rhodocyclus gelatinosus (Jirsakova and Reiss-Husson, 1993) and Chromatium purpuratum (Kerfeld et al., 1994) and the LH2 complex of Phaeospirillum (Phs.) molischianum (Todd et al., 1998). It appears to be the fundamental subunit of light-harvesting complexes of photosynthetic bacteria. B820 has been shown to have a composition of α1β1•2BChl. The two BChls strongly interact at van der Waals distance as determined by absorption (Miller et al., 1987; Chang et al., 1990a; ParkesLoach et al., 1995), circular dichroism (CD) (Miller et al., 1987; Chang et al., 1990a; Parkes-Loach et al., 1995) fluorescence (Chang et al., 1990a), nuclear magnetic resonance (NMR) (Wang et al., 2002), electron paramagnetic resonance (Srivatsan and Norris, 2001), singlet-singlet and singlet-triplet annihilation (van Mourik et al., 1991) spectroscopies. Absorption spectra indicate that one or more intermediates are formed during dissociation and reassociation of LH1 (Miller et al., 1987; van Mourik et al., 1992; Loach and Parkes-Loach, 1995; Pandit et al., 2001, 2003). In the presence of carotenoid, B820 associates to form an LH1-type complex with at least one intermediate species (Davis et al., 1995; Fiedor and Scheer, 2005; Chapter 46, Braun and Fiedor). B820 can also be re-associated in the presence of RCs and phospholipid to form an RC-LH1 complex displaying many of the characteristics of the native PRC (Bustamante and Loach, 1994). III. Reversible Dissociation of B820 to its Fundamental Components Under appropriate conditions of concentration, temperature, pH and detergent concentration, B820 can be reversibly dissociated to its fundamental components and the separately-isolated polypeptides and BChl can be used to reconstitute B820 and LH1 (Fig. 1) (Parkes-Loach et al., 1988, 1994; Chang et al., 1990b; Heller and Loach, 1990). The following equilibria have been demonstrated (Sturgis and Robert, 1994; Loach and Parkes-Loach, 1995; Law et al., 2003).
183
Fig. 1. Reconstitution of B820 and LH1 using the native α- and β-polypeptides of Rba. sphaeroides and BChl. Spectra were recorded at 0.90% β-OG (dashed curve), 0.75% β-OG (diamond symbols), 0.66% β-OG after chilling on ice for 1 hour (solid black curve) and 0.66% β-OG after chilling overnight (grey curve). Concentrations of reactants were: 3.1 µM α-polypeptide, 2.2 µM β-polypeptide, and 1.1 µM BChl, at 0.66% β-OG. Spectra were recorded in 1 cm cuvettes and the data at 0.90% and 0.75% β-OG were multiplied by appropriate factors to compare each spectrum at the polypeptide concentrations of 0.66% β-OG.
α + BChl
α•BChl
(1)
β + BChl
β•BChl
(2)
α•BChl + β•BChl
B820
(3)
Under conditions where KD for eq. (3) is much smaller than the KDs for eqs. (1) and (2), the equilibrium measured at low concentrations (Law et al., 2003) may be expressed by: α + β + 2BChl ↔ B820
(4)
which is the result of addition of eqs. (1), (2) and (3). The KD for eq. (4) equals the product of the KDs for eqs. (1), (2) and (3). The equations for these equilibria are not complete in that they ignore the role of interacting detergent molecules. BChl and each polypeptide, as well as each complex formed, are ‘solvated’ by detergent to form mixed micelles. Thus, as each complex is formed, ‘bound’ detergent is released as the hydrophobic surface area of the complex is less than the sum of those for the separately interacting species. Because of this, the β-OG concentration dependence of KD for the reactions of equations (1) to (4) can be different and the relationship between the KDs may change with detergent concentration. It would
184
Paul A. Loach and Pamela S. Parkes-Loach
be expected that the KD for reaction (3) would have a greater dependence on detergent concentration than those for reactions (1) and (2) because a larger number of β-OG molecules should be released as the B820 complex forms. It should also be noted that the number of bound detergent molecules will depend on the hydrophobic surface of each polypeptide and is therefore dependent on the amino acid sequence and secondary structure. IV. Cofactor Requirements Even though the effects of detergent would be expected to vary somewhat with variations in the structure of the chromophore and protein, this contribution to a change in KD can probably be neglected relative to the effect of small but important changes in structure. Thus, measurements of KD for a series of BChl analogs at constant β-OG concentration can be assumed to reflect the effects of structural change of the analog on the complex irrespective of small changes in βOG interaction. The first measurements of this type examined analogs of BChl with LH1 polypeptides from two different organisms, Rsp. rubrum and Rba. sphaeroides (Parkes-Loach et al., 1990; Davis et al., 1996). These systems showed the following common requirements for formation of the subunit complex and LH1: (1) Mg or a metal of similar size and coordination chemistry (e.g., Zn, Cd, Ni), (2) a bacteriochlorin oxidation state of the macrocyclic ring, (3) a 132 carbomethoxy group (see Fig. 2 for numbering of BChl), and (4) an intact ring V. Some structural features were not as critically important. For example, the subunit complex and LH1 could be formed with both sets of polypeptides and BChl b instead of BChl a, as well as with analogs of BChl a containing either short (ethanol) or long (phytol) esterifying alcohols. Two derivatives were identified that behave differently with the two sets of polypeptides. The 3-acetyl group is required to form LH1 in both bacteria, although a subunit-type complex was readily formed with [3-vinyl] BChl a and the polypeptides of Rsp. rubrum but formed only slightly under special conditions with polypeptides of Rba. sphaeroides. [132OH] BChl ap formed both subunit- and LH1-type complexes with the α and β polypeptides of Rba. sphaeroides but not with those of Rsp. rubrum. Thus, some subtle differences in the BChl binding sites exist in the LH1 complexes of these two bacteria. The greater stability of the B820-type
Fig. 2. Structure of BChl.
complex with the Zn analog of BChl has been used to advantage in many biomimetic studies (Kashiada et al., 2000a,b; Wendling et al., 2002; Nagata et al., 2003a,b; Dewa et al., 2005). V. Cofactor – Protein Interactions A. Mg Coordination of Bacteriochlorophyll From the amino acid sequences of the α and β polypeptides of LH1 and LH2 complexes (Fig. 3), it was predicted that these proteins would have one transmembrane α-helical span and that a conserved His located within this spanning segment would provide the ligand coordinated to the Mg atom of BChl (Zuber and Brunisholz, 1991). Consistent with the importance of this role for His, mutation of this residue to other amino acids resulted in complete loss, or a severely limited expression, of LH1 (Bylina et al., 1988; Olsen, 1994; Davis et al., 1997; Olsen et al., 1997). Further confirmation of this role for His came from the early application of resonance Raman (RR) spectroscopy to studying preparations of LH1 and LH2. From the comparison with model complexes, it was concluded that His side chains may be the Mg ligands (Robert and Lutz, 1985). Also derived from early
Chapter 10
Light-Harvesting Reconstitution Studies
RR studies was the assignment that the coordination number of the central Mg was 5, indicating a squarepyramidal complex (Cotton and van Duyne, 1981). Final confirmation of Mg coordination in LH2 came from the determination of the crystal structure of the LH2 complex from Rps. acidophila (McDermott et al., 1995; Prince et al., 1997) and Phs. molischianum (Koepke et al., 1996). To evaluate the importance of this His residue, mutants of Rba. sphaeroides were prepared in which His0 was changed to Asn, Tyr, or Leu (Olsen, 1994). By spectral examination of the whole cells, only the mutant cells containing βHis0→ Asn exhibited a slight LH1 complex absorbance. A β polypeptide was isolated in very low yield from the βHis0→Asn mutant; when this polypeptide was tested in the reconstitution assay, formation of a sub-
185
unit-type complex could not be observed under our assay conditions (Davis et al., 1997). The inability to demonstrate formation of a subunit-type complex means that the stabilization of this complex by His coordination by the β polypeptide is at least 4.5 kcal mol–1. From model studies, imidazole coordination to BChl to form a 5-coordinate complex is expected to occur with 4.5–5.5 kcal mol–1 binding energy (Cotton, 1976). In the case of the α-polypeptide isolated from the mutant αHis0→Asn, an LH1-type complex could not be formed. Because the conditions used to reconstitute an LH1-type complex employ lowering the β-OG concentration to below the critical micelle concentration, it is surprising that even this driving force is insufficient to overcome the absence of the His ligand.
Fig. 3. Amino acid sequences of the light-harvesting polypeptides of some photosynthetic bacteria. Sources: 1, 2, 4, 7, 8, 10, 11 (Zuber and Cogdell, 1995); 3, 9 (Youvan et al., 1984; Theiler et al., 1985); 5 (Visschers et al., 1995); 6 (Germeroth et al., 1993). For ease of comparison, the sequences have been aligned relative to the BChl-liganding His residue, labeled position 0. Asterisk: The first approximately 7 amino acids of the LH1 β and the last 2 amino acids of the LH1 α polypeptides of Phs. molischianum could not be definitely identified by sequencing. See Visschers et al. (1995).
Paul A. Loach and Pamela S. Parkes-Loach
186 B. Bacteriochlorophyll – Protein Hydrogen Bonds The results obtained from reconstitution experiments using analogs of BChl indicated that the C31 acetyl group and the C132 keto group were important for formation of B820 and LH1. Since these groups would be very good hydrogen bond acceptors, it was assumed that amino acid side chains, or possibly peptide backbone N-H groups, would be the hydrogen bond donors. Using site-directed mutants, Fowler et al. (1994) demonstrated that the tyrosine side chains of the α polypeptide at positions +13 and +14 (see Fig. 3 for numbering) on the α subunit interact with the acetyl carbonyl groups of BChl responsible for the 850-nm absorption in LH2 of Rba. sphaeroides. This was confirmed in the crystal structure of LH2 of Rps. acidophila (McDermott et al., 1995, Prince et al., 1997). Correspondingly, in the LH1 complex of Rba. sphaeroides, the acetyl carbonyls of BChl were shown to interact with Trp residues, one with αTrp+11 and one with βTrp+9 (Olsen et al., 1994; Sturgis et al., 1997) which is consistent with the crystal structure of LH2 from Phs. molischianum (Koepke et al., 1996) whose LH2 shows many common features to LH1 (Germeroth et al., 1993). Reconstitution experiments with α and β LH1
polypeptides isolated from Rba. sphaeroides modified by site-directed mutagenesis also demonstrated the importance of αTrp+11 and βTrp+9 of Rba. sphaeroides in stabilizing B820 as well as LH1 (Davis et al., 1997; Kehoe et al., 1998). Unique to these measurements was the ability to determine the magnitude of the energy of stabilization contributed by these hydrogen bonding interactions (see Table 1 and Kehoe et al., 1998). The role of hydrogen bonding to the C132 carbonyl group of BChl was indicated to also change somewhat in B820 compared with LH1 (Visschers et al., 1993) suggesting a change in hydrogen bonding to the protein. From the crystal structures of LH2 of Phs. molischianum (Koepke et al., 1996; Hu and Schulten, 1998) and mutational studies of Olsen et al. (1997), His0, in addition to providing the coordinated ligand to BChl, was indicated to form a hydrogen bond with the C132 carbonyl group of BChl coordinated to the partner polypeptide. Thus, a kind of cross-hydrogen bonding occurs which would help stabilize the subunit structure. By conducting reconstitution experiments with a chemically-synthesized β polypeptide in which 3-methylhistidine was substituted for His0, the magnitude of this stabilization energy was estimated to be quite significant for subunit formation (Parkes-Loach et al., 2004). In addition, reconstitution experiments with this analog established that B820 has the dimeric
Table 1. Association constants of reconstituted subunit complexes System Rsp. rubrum
Rba. sphaeroides
Polypeptide (µM at 0.75% β−OG) native β (2.0) + native α (1.7)
@0.90% β−OG 126
KA (M3 x 10–16) @0.75% β−OG ≤300
native β (3.7)
2
24
rrβ31 (4.9) (chemically synthesized)
12
120
native β (1.8) + native α (1.2)
38
≤300
native β (6.1)
23
200
enzyme-truncated sphβCP2 (4.9) + native α (2.4)
≤0.1
0.7
sphβ31 (5.5) (chemically synthesized)
40
≤300
sphβ37 (5.4) (chemically synthesized)
26
≤300
sphβ41 (5.6) (chemically synthesized) + native α (6.8)
19
≤300
sphβ41 (5.6) (chemically synthesized)
8.3
79
sphβ31W+6→F (5.5) (chemically synthesized)
1.5
22
sphβ31R+7→L (6.0) (chemically synthesized)
0.5
6
sphβ31W+9→F (14.2) (chemically synthesized)
≤0.1
0.5
Data and polypeptide nomenclature from Kehoe et al., 1998.
Chapter 10
Light-Harvesting Reconstitution Studies
187
Fig. 4. One quarter of the ring structure of Phs. molischianum LH2 (Koepke et al., 1996) showing BChl and the coordinated His side chains.
structure with π-overlap at rings III and V (Fig. 4) which involves cross-hydrogen bonding from His0. This conclusion is also consistent with the NMR experiments of Wang et al. (2002) in which overlap at rings III and V was observed. These experiments provide important information on the structure of B820 because it has not yet been possible to solve its structure by NMR or by crystallization of this subunit form. C. Steric Requirements for Bacteriochlorophyll Binding Binding the relatively flat BChl molecules to the cylindrical α-helical transmembrane segments of the α and β polypeptides requires amino acids at the –4 position to have small side chains. As pointed out in the extensive work of Zuber and co-workers (Zuber and Brunisholz, 1991), this amino acid is usually Ala and changes to amino acids with side chains larger than Val resulted in the lack of expression of LH1 and LH2 (Bylina et al., 1988). The phytyl or geranylgeranyl esterifying alcohol accounts for about 40% of the mass of the cofactor. It would therefore be expected that it would be important for hydrophobic packing of LH1 and LH2. Although this may be true, substitution of an ethyl group for the geranylgeranyl esterifying alcohol made little difference to B820 or LH1 formation, although LH1 was less stable than when native BChl was used (ParkesLoach et al., 1990; Davis et al., 1996). It would appear that detergent molecules (β-OG) can largely replace the hydrophobic tail of BChl in the B820 heterodimer and in oligomerization to form LH1. In the crystal structures of LH2, the esterifying alcohol of BChl may play a more specific role as the esterifying alcohol of the B850 BChls interdigitate with carotenoid and the
esterifying alcohol of the B800 BChl (McDermott et al., 1995; Koepke et al., 1996; Prince et al., 1997; Chapter 8, Gabrielsen et al.). In addition to hydrogen bond formation between BChl and Trp residues in the C-terminal segments of the α and β polypeptides, other structural constraints exist in these C-terminal regions. As the polypeptides progress from the N-termini through the middle α-helical segments, they then turn to protect the edge of BChl from the aqueous environment outside the membrane and provide the Trp side chains for hydrogen bonding. The amino acids in this region (Fig. 3) appear to be specifically chosen to serve this function. In particular, the sequence from +4 to +9 in the β polypeptide of LH1 seems most restrictive and requires the amino acids W/Y X W R/K P W. Several reconstitution experiments support this conclusion. On the other hand, the Phe at +10 of the β polypeptide in Rba. sphaeroides was shown to be unnecessary for reconstitution (Kehoe et al., 1998). Consistent with this observation, the amino acid at this location is not conserved. Although some variability is found at position +4, changing the Tyr (the residue found in the Rba. sphaeroides β polypeptide) to Met (the residue found in the Rba. capsulatus β polypeptide) substantially weakens B820 formation (Davis et al., 1997). The importance of this region from +4 to +9 was also underscored in experiments testing what modifications would need to be made to the LH2 β polypeptide of Rba. sphaeroides to enable it to form homodimeric B820. In this regard, it was found that substitution of the LH1 amino acid sequence from +4 to +10 into the LH2 β polypeptide was sufficient to enable the modified LH2 β polypeptide to form a B820-type complex (Todd et al., 1999). As mentioned earlier, it is especially noteworthy that the LH2 β polypeptide of Phs. molischianum
188 has been demonstrated to form a B820-type complex (Todd et al., 1998). This further underscores the importance of the C-terminal segment in which the LH2 β polypeptide of Phs. molischianum has the same amino acid sequence (W V W K P W) as found in the LH1 β polypeptides. Interestingly, the LH2 β polypeptide of Phs. molischianum, in the absence of other polypeptides, also formed a further red-shifted complex whose spectrum resembled the 850-nm absorption band of LH2. As further indication of the versatility of this polypeptide, in the presence of the LH1 α polypeptide of Rsp. rubrum or Phs. molischianum, an LH1-type complex was formed, but in the presence of the LH2 α polypeptide of Phs. molischianum an LH2 complex was formed (Todd et al., 1998). This is the first reported reconstitution of an LH2 complex using only isolated LH2 polypeptides and BChl. It is also the first example of an LH2 β polypeptide that can form an LH1 subunit-type complex and an LH1-type complex when paired with an LH1 α polypeptide. Somewhat greater variability seems acceptable for the C-terminal segment of the LH1 α polypeptides. The amino acids from +4 to +13 seem most conserved, requiring much of the sequence L S/G T (D) R/K N W M (D/E). The C-terminal amino acids of the Rsp. rubrum α polypeptide could be removed after Glu+13 with no apparent effect on B820 or LH1 formation (Meadows et al., 1995). Also, deletion of amino acids beyond +12 in the Rba. sphaeroides α polypeptide did not affect the ability to form B820 and LH1 although the wavelength maximum was blue-shifted (Davis et al., 1997) and LH1 was not well expressed when residues +13 to the C-terminus were omitted (McGlynn et al., 1996). VI. Cofactor – Cofactor Interactions The purpose of the protein structure of LH1 and LH2 appears to be to organize BChl molecules in such a way that they form a large overlapping array so that energy is absorbed in the near infrared and the excited state is quickly delocalized among the many molecules and easily transferred to other LH complexes or the RC. The mechanism of formation of LH1 and LH2 involves the oligomerization of an αβ heterodimer, B820. The spatial relationship and interaction of the two BChls in B820 were apparent from the shift of λmax compared with monomeric BChl (Miller et al., 1987; Chang et al., 1990a), from
Paul A. Loach and Pamela S. Parkes-Loach the CD spectrum (Miller et al., 1987; Chang et al., 1990a), from fluorescence polarization (Visschers et al., 1991), and from singlet-singlet annihilation spectroscopy (van Mourik et al., 1991). More recent experiments demonstrating close association of the two BChls of B820 include NMR (Wang et al., 2002) and EPR (Srivatsan and Norris, 2001) spectroscopies. The excitonic coupling of the two BChls, their center-to-center distance and parallel or antiparallel orientation of Qy axes were concluded from the spectroscopic data well before the LH2 crystal structures were known. Comparison of this early model for Rsp. rubrum B820 (van Mourik et al., 1991) to a subunit called the αβ heterodimer in the crystal structure of Phs. molischianum LH2 (Koepke et al., 1996) is striking in its accuracy in that the distance between Mg atoms in the αβ heterodimer of the crystal structure was 9.2 Å (compared with 11 Å in the model), the Qy axes of the BChl are antiparallel to each other and parallel to the membrane plane as proposed by the model and the π-overlap of the two macrocycles assures excitonic coupling. Even the rotation of 17o along an axis perpendicular to the membrane plane used to fit CD data for B820 is consistent with a similar angle between BChls in the LH2 structure of Phs. molischianum. The B820 subunit complexes readily associate to form LH1 or LH2, depending on the subtleties of cofactor-protein and protein-protein interactions. The structures of LH1 and LH2 have been modeled based on association of a fundamental subunit (Hu and Schulten, 1998; Fotiadis et al., 2004; Georgakopoulou et al., 2002, 2006). Although it is possible that small changes in hydrogen bonding occur as LH1 is formed (Sturgis and Robert, 1994), the overall B820 structure remains intact as is evident in the crystal structure of Phs. molischianum LH2 (Koepke et al., 1996). Clearly, the BChl ensembles in LH2 and LH1 have been optimized for light absorption and energy transfer. It is impressive that this regularity of structure has been accomplished with minimal protein involvement to provide the scaffold, protect the BChl and achieve an ideal spatial distribution for optimal absorption of light. VII. Protein – Protein Interactions A. B820 As a result of hybrid reconstitution experiments in-
Chapter 10
Light-Harvesting Reconstitution Studies
volving the α and β polypeptides from four different bacteria, little specificity was found to be required for B820 formation in the membrane-spanning middle segment (Loach et al., 1994). This is consistent with a highly variable amino acid sequence in this region except for His0 and Ala-4. In addition, the fact that the two BChls exist as a closely interacting dimer with π-overlap would require the α and β polypeptides to be separated by the BChl dimer in their α-helical middle segments. These conclusions were solidified when the crystal structures of two LH2 complexes were completed (McDermott et al., 1995; Koepke et al., 1996; Prince et al., 1997). With regard to protein-protein interaction at the N-terminus, B820-type complexes could be formed with only the LH1 β polypeptides, even when the Nterminal segment was greatly shortened by proteolytic removal (Meadows et al., 1995) or in chemically-synthesized analogs (Meadows et al., 1998). Also, much of the C-terminal segment of the α polypeptide could be removed by proteolysis without effect on B820 formation (Meadows et al., 1995; Parkes-Loach et al., 2004). Thus, there is little protein-protein interaction required for homodimeric B820 formation using only the β polypeptide in reconstitution experiments. The stabilization of the homodimeric B820 complex formed is the result of coordination of His0 to Mg of BChl, the cross hydrogen bonding from His0 to the 132 keto carbonyl of the other BChl and hydrophobic complementarity between the BChls and the hydrophobic surfaces of the polypeptides. On the other hand, while heterodimeric B820 formation as found in native complexes is also stabilized by interactions similar to those in the homodimeric complex, it is further stabilized by interactions between the N-terminal segments of the α and β polypeptides (Parkes-Loach et al., 2004). On the basis of results with the chemically synthesized polypeptides, protease-modified polypeptides and mutants, the importance of a stretch of 9–13 amino acids at the N-terminal end of the α and β polypeptides is underscored. A progressive loss of interaction with the LH1 β polypeptide was found as the first three N-terminal amino acids of the LH1 α-polypeptide of Rsp. rubrum were removed. The absence of the N-terminal formylmethionine (fMet), or conversion of the sulfur in this fMet to the sulfoxide, resulted in a decrease in LH1 formation. In addition to the removal of fMet, removal of the next two amino acids also resulted in a decrease in KA for B820 formation and nearly eliminated the ability to form LH1. It was suggested
189
that the first three amino acids (fMetTrpArg) of the LH1 α-polypeptide of Rsp. rubrum form a cluster that is most likely involved in close interaction with the side chain of His-18 of the β polypeptide and also participates in hydrophobic interactions with the β polypeptide (Arluison et al., 2004; Parkes-Loach et al., 2004). These results provide evidence that the folding motif of the α- and β-polypeptides in the N-terminal region observed in crystal structures of Phs. molischianum LH2 is also present in LH1 and contributes significantly to stabilizing the complex. In experiments designed to determine the minimal protein structure that would form a B820 complex with BChl, the proteolytic enzyme endoprotease GluC was used to prepare a Rba. sphaeroides β polypeptide in which the N-terminal segment was removed up through Glu-20. This 30 amino acid polypeptide displayed good activity for forming a homodimeric B820-type complex. To further explore this region of the β polypeptide, a chemically-synthesized polypeptide of 31 amino acids with an amino acid sequence identical to that of the Rba. sphaeroides β polypeptide between residues –20 and +10 was prepared and demonstrated to support formation of B820 with as high a binding affinity as the native Rba. sphaeroides β polypeptide (Table 1). From the subsequent use of trypsin, more of the N-terminus of the Rsp. rubrum β polypeptide was removed through Lys-17. This polypeptide would still form a B820type complex but with a lower affinity (Meadows et al., 1995). In agreement with this result, Nango et al. (2002) chemically-synthesized a polypeptide of 27 amino acids reproducing the amino acid sequence of the Rba. sphaeroides β polypeptide from residue –16 to +10 and found that a B820-type complex was readily formed by reconstitution with BChl. Interestingly, an analog of this polypeptide with a C-terminal Cys also readily formed a B820-type complex (Nango et al., 2002). An even shorter polypeptide containing 24 amino acids was chemically-synthesized by Noy and Dutton (2006) and shown to form a B820type complex. This latter polypeptide reproduced the amino acid sequence of the Rba. sphaeroides β polypeptide from residue –8 to +10 and also contained the sequence from –20 to –16 attached to the N-terminus. In further experiments with shorter synthetic polypeptides, our laboratory showed that a synthetic polypeptide of 16 amino acids reproducing the sequence of the Rba. sphaeroides β polypeptide from residue –5 to +10 would not support homodimeric
Paul A. Loach and Pamela S. Parkes-Loach
190 B820 formation (Meadows et al., 1998). This latter polypeptide was also lengthened by adding LysIleSerLys to the N-terminus to improve solubility, but this 20 amino acid polypeptide also failed to support formation of a B820-type complex (Meadows et al., 1998). Nango et al. (2002) also chemicallysynthesized a polypeptide of 22 amino acids with a sequence identical to that of the Rba. sphaeroides β polypeptide from residue –11 to +10 and found that it also would not support formation of a B820-type complex. One might interpret these experiments with variable lengths of the N-terminal part of the Rba. sphaeroides β polypeptide to indicate that the α-helical middle segment must be of sufficient length to stabilize the hydrophobic esterifying alcohols of BChl. This apparently requires at least 13 N-terminal amino acids to His 0 to provide almost 4 complete turns of the helix (about 20 Å long). Consistent with this conclusion is the location of the B850 BChl esterifying alcohol in the crystal structure of Phs. molischianum (Koepke et al., 1996). The minimal requirements to support formation of a B820-type complex are summarized in Table 2. Three of these seem to be essential properties of both the α and β polypeptides (or for only the β polypeptide in forming a homodimeric B820-type complex): (1) an α-helical segment of at least five to six turns in which the amino acid side chains are predominantly hydrophobic groups (Meadows et al., 1995, 1998); (2) His at position 0 (Bylina et al., 1988; Davis et al., 1997; Olsen et al., 1997) whose interaction with BChl provides over half of the stabilization energy (Loach and Parkes-Loach, 1995), and (3) appropriate placement of amino acids with relatively small hydrophobic side chains (e.g., at position –4) to allow appropriate packing of BChl and the α-helical portions of the protein (Bylina et al., 1988). Additional structural features that are very important are Trp+9 in the β polypeptide (Davis et al., 1997; Sturgis et
al., 1997; Kehoe et al., 1998) and Trp+11 in the α polypeptide (Olsen et al., 1994; Sturgis et al., 1997). Other structural features that seem to be collectively important in the β polypeptide are Trp (or Tyr) at position +4, Trp at position +6 (Davis et al., 1997; Sturgis et al., 1997; Kehoe et al., 1998) and an Arg or Lys at position +7 (Davis et al., 1997; Kehoe et al., 1998). Any one of these latter structural features seems to be expendable for subunit-type and LH1 complex formation. B. Light-Harvesting 1 and 2 Complexes Oligomerization of B820 to form LH1 or LH2 requires interaction between BChl dimers as well as protein-protein interactions. In the structure of LH2 of Phs. molischianum, the BChls of the B820 subunit overlap at ring I to about the same extent that the two BChls within B820 overlap at rings III and V. Thus, an extensive array of BChls is formed which effectively absorbs and transfers quanta of light. In addition to the π-π interactions of BChl, the α and β polypeptides interact to help stabilize the complex. His -18 of the β polypeptide of Rba. sphaeroides has been shown to be important in this regard. Changing this amino acid to Val by site-directed mutagenesis only slightly affected the formation of B820 but eliminated the ability to form LH1 under the usual reconstitution conditions (Todd et al., 1999). An LH1type complex could be formed at higher protein and BChl concentrations. As discussed earlier, the N-terminal regions of the α and β polypeptides interact to help stabilize heterodimeric B820 (Parkes-Loach et al., 2004). This region of B820 also interacts upon association to stabilize LH1 formation (Parkes-Loach et al., 2004). Several lines of evidence demonstrate that it is the structure of the α polypeptide that determines the λmax of the LH complex upon oligomerization. For
Table 2. Energies of stabilizing interactions in subunit complexes Structure β His 0
Stabilization energy > 6 kcal mol–1
4 – 6 turns of α-helix Small amino acid at –4 β Trp or Tyr at +4 β Trp at +6 β Arg or Lys at +7 β Trp at + 9
Necessary Necessary 1.0 kcal mol–1 1.4 kcal mol–1 2.0 kcal mol–1 3.7 kcal mol–1
References Bylina et al., 1988; Davis et al., 1997; Loach and Parkes-Loach, 1995 Meadows et al., 1995 Bylina et al., 1988 Davis et al., 1997 Sturgis et al., 1997; Kehoe et al., 1998 Kehoe et al., 1998 Kehoe et al., 1998
Chapter 10
Light-Harvesting Reconstitution Studies
example, in hybrid reconstitution experiments, addition of the α polypeptide of Phs. molischianum LH1 with the LH2 β polypeptide of Phs. molischianum resulted in a LH1 type-complex with an absorption maximum at 869 nm whereas the reconstitution of the same Phs. molischianum LH2 β polypeptide with the Phs. molischianum LH2 α polypeptide resulted in a LH2-type complex with an absorption maximum at 848 nm (Todd et al., 1998). It is unclear how the α polypeptide steers the associative process but it might provide an explanation for why the LH complexes are composed of heterodimers instead of homodimers. That is, the α polypeptide may have evolved to fine tune the absorption maximum of the associated complex in order to optimize light absorption in its environment. VIII. Effect of PufX on Reconstitution of Light-Harvesting 1 Complexes The importance of the PufX protein in bacterial photosynthesis became evident when it was found that deletion of the pufX gene in Rba. capsulatus caused changes in the LH1 content (Klug and Cohen, 1988) and also resulted in photosynthetic incompetence (Farchaus et al., 1990b; see also Chapter 9, Bullough et al.). Moreover, adding pufX back to pufX– strains resulted in a recovery of photosynthetic competence (Farchaus et al., 1990a). Using mutant strains of Rba. capsulatus and Rba. sphaeroides in which the pufX gene had been deleted, it was possible to identify, by high performance liquid chromatography (HPLC), membrane protein components present in pufX + cells but absent in pufX– cells. Using a combined approach of organic solvent extraction and HPLC, PufX was isolated from these two bacteria (Recchia et al., 1998). From the behavior of the PufX protein and the αpolypeptide of LH1 on HPLC, qualitative evidence was obtained that the two proteins have a high affinity for each other. In reconstitution assays with BChl and the LH1 α and β polypeptides of Rba. capsulatus, the PufX protein of Rba. capsulatus was inhibitory to LH1 formation at low concentration. A similar inhibition was exhibited by Rba. sphaeroides PufX protein for reconstitution of LH1 with BChl and the LH1 α and β polypeptides of Rba. sphaeroides. In both cases, the ratios of concentrations of the PufX protein to the α polypeptide causing 50% inhibition were approximately 0.5.
191
The mature forms of these PufX proteins contain 9 and 12 fewer amino acids, respectively, at the Cterminal end of the protein than are encoded by their pufX genes. To identify the portion of PufX responsible for inhibition of LH1 formation in reconstitution experiments, different regions (N-terminus and several core regions containing different lengths of the C-terminus) of Rba. sphaeroides and Rba. capsulatus PufX were chemically synthesized. Neither the N- nor C- terminal polypeptides of Rba. sphaeroides were inhibitory to LH1 reconstitution. In order to examine the role of the core regions of PufX, two large segments of this protein were chemically synthesized, one reproducing the amino acid sequence of the core segment predicted for Rba. sphaeroides PufX and the other reproducing the amino acid sequence predicted for the core segment of Rba. capsulatus PufX. Both of these core segments were found to be inhibitory to LH1 formation although higher concentrations were required relative to the native PufX proteins (ParkesLoach et al., 2001). CD measurements indicated that the core segment containing 39 amino acids of Rba. sphaeroides PufX exhibited 47% α-helix in trifluoroethanol while the core segment containing 43 amino acids of Rba. capsulatus PufX exhibited 59 and 55% α-helix in trifluoroethanol and in 0.80% β-OG in water, respectively. Thus, as predicted from the amino acid sequence, the hydrophobic middle segment is largely α-helical and presumed to span the membrane. Each chemically-synthesized core segment was covalently labeled with a fluorescent probe and tested for energy transfer to BChl (Law et al., 2003). Each was found to bind BChl with an affinity similar to the affinity of the LH1 polypeptides for BChl. It is suggested that PufX binds BChl and interacts with a BChl α-polypeptide component of LH1 to truncate, or interrupt, the LH1 ring adjacent to the location of the QB binding site of the RC. In other biochemical experiments evaluating the role of PufX, Francia et al. (1999) found that PufX was present in an isolated RC-LH1 complex in a 1:1 stoichiometry with the RC, and that a dimeric form of the core complex depended on the presence of PufX. In further mutagenic studies, it was shown that the N-terminal segment of PufX was required for dimerization of the core complex, the middle segment was required for cell growth and the C-terminal segment seemed to be involved in assembly (Francia et al., 2002). The possible location of PufX in the membrane of
192 Rba. sphaeroides has been studied by cryoelectron microscopy by Qian et al. (2005) in which 8.5 Å resolution was obtained. A dimeric RC-LH1-PufX structure was observed with an S-shaped LH1 array containing 28 LH1 αβ subunits within the dimer. An additional electron dense area was attributed to PufX which was located between LH1 and the RC near the QB binding site of the RC. This structure was similar to the results of Jungas et al. (1999) that first found evidence by negative staining EM for a dimeric structure for the RC-LH1 complex in natural tubular membranes. It is presumed that the role of PufX is to interrupt the cyclization of the LH1 αβ subunits at the QB site of the RC to enable reduced QB to communicate with the Q-pool in the membrane. A somewhat similar result was found by Scheuring et al. (2005) studying Rba. blasticus membranes using atomic force microscopy. Once again, evidence was provided that an S-shaped LH1 array exists, but with 13 LH1 αβ subunits per RC connected in the middle by what was assumed to be a PufX dimer. IX. In vitro versus In vivo Assembly of Complexes A. Oligomerization of B820 When B820 is exposed to in vitro conditions in which it reassociates to form LH1, several intermediate species are presumed to be formed which would reflect different numbers of subunits associated as the complex grows and eventually forms a closed ring. In addition, once a cyclic structure is formed, some rearrangement of interacting segments of the protein may occur on a slower time scale. Evidence for the formation of intermediates was first provided in the dissociation and reassociation experiments where separate sets of isosbestic points were observed during the course of the kinetic process (Miller et al., 1987). In stopped-flow experiments following B820 association to LH1, van Mourik et al. (1992) could fit the reassociation kinetics assuming a two-step process involving a species with a λmax at 850 nm, but it was suggested that more intermediates might be present. Evidence for a second spectral intermediate was obtained in further studies and the 850-nm species was suggested to be due to a tetrameric aggregate consisting of two B820 subunits (Pandit et al., 2003). Attempts to observe the B820 subunit as LH1 is assembled in vivo has so far not been successful.
Paul A. Loach and Pamela S. Parkes-Loach This laboratory has studied the formation of LH1 in whole cells of the G-9 mutant of Rsp. rubrum at very early periods of cell growth, but did not find evidence of B820 as an intermediate (P. Loach, unpublished results). It can be assumed that the equilibria lie far toward LH1 formation when precursors are present in the membrane so that any B820 subunits would be at very low concentration. Interaction of LH1 and its precursors with the RC may also stabilize the oligomeric state, further minimizing the concentration of any intermediates. B. Reconstitution of B820 and Reaction Center Complexes to Form a Photoreceptor Complex B820 and isolated RC of Rsp. rubrum have been reassociated to form the RC-LH1 core complex (Bustamante and Loach, 1994). The resulting complex displayed the same wavelength maxima, CD spectrum and high quantum yield for photochemistry as the in vivo complex. The experimental conditions for this reconstitution began with B820 in 0.80% β-OG to which phospholipids and RC were added. The mixture was then cooled or diluted to allow B820 to oligomerize around the RC as lipid vesicles were formed. From these experiments, it is clear that the LH1 subunit can reassociate with RC to form nativelike core complexes. The in vivo assembly of the photoreceptor RCLH1 complex may or may not follow a similar pattern. It would perhaps be expected that the RC would assemble before the B820 subunits in order that they could then oligomerize around it, otherwise complete LH1 circles might be formed without an RC inside. Consistent with this idea, in the case of Rba. sphaeroides where PufX is involved, Pugh et al. (1998) found that PufX appeared to interact with the RC before encirclement with LH1. However, this assumes that LH1 (and LH2) are static once a complete circular structure is formed. These complexes may, in fact, be quite flexible and even dynamic (Jamieson et al., 2002; Bahatyrova et al., 2004). The forces involved in maintaining a B820 subunit within a circular structure in a membrane environment are probably modest except where interaction with the RC is significant. Indeed, the complete encirclement of the RC in Rsp. rubrum and Rps. viridis may not be a problem for reduced QB of the RC to communicate with the Q-pool because a B820 subunit may readily dissociate from the circle at the QB binding site.
Chapter 10
Light-Harvesting Reconstitution Studies
C. Role of Carotenoids All wild-type photosynthetic bacteria contain carotenoids. It has long been known that the major functions of carotenoids, in addition to serving as accessory light absorbing pigments, are to stabilize LH and RC complexes and to retard photodegradation of BChl (Frank and Cogdell, 1995). Carotenoidless mutants such as G-9 of Rsp. rubrum are much more susceptible to degradation of BChl by conditions of combined light and oxygen than wild-type Rsp. rubrum. In LH complexes, carotenoid removal, or use of carotenoidless mutants, is usually required in order to reversibly dissociate the complex to B820 (Loach and Parkes-Loach, 1995). Moreover, addition of carotenoid to the LH1 α and β polypeptides and BChl under reconstitution conditions leads to a decrease in the concentration of the free B820 subunit because of enhanced oligomerization to form LH1 (Davis et al., 1995; Fiedor and Scheer, 2005; Chapter 46, Braun and Fiedor). From the crystal structures of LH2 (McDermott et al., 1995; Koepke et al., 1996; Prince et al., 1997), carotenoid molecules extend across the membrane between subunits. Thus, they act as a kind of glue stabilizing LH1, but also making intimate contact with BChl. In the absence of carotenoid, it is assumed that lipid replaces the carotenoid in vivo in carotenoidless mutants and that detergent replaces carotenoid in isolated B820 complexes. X. Reconstitution of the Reaction Center The foregoing discussion summarizes the many contributions reconstitution experiments have added to understanding structure-function relationships in light-harvesting and photoreceptor complexes. Complex LH systems have been comprehensively described; this includes evaluations of BChl structural requirements, thermodynamic properties, BChlBChl, BChl-protein and protein-protein interactions, as well as developing insights into the assembly process. If it were possible to achieve a comparable reconstitution of the RC, many unresolved questions could be addressed and groundwork laid for building nanodevices (Chapter 43, Iida et al). Toward this end, this laboratory has developed non-detergent methodology to isolate the RC-L, RC-M and RC-H polypeptides, a prerequisite for reconstitution studies (P. Loach, unpublished). Although these proteins can be isolated by SDS-
193
polyacrylamide gel electrophoresis (SDS-PAGE) and other procedures requiring detergents such as lauryl dimethyl amine oxide (Sutton et al., 1982), proteins are often irreversibly denatured by these methods. The organic solvent extraction and HPLC methodology developed for isolation of LH polypeptides (Tonn et al., 1977; Brunisholz et al., 1981, 1984; Miller et al., 1987; Loach and Parkes-Loach, 1995) has been extended to the isolation of the RC polypeptides. As can be seen by HPLC and SDS-PAGE analyses (Fig. 5), such a preparation of pure RC-H from Rsp. rubrum has been achieved. RC-M has also been isolated utilizing solvent extraction methodology, although the preparation still contains a small amount of RC-L (Fig. 6) A comparable preparation of RC-L of Rsp. rubrum is still under development, but the preparation containing RC-M and a small amount of RC-L may be useful for subsequent separation of RC-M and RC-L. RC reconstitution experiments have been initiated using the protocol that was successful in the reconstitution of LH1 and LH2. In the case of the RC, not only are BChl, the RC polypeptides, bacteriopheophytin, ubiquinone and iron required, but also lipids such as cardiolipin and possibly metals, such as Zn. Thus, there are many variables that will need to be tested. The results of preliminary experiments are shown in Fig. 7. The absorption spectrum shows a red-shifted absorption band at 843 nm with shoulders at about 780 and 900 nm. Control experiments are essential in this kind of work. Neither RC-H nor RC-M alone with BChl form species with a red-shifted absorption band. On the other hand, BPhe alone oligomerizes to form a species absorbing at 840 nm, but without the long wavelength shoulder. To evaluate whether a RC-type complex might be formed, the sample of Fig. 7 was illuminated and a light-dark difference spectrum recorded. A very small change was noted which showed some of the changes expected for a native RC (Fig. 8A). The experiment shown in Fig. 7 contained adequate RC-H and RC-M to interact fully with the BChl, BPhe and UQ. However, only about 0.1 equivalent of RC-L was present. It is possible that the light-induced absorption change of Fig. 8A reflects a RC-type complex formed with the small amount of RC-L present. If a symmetrical RC was reconstituted incorporating only RC-M and RC-H, one would expect its absorption spectrum to be similar to that of native RC (Fig. 8B). However, the quantum yield for its photochemical charge separation would
194
Paul A. Loach and Pamela S. Parkes-Loach
a
Fig. 6. RC-M containing fraction after separation on a Sephadex LH60 column. From left to right, lanes 1 & 5, chromatophores; lane 3 & 4, LH60 void volume at two concentrations; lane 2, second protein band (slower) containing small polypeptides.
b
Fig. 5. (a) HPLC of the RC-H polypeptide. Top, supernatant from solvent extraction. Middle, second HPLC of 14 min peak (RC-H). Bottom, third HPLC of 14 min peak. (b). SDS-PAGE of RC-H fractions. Lanes 1 and 5, standards of 194, 115, 97, 53, 37, 29, 20 and 7 kDa (top to bottom). Lane 4, chromatophores. Lane 3, 14 min peak after the first HPLC (Fig. 5a Top). Lane 2, 14 min peak after second HPLC (Fig. 5a middle).
Fig. 7. Absorption spectrum of an RC reconstitution trial in which RC-H, RC-M, small amount of RC-L, BChl, Bph, UQ and Mn(Ac)2 were present in 0.66% β-OG at room temperature.
be predicted to be substantially lower than for native RC since electron transfer apparently proceeds down the L-side much more efficiently (Parson, 1991; Okamura and Feher, 1992). In order to measure the light-induced charge separation of such a RC-M-only RC-type complex, it might require a more intense exciting light and faster kinetic capability than used in the preliminary experiments. Clearly, the preparation of RC-L must be refined to allow much greater quantities of this protein to be obtained. Although the results of our preliminary experiments should be interpreted with considerable caution, it is clear that the reconstitution procedure
Chapter 10
Light-Harvesting Reconstitution Studies
a
195
b
Fig. 8. (a) Light minus dark spectrum of reaction mixture of Fig. 7. Continuous illumination with tungsten light passed through a Corning 4-96 filter. Recorded with a Cary 50 spectrophotometer using 90o illumination. (b) Light minus dark absorbance spectrum of chromatophores of Rsp. rubrum. Same set up as for Fig. 8A. Absorbance of sample was 1.0 at 881 nm.
will allow effective mixing of all components, will provide a basis for systematic variation of conditions and will allow for running extensive controls. It should be noted that our first experiments in the reconstitution of B820 and LH1 provided only very small signs of success, but allowed us to refine conditions to the point that anyone can now easily reconstitute B820 and LH1. In conclusion, development of methodology that will allow reconstitution of the RC and the RC-LH photoreceptor complex would be of immense benefit to studies of structure-function relationships, just as it has been for LH1. In addition, important questions about the assembly of the PRC can be addressed which is expected to provide insights into the assembly of supramolecular complexes in general. Is it possible to reconstitute the PRC from individual components? Since the photosynthetic bacteria have been doing it for almost four billion years, it is certainly possible. Whether it can soon be accomplished under in vitro conditions in the laboratory will depend on producing the right conditions with the right components. Acknowledgments This work was supported by research grants from the National Science Foundation (MCB-0111255) and Northwestern University. References Arluison V, Seguin J, Le Caer J-P, Sturgis JN and Robert B (2004) Hydrophobic pockets at the membrane interface: An original
mechanism for membrane protein interactions. Biochemistry 43: 1276–1282 Bahatyrova S, Frese RN, van der Werf KO, Otto C, Hunter CN and Olsen JD (2004) Flexibility and size heterogeneity of the LH1 light harvesting complex revealed by atomic force microscopy — functional significance for bacterial photosynthesis. J Biol Chem 279: 21327–21333 Brunisholz RA, Cuendet PA, Theiler R and Zuber H (1981) The complete amino acid sequence of the single light harvesting protein from chromatophores of Rhodospirillum rubrum G-9+. FEBS Lett 129: 150–154 Brunisholz RA, Suter F and Zuber H (1984)The light-harvesting polypeptides of Rhodospirillum rubrum.1. The amino-acidsequence of the second light-harvesting polypeptide B880-β (B-870-β) of Rhodospirillum rubrum S-1 and the carotenoidless mutant G-9+. Aspects of the molecular structure of the two lightharvesting polypeptides B880-α (B870-α) and B880-β (B870-β) and of the antenna complex B880 (B870) from Rhodospirillum rubrum. Hoppe-Seyler’s Z Physiol Chem 365: 675–688 Bustamante PL and Loach PA (1994) Reconstitution of a functional photoreceptor complex with isolated subunits of core light-harvesting complex and reaction centers. Biochemistry 33: 13329–13339 Bylina E J, Robles S J, and Youvan DC (1988) Directed mutations affecting the putative bacteriochlorophyll-binding sites in light-harvesting I antenna of Rhodobacter capsulatus. Israel J Chem 28: 73–78 Chang MC, Callahan PC, Parkes-Loach PS, Cotton TM and Loach PA (1990a) Spectroscopic characterization of the lightharvesting complex of Rhodospirillum rubrum and its structural subunit. Biochemistry 29: 421–429 Chang MC, Meyer L and Loach PA (1990b) Isolation and characterization of a structural subunit from the core light-harvesting complex of Rhodobacter sphaeroides 2.4.1 and puc705-BA. Photochem Photobiol 52: 873–881 Cotton TM (1976) Spectroscopic investigations of chlorophyll a as donor and acceptor: A basis for chlorophyll a interaction in vivo. PhD Thesis, Northwestern University, Evanston, Illinois Cotton TM and van Duyne RP (1981) Characterization of bacteriochlorophyll interactions in vitro by resonance Raman spectroscopy. J Am Chem Soc 103: 6020–6024
196 Davis CM, Bustamante PL and Loach PA (1995) Reconstitution of the bacterial core light-harvesting complexes of Rhodobacter sphaeroides and Rhodospirillum rubrum with isolated α- and β-polypeptides, bacteriochlorophyll and carotenoid. J Biol Chem 270: 5793–5804 Davis CM, Parkes-Loach PS, Cook CK, Meadows KA, Bandilla M, Scheer H and Loach PA (1996) Comparison of the structural requirements for bacteriochlorophyll binding in the core light-harvesting complexes of Rhodospirillum rubrum and Rhodobacter sphaeroides using reconstitution methodology with bacteriochlorophyll analogs. Biochemistry 35: 3072–3084 Davis CM, Bustamante PL, Todd JB, Parkes-Loach PS, McGlynn P, Olsen JD, McMaster L, Hunter CN and Loach PA (1997) Evaluation of structure-function relationships in the core light-harvesting complex (LH1) of photosynthetic bacteria by reconstitution with mutant polypeptides. Biochemistry 36: 3671–3679 Dewa T, Yamada T, Ogawa M, Sugimoto M, Mizuno T, Yoshida K, Nakao Y, Kondo M, Iida K, Yamashita K, Tanaka T and Nango M (2005) Design and expression of cysteine-bearing hydrophobic polypeptides and their self-assembling properties with bacteriochlorophyll a derivatives as a mimic of bacterial photosynthetic antenna complexes. Effect of steric confinement and orientation of the polypeptides on the pigment/polypeptide assembly process. Biochemistry 44: 5129–5139 Farchaus JW, Gruenberg H, Gray KA, Wachveitl J, DeHoff B, Kaplan S and Oesterhelt D (1990a) The puf L, M, X deletion strain. In: Drews G and Dawes EA (eds) Molecular Biology of Membrane-Bound Complexes in Phototropic Bacteria, pp 65–76. Plenum Press, New York Farchaus JW, Gruenberg H and Oesterhelt D (1990b) Complementation of a reaction center-deficient Rhodobacter sphaeroides pufLMX deletion strain in trans with pufBALM does not restore the photosynthesis-positive phenotype. J Bacteriol 172: 977–985 Fiedor L and Scheer H (2005) Trapping of an assembly intermediate of photosynthetic LH1 antenna beyond B820 subunit. J Biol Chem 280: 20921–20926 Fotiadis D, Qian P, Pilippsen A, Bullough PA, Engel A and Hunter CN (2004) Structural analysis of the RC-LH1 photosynthetic core complex of Rhodospirillum rubrum using atomic force microscopy. J Biol Chem 279: 2063–2068 Fowler GJS, Sockalingum GD, Robert B and Hunter CN (1994) Blue shifts in bacteriochlorophyll absorbance correlate with changed hydrogen bonding patterns in light-harvesting 2 mutants of Rhodobacter sphaeroides with alterations at α-Tyr-44 and α-Tyr-45. Biochem J 299: 695–700 Francia F, Wang J, Venturoli G, Melandri BA, Barz WP and Oesterhelt D (1999) The reaction center-LH1 antenna complex of Rhodobacter sphaeroides contains one PufX molecule which is involved in dimerization of this complex. Biochemistry 38: 6834–6845 Francia F, Wang J, Venturoli G and Oesterhelt D (2002) Role of the N- and C-terminal regions of the PufX protein in the structural organization of the photosynthetic core complex of Rhodobacter sphaeroides. Eur J Biochem 269: 1877–1885 Francke C and Amesz J (1995) The size of the photosynthetic unit in purple bacteria. Photosynth Res 46: 347–352 Frank HA and Cogdell RJ (1995) Carotenoids in photosynthesis. Photochem Photobiol 63: 257–264 Georgakopoulou S, Frese RN, Johnson E, Koolhaas C, Cogdell R,
Paul A. Loach and Pamela S. Parkes-Loach van Grondelle R and van der Zwan G (2002) Absorption and CD spectroscopy and modeling of various LH2 complexes from purple bacteria. Biophys J 82: 2184–2197 Georgakopoulou S, van Grondelle R and van der Zwan G (2006) Explaining the visible and near-infrared circular dichroism spectra of light-harvesting 1 complexes from purple bacteria: A modeling study. J Phys Chem B 110: 3344–3353 Germeroth L, Lottspeich F, Robert B and Michel H (1993) Unexpected similarities of the B800-850 light-harvesting complex from Rhodospirillum molischianum to the B870 light-harvesting complexes from other purple photosynthetic bacteria. Biochemistry 32: 5615–5621 Heller BA and Loach PA (1990) Isolation and characterization of a subunit form of the B875 light-harvesting complex from Rhodobacter capsulatus. Photochem Photobiol 51: 621–627 Hu X and Schulten K (1998) Model for the light-harvesting complex I B875 of Rhodobacter sphaeroides. Biophys J 75: 683–694 Jamieson SJ, Wang P, Qian P, Kirkland JY, Conroy M, Hunter CN and Bullough PA (2002) Projection structure of the photosynthetic reaction centre-antenna complex of Rhodospirillum rubrum at 8.5 Å resolution. EMBO J 21: 3927–3935 Jirsakova V and Reiss-Husson F (1993) Isolation and characterization of the core light-harvesting complex B875 and its subunit form, B820, from Rhodocyclus gelatinosus. Biochim Biophys Acta 1138: 301–308 Jungas C, Ranck J-L, Rigaud J-L, Joliot P and Verméglio A (1999) Supramolecular organization of the photosynthetic apparatus of Rhodobacter sphaeroides. EMBO J 18: 534–542 Kashiwada A, Watanabe H, Mizuno T, Iida K Miyatake T, Tamiaki H, Kobayashi M and Nango M (2000a) Structural requirements of zinc porphyrin derivatives on the complex-forming with light-harvesting polypeptides. Chem Lett 29: 158–159 Kashiwada A, Watanabe H, Tanaka T and Nango M (2000b) Molecular assembly of zinc bacteriochlorophyll a by synthetic hydrophobic 1α-helix polypeptides. Chem Lett 29: 24–25 Kehoe JW, Meadows, KA, Parkes-Loach PS and Loach PA (1998) Reconstitution of light-harvesting complexes of photosynthetic bacteria using chemically synthesized polypeptides: II. Determination of structural features that stabilize complex formation and implications on the structure of the subunit complex. Biochemistry 37: 3418–3428 Kerfeld CA, Yeates TO and Thornber JP (1994) Biochemical and spectroscopic characterization of the reaction center-LHI complex and the carotenoid-containing B820 subunit of Chromatium purpuratum. Biochim Biophys Acta 1185: 193–202 Klug G and Cohen SN (1988) Pleiotropic effects of localized Rhodobacter capsulatus puf operon deletions on production of light-absorbing pigment-protein complexes. J Bacteriol 170: 5814–5821 Koepke J, Hu X, Muenke C, Schulten K and Michel H (1996) The crystal structure of the light-harvesting complex II (B800-850) from Rhodospirillum molischianum. Structure 4: 581–597 Law CJ, Chen J, Parkes-Loach PS and Loach PA (2003) Interaction of bacteriochlorophyll with the LH1 and PufX polypeptides of photosynthetic bacteria: Use of chemically synthesized analogs and covalently attached fluorescent probes. Photosynth Res 75: 193–210 Loach PA and Parkes-Loach PS (1995) Structure-function relationships in core light-harvesting complexes (LHI) as determined by characterization of the structural subunit and by reconstitution
Chapter 10
Light-Harvesting Reconstitution Studies
experiments. In: Blankenship RE, Madigan MT and Bauer CD (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 437–471. Kluwer Academic Publishers, Dordrecht Loach PA and Sekura DL (1968) Primary photochemistry and electron transport in Rhodospirillum rubrum. Biochemistry 7: 2642–2649 Loach PA, Parkes-Loach PS, Davis CM and Heller BA (1994) Probing protein structural requirements for formation of the core light-harvesting complex of photosynthetic bacteria using hybrid reconstitution methodology. Photosynth Res 40: 235–246 McDermott G, Prince SM., Freer AA., Hawthornthwaite-Lawless AM, Papiz MZ, Cogdell RJ and Isaacs NW (1995) Crystal structure of an integral membrane light-harvesting complex from photosynthetic bacteria. Nature 374: 517–521 McGlynn P, Westerhuis WHJ, Jones MR and Hunter CN (1996) Consequences for the organization of reaction center-light harvesting antenna 1 (LH 1) core complexes of Rhodobacter sphaeroides arising from deletion of amino acid residues from the C terminus of the LH1 α polypeptide. J Biol Chem 271: 3285–3292 Meadows KA, Iida K, Kazuichi T, Recchia PA, Heller BA, Antonio B, Nango M and Loach P (1995) Enzymatic and chemical cleavage of the core light-harvesting polypeptides of photosynthetic bacteria: Determination of the minimal polypeptide size and structure required for subunit and light-harvesting complex formation. Biochemistry 34: 1559–1574 Meadows KA, Parkes-Loach PS, Kehoe JW and Loach PA (1998) Reconstitution of light-harvesting complexes of photosynthetic bacteria using chemically synthesized polypeptides: I. Minimal requirements for subunit formation. Biochemistry 37: 3411–3417 Meckenstock RU, Brunisholz RA and Zuber H (1992) The lightharvesting core-complex and the B820-subunit from Rhodopseudomonas marina. Part 1. Purification and characterization. FEBS Lett 311: 128–134 Miller JF, Hinchigeri SB, Parkes-Loach PS, Callahan PM, Sprinkle JR, Riccobono JR and Loach PA (1987) Isolation and characterization of a subunit form of the light-harvesting complex of Rhodospirillum rubrum. Biochemistry 26: 5055–5062 Nagata M, Nango M, Kashiwada A, Yamada S, Ito S, Sawa N, Ogawa M, Iida K, Kurono Y and Ohtsuka T (2003a) Construction of photosynthetic antenna complex using light-harvesting polypeptide-α from photosynthetic bacteria, R. rubrum with zinc substituted bacteriochlorophyll α. Chem Lett 32: 216–217 Nagata M, Yoshimura Y, Inagaki J, Suemori Y, Iida K, Ohtsuka T and Nango M (2003b) Construction and photocurrent of lightharvesting polypeptides/zinc bacteriochlorophyll a complex in lipid bilayers. Chem Lett 32: 852–853 Nango M, Kashiwada A, Watanabe H, Yamada S, Yamada T, Ogawa M, Tanaka T and Iida K (2002) Molecular assembly of bacteriochlorophyll a using light-harvesting model 1α-helix polypeptides and 2α-helix polypeptide with disulfide-linkage. Chem Lett 2002: 312–314 Noy D and Dutton PL (2006) Design of a minimal polypeptide unit for bacteriochlorophyll binding and self-assembly based on photosynthetic bacterial light-harvesting proteins. Biochemistry 45: 2103–2113 Okamura MY and Feher G (1992) Proton transfer in reaction
197
centers from photosynthetic bacteria. Annu Rev Biochem 61: 861–896 Olsen JD (1994) The role of highly conserved amino-acid residues in the light-harvesting 1 complex of Rhodobacter sphaeroides. PhD Thesis, University of Sheffield Olsen JD, Sockalingum GD, Robert B and Hunter CN (1994) Modification of a hydrogen bond to a bacteriochlorophyll a molecule in the light-harvesting 1 antenna of Rhodobacter sphaeroides. Proc Natl Acad Sci USA 91: 7124–7128 Olsen JD, Sturgis JN, Westerhuis WHJ, Fowler GJS, Hunter CN and Robert B (1997) Site-directed modification of the ligands to the bacteriochlorophylls of the light-harvesting LH1 and LH2 complexes of Rhodobacter sphaeroides. Biochemistry 36: 12625–12632 Pandit A, Visschers RW, van Stokkum IHM, Krayenhof R and van Grondelle R (2001) Oligomerization of light-harvesting I antenna peptides of Rhodospirillum rubrum. Biochemistry 40: 12913–12924 Pandit A, van Stokkum IHM, Georgakopoulou S, van der Zwan G and van Grondelle R (2003) Investigations of intermediates appearing in the reassociation of the light-harvesting 1 complex of Rhodospirillum rubrum. Photosynth Res 75: 235–248 Parkes-Loach PS, Sprinkle JR and Loach PA (1988) Reconstitution of the B873 light-harvesting complex of Rhodospirillum rubrum from the separately-isolated α- and β-polypeptides and bacteriochlorophyll a. Biochemistry 27: 2718–2727 Parkes-Loach PS, Michalski TJ, Bass WJ, Smith U and Loach PA (1990) Probing the bacteriochlorophyll binding site by reconstitution of the light-harvesting complex of Rhodospirillum rubrum with bacteriochlorophyll a analogues. Biochemistry 29: 2951–2960 Parkes-Loach PS, Jones SM and Loach PA (1994) Probing the structure of the core light-harvesting complex (LHI) of Rhodopseudomonas viridis by dissociation and reconstitution methodology. Photosynth Res 40: 247–261 Parkes-Loach PS, Law CJ, Recchia PA, Kehoe J, Nehrlich S, Chen J and Loach PA (2001) Role of the core region of the PufX protein in inhibition of reconstitution of the core light-harvesting complexes of Rhodobacter sphaeroides and Rhodobacter capsulatus. Biochemistry 40: 5593–5601 Parkes-Loach PS, Majeed AP, Law CJ and Loach PA (2004) Interactions stabilizing the structure of the core light-harvesting complex (LH1) of photosynthetic bacteria and its subunit (B820). Biochemistry 43: 7003–7016 Parson WW (1991) Reaction centers. In: Scheer H (ed) Chlorophylls, pp 1153–1180. CRC Press, Boca Raton Prince SM, Papiz MZ, Freer AA, McDermott G, HawthornthwaiteLawless AM, Cogdell RJ and Isaacs NW (1997) Apoprotein structure in the LH2 complex from Rhodopseudomonas acidophila strain 10050: Modular assembly and protein pigment interactions. J Mol Biol 268: 412–423 Pugh RJ, McGlynn P, Jones MR and Hunter CN (1998) The LH 1-RC core complex of Rhodobacter sphaeroides: Interaction between components, time-dependent assembly, and topology of the PufX protein. Biochim Biophys Acta 1366: 301–316 Qian P, Hunter CN and Bullough PA (2005) The 8.5 Å projection structure of the core RC-LH1-PufX dimer of Rhodobacter sphaeroides. J Mol Biol 349: 948–960 Recchia PA, Davis CM, Lilburn TG, Beatty JT, Parkes-Loach PS, Hunter CN and Loach PA (1998) Isolation of the PufX protein from Rhodobacter capsulatus and Rhodobacter sphaeroides:
198 Evidence for its interaction with the α-polypeptide of the core light harvesting complex. Biochemistry 37: 11055–11063 Robert B and Lutz M (1985) Structure of antenna complexes of several Rhodospirillales from their resonance Raman spectra. Biochim Biophys Acta 807: 10–23 Scheuring S, Busselez J and Levy D (2005) Structure of the dimeric PufX-containing core complex of Rhodobacter blasticus by in situ atomic force microscopy. J Biol Chem 280: 1426–1431 Srivatsan N and Norris JR (2001) Electron paramagnetic resonance study of oxidized B820 complexes. J Phys Chem B 105: 12391–12398 Sturgis J and Robert B (1994) Thermodynamics of membrane polypeptide oligomerization in light-harvesting complexes and associated structural changes. J Mol Biol 238: 445–454 Sturgis JN, Olsen JD, Robert B and Hunter CN (1997) Functions of conserved tryptophan residues of the core light-harvesting complex of Rhodobacter sphaeroides. Biochemistry 36: 2772–2778 Sutton MR., Rosen D, Feher G and Steiner LA (1982) Aminoterminal sequences of the L, M, and H subunits of reaction centers from the photosynthetic bacterium Rhodopseudomonas sphaeroides R-26. Biochemistry 21: 3842–3849 Todd JB, Parkes-Loach PS, Leykam JF and Loach PA (1998) In vitro reconstitution of the core and peripheral light-harvesting complexes of Rhodospirillum molischianum from separately isolated components. Biochemistry 37: 17458–17468 Todd JB, Recchia PA, Parkes-Loach PS, Olsen JD, Fowler GJS, McGlynn P, Hunter CN and Loach PA (1999) Minimal requirements for in vitro reconstitution of the structural subunit of light-harvesting complexes of photosynthetic bacteria. Photosynth Res 62: 85–98 Tonn SJ, Gogel GE and Loach PA (1977) Isolation and characterization of an organic solvent soluble polypeptide component from photoreceptor complexes of Rhodospirillum rubrum.
Paul A. Loach and Pamela S. Parkes-Loach Biochemistry 16: 877–885 van Mourik F, van der Ord JR, Visscher KJ, Parkes-Loach PA, Loach PA, Visschers, RW, and van Grondelle R (1991) Exciton interactions in the light-harvesting antenna of photosynthetic bacteria studied with triplet-singlet spectroscopy and singlettriplet annihilation on the B820 subunit form of Rhodospirillum rubrum. Biochim Biophys Acta 1059: 111–119 van Mourik F, Corten EPM, van Stokkum IHM, Visschers RW, Loach PA, Kraayenhof R and van Grondelle R (1992) Self assembly of the LH-1 antenna of Rhodospirillum rubrum, a time-resolved study of the aggregation of the B820 subunit form. In: Murata N (ed) Research in Photosynthesis, pp 101–104. Kluwer Academic Publishers, Dordrecht Visschers RW, Chang MC, van Mourik F, Parkes-Loach PS, Heller BA, Loach PA and van Grondelle (1991) Fluorescence polarization and low-temperature absorption spectroscopy of a subunit form of light-harvesting complex I from purple photosynthetic bacteria. Biochemistry 30: 5734–5742 Visschers RW, van Grondelle R and Robert B (1993) Resonance Raman spectroscopy of the B820 subunit of the core antenna from Rhodospirillum rubrem G9. Biochim Biophys Acta 1183: 369–373 Wang Z, Muraoka Y, Shimonaga M, Kobayashi M and Nozawa T (2002) Selective detection and assignment of the solution NMR signals of bacteriochlorophyll a in a reconstituted subunit of a light-harvesting complex. J Am Chem Soc 124: 1072–1078 Wendling M, Lapouge K, van Mourik F, Novoderezhkin V, Robert B and van Grondelle R (2002) Steady-state spectroscopy of zinc-bacteriopheophytin containing LH1—an in vitro and in silico study. Chem Physics 275: 31–45 Zuber H, & Brunisholz RA (1991) Structure and function of antenna polypeptides and chlorophyll-protein complexes: Principles and variability. In: Scheer H (ed) Chlorophylls, pp 627–703. CRC Press, Boca Raton
Chapter 11 Spectroscopic Properties of Antenna Complexes from Purple Bacteria Bruno Robert* CEA, Institut de Biologie et de Technologie de Saclay, and CNRS, 91191 Gif/Yvette, France
Summary ............................................................................................................................................................... 199 1. Introduction..................................................................................................................................................... 200 II. The Different Spectral Forms of Antenna Proteins from Purple Bacteria ...................................................... 200 III. Antenna Proteins from Purple Bacteria: Variations Around a Structural Theme ........................................... 201 IV. The Role of Ground State Interactions in Tuning the Antenna Absorption Transition ................................... 203 V. Excitonic Interactions and Disorder in Light-Harvesting Complexes ............................................................. 205 VI. Chromophore Interactions in Light-Harvesting Proteins: Additional Effects .................................................. 207 VII. Conclusions .................................................................................................................................................... 208 Acknowledgments ................................................................................................................................................. 209 References ............................................................................................................................................................ 209
Summary This chapter describes the known factors which shape the electronic (and thus the functional) properties of light-harvesting (LH) complexes of purple bacteria. Although a variety of high- and low-resolution structures from LH complexes are now available, they do not provide, per se, a detailed picture of the electronic properties of the pigments in these complexes. However they constitute a framework which has helped, through the use of wide range of techniques, most often combining site-selection mutagenesis and advanced spectroscopies, to make progress in determining which parameters are important for the function of these proteins. Today, antenna proteins from purple photosynthetic bacteria are probably the best-understood photosynthetic LH proteins, and among the best-characterized membrane proteins in any biological field. However, there are still some discrepancies about the precise relationship between their structure and function. This chapter attempts to describe, as simply as possible, the physical mechanisms which are thought to underlie the tuning of the absorption properties of bacteriochlorophyll molecules in these proteins, as well as the results which are at the origin of the running arguments.
*Email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 199–212. © 2009 Springer Science + Business Media B.V.
200 1. Introduction More than hundred years ago, it was noticed that purple bacteria are able to grow photosynthetically under near infrared illumination, at wavelengths much higher than the maximum absorbance of the isolated pigments (Molisch, 1907). As early as 1921, Lubimenko interpreted this red-shift as evidence that these pigments were bound to proteins (cited in Pardee et al., 1952). In the following years, the unusual absorption of the bacteriochlorophyll (BChl) molecules, as observed in vivo, was extensively used to monitor and control the harshness of the first biochemical protocols designed to isolate the components of the photosynthetic apparatus from these organisms, as well as the intactness of the resulting purified fractions. Because of this very useful internal control, the biochemistry of pigment protein complexes of purple bacteria developed early: French (1940) reported the first extraction of intracytoplasmic membranes of various strains retaining their native absorption properties and, less than twenty five years later, the first successful attempt to purify photoactive protein fractions from these membranes (Clayton, 1963). In addition to their very important role in guiding the pioneer biochemists as they took these first steps in membrane protein science, these red-shifted electronic transitions were intriguing: depending on the bacterial strain, they could exhibit completely different shapes and positions, always different from those of isolated BChl molecules, and sometimes redshifted by as much as 1500 cm–1 or 130 nm relative to the purified pigment. If the antenna proteins for purple photosynthetic bacteria are nowadays probably the best-characterized membrane proteins, it is likely that this originates from these fascinating electronic transitions, which are so easy to observe in intact bacteria. In the last twenty five years, light-harvesting (LH) proteins from purple bacteria have been extensively studied, both from experimental and theoretical points of view. These studies have developed far beyond the field of photosynthesis, and the detailed analysis of the bacterial antenna complexes has led to the development of concepts which are of interest for most fields in molecular biology. For Abbreviations: BChl – bacteriochlorophyll; CD – circular dichroism; Gpa – GigaPascal; LH – light-harvesting; Rba. – Rhodobacter; RC – reaction center; Rps. – Rhodopseudomonas; Rsp. – Rhodospirillum
Bruno Robert instance, calculating the electronic properties of the peripheral light-harvesting 2 (LH2) complex from its crystallographic structure leads to predictions that are obviously wrong. The electronic properties may actually be predicted correctly, but only by taking into account molecular disorder, a phenomenon inherently present in the macromolecular structure of any complex. In this particular case, one may thus consider that theoretical advances on light-harvesting complexes from purple photosynthetic bacteria have yielded one of the most accurate descriptions of disorder and its effects in any biological structure. In this short review article, I will summarize the present state of our knowledge on the electronic properties of these proteins, with the constant aim of linking complex theoretical developments and the structure of these complexes. II. The Different Spectral Forms of Antenna Proteins from Purple Bacteria In photosynthetic purple bacteria, the LH system generally contains a core antenna, also called LH1, which transfers excitation energy directly to the reaction centers (RCs). In BChl a-synthesizing bacteria, LH1 complexes typically absorb at 870–890 nm (Fig. 1). BChl b-containing LH1 complexes exhibit near infrared transitions above 1000 nm. Many bacterial species also contain peripheral antenna complexes, LH2, which transfer excitation energy to the RCs via LH1. LH2 complexes usually exhibit two major absorption transitions in the near infrared, at 800 and 850 nm in BChl a-containing bacteria (Fig. 1). In these organisms, the electronic properties of the LH complexes funnel excitation energy towards the RC, a process driven by the energy gradient of the electronic transitions of the different LH complexes. In that sense, understanding the electronic properties of the LH proteins of purple bacteria is particularly important, as they are the major factor underlying the vectorial flow of excitation energy transfer towards the RCs. Some bacterial species, in low illumination conditions or when grown at low temperature, are able to synthesize additional LH2, the lower transition of which is downshifted to 820 nm (Fig. 1). LH1 and LH2 thus generally exhibit electronic transitions considerably red-shifted relative to the absorption of isolated, monomeric BChl a or b (770 and 800 nm in most organic solvents, respectively). In LH complexes, BChl molecules responsible for a given
Chapter 11
Spectroscopic Properties of Antenna Complexes
201
also the case for complexes of thermophilic bacteria: a 920 nm-absorbing LH1 complex has been isolated from the thermophilic Chromatium tepidum species (Garcia et al., 1986; Fathir et al., 1998). III. Antenna Proteins from Purple Bacteria: Variations Around a Structural Theme
Fig. 1. Electronic absorption spectra (Qy transitions) of LH1 from Rsp. rubrum, strain G9+ (top), and of LH2 from Rps. acidophila, strain 10050 (middle trace: B800-850, bottom : B800-820). It must be emphasized that the spectra chosen here correspond to those of the most-commonly studied types of bacterial LH complexes, but that it is unclear whether they represent the most commonly found in purple bacteria.
Qy absorption transition are usually referred to as B followed by the position of this transition, e.g., B800, B850. In addition to these proteins with ‘classical’ absorption properties, a number of strains are able to synthesize antenna proteins with unusual absorption transitions. This is the case, for example, with Rhodopseudomonas (Rps.) palustris strain 2.6.1, which, when grown at low light intensity, produces a modified peripheral antenna complex mainly absorbing at 800 nm (Robert et al., 1985). Chromatium purpuratum, also produces a peripheral antenna complex with a single strong absorption band at 830 nm and a small shoulder at 800 nm (Cogdell et al., 1990). Antenna complexes from marine, aerobic species of purple photosynthetic bacteria often exhibit unusual properties (Chapter 3, Yurkov and Csotonyi). For instance, a 865 nm-absorbing LH1 has been reported in Roseococcus thiosulfatophilus (Gall et al., 1999). This is
As discussed elsewhere in this book (Chapter 8, Gabrielsen et al.; Chapter 9, Bullough et al.; Chapter 10, Loach and Parkes-Loach) antenna complexes from purple photosynthetic bacteria are oligomers of an elementary unit, composed of a pair of small (5–7 kDa), very hydrophobic, apoproteins, called α and β, each of which contains a single transmembrane helix as the major structural element (Zuber and Brunisholz, 1991). Since the cryo-electron microscopy studies on 2-D crystals of LH1 from Rhodospirillum (Rsp.) rubrum (Karrasch et al., 1995), and the determination of the 3-D structure of LH2 from Rps. acidophila (McDermott et al., 1995), we know that LH complexes are composed of circular associations of αβ subunits. Fully-assembled LH complexes are comprised of two concentric rings of helical α- (inside ring) and β- (outside ring) polypeptides (Isaacs et al., 1995). Sixteen αβ subunits are required to produce a complete, closed LH1 ring (Karrasch et al., 1995). The central hole of these rings is large enough to contain the RC, and 2-D crystals of RCLH1 complexes clearly show that it does (Walz and Ghosh, 1997; Walz et al., 1998; Jamieson et al., 2002). Whether these closed, 16-membered rings truly exist in vivo has been a matter of debate. It has become clear that the precise structure of LH1 depends on the bacterial species considered. In Rps. viridis, for instance, evidence for closed rings has been provided by atomic force microscopy on native membranes (Scheuring et al. 2003; Chapter 47, Scheuring). In contrast, in Rhodobacter (Rba.) sphaeroides, it is now well established that LH1 comprises two incomplete αβ rings that associate to form a dimer, which surrounds two RCs (Jungas et al., 1999; Siebert et al., 2004; Bahatyrova et al., 2004b; Qian et al., 2005). This organization is dependent on the presence of an additional component in the RC-LH1 complex, named PufX (Francia et al., 1999; Frese et al., 2000; Siebert et al., 2004). In Rps. palustris, the LH1 complex is also a ring, nearly closed, but interrupted by the presence of a hydrophobic polypeptide, located adjacent to the RC on the side opposite to the transmembrane
202 helix of the RC-H subunit (Roszak et al., 2003). In the case of LH2, structural studies, involving either X-ray crystallography of 3-D crystals or cryo-electron microscopy on 2-D crystals, have shown that these complexes are formed by a smaller ring, involving a smaller number of α and β polypeptides. The 2.5 Å crystal structure of LH2 from Rps. acidophila (Isaacs et al., 1995; Prince et al., 1997) revealed that it consists of a nonameric circular association of αβ heterodimers. A similar organization has been shown for LH2 from Rhodovulum sulfidophilus (Savage et al., 1996) as well as for the 820 nm-absorbing LH2 from Rps. acidophila (McLuskey et al., 1999). In contrast, the structure of the LH2 complex from Rsp. molischianum is an octameric ring of α and β apoproteins (Koepke et al., 1996). This difference in quaternary structure is likely to be due to the peculiarity of the sequences of the polypeptide which constitute this protein (Germeroth et al., 1993). Parallel to this rather simple view of the structure of these complexes, it has been shown recently that in vivo LH complexes exhibit some structural variability around these canonical structures (Bahatyrova et al., 2004a; Scheuring et al., 2004). The extent to which complex-to-complex structural heterogeneity is important for the functional properties of the LH proteins has not yet been measured experimentally. This is, however, an essential point which has to be addressed in the near future (see below). In all antenna complexes, the α and β polypeptides each bind a single BChl molecule, through a conserved histidine in the hydrophobic phase of the membrane. In the fully assembled LH complex, this pattern of BChl binding results in a ring of closely interacting BChl molecules, sandwiched between the two polypeptide rings (see Fig. 2, and Chapter 8, Gabrielsen et al.), and oriented perpendicular to the membrane plane. In LH1 and LH2, these rings (or portions of rings, in the case of LH1) are thus formed by 24 to 32 and 16 to 18 BChl molecules, respectively. Such associations of many BChl molecules within van der Waals contact gives rise to the red-most absorption transition of the complexes; see, for example, van Grondelle et al. (1994). In LH2 complexes, each αβ subunit binds one additional BChl molecule near the interface between the membrane and the cytosol. The structure of the binding site and the orientation of these molecules is clearly species dependent. In Rps. acidophila, these molecule lie close to parallel to the membrane plane, and their central Mg ion is bound by the oxygen atom of the formylated N-ter-
Bruno Robert
Fig. 2. Organization of the BChl arrays in LH2 from Rps. acidophila. Top: 850-nm absorbing array. Bottom: 800-nm absorbing array.
minal of the α polypeptide. This binding site is very polar and is mainly composed of aminoacids from the β polypeptide. These BChls form an additional ring of nine molecules, spaced by about 21 Å and thus weakly interacting with each other. They give rise to the 800 nm absorption band of these proteins. In Rsp. molischianum LH2, the plane of these B800 molecules is rotated by 90 degrees compared with the LH2 from Rps. acidophila, and is tilted away from
Chapter 11
Spectroscopic Properties of Antenna Complexes
203
the membrane plane (Koepke et al., 1996). In addition, the nature of the aminoacids interacting with the B800 molecules is different. LH2 from purple sulfur bacteria exhibit not one but two transitions around 800 nm (Krikunova et al., 2002). It is not yet clear whether an additional binding site is present in these complexes. In LH2 synthesized in low light by Rps. palustris, the transition at 800 nm much larger and broader as compared to Rps. acidophila LH2 (Robert et al., 1985). It is clear that, in these complexes also, the overall organization of the B800 molecules must differ from that observed in the crystal structure of Rps. acidophila LH2 (Gall and Robert, 1999; Hartigan et al., 2002). IV. The Role of Ground State Interactions in Tuning the Antenna Absorption Transition As described above, the lowest energy transition (Qy) of BChl molecules in LH proteins are generally red-shifted in comparison with that of BChl in organic solvents. The least red-shifted Qy transition is that of the B800 molecules in the LH2 complexes, which represents a red-shift just a bit less than 500 cm–1, compared to the 770 nm position observed for BChl in acetone. The B800 BChl molecules interact only weakly with each other. From the structure, the calculated B800/B800 coupling is only about 30 cm–1. The 770 to 800 nm red-shift experienced by the Qy transition of these molecules can thus only arise from BChl/protein interactions. The protein matrix may influence pigment absorption in different ways, i) by providing aminoacids able to interact with the pigments, ii) through dispersive forces, which are the result of the protein environment with anisotropic dielectric properties and iii) through steric hindrance, by inducing pigment conformational changes, which in turn influence their electronic properties. In the Rps. acidophila LH2 complex, the acetyl carbonyl of each B800 molecule interacts with the amine group of the arginine β21 (McDermott et al., 1995). The C=O stretching frequency of this group, as observed by resonance Raman spectroscopy, indicates that this group is strongly H-bonded by this aminoacid (Robert and Lutz, 1985; Gall et al., 1997). In mutants of LH2 from Rba. sphaeroides where this aminoacid has been altered, the Qy transition of the B800 BChl molecules blue-shifts by about 10 nm, towards 790 nm (Fowler et al., 1997; Gall et al., 1997;
Fig. 3. Influence of the β21 arginine on the 800 nm electronic transition of LH2 from Rba. sphaeroides. Electronic absorption spectra at room temperature of LH2 from (top to bottom) wildtype, β21 asparagine, β21 leucine, β21 glutamic acid, β21 lysine, β21 histidine, β21 methionine (redrawn from Gall et al., 1997).
see Fig. 3). The strong H-bond between the arginine β21 and the acetyl carbonyl of the B800 molecule is thus responsible of a ca 180 cm–1 red-shift of the Qy of these molecules. Such an effect of H-bonds involving the acetyl carbonyl of BChl molecules had actually been predicted by rather simple molecular
204 calculations (Gudowska-Nowak et al., 1990). In LH2 mutants where this H-bond has been broken, the Qy absorption of the B800 BChl is observed to vary between 780 and 792 nm. This reflects an additional influence from the protein environment, and more particularly from the precise chemical nature of the amino acid sidechain at position β21. For instance, a leucine residue at position β21 results in a Qy position of the B800 molecules at 787 nm, while this transition is observed at 782 nm, in the presence of the more polar asparagine residue at this position (Fowler et al., 1997). Resonance Raman measurements led to the proposal that this additional 12 nm (160 cm–1) shift was due to local changes in the local dielectric constant around the BChl (Gall et al., 1997). The combined effects of the H-bond together with the low dielectric constant around the B800 molecules are almost sufficient to account for the Qy red-shift of B800 molecules in LH2 (Gall et al., 1997). These experiments provide direct evidence that the H-bonding state, as well as the dielectric properties of the environment affect the BChl absorption in proteins. In excitonically coupled BChl systems, these parameters should influence the energy site of each participating molecule, and thus in turn affect the excitonically-coupled levels of the ensemble. This was elegantly shown by site-mutagenesis, performed on LH2 from Rba. sphaeroides: when the residues Tyr α44 and Trp α45 are replaced by phenylalanine and leucine, respectively, the lower energy transition of these complexes is shifted from 850 to 826 nm (Fowler et al., 1992). Both these aminoacids are involved in H-bond interactions with the acetyl carbonyl of the 850 nm-absorbing BChl molecules (Fowler et al.¸ 1994). It is of note that these mutations are naturally observed between the sequences of the α polypeptides of 820 nm-absorbing LH2 complexes (Brunisholz et al., 1987). The tuning of the absorption of these complexes should thus naturally occur through variations of the H-bonding state of the BChl molecules. The study of a number of natural and artificial variants of LH complexes, including naturally 820-nm absorbing LH2 (Sturgis et al., 1995a,b; Sturgis and Robert, 1997), genetically engineered LH1 proteins (Olsen et al., 1994; Sturgis et al., 1997), and unusual complexes from Roseococcus thiosulfatophilus LH1 proteins (Gall et al., 1999) unambiguously showed that this is actually the case. Crystallographic studies of 820 nm-absorbing complexes from Rps. acidophila have led to somewhat different conclusions. From the electronic density map obtained from X-ray
Bruno Robert studies of these proteins, it was proposed that, in these complexes, the acetyl carbonyl of the B850 BChl molecules would be rotated out of the plane of the dihydrophorbin macrocycle (McLuskey et al., 1999). It was proposed that this could be the main reason of the observed blue-shift. However, for such a blue-shift to occur, the rotation should induce the deconjugation of the acetyl carbonyl. This should in turn result in the disappearance of the contribution of its stretching vibrations from resonance Raman spectra. Such effect has been observed in resonance Raman spectra of BChl bound to the bacterial RC (Ridge et al., 2000). However, in resonance Raman spectra from the 820 nm-absorbing complexes from Rps. acidophila, there is a clear, and rather intense, contribution arising from the stretching mode of these carbonyl groups (Sturgis et al., 1995). This suggests that this group is still enough in plane to be conjugated with the electronic transition of the BChl molecule. Under these conditions, the observed rotation should have a very limited effect on the Qy electronic absorption of the LH2 complexes, and it is unlikely that it is alone at the origin of the observed blue shift. In any case, the description of the structure of the 820 nm-absorbing complexes from X-ray crystallographic studies (McLuskey et al., 1999) fully validated the idea that the difference in absorption between the 820 nm- and 850 nm-absorbing LH2 complexes is mainly due to differences in BChl-protein ground state interaction, and not to differences in excitonic coupling between the BChl molecules. This structure shows that both the B820 and the B850 rings have similar, if not identical, structures, thus ruling out the possibility of major differences in BChl/BChl interactions between the two complexes. Although other parameters could play a role in tuning the BChl energy site, such as the H-bonding state of the other, conjugated carbonyl of these molecules, namely the keto one, it was clearly shown in two cases that this is not the case. In both the LH1 and LH2 complexes from Rba. sphaeroides, it was possible to break an H-bond between this carbonyl and a neighboring aminoacid, by site-selected mutagenesis, without any detectable influence on the position of the main Qy transition of these complexes (Olsen et al., 1997; Kwa et al., 2004). The interaction state of the keto carbonyl group of the BChl molecules is thus not a major parameter for tuning BChl absorption. This is probably due to the fact that the keto carbonyl group is less conjugated with the conjugated macrocycle of the BChl molecule, as indicated by its lower
Chapter 11
Spectroscopic Properties of Antenna Complexes
stretching frequency (Robert and Lutz, 1985). The conformation of the BChl dihydrophorbin macrocycle is also likely to play a role in tuning the absorption of BChl molecules (Gudowska-Nowak et al., 1990). Up to now, no steric hindrance large enough could be created by site-selected mutagenesis to perturb the conformation of any of the LH-bound BChl molecules. Crystallographic studies showed that the BChl bound to the β polypeptide of the 850 nm-absorbing pair in LH2 from Rps. acidophila could be significantly distorted (Prince et al., 1997). However, determining the precise conformation of a proteinbound cofactor is a particularly difficult task for crystallographers. Spectroscopic studies confirmed that some of the BChl molecules responsible for the lower electronic transition in both LH2 and LH1 were indeed distorted, as compared to the relaxed conformation observed in organic solvent (Lapouge et al., 1999). This distortion may also be observed when modified cofactors are artificially introduced in these complexes (Lapouge et al., 2000). In the absence of direct experimental evidence of the effect of such distortion, it is unclear yet whether it has any influence on the BChl absorption, and even less at which extent it perturbs its electronic properties. However, interpreting the dichroic properties of the Qy electronic transitions of the LH2 complexes requires the two BChl molecules bound to the α and β polypeptides to be spectrally different, with the β BChl about 300 cm–1 redder than the α BChl (Koolhaas et al., 1998). A similar difference was predicted by quantum chemical calculations (Alden et al., 1997). These calculations suggest, although indirectly, that the conformation of BChls influence their electronic properties in LH complexes. V. Excitonic Interactions and Disorder in Light-Harvesting Complexes We have seen that ground-state interactions play an important role in tuning the position of the lower energy transition of antenna complexes of purple bacteria. However, ground-state interactions alone do not explain more than a fraction of the observed red-shift experienced by this transition. Early fluorescence depolarization studies (i.e., more than a decade before the structure was known) led to the prediction of the cylindrical symmetry of the involved cluster of BChl molecules (Kramer et al., 1984). These experiments also indicated the efficiency of excitation
205
energy transfer between the 850 nm-absorbing BChl molecules. Since then, a wide range of spectroscopic measurements have established that excitation energy transfers between the 850 nm-absorbing BChl molecules occur on an ultrafast timescale (Visser et al., 1995; Jimenez and Fleming, 1996; Jimenez et al., 1997; Kennis et al., 1997b; Chachisvilis et al., 1997; Nagarajan et al., 1999). This, together with the intense dichroic signal of the 850 nm transition indicates that the lower energy electronic transition arises from excitonically coupled BChl molecules. Spectroscopic measurements, as well as calculations based on the atomic structure of the LH2 complex, estimate the coupling strength between adjacent BChl molecules as ca 300 cm–1. In the structure of LH complexes from purple bacteria the large circular arrays of BChl cofactors in close contact must be considered as excitonically coupled systems. Taking into account the strength of the BChl/BChl excitonic interactions, these circular arrays should formally give rise to absorption transitions corresponding not to the sum of the absorption transitions of each of the individual BChls, but to one giant electronic transition reflecting their collective property. This pure exciton model is able to predict the correct position and shape of the absorption transition of the LH complex (see e.g., McDermott et al. (1995) or Sauer et al (1996)). However, such a prediction alone does not validate the pure exciton model (Sturgis and Robert, 1996; Monshouwer and van Grondelle, 1996). Moreover, in the pure exciton model, the lowest excitonic level must have close to zero dipole strength. A straightforward consequence of this conclusion is that LH2 complexes should not be fluorescent at very low temperature. However a number of experiments show clearly that this prediction is wrong; LH2 complexes from Rba. sphaeroides and Rba. capsulatus both fluoresce with a rate 2–3 times faster than monomeric BChl in solvent (Monshouwer et al., 1997). It may thus be concluded that the direct use of the molecular structure of LH protein for calculating its functional properties leads to wrong predictions. This is a fascinating feature of the LH proteins, the properties of which can only be understood if defaults are introduced in their molecular structure. The first experiments which showed that disorder was an important factor for explaining the functional properties of LH were actually performed in the 1980s, when it was discovered that the position of their fluorescence electronic transition at low temperature depends on the position of the excita-
206 tion wavelength. These experiments suggested the existence of subpopulations of LH complexes, but, as no such subpopulation could be detected, they had to be taken as evidence that, in a biochemically homogeneous set of LH proteins, not all share the same electronic properties (van Mourik et al., 1992). Such heterogeneity may either be due to differences in the local physico-chemical properties of the BChl molecules, arising either from slight differences in conformation and/or the conformation of the immediate surroundings, or from dynamic fluctuations of the system (these two effects are usually referred to as static and dynamic disorder). Because of the existence of disorder in the structure, each of the otherwise chemically identical BChl molecules shares slightly different properties, and they will tend to retain their individual physico-chemical properties despite their interactions. This, in turn, will tend to localize the excited state onto a subset of BChls within the interacting ensemble. It is worth underlining here that an appropriate estimation of the disorder in the LH complexes from purple bacteria constitutes a crucial point in many respects. Firstly, it is essential for the proper understanding and modeling of the electronic properties of these proteins (in particular their fluorescence properties). Second, the whole physical description of the dynamics of the excitation energy in these complexes will depend on the extent of disorder. Depending on whether the excitation energy is localized on a subset of molecules, or whether it is entirely delocalized on the BChl ring, the physical processes which will underlie its evolution will be entirely different. In the first case, absorption of the photon will be followed by an ultrafast relaxation process, and then by incoherent hopping of the excitation energy from a subset to another. In the second case, the whole system may be described as one unique supermolecule, with electronic levels given by diagonalization of the interaction Hamiltonian. On this one supermolecule, the only phenomenon possible is the relaxation between these energy levels, through energy exchange with vibrations of the system (phonon-induced relaxation). The LH complexes are probably the biological macromolecules best-suited for studying disorder in biological system. This is due to the fact that i) the events in these proteins are on the same timescale as the protein motion and ii) the interacting BChl molecules constitute a system exquisitely sensitive to disorder. As stated above, in the absence of disorder, LH2 complexes should exhibit no fluorescence at very low temperature. Measuring the fluorescence properties
Bruno Robert of LH2 as a function of temperature is probably one of the best methods to evaluate the consequence of disorder on the excitonically-coupled systems in these proteins. This yields a value for the disorder about two times larger than the pigment-pigment coupling strength (van Grondelle et al., 1997), and a localization of the excitation in the 850-nm absorbing ring over a few (2–4) BChl molecules (Monshouwer et al., 1997). Similar values have been deduced from a variety of spectroscopic measurements ( Pullerits et al., 1996; Kennis et al., 1997a; Kuhn and Sundström, 1997). The LH rings thus constitute systems of strongly coupled pigments, where the disorder is large enough to destroy the delocalization of the exciton over the whole set of molecules. However, neither can they be fully considered as assemblies of these small sets of pigments, as each of these is still part of the whole circular array. Circular dichroism (CD) measurements clearly reveal the dual character of these rings. In the strong, conservative CD signal of LH2 (Cogdell and Scheer, 1985) the zero crossing point does not coincide with the absorption maximum but is slightly shifted to the red (Sauer et al., 1996; Koolhaas et al., 1997, 2000). Furthermore, in a mutant of Rba. sphaeroides lacking the B800 BChls, a broad, weak, negative CD-band around 780 nm (Koolhaas et al., 1998) was observed which was ascribed to the upper exciton band of the B850 ring. The contribution to the LH2 CD signal from neighboring BChl molecules is small, as they have almost parallel transition dipoles. Modeling the CD from LH2 thus requires taking into account all of the BChl molecules of the 850 nm ring, as the maximum CD contribution originates from interactions between dipoles separated by about a quarter of the ring (Koolhaas et al., 1997). Thus, although disorder induces a localization of the excitation on a few molecules, the ensemble of BChl/BChl interactions in the whole ring is necessary to model the CD of the LH2 (Somsen et al., 1996; Koolhaas et al., 1998; Georgakopoulou et al., 2002). LH1 proteins, in which similar or stronger BChl-BChl interactions would be expected, exhibit a more complex, weaker, non-conservative signal in the 875 nm region, which may vary from complex to complex. Modeling the CD of LH1 has proven to be an extremely difficult task (Georgakopoulou et al., 2006), probably because the transition dipoles in LH1 are oriented even more in the plane of the ring than in LH2. As a consequence, all contributions to the CD cancel and the signal becomes sensitive to small variations in the orientation and position of the transition dipoles. Of course, disorder-induced localization of the
Chapter 11
Spectroscopic Properties of Antenna Complexes
excitation should wipe out most of the details of the exciton manifold which can be deduced from the symmetry of the LH structure. In addition, structural differences such as i) ellipiticity of the LH1 induced by the interaction with the reaction centers (Jamieson et al., 2002; Scheuring et al., 2003), or ii) interruption of the ring by the PufX component (Roszak et al., 2003) or even iii) absence of part of the subunits in an LH1 ring (Bahatyrova et al., 2004a) should not have any detectable influence on the electronic properties of the complexes, which should be primarily dominated by the disorder. This view has been challenged by fluorescence experiments on single LH complexes performed at very low temperature. Fluorescence excitation spectra of single LH2 in the 850 nm region suggest the presence of two perpendicularly polarized transitions (Bopp et al., 1999; van Oijen et al., 1999a; see also Chapter 44, Köhler, for a discussion of single molecule methodology and the results on LH2 complexes). This would indicate that disorder, either static or dynamic, is not large enough to fully destroy excitation collectiveness in LH2. However, in these experiments, the splitting between the two transitions is too large to be accounted by the X-ray crystallographic structure of the complexes. It was thus further concluded that the BChl/BChl coupling was modulated by an elliptical deformation of the B850 ring (van Oijen et al., 1999b; Ketelaars et al., 2001; Matsushita et al., 2001). In the complete absence of experimental evidence of such LH2 structures (see Chapter 8, Gabrielsen et al.; Chapter 14, Sturgis and Niederman, and references therein) definitive interpretation of these experiments is probably still quite open. More recently it has been proposed, still from single protein fluorescence experiments at low temperature, that interrupting the LH1 ring by the PufX protein would result in a change in the LH1 symmetry which would, in turn, affect the exciton manifold of the 32 (or less) B875 BChl molecules (Richter et al., 2007). Indeed, in the exciton picture, breakage of the perfect symmetry should result in a gain of oscillator strength by the lowest excitonic level. A sharp transition was observed in the LH1 from Rps. palustris, where crystallographic studies suggest that the LH1 ring is indeed interrupted by an equivalent of the PufX protein is present (Roszak et al., 2003). Such a sharp transition was attributed to a gain of oscillator strength by the lowest excitonic level, due to symmetry breaking. These experiments tend to favor a picture of LH complexes where the disorder has little influence, if at all, on the exciton manifold of the excitonically coupled BChl molecules. However, one may wonder whether such a detailed conclusion
207
may be drawn from these experiments. LH complexes from Rps. palustris are poorly characterized. The fact that it is possible to obtain a crystallographic structure from a LH preparation does not, in the absence of detailed spectroscopic characterization, indicate that the obtained structure is representative of the average structure present in biochemical preparations of LH1, and even less that it is relevant for describing the structure of the few LH1 molecules studied by single molecule fluorescence. Again, single molecule fluorescence experiments challenge the commonly accepted view of LH as a disorder-dominated structure, but not in an undisputable way. In contrast, room temperature fluorescence experiments performed on LH complexes have reported sudden changes in the fluorescence spectra (Rutkauskas et al., 2004, 2005, 2006). Such large changes strongly suggest that LH2 goes through a number of conformational substates at room temperature, thus reinforcing the idea that the dynamic disorder is large in these complexes. A quantitative analysis of these results was recently proposed, based on the modified Redfield theory and the disordered exciton model (Novoderezhkin et al., 2006; see also Chapter 13, van Grondelle and Novoderezhkin). Interestingly, from such a model, developed for interpreting single molecule fluorescence, it may be predicted that the electronic absorption transition of the LH2 ensemble should exhibit a temperature dependence between 0 and 60 °C (L. Valkunas, personal communication), which has been recently observed (J. Seguin and M. Paternostre, personal communication). VI. Chromophore Interactions in LightHarvesting Proteins: Additional Effects In the previous paragraphs, we have only considered the interactions between BChl molecules from an excitonic point of view. Excitonic interactions are likely to be one of the dominant parameters for tuning the electronic absorption of the LH complexes. However, other aspects of BChl/BChl interactions must be considered. In the B850/B875 BChl rings from LH2/LH1, each BChl molecule provides an environment for its neighbors. In these rings, the molecular situation is thus different from, for example, that of the B800 BChl molecules. As the spectroscopic properties of the BChl molecules are expected to be sensitive to influences from their surrounding environment, solvation mechanisms may play a role in tuning the electronic properties of LH proteins. It is clear
Bruno Robert
208 that BChl/BChl interactions influence the shape of the lower energy transition in these proteins. Rapid fluctuations in the environment, related to vibrations of the pigments and the surrounding protein, indeed define the homogeneous line shape of the electronic transitions (see Chapter 13, van Grondelle and Novoderezhkin). The nature of the environment of the B800 and B850 BChl molecules being different, the homogeneous line shapes of their electronic transitions are also different (Urboniene et al., 2007). In a sense, BChl/BChl interactions in BChl rings partly ‘shape’ the lowest electronic transitions in LH complexes. On the other hand, the very tight packing of the BChl in LH1 and LH2 rings may lead to a quite complex situation. The BChl/BChl interactions, eventually in combination with the asymmetric electrostatic environment provided by the protein, may perturb the properties of the excited states. The asymmetry of the electron density in the excited state may become larger, and the excited state may, in these conditions exhibit charge-transfer character. For both LH1 and LH2, it actually was shown by Stark spectroscopy that the Qy transitions exhibit a strong charge-transfer character, with LH1 significantly larger than LH2 (Gottfried et al., 1991; Beekman et al., 1997a,b). This charge-transfer character is not present in the Qy transitions of monomeric pigments, and must thus arise from the influence of αβ polypeptides on the BChl molecule array. It was proposed that the existence of such a charge-transfer character of the Qy transition of LH1 and LH2 is one of the reasons why this transition experiences such a large red-shift in these proteins (Somsen et al., 1998). Measurements conducted at high hydrostatic pressure, have shown that the Qy transition of LH1 and LH2 gradually shift to the red according to pressure (Timpmann et al., 2001), This pressure dependence of the Qy transition may be very large (as large as 1000 cm–1/GPa in LH1 (Sturgis et al., 1998)) although the conformations and the specific protein/BChl interactions are generally not affected below 0.7 GPa (Sturgis et al., 1998; Gall et al., 2003) (Fig. 4). It is worth noting that there seems to be a correlation between the sensitivity to pressure of the Qy transition of the complex and its charge-transfer character, as deduced from Stark measurements. This suggests that charge-transfer character is an important parameter in the pressure behavior of these transitions. Additional experiments are however needed to clarify the causes of the pressure-induced red-shift of the LH complexes, and they
Fig. 4. Pressure-induced redshift on the lower energy transition of LH1 from Rsp. rubrum, strain S1. Pressures at which each spectrum was recorded is indicated in GPa on the figure.
may clarify the role of the charge-transfer character in tuning the absorption of LH complexes. VII. Conclusions LH complexes of purple photosynthetic bacteria are biological molecules which have been characterized at an unprecedented level of physics and physical chemistry. In recent years, theoretical and experimental developments have reached such a point that most of the electronic properties of these proteins are understood. There are, however, a number of open questions which require further developments. Among these questions, the most important one is probably which structural parameters determine that an LH protein is an LH1 or an LH2, and how precisely the difference in the structure of these complexes determines the difference in their spectroscopic properties. The few experiments conducted in this direction (Hu et al., 1998) have not yet yielded any clearly confirmed conclusion. Many structural factors may play a role in determining the nature of LH complexes, such as the presence of an additional BChl ring in LH2, or the different organization of the H-bond network between the polypeptides and the BChl responsible for the lower energy transition. If it is now quite clear that most of the differences between the lower energy electronic transitions of LH1 and LH2 arise from the difference in the orientation of the BChl dipoles both relative to each other and in the plane of the LH ring, it is often difficult to accurately (and uniquely) relate these parameters with the spectroscopic properties of large and small circular arrays of
Chapter 11
Spectroscopic Properties of Antenna Complexes
BChl molecules. This is due, in particular, to the fact that only small (16–18) and large (32) circular arrays of BChl molecules have been accessible to experimental work. We are clearly at a point, in this field, where new proteins have to be characterized, which would comprise intermediate circular oligomers of polypeptides, and in which would exist circular arrays of BChl molecules of intermediate size, with symmetries that differ from those already described. There have been some preliminary reports that such LH complexes do exist. Their extensive characterization will certainly constitute, in a near future, the basis for new breakthroughs in the description and modeling of the spectroscopic properties of light-harvesting proteins from purple photosynthetic bacteria. Acknowledgments The author thanks the French Agence Nationale de la Recherche (programs Mastrit and Caroprotect) and the European Union (program INTRO2) for financial support. References Alden RG, Johnson E, Nagarajan V, Parson WW, Law CJ, and Cogdell RJC (1997) Calculations of spectroscopic properties of the LH2 bacteriochlorophyll-protein antenna complex from Rhodopseudomonas acidophila. J Phys Chem B 101: 4667–4680 Bahatyrova S, Frese RN, van der Werf KO, Otto C, Hunter CN and Olsen JD (2004a) Flexibility and size heterogeneity of the LH1 light harvesting complex revealed by atomic force microscopy Functional significance for bacterial photosynthesis. J Biol Chem 279: 21327–21333 Bahatyrova S, Frese RN, Siebert, CA, van der Werf KO, van Grondelle R, Niederman RA, Bullough PA, Otto C, Olsen JD and Hunter CN (2004b) The native architecture of a photosynthetic membrane. Nature 430: 1058–1062 Beekman LMP, Frese RN, Fowler GJS, Ortiz de Zarate I, Cogdell RJ, van Stokkum I, Hunter CN and van Grondelle R (1997a) Characterization of the light-harvesting antennas of photosynthetic purple bacteria by Stark spectroscopy. 2. LH2 complexes: Influence of the protein environment. J Phys Chem B 101: 7293–7301 Beekman LMP, Steffen M, van Stokkum I, Olsen JD, Hunter CN, Boxer SG and van Grondelle R (1997b) Characterization of the light-harvesting antennas of photosynthetic purple bacteria by Stark spectroscopy. 1. LH1 antenna complex and the B820 subunit from Rhodospirillum rubrum. J Phys Chem B 101: 7284–7292 Bopp MA, Sytnik A, Howard TD, Cogdell RJ and Hochstrasser RM (1999) The dynamics of structural deformations of im-
209
mobilized single light-harvesting complexes. Proc Natl Acad Sci USA 96: 11271–11276 Brunisholz RA, Bissig I, Niederer E, Suter F and Zuber H (1987) Structural studies on the light-harvesting polypeptides of Rp. acidophila. In: Biggins J (ed) Progress in Photosynthesis Research, Proceedings 7th International Congress of Photosynthesis, pp 13–16. Nijhoff, Dordrecht Clayton RK (1963) Towards the isolation of a photochemical reaction center in Rhodopseudomonas spheroides. Biochim Biophys Acta 75: 312–323 Chachisvilis M, Kuehn O, Pullerits T and Sundström V (1997) Excitons in photosynthetic purple bacteria: Wavelike motion or incoherent hopping? J Phys Chem B 101: 7275–7283 Cogdell RJ and Scheer H (1985) Circular dichroism of lightharvesting complexes from purple photosynthetic bacteria. Photochem Photobiol 42: 669–678 Cogdell RJ, Hawthornthwaite AM, Evans MB, Ferguson LA, Kerfeld C, Thornber JP, Van Mourik F and van Grondelle R (1990) Isolation and characterization of an unusual antenna complex from the marine purple sulfur photosynthetic bacterium Chromatium purpuratum BN5500. Biochim Biophys Acta 1019: 239–244 Fathir I, Ashikaga M, Tanaka K, Katano T, Nirasawa T, Kobayashi M, Wang Z-Y and Nozawa T (1998) Biochemical and spectral characterization of the core light harvesting complex 1 (LH1) from the thermophilic purple sulfur bacterium Chromatium tepidum. Photosynth Res 58: 193–202 Fowler GJS, Visschers RW, Grief GG, van Grondelle R and Hunter CN (1992) Genetically modified photosynthetic antenna complexes with blueshifted absorbance bands. Nature 355: 848–850 Fowler GJS, Sockalingum GD, Robert B and Hunter CN (1994) Blue shifts in bacteriochlorophyll absorbance correlate with changed hydrogen bonding patterns in light-harvesting 2 mutants of Rhodobacter sphaeroides with alterations at α-Tyr-44 and α-Tyr-45. Biochem J 299: 695–700 Fowler GJS, Hess S, Pullerits T, Sundström V and Hunter CN (1997) Role of β Arg-10 in the B800 bacteriochlorophyll and carotenoid pigment environment within the light-harvesting LH2 complex of Rhodobacter sphaeroides. Biochemistry 36: 11282–11291 Francia F, Wang J, Venturoli G, Melandri BA, Barz WP and Oesterhelt D (1999) The reaction center-LH1 antenna complex of Rhodobacter sphaeroides contains one PufX molecule which is involved in dimerization of this complex. Biochemistry 38: 6834–6845 French CS (1940) The pigment protein compound of photosynthetic bacteria. J Gen Physiol 23: 483–494 Frese RN, Olsen JD, Branvall R, Westerhuis WHJ, Hunter CN and van Grondelle R (2000) The long-range supraorganization of the bacterial photosynthetic unit: A key role for PufX. Proc Natl Acad Sci USA 97: 5197–5202 Gall A and Robert B (1999) Characterization of the different peripheral light-harvesting complexes from high- and low-light grown cells from Rhodopseudomonas palustris. Biochemistry 38: 5185–5190 Gall A, Fowler GJS, Hunter CN and Robert B (1997) Influence of the protein binding site on the absorption properties of the monomeric bacteriochlorophyll in Rhodobacter sphaeroides LH2 complex. Biochemistry 36: 16282–16287 Gall A, Yurkov V, Verméglio A and Robert B (1999) Certain species
210 of the Proteobacteria possess unusual bacteriochlorophyll a environments in their light-harvesting proteins. Biospectroscopy 5: 338-345 Gall A, Ellervee A, Sturgis JN, Fraser NJ, Cogdell RJ, Freiberg A and Robert B. (2003) Membrane protein stability: High pressure effects on the structure and chromophore-binding properties of the light-harvesting complex LH2. Biochemistry 42: 13019–13026 Garcia D, Parot P, Verméglio A and Madigan MT (1986) The light-harvesting complexes of a thermophilic purple sulfur photosynthetic bacterium Chromatium tepidum. Biochim Biophys Acta 850: 390–395 Georgakopoulou S, Frese RN, Johnson E, Koolhaas MHC, Cogdell RJ, van Grondelle R and van der Zwan G (2002) Absorption and CD spectroscopy and modeling of various LH2 complexes from purple bacteria. Biophys J 82: 2184–2197 Georgakopoulou S, van der Zwan G, Olsen JD, Hunter CN, Niederman RA and van Grondelle R (2006) Investigation of the effects of different carotenoids on the absorption and CD signals of light harvesting 1 complexes. J Phys Chem B. 110: 3354–3361 Germeroth L, Lottspeich F, Robert B and Michel H (1993) Unexpected similarities of the B800-850 light-harvesting complex from Rhodospirillum molischianum to the B870 light-harvesting complexes from other purple photosynthetic bacteria. Biochemistry 32: 5615–21 Gottfried DS, Stocker JW and Boxer SG (1991) Stark effect spectroscopy of bacteriochlorophyll in light-harvesting complexes from photosynthetic bacteria. Biochim Biophys Acta 1059: 63–75 Gudowska-Nowak E, Newton MD and Fajer J (1990) Conformation and environmental effects on bacteriochlorophyll optical spectra: Correlations of calculated spectra with structural studies. J Phys Chem 94: 5795–5801 Hartigan N, Tharia HA, Sweeney F, Lawless AM and Papiz MZ (2002) The 7.5-Å electron density and spectroscopic properties of a novel low-light B800 LH2 from Rhodopseudomonas palustris. Biophys J. 82: 963–977 Hu Q, Sturgis JN, Robert B, Delagrave S, Youvan DC and Niederman RA (1998) Hydrogen bonding and circular dichroism of bacteriochlorophylls in the Rhodobacter capsulatus lightharvesting II complex altered by combinatorial mutagenesis. Biochemistry 37: 10006–10015 Isaacs NW, Cogdell RJ, Freer AA and Prince SM (1995) Lightharvesting mechanisms in purple photosynthetic bacteria. Curr Opin Struct Biol 5:794–797 Jamieson SJ, Wang P, Qian P, Kirkland JY, Conroy MJ, Hunter CN and Bullough PA. (2002) Projection structure of the photosynthetic reaction centre-antenna complex of Rhodospirillum rubrum at 8.5 Å resolution. EMBO J 21: 3927–3935 Jimenez R and Fleming GR (1996) Ultrafast spectroscopy of photosynthetic systems. Adv Photosynth 3: 63–73 Jimenez R, van Mourik F, Yu JY and Fleming GR (1997) Threepulse photon echo measurements on LH1 and LH2 complexes of Rhodobacter sphaeroides: A nonlinear spectroscopic probe of energy transfer. J Phys Chem B 101: 7350–7359 Jungas C, Ranck J-L, Rigaud J-L, Joliot P and Verméglio A (1999) Supramolecular organization of the photosynthetic apparatus of Rhodobacter sphaeroides. EMBO J 18: 534–542 Karrasch S, Bullough PA and Ghosh R (1995) The 8.5 Å projection map of the light-harvesting complex I from Rhodospiril-
Bruno Robert lum rubrum reveals a ring composed of 16 subunits. EMBO J 14: 631–368 Kennis JTM, Streltsov AM, Permentier H, Aartsma TJ and Amesz J (1997a) Exciton coherence and energy transfer in the LH2 antenna complex of Rhodopseudomonas acidophila at low temperature. J Phys Chem B 101: 8369–8374 Kennis JTM, Streltsov AM, Vulto SIE, Aartsma TJ, Nozawa T and Amesz J (1997b) Femtosecond dynamics in isolated LH2 complexes of various species of purple bacteria. J Phys Chem B 101: 7827–7834 Ketelaars M, Van Oijen AM, Matsushita M, Köhler J, Schmidt J and Aartsma TJ (2001) Spectroscopy on the B850 band of individual light-harvesting 2 complexes of Rhodopseudomonas acidophila I. experiments and Monte Carlo simulations. Biophys J 80: 1591–1603 Koepke J, Hu X, Muenke C, Schulten K and Michel H (1996) The crystal structure of the light-harvesting complex II (B800-850) from Rhodospirillum molischianum. Structure 4: 581–597 Koolhaas MHC, van der Zwan G, Frese RN and van Grondelle R (1997) Red shift of the zero crossing in the CD spectra of the LH2 antenna complex of Rhodopseudomonas acidophila: A structure-based study. J Phys Chem B 101: 7262–7270 Koolhaas MHC, Frese RN, Fowler GJS, Bibby TA, Georgakopoulou S, van der Zwan G, Hunter CN, and van Grondelle R (1998) Identification of the upper exciton component of the B850 bacteriochlorophylls of the LH2 antenna complex, using a B800-free mutant of Rhodobacter sphaeroides. Biochemistry 37: 4693–4698 Koolhaas MHC, van der Zwan G and van Grondelle R (2000) Local and nonlocal contributions to the linear spectroscopy of light-harvesting antenna systems. J Phys Chem B 104: 4489–4502 Kramer HJM, van Grondelle R, Hunter CN, Westerhuis WHJ and Amesz J (1984) Pigment organization of the B800-850 antenna complex of Rhodopseudomonas sphaeroides. Biochim Biophys Acta 765: 156–165 Kuhn O and Sundström V (1997) Pump-probe spectroscopy of dissipative energy transfer dynamics in photosynthetic antenna complexes: A density matrix approach. J Chem Phys 107: 4154–4164 Kwa LG, Garcia-Martin A, Vegh AP, Strohmann B, Robert B and Braun P. (2004) Hydrogen bonding in a model bacteriochlorophyll-binding site drives assembly of light harvesting complex. J Biol Chem. 279:15067–15075 Krikunova M, Kummrow A, Voigt B, Rini M, Lokstein H, Moskalenko, AA, Scheer H, Razjivin A and Leupold D (2002) Fluorescence of native and carotenoid-depleted LH2 from Chromatium minutissimum originating from simultaneous two-photon absorption in the spectral range of the presumed (optically ‘dark’) S1 state of carotenoids. FEBS Lett 530: 227–229 Lapouge K, Näveke A, Gall A, Seguin J, Scheer H, Sturgis J and Robert B (1999) Conformation of bacteriochlorophyll molecules in photosynthetic proteins from purple bacteria. Biochemistry 38: 11115–11121 Lapouge K, Näveke A, Robert B, Scheer H and Sturgis JN (2000) Exchanging cofactors in the core antennae from purple bacteria: structure and properties of Zn-bacteriopheophytin-containing LH1. Biochemistry 39: 1091–1099 Matsushita M, Ketelaars M, Van Oijen AM, Köhler J, Aartsma TJ and Schmidt J (2001) Spectroscopy on the B850 band of
Chapter 11
Spectroscopic Properties of Antenna Complexes
individual light-harvesting 2 complexes of Rhodopseudomonas acidophila II. Exciton states of an elliptically deformed ring aggregate. Biophys J 80: 1604–1614 McDermott G, Prince SM, Freer AA, Hawthornthwaite-Lawless AM, Papiz MZ, Cogdell RJ and Isaacs NW (1995) Crystal structure of an integral membrane light-harvesting complex from photosynthetic bacteria. Nature 374: 517–521 McLuskey K, Prince SM, Cogdell RJ and Isaacs NW (1999) Crystallization and preliminary X-ray crystallographic analysis of the B800-820 light-harvesting complex from Rhodopseudomonas acidophila strain 7050. Acta Cryst D55: 885–887 Molisch H (1907) Die Purpurbakterien nach neuen Untersuchungen. Gustav Fisher, Jena Monshouwer R and van Grondelle R (1996) Excitations and excitons in bacterial light-harvesting complexes. Biochim Biophys Acta 1275: 70–75 Monshouwer R, Abrahamsson M, van Mourik F and van Grondelle R (1997) Superradiance and exciton delocalization in bacterial photosynthetic light-harvesting systems. J Phys Chem B 101: 7241–7248 Nagarajan V, Johnson ET, Williams JC and Parson WW (1999) Femtosecond pump-probe spectroscopy of the B850 antenna complex of Rhodobacter sphaeroides at room temperature. J Phys Chem B 103: 2297–2309 Novoderezhkin VI, Rutkauskas D and van Grondelle R (2006) Dynamics of the emission spectrum of a single LH2 complex: Interplay of slow and fast nuclear motions. Biophys J 90: 2890–2902 Olsen JD, Sockalingum GD, Robert B and Hunter CN (1994) Modification of a hydrogen bond to a bacteriochlorophyll a molecule in the light-harvesting 1 antenna of Rhodobacter sphaeroides. Proc Natl Acad Sci USA 91: 7124–7128 Olsen JD, Sturgis JN, Westerhuis WHJ, Fowler GJS, Hunter CN and Robert B (1997) Site-directed modification of the ligands to the bacteriochlorophylls of the light-harvesting LH1 and LH2 complexes of Rhodobacter sphaeroides. Biochemistry 36: 12625–12632 Pardee AB, Schachman HK and Stanier RY (1952) Chromatophores of Rhodospirillum rubrum. Nature 169: 282–283 Prince SM, Papiz MZ., Freer AA, McDermott G, Hawthornthwaite-Lawless AM, Cogdell RJ, and Isaacs NW (1997) Apoprotein structure in the LH2 complex from Rhodopseudomonas acidophila strain 10050: Modular assembly and protein pigment interactions. J Mol Biol 268: 412–423 Pullerits T, Chachisvilis M and Sundström V (1996) Exciton delocalization length in the B850 antenna of Rhodobacter sphaeroides. J Phys Chem 100: 10787–10792 Qian P, Hunter CN and Bullough PA (2005) The 8.5Å projection structure of the core RC-LH1-PufX dimer of Rhodobacter sphaeroides. J Mol Biol 349: 948–960 Richter MF, Baier J, Prem T, Oellerich S, Francia F, Venturoli G, Oesterhelt D, Southall J, Cogdell RJ and Köhler J (2007) Symmetry matters for the electronic structure of core complexes from Rhodopseudomonas palustris and Rhodobacter sphaeroides PufX−. Proc Natl Acad Sci USA. 104: 6661–6665 Ridge JP, Fyfe PK, McAuley KE, van Brederode ME, Robert B, van Grondelle R, Isaacs NW, Cogdell RJ and Jones MR (2000) An examination of how structural changes can affect the rate of electron transfer in a mutated bacterial photoreaction centre. Biochem J 352: 567–578 Robert B and Lutz M (1985) Structures of antenna complexes of
211
several Rhodospirillales from their resonance Raman spectra. Biochim Biophys Acta 807: 10–23 Robert B, Andrianambinintsoa S and Lutz M. (1985) Structural characterization of high 800 nm-absorbing light-harvesting complexes from Rhodospirillales from their resonance Raman spectra. J. Biochem. 98: 349–354 Roszak AW, Howard TD, Southall J, Gardiner AT, Law CJ, Isaacs, NW and Cogdell RJ (2003) Crystal structure of the RC-LH1 core complex from Rhodopseudomonas palustris. Science 302: 1969–1972 Rutkauskas D, Novoderezhkin V, Cogdell RJ and van Grondelle R (2004) Fluorescence spectral fluctuations of single LH2 complexes from Rhodopseudomonas acidophila strain 10050. Biochemistry 43: 4431–4438 Rutkauskas D, Novoderezhkin V, Cogdell RJ and van Grondelle R (2005) Fluorescence spectroscopy of conformational changes of single LH2 complexes. Biophys J 88: 422–435 Rutkauskas D, Novoderezhkin V, Gall A, Olsen J, Cogdell RJ, Hunter CN and van Grondelle R (2006) Spectral trends in the fluorescence of single bacterial light-harvesting complexes: Experiments and modified Redfield simulations. Biophys J 90: 2475–2485 Sauer K, Cogdell RJ, Prince SM, Freer A, Isaacs NW and Scheer H (1996) Structure-based calculations of the optical spectra of the LH2 bacteriochlorophyll-protein complex from Rhodopseudomonas acidophila. Photochem Photobiol 64: 564–576 Savage H, Cyrklaff M, Montoya G, Kühlbrandt W and Sinning I (1996) Two-dimensional structure of light harvesting complex II (LHII) from the purple bacterium Rhodovulum sulfidophilum and comparison with LHII from Rhodopseudomonas acidophila. Structure 4: 243–252 Scheuring, S, Seguin J, Marco S, Lévy D, Robert B and Rigaud J-L (2003) Nanodissection and high-resolution imaging of the Rhodopseudomonas viridis photosynthetic core complex in native membranes by AFM. Proc Natl Acad Sci USA 100: 1690–1693 Scheuring S, Rigaud JL and Sturgis JN (2004) Variable LH2 stoichiometry and core clustering in native membranes of Rhodospirillum photometricum. EMBO J 23: 4127–413 Siebert CA, Qian P, Fotiadis D, Engel A, Hunter CN and Bullough PA (2004) Molecular architecture of photosynthetic membranes in Rhodobacter sphaeroides: The role of PufX. EMBO J 23: 690–700 Somsen OJG, Valkunas L and van Grondelle R (1996) A perturbed two-level model for exciton trapping in small photosynthetic systems. Biophys J 70: 669–683 Somsen OJG, Chernyak V, Frese RN, van Grondelle R and Mukamel S (1998) Excitonic interactions and Stark spectroscopy of light harvesting systems. J Phys Chem B 102: 8893–8908 Sturgis JN and Robert B (1996) The role of chromophore coupling in tuning the spectral properties of peripheral light-harvesting protein of purple bacteria. Photosynth Res 50: 5-10 Sturgis JN and Robert B (1997) Pigment binding-site and electronic properties in light-harvesting proteins of purple bacteria. J Phys Chem B 101: 7227–7231 Sturgis JN, Hageman G, Tadros MH, and Robert B (1995a) Biochemical and spectroscopic characterization of the B800-850 light-harvesting complex from Rhodobacter sulfidophilus and its B800-830 spectral form. Biochemistry 34: 10519–10524 Sturgis JN, Jirsakova V, Reiss-Husson F, Cogdell RJ and Robert B (1995b) Structure and properties of the bacteriochlorophyll
212 binding site in peripheral light-harvesting complexes of purple bacteria. Biochemistry 34:517–523 Sturgis JN, Olsen JD, Robert B and Hunter CN (1997) Functions of conserved tryptophan residues of the core light-harvesting complex of Rhodobacter sphaeroides. Biochemistry 36:2772–2778 Sturgis JN, Gall A, Ellervee A, Freiberg A and Robert B (1998) The effect of pressure on the bacteriochlorophyll a binding sites of the core antenna complex from Rhodospirillum rubrum. Biochemistry 37: 14875–14880 Timpmann K, Ellervee A, Pullerits T, Ruus R, Sundström V and Freiberg A (2001) Short-range exciton couplings in LH2 photosynthetic antenna proteins studied by high hydrostatic pressure absorption spectroscopy. J Phys Chem B 105: 8436–8444 Urboniene V, Vrublevskaja O, Trinkunas G, Gall A, Robert B and Valkunas L. (2007) Solvation effect of bacteriochlorophyll excitons in light-harvesting complex LH2 Biophys J 93: 2188–2198 van Grondelle R, Dekker JP, Gillbro T and Sundström V (1994) Energy transfer and trapping in photosynthesis. Biochim Biophys Acta 1187: 1–65 van Grondelle R, Monshouwer R, and Valkunas L (1997) Photosynthetic light-harvesting. Pure Appl Chem 69: 1211–1218 Van Mourik F, Visschers RW and van Grondelle R (1992) Energy transfer and aggregate size effects in the inhomogeneously
Bruno Robert broadened core light-harvesting complex of Rhodobacter sphaeroides. Chem Phys Lett 193: 1–7 Van Oijen AM, Ketelaars M, Köhler J, Aartsma TJ and Schmidt J (1999a) Unraveling the electronic structure of individual photosynthetic pigment-protein complexes. Science 285: 400–402 Van Oijen AM, Ketelaars M, Köhler J, Aartsma TJ and Schmidt J (1999b) Spectroscopy of individual LH2 complexes of Rhodopseudomonas acidophila: Localized excitations in the B800 band. Chem Phys 247: 53–60 Visser HM, Somsen OJG, von Mourik F, Lin S, van Stokkum IHM and van Grondelle R (1995) Direct observation of sub-picosecond equilibration of excitation energy in the light-harvesting antenna of Rhodospirillum rubrum. Biophys J 69: 1083–1099 Walz T and Ghosh R (1997) Two-dimensional crystallization of the light-harvesting I-reaction center photounit from Rhodospirillum rubrum. J Mol Biol 265:107–111 Walz T, Jamieson SJ, Bowers CM, Bullough PA and Hunter CN (1998) Projection structures of three photosynthetic complexes from Rhodobacter sphaeroides: LH2 at 6 Å, LH1 and RC-LH1 at 25 Å. J Mol Biol 282: 833–845 Zuber H, and Brunisholz RA (1991) Structure and function of antenna polypeptides and chlorophyll-protein complexes: principles and variability. In Scheer H (ed) Chlorophylls, pp 627–703. CRC, Boca Raton
Chapter 12 Energy Transfer from Carotenoids to Bacteriochlorophylls Harry A. Frank Department of Chemistry, University of Connecticut, Storrs, CT 06269-3060 U.S.A.
Tomáš Polívka* Institute of Physical Biology, University of South Bohemia, Zamek 136, CZ-373 33 Nove Hrady Czech Republic; and Biological Centre, Czech Academy of Sciences, Czech Republic
Summary ............................................................................................................................................................... 213 I. Introduction..................................................................................................................................................... 214 II. Carotenoid Excited States .............................................................................................................................. 215 III. Energy Transfer in Light-Harvesting 2 Complexes......................................................................................... 216 A. Energy Transfer via the S2 State ...................................................................................................... 216 B. Energy Transfer via the S1 State ...................................................................................................... 219 C. The S* State ..................................................................................................................................... 222 D. Other Pathways................................................................................................................................ 224 E. The Role of B800 ............................................................................................................................. 225 IV. Energy Transfer in Light-Harvesting 1 Complexes and Reaction Centers ..................................................... 226 V. Outlook ........................................................................................................................................................... 227 Acknowledgments ................................................................................................................................................. 227 References ............................................................................................................................................................ 227
Summary The photosynthetic apparatus contains light-harvesting (LH) pigment-protein complexes that capture light energy from the sun and transfer it efficiently to the reaction center. In photosynthetic bacteria, carotenoids supplement the non-optimal LH capacity of bacteriochlorophyll (BChl) in the 400–500 nm region of the visible spectrum. Thus, carotenoid-to-BChl energy transfer provides an essential process for enhancing the ability of these systems to capture light energy and convert it into useful work. Carotenoids have at least two states involved in energy transfer to BChl. These are the S2 state into which absorption from the ground state, S0, is strongly allowed, and a low-lying, S1 state into which absorption is forbidden by symmetry. These two states represent the primary energy donors for carotenoid-to-BChl energy transfer. The S2 state transfers energy with an efficiency between 30 and 70%, the value of which is only slightly dependent on the structure of the carotenoid. The S1-mediated energy transfer pathway depends strongly on the π-electron conjugation length of the carotenoid. This route is essentially closed for carotenoids with eleven or more conjugated carbon-carbon double bonds because in these cases the S1 energy of the carotenoid lies too low to enable transfer to BChl. *Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 213–230. © 2009 Springer Science + Business Media B.V.
214
Harry Frank and Tomáš Polívka
Besides the main S2 and S1 pathways, the past few years of investigations have raised the prospect of other carotenoid excited states participating in energy transfer. The possibilities include vibrationally hot S1 states, a state denoted S* thought to be formed by a branched deactivation pathway from S2, and a state with symmetry representation 1Bu– predicted on the basis of theoretical computations to lie between S1 and S2. This chapter reviews the evidence for these states and discusses their possible involvement as energy donors in the process of light-harvesting in photosynthetic bacteria. I. Introduction Carotenoids in antenna complexes of purple bacteria have multiple functions. Besides their key roles in photoprotection and structure stabilization, purple bacteria utilize carotenoids as efficient LH agents covering the 400–550 nm spectral region. This part of the spectrum is of crucial importance, because it matches the maximum of the solar irradiance curve at the Earth’s surface, and it is where the main photosynthetic pigment of purple bacteria, BChl a, has very low absorption (Fig. 1). Purple bacterial antenna systems are ideal for studying carotenoids as LH pigments in photosynthetic pigment-protein complexes. This is because detailed structural knowledge exists and provides an ideal platform for experimental and theoretical investigations of mechanisms and pathways of energy transfer between carotenoids and BChl. To date, the structures of a few LH complexes have been reported. In 1995, the 2.5 Å structure of an outer antenna complex called LH2 from Rhodopseudomonas (Rps.) acidophila containing the carotenoid rhodopin glucoside (McDermott et al., 1995) became a landmark study of purple bacterial antenna complexes. A 2.4 Å structure of the LH2 complex from Rhodospirillum (Rsp.) molischianum containing lycopene followed shortly after (Koepke et al., 1996). Subsequently, the structure of the related complex LH3 (McLuskey et al., 2001) was reported, as well as the structure of the LH1 antenna complex surrounding the reaction center (Roszsak et al., 2003). These studies revealed the common theme of a ring-shaped protein motif containing either two (LH1) or three (LH2, LH3) BChl a molecules, and one carotenoid molecule per fundamental unit (Gall et al., 2006). The BChl a molecules in LH2 are arranged in two layers. One consists of two strongly coupled BChl a molecules absorbing at 850 nm (B850). The Abbreviations: BChl a – bacteriochlorophyll a; LDS – lithium dodecyl sulfate; CD – circular dichroism; LH – light harvesting; RC – reaction center
second is a monomeric BChl a absorbing at 800 nm (B800). For other complexes, the strongly-coupled BChl a absorb at different wavelengths, giving rise to a B875 band in LH1 or a B820 band in the LH3 complex. The monomeric BChl a molecule absorbs at 800 nm in all complexes except LH1 which lacks this BChl a completely. The carotenoids are in close contact with both types of BChl a molecules allowing them to serve as efficient energy donors in the light-harvesting process. The detailed knowledge about the arrangement of pigments in purple bacterial antenna complexes provided a structural basis for both experimental and theoretical studies of energy transfer between carotenoids and BChl a. The studies of carotenoid to BChl a energy transfer are also facilitated by the fact that many antenna complexes from purple bacteria accumulate only one major carotenoid. However, different species of purple bacteria utilize distinct ca-
Fig. 1. Absorption spectra of LH2 complexes from Rba. sphaeroides containing carotenoids with different conjugation lengths: neurosporene, N = 9 (solid); spheroidene, N = 10 (dotted); spheroidenone N = 10 + C=O (dashed). All spectra are normalized to the maximum of the B850 band.
Chapter 12
Carotenoid-to-BChl Energy Transfer
215
rotenoid types in their antenna complexes (Fig. 2). It was known long before X-ray structures were resolved that the structure of a particular carotenoid affects its ability to carry out energy transfer to BChl a. Using fluorescence excitation spectroscopy, carotenoid-toBChl energy transfer efficiencies of 80–100% have been reported for Rhodobacter (Rba.) sphaeroides containing spheroidene (Cogdell et al., 1981; van Grondelle et al., 1982; Angerhofer et al., 1995), while values between 35–70% have been obtained for Rps. acidophila containing rhodopin glucoside (Angerhofer et al., 1986, 1995; Chadwick et al., 1987). The carotenoid-BChl energy transfer yield drops to ~30% for LH1 of Rsp. rubrum containing spirilloxanthin (Rademaker et al., 1980). These values were later confirmed and/or refined by a variety of time-resolved spectroscopic techniques that enabled the pathways of energy transfer between carotenoids and BChl a to be followed. Despite the fact that this topic has been studied extensively for more than ten years, and summaries of the achievements obtained during this period can be found in specialized reviews (Frank and Cogdell, 1996; Ritz et al., 2000; Fraser et al., 2001; Polívka and Sundström, 2004), the details concerning the mechanisms, pathways, participating donor and acceptor states and the molecular factors controlling energy transfer are still a matter of considerable debate. This is mainly because the excited state properties of carotenoids in solution, which serve as models for interpreting the data obtained from carotenoid excited-state dynamics in purple bacterial antenna complexes, are still not sufficiently well known. Fig. 2. Structures of carotenoids commonly occurring in LH complexes of purple bacteria.
II. Carotenoid Excited States The LH function of carotenoids in the LH2 complex relies on their ability to transfer energy from the low-lying excited states to both Qx and Qy states of BChl a. Consequently, knowledge of the details of the carotenoid excited state manifold, including energies of the excited states and dynamics following the excitation is a necessary prerequisite to understanding carotenoid-BChl energy transfer in antenna complexes of purple bacteria. Much of our current understanding of mechanisms and pathways of carotenoid-BChl energy transfer stems from numerous experimental studies of carotenoids in solution. Until the late nineties much of the photophysics of carotenoids was interpreted in terms of the allowed
S2 state (1B+u in C2h symmetry) responsible for the strong absorption of carotenoids in the 400–550 nm region, and the symmetry-forbidden S1 state (2Ag–), which is located a few thousand reciprocal centimeters below the S2 state. Thus, in contrast to other naturally-occurring dyes, the lowest excited state is a dark state. A number of spectroscopic studies have demonstrated that after being promoted to the S2 state a carotenoid molecule will relax in less than 300 fs to the S1 state. The lifetime of the S1 state is determined by conjugation length of carotenoid and varies from 300 ps for short conjugated chains to ~1 ps for the longest (Polívka and Sundström, 2004). Recent reports, however, have suggested that a
216 few other states denoted as 1Bu– and 3Ag– (Furuichi et al., 2002) and predicted by calculations on polyenes (Tavan and Schulten, 1987), or S* and S‡ (Gradinaru et al., 2001; Papagiannakis et al., 2002; Larsen et al., 2003), and whose origin and properties remain to be firmly established, may be located either between or in proximity to the S1 and S2 states. All of these states represent potential donors in the process of carotenoid-BChl energy transfer in LH2 complexes. In addition, novel excited state properties were recently discovered for carotenoids containing a conjugated carbonyl group. Unlike other carotenoids their photophysics is strongly affected by the polarity of the environment. This behavior was attributed to a presence of another excited state having a charge transfer character (Frank et al., 2000; Zigmantas et al., 2004), and added to the complexity of the carotenoid excited state manifold. Because some carbonyl carotenoids, e.g., spheroidenone or okenone (Fig. 2), are constituents of purple bacterial antenna complexes and reaction centers, the charge transfer state may also be involved in energy transfer processes taking place in these organisms. A scheme showing the current view of the carotenoid excited state ordering together with possible energy transfer pathways to BChl is depicted in Fig. 3. A detailed description of carotenoid excited states in solution is beyond the scope of this chapter and can be found in specialized reviews (Frank et al., 2001; Hashimoto et al., 2004; Koyama et al., 2004; Polívka and Sundström, 2004). III. Energy Transfer in Light-Harvesting 2 Complexes The large complexity of carotenoid excited states described in the previous section was not revealed until the end of the last century. Thus prior suggestions regarding mechanisms and pathways of carotenoidBChl energy transfer in LH2 complexes involved only the S1 and S2 states. Moreover, because of the extremely fast deactivation of the S2 state, the S1 state was suggested to be the primary donor responsible for carotenoid-BChl energy transfer. The forbidden nature of transitions to and from the S1 state imply a negligible transition dipole moment of the S1 state and led to the suggestion that the Dexter electron exchange mechanism should be active in this process (Naqvi, 1980; van Grondelle, 1985; Cogdell and Frank, 1987). However, time-resolved experiments performed in the early nineties showed that following excitation of ca-
Harry Frank and Tomáš Polívka
Fig. 3. Scheme of energy levels of carotenoid and BChl in light harvesting complexes of purple bacteria. Wavy arrows denote intramolecular relaxation processes, while the dashed arrow represents the long-lived BChl a fluorescence. Solid arrows represent the dominating energy-transfer channels involving the S2 and S1 states. The dotted lines represent minor energy-transfer channels that usually contribute only fractionally to the total energy transfer: the pathway via higher vibrational levels of the S1 state, the pathway via the S* state and the Bu– pathway were proposed to be active for some LH2 complexes. See text for details.
rotenoids, the B850 Qy state was populated in less than 200 fs, indicating that, despite its very short lifetime, the S2 state participates as a donor in carotenoid-BChl energy transfer (Trautman et al., 1990; Shreve et al., 1991). This suggestion was supported by calculations carried out by Nagae et al. (1993) who showed that for the carotenoid neurosporene both the S1 and S2 states can efficiently participate in energy transfer. When LH2 structures at atomic resolution became available, a number of experimental and theoretical investigations of energy transfer mechanisms and pathways between carotenoids and BChls bolstered the initial proposal that both the S1 and S2 states are involved in energy transfer. It was also concluded that the precise pathways and directions of energy flow are governed mainly by the conjugation length of the carotenoid. In addition, new experimental approaches and improved methods of data analysis have revealed alternate pathways that may contribute substantially to the overall carotenoid-BChl energy transfer in LH complexes of purple bacteria. A. Energy Transfer via the S2 State Absorption spectra of carotenoids in LH2 complexes resemble those obtained for carotenoids in solution,
Chapter 12
Carotenoid-to-BChl Energy Transfer
except for a red shift of ~1000 cm–1 caused by interaction with the protein environment (Fig. 1). The vibrational sub-structure of the carotenoid S0-S2 spectra of LH2 complexes is usually very similar to that obtained in solution, although for some carotenoids the protein environment represents a confinement of the carotenoid structure, leading to a better resolution of vibrational bands of the spectra (Polívka et al., 2002; Georgakopolou et al., 2004). Early experimental data on energy transfer via the S2 state was obtained by fluorescence up-conversion (Ricci et al., 1996). S2 emission decays of spheroidene in LH2 complexes of Rba. sphaeroides yielded a time constant of 80 fs. Comparing this result with the markedly longer decays in solution (150–250 fs), it was concluded that energy transfer via the S2 state takes place with a time constant of 170 fs, which corresponded to an efficiency of 47%. Thus, these experiments clearly demonstrated that energy transfer via the S2 state can successfully compete with fast S2-S1 internal conversion. On the basis of spectral overlap, it was also concluded that energy transfer via the S2 state occurs to the Qx state of BChl a (Ricci et al., 1996). Similar results were reported for the LH3 complex from Rps. acidophila containing rhodopin glucoside, which has a longer conjugation length than spheroidene (Fig. 2). Analysis of up-conversion kinetics supported by calculations yielded an S2-Qx energy transfer rate of (120–150 fs)–1. Based on this, the S2-Qx channel was concluded to be the dominating one, accounting for 60% of the total energy transfer efficiency (Krueger et al., 1998). Later, a similar value was obtained when up-conversion experiments were carried out on the LH2 complex of Rps. acidophila (Macpherson et al., 2001). An observed 51% efficiency of S2-mediated energy transfer is close to that obtained for LH3, suggesting that small structural differences between LH2 and LH3 complexes have no effect on carotenoid-BChl energy transfer. Instead, energy transfer efficiency appears to be primarily determined by the structure of the carotenoid. The fast and efficient S2 energy transfer pathway was also observed by transient absorption measurements. Although the data obtained from this method are more complicated to analyze than the data obtained from fluorescence up-conversion, recently introduced advanced methods of global fitting data analysis (van Stokkum et al., 2004) allowed for reliable separation of signals originating from different excited states. Application of these methods to transient absorp-
217 tion data measured in a broad spectral range for the LH2 complex from Rps. acidophila determined the efficiency of the S2-mediated energy transfer to be 45–50%; the range is given by the model used for the global fitting of the transient absorption data (Papagiannakis et al., 2006a). Essentially the same value was obtained by Rondonuwu et al. (2004) who concluded that S2 state of rhodopin glucoside in LH2 from Rps. acidophila transfers energy to BChl a with 48% efficiency. A different method of data treatment using evolutionary target analysis determined the efficiency of the S2 pathway in the same system to be 42% (Wohlleben et al., 2003). Thus, for the LH2 complex of Rps. acidophila, independent experiments utilizing different types of data analysis gave values of S2-mediated energy transfer efficiency in the range of 42–51%. This mild discrepancy between different experiments points to an intrinsic problem in determining the S2mediated energy transfer efficiencies. The calculation of efficiency requires a precise knowledge of the rate constant for S2-S1 internal conversion in LH2. However, this depends on the environment, and it is thus impossible to determine an accurate value for the intrinsic (without energy transfer) lifetime of the S2 state in the protein environment. This problem, discussed in detail by Macpherson et al. (2001), puts a limitation on the precision with which the S2 energy transfer efficiency may be determined. Another limitation is the time resolution of the experiments. The measured S2 lifetimes of rhodopin glucoside in LH2 complexes are on the sub-100 fs time scale, which in most cases is faster than the limits of time resolution of the instrument. However, recent experiments employing 10 fs pulses to study carotenoid to BChl a energy transfer in the LH2 complex of Rps. acidophila yielded a value for the efficiency of the S2-mediated energy transfer of 49% (Polli et al., 2006), confirming the values obtained from earlier experiments. It is also worth noting that the 70 fs value of the S2 lifetime of rhodopin glucoside in LH2 was also obtained on the basis of modeling homogeneous linewidths of the absorption and CD bands (Georgakopolou et al., 2004), further supporting the results obtained from time-resolved experiments. Besides the LH2 complex from Rps. acidophila, Papagiannakis et al. (2002) studied S2-mediated energy transfer from spheroidene in the LH2 from Rba. sphaeroides and obtained an energy transfer efficiency of 57%, which is in good agreement with the up-conversion data obtained by Ricci et al. (1996).
218 Another system studied by transient absorption spectroscopy is the LH2 complex from Chromatium purpuratum which binds the carotenoid okenone (Fig. 2). The data obtained on this complex are more complicated to analyze, because the long conjugated chain of okenone (Fig. 2) shifts the 0-0 band of the S2 state so much to the red that it partially overlaps with the Qx band of BChl a (Andersson et al., 1996). Consequently, partial excitation of the Qx band makes the determination of energy transfer efficiency difficult. Moreover, okenone possesses a conjugated carbonyl group that makes the excited state properties sensitive to the polarity of environment (Frank et al., 2000b; Zigmantas et al., 2004). Andersson et al. (1996) reported the S2-mediated energy transfer time from okenone to be 50–100 fs, which taking into account the ~150 fs S2 lifetime of okenone in solution, gave energy transfer efficiencies over 50%. These experiments were recently repeated by Polli et al. (2006) with the significantly better time resolution of 10 fs. These authors compared the excited state dynamics of okenone in a few solvents and in LH2 and concluded that the protein environment is best mimicked by CS2. By comparing the S2 lifetimes of okenone in CS2 and in LH2 they obtained an energy transfer rate constant of (56 fs)–1 yielding a 65% efficiency of the S2 channel (Polli et al., 2006). An even higher efficiency of 70% for the S2-mediated energy transfer was reported by Zhang et al. (2001) for the LH2 complex from Rba. sphaeroides G1C which contains neurosporene (conjugation length N=9). Comparing the results on LH2 complexes containing carotenoids with N=9-13, there is a trend to higher efficiency of S2 energy transfer for shorter carotenoids. This trend is supported by experiments on LH2 of the carotenoidless mutant Rba. sphaeroides R26.1 incorporated with spheroidene analogs of different conjugation lengths (Desamero et al., 1998). Using S2-S1 internal conversion rates measured in solution, overall carotenoid-BChl energy transfer efficiencies, and S1-mediated energy transfer efficiencies, the efficiencies of S2 energy transfer were calculated. Although the LH2s with spheroidene analogs having N=8,9 were observed to have slightly less efficient S2 energy transfer than LH2 with spheroidene (N=10), the systematic decrease of S2 efficiency from 50% (spheroidene) to 12% for the analog having N=13, supported the observed trend of decreasing S2-mediated energy transfer efficiency with increasing conjugation length (Desamero et al., 1998). On the other hand, essentially no change in
Harry Frank and Tomáš Polívka the efficiency of the S2 channel was found by Rondonuwu et al. (2004), who explored a series of LH2 complexes possessing different carotenoids. For the S2-mediated energy transfer, no dependence on carotenoid structure and conjugation length was found. A narrow 46–48% range of values was determined for neurosporene (N=9) and spheroidene (N=10) in LH2 from Rba. sphaeroides, lycopene (N=11) in LH2 from Rsp. molischianum, and rhodopin glucoside (N=11) in LH2 from Rps. acidophila (Koyama et al., 2004; Rondonuwu et al., 2004). For neurosporene and spheroidene these values are significantly lower than those obtained in other experiments (Zhang et al., 2001; Papagiannakis et al., 2002), which is likely caused by additional decay channels included in the analysis performed by Rondonuwu et al. (2004) who assumed other carotenoid excited states act as energy donors in addition to the S2 state. These potential additional energy transfer channels are discussed below. Although the precise values of the S2 transfer efficiencies are still a matter of debate, mainly because of possible involvement of other states, it is clear that the S2 energy transfer route operates with efficiencies in the range 30–70% in all purple bacterial antenna complexes studied so far. Thus, the S2 channel successfully competes with S2-S1 internal conversion for a number of LH2 complexes. This competition of S2 energy transfer with the S2-S1 internal conversion has been rationalized theoretically. The contribution of the electron-exchange Dexter energy transfer mechanism to the S2 energy transfer pathway has been computed to be negligible with the Förster-type mechanism dominating (Nagae et al., 1993; Krueger et al., 1998; Damjanovic et al., 1999). Even without detailed structural information, Nagae et al. (1993) calculated the Coulombic couplings between the S0-S2 transition of neurosporene and the Qx transition of BChl a for a few hypothetical configurations. They concluded that S2 energy transfer can be faster than 100 fs, if the proper orientation of donor and acceptor is realized (Nagae et al., 1993). Determination of the LH2 structure provided detailed information about the mutual orientation of pigments within LH2, allowing more precise calculations of couplings between carotenoids and BChl a. Krueger et al., applied an advanced method, the so-called transition density cube method, to calculate full Coulombic couplings between pigments in both LH3 (Krueger et al., 1998a) and LH2 (Krueger et al., 1998b) complexes from Rps. acidophila. This method replaces the vector
Chapter 12
Carotenoid-to-BChl Energy Transfer
description of the transition dipole moments by threedimensional transition density volumes, and gives a more accurate account of the interaction between molecules (Krueger et al., 1998b). The couplings of the S0-S2 transition of rhodopin glucoside with all possible transitions of neighboring BChl a were calculated for the LH3 complex, yielding values larger than 100 cm–1 for β-B820 Qx, B800 Qy and also for the B800 Qy transition of BChl a located in the neighboring building unit. However, due to the small value of the spectral overlap integral for the Qy states, only the S2-B820 Qx yielded an appreciable energy transfer rate constant of (240 fs)–1 (Krueger et al., 1998a). Similar results were obtained for the LH2 complex. Although the actual couplings were slightly different from those calculated for the LH3 complex, the S2-B850 Qx channel again represented the dominant route with an energy transfer rate constant of (135 fs)–1 (Krueger et al., 1998b). The results of these calculations are in very good agreement with the observed depopulation rates of the S2 state, but predict no S2 transfer to the B800 BChl a molecule which contradicts experimental observation (Macpherson et al., 2001). On the other hand, similar calculations using full Coulombic couplings performed on the basis of the LH2 structure of Rsp. molischianum containing the carotenoid lycopene, yielded appreciable couplings of the S0-S2 transition with both the B850 and B800 BChls, resulting in comparable S2 energy transfer rate constants of ~(200 fs)–1 for both B850 and B800 acceptors (Damjanovic et al., 1999). A similar result was obtained by calculations of the lycopene-BChl couplings in LH2 by means of the collective electronic oscillators algorithm (Tretiak et al., 2000). It is interesting to note that these calculations also proposed strong coupling between the Bx (Soret) band of both B800 and B850 BChls and the S2 state of lycopene, indicating a possibility of efficient energy transfer from BChl a Soret to lycopene (Tretiak et al., 2000), but no experimental evidence for such transfer has been given so far. Also, calculations of couplings were carried out for either isolated molecules in LH2 complexes or in a dielectric medium simulating the mean field created by the protein environment. The effect of the dielectric medium on couplings involving the S2 state were quite significant, underlining the importance of protein effects (Tretiak et al., 2000). An interesting aspect of the carotenoid-BChl energy transfer in LH2 is the fact that by proper shaping of excitation pulses it is possible to control
219 the ratio between S2-S1 internal conversion and energy transfer. However, as was shown for rhodopin glucoside in LH2 from Rps. acidophila, only a decrease in energy transfer efficiency can be achieved. Herek et al. (2002) showed that the S2 population of rhodopin glucoside can be steered to the S2-S1 channel, as demonstrated by increase of the signal due to the S1 state by a factor 1.4. In the coherent control experiments carried out by Herek et al. (2002) the increase in proportion of S2-S1 internal conversion was achieved by optimizing the envelope and phase of the excitation pulses (Herek et al., 2002; Wohlleben et al., 2005). Recently, Papagiannakis et al. (2006a) showed that also incoherent effects, such as annihilation and saturation can lead to similar control of the S2-S1/S2-BChl ratio, demonstrating that the actual value of S2 energy transfer efficiency depends also on the intensity of excitation pulses. B. Energy Transfer via the S1 State Despite the forbidden nature of the S1 state, energy transfer via the S1 route is straightforward to study by means of transient absorption spectroscopy because of the strong S1-Sn band readily observed in the visible spectral region (Polívka and Sundström, 2004). The position and shape of this band is characteristic of each carotenoid (Fig. 4) and the dynamics of this band can be used to determine the S1 lifetimes of carotenoids in various environments. Thus, comparison of S1 lifetimes in solution and in antenna complexes allows for determining energy transfer efficiencies via the S1 state. To study the effect of conjugation length on energy transfer in LH2 complexes, this method was applied to the Rba. sphaeroides R26.1 mutant with incorporated spheroidene analogs of different conjugation lengths (Desamero et al., 1998). Although for conjugation lengths N≥11 the S1 energy transfer was undetectable, quenching of the S1 state due to energy transfer was observed for shorter carotenoids. This result was explained in terms of spectral overlap between hypothetical S1 fluorescence and the B850 absorption band, which becomes small for longer carotenoids. A similar study was carried out by Zhang et al. (2000) using three LH2 complexes with different carotenoids. On the basis of measured S1 lifetimes of neurosporene (N=9), spheroidene (N=10) and lycopene (N=11) in solution and LH2, a significant decrease of S1 energy transfer efficiency was observed; it dropped from 94% (neurosporene) to 82 % (spheroidene) and further to
220
Harry Frank and Tomáš Polívka
Fig. 4. Transient absorption spectra in the S1-Sn region measured 1 ps after excitation into the lowest vibrational band of the carotenoid S2 state for LH2 complexes from Rba. sphaeroides with neurosporene (Neu), spheroidene (Sph) and spheroidenone (Spn). All spectra are normalized to the S1-Sn maximum.
less than 30 % (lycopene), as the conjugation length increased from 9 to 11. The significant drop in energy transfer via the S1 state was also confirmed for spheroidene and rhodopin glucoside in LH2 complexes from Rba. sphaeroides and Rps. acidophila by means of measurements of S1-Sn kinetics either after two-photon excitation of the S1 state (Walla et al., 2000) or after one-photon excitation of the S2 state (Polívka et al., 2002). These authors confirmed the ~1.7 ps S1 lifetime of spheroidene in LH2 indicating energy transfer. For rhodopin glucoside, however, no change in S1 lifetime was observed upon going from solvent to LH2, demonstrating that in this complex the S1 state does not transfer energy (Fig. 5). A few other studies confirmed this trend and established that LH2 complexes utilizing carotenoids with N≥11 are incapable of energy transfer via the S1-route as its efficiency does not exceed 5% (Macpherson et al., 2001; Wohlleben et al., 2003; Koyama et al., 2004; Rondonuwu et al., 2004; Polli et al., 2006). The importance of the carotenoid structure was shown in a study of a Rba. sphaeroides mutant incorporating the longer (N=11) lycopene instead of spheroidene (N=10) naturally present in wild type LH2. When lycopene is present in the complex, the efficiency of S1-mediated energy transfer drops to a few percent, while more than 80% is achieved when spheroidene is present in the wild
Fig. 5. Kinetics recorded at the maximum of the S1-Sn transition after excitation of the 0-0 band of the S0-S2 transition of the carotenoids. The S1 decays of the carotenoids in solution (open symbols) and in the LH2 complex (full symbols) are compared for the LH2 complexes from Rba. sphaeroides (a) and Rps. acidophila (b). The corresponding fits of the kinetics are represented by solid lines. Excitation wavelengths are 515 nm (Rba. sphaeroides) and 525 nm (Rps. acidophila).
type (Billsten et al., 2002a). Further evidence that carotenoids with N=11 are on the edge of capability to transfer energy via the S1 state was provided by studies of LH2 complexes from Rba. sphaeroides R26.1 reconstituted with spirilloxanthin (N=13) (Papagiannakis et al., 2003a) or LH2 complexes from Chromatium purpuratum containing okenone (N~12) (Polli et al., 2006). These studies showed that no (or very little) S1-mediated energy transfer is achieved by these long carotenoids as essentially no S1-mediated energy transfer was observed for spirilloxanthin, and an upper limit of 6% was found for okenone. The obvious relation between the conjugation length and efficiency of the S1-mediated energy transfer can be explained by a decrease of the S1 energy with increasing conjugation length. To achieve efficient energy transfer the S1 energies of carotenoids must be higher than those of the acceptor states.
Chapter 12
Carotenoid-to-BChl Energy Transfer
While the determination of the energy is a trivial task for the S2 state, the forbidden nature of the S1 state prevents a direct measurement of the S1 energy. In the late nineties a few papers appeared describing different methods for determining S1 energies of carotenoids: 1) measurement of the extremely-weak S1 emission (Fujii et al., 1998); 2) measurements of resonance Raman profiles (Sashima et al., 1998); 3) time-resolved measurements of the S1-S2 spectra (Polívka et al., 1999); and 4) two-photon absorption (Krueger et al., 1999). Detection of S1 fluorescence and measurements of resonance Raman profiles, however, could not provide unambiguous information about S1 energies in LH2 and related LH systems. However, since it was known that S1 energies are usually insensitive to solvent properties (except for carbonyl carotenoids that are not typical for purple bacterial LH complexes; Polívka and Sundström, 2004), it was reasonable to assume that the S1 energies in LH2 complexes should be very close to those determined for carotenoids in solution. Under this assumption, S1 energies of carotenoids with N≤10 were calculated to be high enough to allow sufficient overlap between carotenoid S1 emission and Qy bands of BChl a. This conclusion was later verified by direct measurements of the S1 energies of neurosporene, spheroidene and rhodopin glucoside in LH2 complexes by recording the S1-S2 spectra. S1 energies of 14500 cm–1, 13400 cm–1, and 12550 cm–1 were determined for neurosporene, spheroidene, and rhodopin glucoside, respectively (Polívka et al., 2002, 2004), showing that the difference of 850 cm–1 between S1 energies of spheroidene and rhodopin glucoside makes a significant change in the energy transfer pathways in LH2 complexes. These experiments also justified the approximation of S1 energies in LH2 complexes using energies obtained in solution. In the case of spheroidene, the S1 energy extracted from the S1-S2 spectra is the same as that determined by this method in solution (Polívka et al., 2001). For rhodopin glucoside the S1 energy is 250 cm–1 lower than that in solution. This difference was explained by confinement of rhodopin glucoside in the LH2 structure, which narrows the distribution of conformers compared to that in solution. This was also supported by modeling CD and absorption spectra of carotenoids in LH2 complexes (Georgakopoulou et al., 2004). Another direct measurement of S1 energies in the LH2 complex from Rba. sphaeroides was carried out by means of two-photon fluorescence excitation
221 (Krueger et al., 1999). Using this technique the S1 state can be excited directly because the S0-S1 transition is allowed for a two-photon transition. By measuring B850 emission after two-photon excitation of the S1 state achieved by exciting in the 1200–1500 nm spectral range, the two-photon excitation spectrum placed the 0-0 energy of spheroidene in LH2 at 13900 cm–1, confirming the similarity of the S1 energies in LH2 and in solution. Together with measurements of S1 energies, calculations of the couplings between the S0-S1 transition and Qy transitions of both B800 and B850 demonstrated that the S1 transfer rates can be explained in terms of the same energy transfer mechanism as for the S2 route. Although the very small transition dipole moment of the S0-S1 transition led initially to a suggestion that the Dexter mechanism had to be invoked (Naqvi, 1980; van Grondelle, 1985; Cogdell, 1987), more detailed calculations later showed that the Dexter contribution is negligible and that higher-order Coulombic and polarization interactions dominate the S1-mediated energy transfer (Nagae et al., 1993; Scholes et al., 1997; Krueger et al., 1998; Damjanovic et al., 1999; Tretiak et al., 2000). In some studies (Damjanovic et al., 1999; Zhang et al., 2000), the Coulombic couplings for the S1 state were obtained by scaling down the couplings calculated for the S2 state and using an estimate that the transition dipole moment of the S0-S1 transition is about 4–6% of the S0-S2 transition. Although the results of this approach were promising, the calculated rates did not reproduce the measured energy transfer rates. To resolve this problem, an increase of S0-S1 Coulombic coupling via intensity borrowing from the allowed S2 state due to S2-S1 mixing was proposed (Damjanovic et al., 1999; Zhang et al., 2000). Under the assumption that the degree of mixing is inversely proportional to the square of the S1-S2 energy gap, Zhang et al. (2000) calculated S1 energy transfer rates for LH2 from Rsp. molischianum that were reasonably close to the measured values. Further improvement of the agreement between experiment and theory was obtained by calculations of Coulombic couplings by means of time-dependent density functional theory (TDDFT) (Walla et al., 2000; Hsu et al., 2001). This method, which allows ab initio calculations of the S1 couplings with Qy states of BChl, confirmed that small mixing between S2 and S1 states plays an important role in the Coulombic coupling. S1 energy transfer rates for LH2 complexes from three different species of purple bacteria were
222 calculated using the Coulombic couplings obtained from TDDFT, and the obtained values (9 ps)–1 (Rsp. molischianum), (1.2 ps)–1 (Rba. sphaeroides) and (3 ps)–1 (Rba. sphaeroides G1C) (Hsu et al., 2001) are very close to the experimental ones of (12 ps)–1, (2.4 ps)–1 and (1.4 ps)–1 (Zhang et al., 2001). For LH2 from Rps. acidophila, an S1 energy transfer time >25 ps was calculated (Hsu et al., 2001), also in agreement with the very low efficiency obtained experimentally (Macpherson, 2001; Polívka et al., 2002; Wohlleben, 2003). Although the trend of decreasing efficiency of the S1-mediated energy transfer was confirmed by experiments using LH2 complexes from Rba. sphaeroides having the different carotenoids, neurosporene, spheroidene, and spheroidenone, the drop in efficiency when going from spheroidene to spheroidenone was less than expected on the basis of a change in spectral overlap. The S1 lifetimes of carotenoids in LH2 complexes of 1.4, 1.5 and 1.4 ps for neurosporene, spheroidene, and spheroidenone were compared with their lifetimes in solution (24, 8.5 and 6 ps), resulting in energy transfer efficiencies of 94, 82 and 76% (Polívka et al., 2004). Knowing that the S1 energy of spheroidenone is around 13000 cm–1 (Zigmantas et al., 2004), the efficiency of energy transfer from spheroidenone does not match that expected from calculations by Hsu et al. (2001). Also, comparison of 76% with the nearly zero efficiency of the S1-mediated energy transfer obtained for rhodopin glucoside that has a LH2 S1 energy of 12550 ± 150 cm–1 (Polívka et al., 2002) would imply that ~500 cm–1 decrease in energy is enough to decrease efficiency from 76 to 5%. Taking into account that S1 emission of carotenoids is quite broad (Fujii et al., 1998; Frank et al., 2000a), such a drop can be hardly explained as due solely to change in the spectral overlap (Ritz et al., 2000). Thus, it indicates that other factors besides conjugation length may play a role. One possibility is that the high efficiency of the S1mediated energy transfer of spheroidenone is related to the fact that it belongs to the family of carbonyl carotenoids, which possess an excited state with chargetransfer character (Frank et al., 2000b; Zigmantas et al., 2004) that may enhance the spheroidenone-BChl coupling. Another explanation was offered by Ritz et al. (2000) who noted that the conjugated systems of neurosporene and spheroidene, which systematically exhibit highly-efficient energy transfer via the S1 state, have a non C2h-symmetrical arrangement of their methyl side groups. This is, however, not
Harry Frank and Tomáš Polívka the case for lycopene and rhodopin glucoside, both exhibiting very low efficiencies of the S1-mediated energy transfer. Consequently, Ritz et al. (2000) suggested that symmetry breaking of neurosporene and spheroidene may be an important factor in explaining why energy transfer via the S1 state is much more efficient for these carotenoids compared to lycopene and rhodopin glucoside. Since the conjugated system of spheroidenone has also asymmetric arrangement of methyl groups (Fig. 2), the observation of efficient S1-mediated energy transfer, despite its low S1 energy, provides further support for the conjecture proposed by Ritz et al. (2000). In addition to the S1-mediated energy transfer occurring from a thermalized S1 state discussed above, it has been proposed that a portion of the pathway may involve a vibrationally hot S1 state. Using global analysis, Papagiannakis et al. (2002) reported the decay of a species associated spectrum corresponding to the vibrationally hot S1 state, suggesting the presence of an energy transfer channel via this route. However, the contribution of this pathway to the total energy transfer efficiency was only 5%. Indirect evidence for the presence of this channel also could be found in experiments carried out by Krueger et al. (1999) who measured the spectral profile of the S0-S1 transition in the LH2 complex by two-photon excitation techniques that detect emission from BChl a. The 0-0 band of the S0-S1 transition was very weak but the intensity of higher vibrational bands was high which may be due to energy transfer occurring from the hot vibrational states. In any case, this energy transfer channel plays only a minor role in the overall carotenoid-BChl energy transfer in LH2. An upper limit of 3% was found in the LH2 complex from Rps. acidophila for this channel (Wohlleben et al., 2003). Moreover, as no evidence for energy transfer via a hot S1 state was provided by a number of other experiments using various LH2 complexes (Papagiannakis et al., 2003a; Koyama et al., 2004; Rondonuwu et al., 2004; Polli et al., 2006), further experiments using both advanced experimental approaches and sophisticated data analysis will be needed to verify the presence of this energy transfer channel. C. The S* State Reports of other carotenoid excited states located between the S2 and S1 states (see Polívka and Sundström, 2004, for a review) initiated lively debates regarding whether these states act as energy donors
Chapter 12
Carotenoid-to-BChl Energy Transfer
in carotenoid-BChl energy transfer. New ways of data analysis allowed for a more rigorous assignment of various excited state species (van Stokkum et al., 2004), and it was shown in a number of cases that the two-state (S2 and S1) model is not sufficient to describe all features revealed in experimental data. The so-called S* state is the best studied in this respect; it was first reported in the excited state manifold of the carotenoid spirilloxanthin in both solution and the LH1 complex of Rsp. rubrum (Gradinaru et al., 2001). Using global analysis of data in the 470–720 nm spectral region, it was shown that the S1-Sn band of spirilloxanthin in solution (peaking at 590 nm) possessed a distinct shoulder at ~540 nm. While the 590 nm band decayed with 1.4 ps corresponding to the S1 lifetime of spirilloxanthin, the 540 nm shoulder exhibited a much longer decay time of ~6 ps (Gradinaru et al., 2001). This result was explained in terms of two parallel pathways of S2 depopulation; a major part (70%) decaying to form the S1 state, while a minor pathway (30%) leads to population of the S* state which then decays to the ground state with a 6 ps lifetime. Interestingly, as shown in subsequent studies (Papagiannakis et al., 2002; Papagiannakis et al., 2003a; Wohlleben et al., 2003), the S* state is formed with much higher yield when carotenoids are incorporated into purple bacterial LH complexes. Moreover, in LH1 and LH2 complexes the S* state was found to be a precursor of fast carotenoid triplet state formation. A relatively high triplet yield of 25–30% was explained to be a result of a conformational distortion of spirilloxanthin in the LH1 complex, promoting triplet formation via singlet homofission from the S* state (Gradinaru et al., 2001). Further studies confirmed the triplet state formation via the S* state for other complexes containing different carotenoids. The yields of triplet formation varied from nearly 40% for rhodopin glucoside in LH2 from Rps. acidophila (Wohlleben et al., 2003) to less than 10% for spheroidene in both native LH2 from Rba. sphaeroides (Papagiannakis et al., 2002) and incorporated into the Rb. sphaeroides R26.1 mutant (Papagiannakis et al., 2003a). The fact that no triplet formation is observed in solution suggests that the protein environment is a crucial factor governing the formation of the triplet via singlet homofission, and it strongly supports the conclusion that the deviation from planar conformation in antenna complexes of purple bacteria is a necessary condition for efficient triplet formation (Papagiannakis et al., 2003a).
223 However, it turns out that the decay to the S0 state and triplet formation are not the only possible fates of the S* state in LH2 complexes. In LH2 complexes containing either spheroidene or rhodopin glucoside, the S* state also contributes to carotenoid-BChl energy transfer. For spheroidene in the LH2 of Rba. sphaeroides, S*-mediated energy transfer contributes 10–15% to the total spheroidene-BChl energy transfer (Papagiannakis et al., 2002). For rhodopin glucoside in Rps. acidophila LH2, the S*-mediated channel was found to be ~10% efficient; i.e., even higher than the energy transfer efficiency from the S1 state (Wohlleben et al., 2003). The possibility of carotenoid-BChl energy transfer via the S* state provided important information for determining the origin of the S* state, which is still a matter of debate. The presence of the S*-mediated energy transfer puts the S* energy above the Qy bands of BChl a, eliminating one of the proposed origins, a vibrationally hot ground state (Wohlleben et al., 2004). Instead, it seems that S* is indeed a separate excited state as originally suggested by Gradinaru et al. (2001), but its symmetry and relation to other states remains unknown. It also must be noted that although only a few carotenoids were subject to studies focusing on the S* state so far, it seems obvious that the lifetime of the S* state does not follow any clear dependence on conjugation length. While the S1 lifetime is changed systematically from ~9 ps (spheroidene) to 4.1 ps (rhodopin glucoside) and 1.5 ps (spirilloxanthin) as a result of increased conjugation length from 10 to 13 (Polívka and Sundström, 2004), the intrinsic S* lifetimes (in the absence of energy transfer) in LH2 and LH1 complexes are scattered in the 6 to 30 ps range without any obvious relation to the conjugation length (Gradinaru et al., 2001; Papagiannakis et al., 2002, 2003a, 2006a; Wohlleben et al., 2003). The lack of correlation between lifetime and conjugation length also casts doubts on assignment of the S* state to the 1Bu– state proposed by some authors (Papagiannakis et al., 2002; Wohlleben et al., 2003). A recent thorough study of the S* state in LH2 complexes from Rps. acidophila and Rba. sphaeroides revealed another aspect of the spectroscopic properties of the S* state in LH2 complexes. Population of the S* state exhibits a dependence on intensity of excitation pulses that differs from that observed for the S1 state (Papagiannakis et al., 2006a), indicating that the S* and S1 state cannot have a common precursor, the S2 state, as previously thought. These authors proposed
224 two models to explain their data. One assumes two different ground-state populations each leading to population of either S1 or S* state, the other involves higher excited states. These upper excited states are populated via excited state absorption from the S2 state that is resonant with the excitation pulse, thus the S2 population created by the front of the excitation pulse can be re-excited into higher excited states by photons arriving in the tail of the pulse. These higher excited states exhibit a relaxation pattern that favors population of the S* state. Because increasing excitation intensity increases the probability of the re-excitation, the S* state gets more populated with higher excitation intensities (Papagiannakis et al., 2006a). This model involving higher excited states has gained support from another study showing that direct excitation of high-lying excited states enhances population of the S* state (Billsten et al., 2005). These authors also suggested that the S* state in solution may be related to a conformational change of the carotenoid molecule. This possibility was later supported by a study on a series of carotenoids with different structures, indicating that the S* state may be a minimum in the S1 potential surface corresponding to a conformational change (Niedzwiedzki et al., 2006). This conclusion also explains why the S* state is preferentially populated in LH2 complexes where such distortions have been confirmed by X-ray crystallography. D. Other Pathways Another excited state widely discussed as a potential energy donor in carotenoid-BChl energy transfer in LH2 complexes is the 1Bu– state. Its presence in the excited state manifold was predicted two decades ago by calculations on polyenes carried out by Tavan and Schulten (1987), and it was shown that for conjugation lengths corresponding to naturally-occurring carotenoids (N = 9-13) it may be located below the strongly absorbing S2 state. Due to the forbidden nature of the 1Bu– state (it is forbidden for both oneand two-photon transitions from the ground state), experimental verification of the presence of the 1Bu– state between the S2 and S1 states is very difficult. In the late nineties Sashima et al. (1999) detected this state using measurements of resonance Raman profiles. Later, the 1Bu– state was also proposed to be active in energy transfer between carotenoids and BChl in LH2 complexes (Rondonuwu et al., 2004; Koyama et al., 2004). Based on previous assignments
Harry Frank and Tomáš Polívka of spectral signatures thought to be associated with the 1Bu– state in transient absorption spectra recorded in the near-infrared and visible ranges (Koyama et al., 2004), Rondonuwu et al. (2004) used global analysis of data taken on a few LH2 complexes containing different carotenoids and concluded that the 1Bu– state transfers energy to BChl with ~20% efficiency. This corresponds to an energy transfer time of ~0.6 ps for neurosporene and spheroidene in the LH2 complexes from Rba. sphaeroides. This pathway was proposed to be inactive for LH2 complexes accommodating the longer carotenoids lycopene and rhodopin glucoside. The absence of this channel in these complexes was explained by the 1Bu– state lying below the Qx state of BChl a, which was assumed to be the energy acceptor. For these carotenoids, however, it was hypothesized that another dark excited state, the 3Ag– state, may be active in carotenoid-BChl energy transfer because the expected 3Ag– energies may be favorable for this state to act as an energy donor. However, no experimental evidence for such a pathway was given (Rondonuwu et al., 2004). It must be noted that both 1Bu– lifetimes and spectral signatures in the visible region obtained by Rondonuwu et al. (2003) are essentially identical to those assigned earlier to the hot S1 state (Billsten et al., 2002b; de Weerd et al., 2002). Consequently, the data by Rondonuwu et al. (2004) may also be interpreted in terms of the hot S1 state being the energy donor instead of the 1Bu– state. Another issue that awaits further clarification is the relationship between the 1Bu– and S* states. It has been proposed that the 1Bu– state is identical with the S* state (Papagiannakis et al., 2002, Wohlleben et al., 2003; Rondonuwu et al., 2003). This assignment may seem correct as the symmetry and origin of the S* state are still unknown, and the 1Bu– state was predicted to be a precursor of ultrafast triplet formation (Rondonuwu et al., 2004; Koyama et al., 2004), the same process reported for the S* state (Gradinaru et al., 2001; Papagiannakis et al., 2002). On the other hand, for the S* state it was shown that there is no S1 ↔ S* conversion (Gradinaru et al., 2001; Papagiannakis et al., 2002; Wohlleben et al., 2003), which is in contradiction with the 1Bu– state being an intermediate state in S2-S1 internal conversion (Koyama et al., 2004). Similarly, the 1Bu– lifetimes (see Koyama et al., 2004, for a review) are about an order of magnitude shorter than those measured for the S* state, which argues that the 1Bu– and S* states are not equivalent.
Chapter 12
Carotenoid-to-BChl Energy Transfer
E. The Role of B800 For the LH2 and LH3 complexes, B850 (B820 in LH3) and B800 BChl a molecules may be acceptors in carotenoid-BChl energy transfer. The question of partitioning between these two possible acceptors has been addressed in a few studies. Macpherson et al. (2001) investigated S2-mediated energy transfer by fluorescence up-conversion in B800-B850 and B850only LH2 complexes from Rps. acidophila. Upon combining the results for these two LH2 complexes, they concluded that the S2 state of rhodopin glucoside transfers energy with 20% efficiency to B800 and with 31% efficiency to B850, leading to the total efficiency of 51% (Macpherson et al., 2001). Similar analysis was carried out by Papagiannakis et al. (2003a) who used transient absorption spectroscopy to study LH2 from a carotenoidless Rba. sphaeroides R26.1 mutant lacking the B800 BChl but incorporated with spheroidene. It was reported that the S2 pathway operates with 25% efficiency, which, when compared with the 57% efficiency of this pathway known for the wild type B800-B850 complex, allowed them to conclude that approximately half of the S2 population transfers energy to B800 (Papagiannakis et al., 2003a). It should be noted that although this work clearly showed that B800 accepts up to 50% of energy from the S2 state, calculations based on the X-ray structure of the LH2 complex from Rps. acidophila gave appreciable couplings only for the S2-B850 Qx channel (Krueger et al., 1998). On the other hand, similar calculations using full Coulombic couplings performed on the basis of the LH2 structure of Rsp. molischianum, yielded appreciable couplings of the S0-S2 transition with both the B850 and B800 BChls, resulting in close to a 1:1 branching ratio between the B800 and B850 acceptors (Damjanovic et al., 1999). Essentially the same results were obtained by calculations of the lycopene-BChl couplings by means of the collective electronic oscillators algorithm (Tretiak et al., 2000). Regarding energy acceptors in the S1-mediated energy transfer route, both B800 and B850 BChls are capable of accepting energy from carotenoids, but the S1-B800 channel seems to dominate. In experiments employing samples having spheroidene incorporated into LH2 from the carotenoidless Rba. sphaeroides R26.1 mutant lacking B800 BChl, the efficiency of energy transfer via the S1 pathway reached only 35% (Papagiannakis et al., 2003a). This is significantly less than the ~80% observed for the LH2 containing both
225 B800 and B850 (Walla et al., 2000; Zhang et al., 2000; Polívka et al., 2002), signaling that the main pathway involves B800 as an acceptor. The same conclusion was reached for LH2 from Rps. acidophila. Although the S1 efficiency is only 4–5% in the wild type complex, selective removal of the B800 BChls led to a complete absence of S1 energy transfer (Macpherson et al., 2001). Polívka et al. (2007) used lithium dodecyl sulfate (LDS) which selectively perturbs the B800 site (Chadwick et al., 1987). The LDS-treated complexes lack the B800 band and can be therefore used to investigate the role of B800 in energy transfer (Fig. 6). By measuring the S1 lifetimes of neurosporene, spheroidene and spheroidenone in LDS-treated LH2 complexes and comparing them with data taken on untreated complexes (Fig. 6), Polívka et al. (2007) showed that the carotenoid S1 lifetime has values in the range 1.4–1.8 ps for the untreated LH2 complexes, but it is prolonged to 2.7–3.5 ps for LDS-treated complexes. This is consistent with slower energy transfer via the S1 state. The S1 state branching ratios for energy transfer to the B850 and B800 BChls were calculated using the S1 lifetimes of the carotenoids in solution and in untreated and LDS-treated LH2 complexes. The calculations showed that the B800: B850 branching ratio remained essentially the same regardless of the conjugation length of the carotenoid. The values were 65:35 for LH2 with spheroidene and 60:40 for the complexes containing neurosporene
Fig. 6. Kinetics recorded at the maximum of the S1-Sn band for LH2 complexes from Rba. sphaeroides containing spheroidene. Kinetics are shown for spheroidene in solution (open triangles), LDS-treated LH2 complexes (open circles) and untreated LH2 complexes (full squares). The LH2 complexes were excited in the 0-0 band of the carotenoid S2 state at 515 nm. All kinetics are normalized to maximum. Solid lines represent fits.
226 or spheroidenone. This confirmed further that the B800 BChl a represents the dominant acceptor in carotenoid-BChl energy transfer via the S1 state. Interestingly, the B800:B850 branching ratio was not affected by conjugation length, indicating that even for spheroidenone having an S1 energy of ~13000 cm–1, the S1-B800 pathway remains dominant. The significance of the B800 BChl a molecule is further underlined by reports that it is likely the only acceptor in the minor energy transfer routes via the hot S1 state or the S* state (Papagiannakis et al., 2003a). IV. Energy Transfer in Light-Harvesting 1 Complexes and Reaction Centers Carotenoid-BChl energy transfer in LH1 complexes has been much less studied, and most of our current knowledge is limited to information provided by time-resolved studies of the LH1 complex from Rsp. rubrum. First experiments, however, were carried out by Ricci et al. (1996) who demonstrated efficient energy transfer via the S2 state of spheroidene in the LH1 complex from Rba. sphaeroides. Based on the comparison between fluorescence up-conversion data on spheroidene in solution and in LH1 they concluded that S2-mediated energy transfer operates with 65% efficiency (energy transfer time of 90 fs), which is even more efficient than for the spheroidenecontaining LH2 complex (Ricci et al., 1996). Later experiments on LH1 complexes from Rsp. rubrum containing the long (N=13) carotenoid, spirilloxanthin, revealed only 35% efficient S2-mediated energy transfer (Gradinaru et al., 2001). Moreover, it was shown that no energy transfer proceeds via the S1 state of spirilloxanthin. Thus, 35% is the total efficiency of the spirilloxanthin-BChl energy transfer, in agreement with earlier results based on measurements of fluorescence excitation spectra (Rademaker et al., 1980). The absence of the S1-mediated energy transfer pathway for spirilloxanthin likely results from its very long conjugated system which pushes its S1 energy too low to transfer energy to BChl a. Measurements of the S1-S2 spectra of spirilloxanthin in the LH1 complex determined the S1 energy of the carotenoid to be 11500 cm–1, which is below the energy of both B800 and B850 Qy transitions (Papagiannakis et al., 2003b). An investigation of LH1 complexes reconstituted with carotenoids with conjugation lengths in the range 9–13 showed that carotenoid-BChl energy transfer
Harry Frank and Tomáš Polívka in LH1 obeys the same trend as shown earlier for LH2. The overall carotenoid-BChl energy transfer decreases with increasing conjugation length, dropping from 78% for LH1 with neurosporene (N=9) to 36% for the spirilloxanthin-containing LH1 complex (Akahane et al., 2004). These are slightly less than values obtained for LH2. These authors found efficiencies for the S1-mediated energy transfer of 20 and 19% for LH1 complexes reconstituted with neurosporene or spheroidene, respectively. No transfer via the S1 state (efficiency <3%) was observed for carotenoids with N≥11, in agreement with previous studies (Akahane et al., 2004). The values of 20 and 19% for neurosporene and spheroidene are much lower than 94 and 82% obtained for LH2 complexes (Zhang et al., 2000; Polívka et al., 2004). However, it should be noted that the low efficiency of the S1-mediated energy transfer obtained by Akahane et al. (2004) may be attributed to involvement of the 1Bu– state that, according to the data analysis by Akahane et al. (2004), both transfers energy to BChl a and forms a triplet state with appreciable efficiency. Thus, as the 1Bu– state was not included in the analysis carried out by Zhang et al. (2000) or Polívka et al. (2004), the values are not directly comparable. Besides the 1Bu– pathway suggested by Akahane et al. (2004), the LH1 complex of Rsp. rubrum is also a system in which the S* state yield is pronounced. In this case, S* does not transfer energy to BChl a, but it is a precursor for the fast (~12 ps) formation of a spirilloxanthin triplet state. A relatively high triplet yield of 25–30% was explained on the basis of a conformational distortion of spirilloxanthin in the LH1 complex, promoting triplet formation via singlet homofission from the S* state (Gradinaru et al., 2001). Carotenoid-BChl energy transfer has also been studied in reaction centers (RC) of purple bacteria. Unlike the antenna complexes, RCs bind carotenoid in a 15,15´-cis configuration. Each RC has one carotenoid that is located in proximity to the BChl a monomer (BB) in the B branch (Allen et al., 1987). Earlier studies of RC from Rba. sphaeroides employing fluorescence excitation measurements reported ~75% efficiency of energy transfer from spheroidene to the primary donor (P), and BB BChl a was suggested to play a role of a mediator in this process (Frank et al., 1993). Later, these findings were elaborated by means of femtosecond time-resolved spectroscopy. Lin et al. (2003) studied the RC from aerobically grown Rba. sphaeroides that contains the carotenoid spheroidenone. By comparing the spheroidenone S1
Chapter 12
Carotenoid-to-BChl Energy Transfer
lifetime in RC (1.6 ps) with that in solution (6 ps), they concluded that energy transfer occurs preferentially from the S1 state of the carotenoid with an efficiency of 75%. This work was recently extended to RCs accommodating neurosporene and spheroidene (Lin et al., 2006). For these two carotenoids the S1mediated energy transfer operated with 96 and 84% efficiencies, respectively, matching perfectly the values obtained for the S1-mediated energy transfer in LH2 complexes (Zhang et al., 2000, Polívka et al., 2004). Thus, regardless of whether the carotenoid is in all-trans (LH2) or 15,15´-cis (RC) configuration, the S1-mediated energy transfer proceeds with the same efficiency. Lin et al. (2006) also found that at least 30% of energy must proceed via the S2 channel. This group also studied a mutant where the BB pigment was replaced by bacteriopheophytin which has a higher S1 energy than BB. In this mutant, the efficiency of total energy transfer from spheroidenone to P decreased substantially, supporting the earlier notion that the BB BChl a is a mediator for carotenoid to P energy transfer in RCs (Lin et al., 2006). V. Outlook Structural information in combination with advances in time-resolved spectroscopic techniques and data analysis during the past few years has vastly improved our knowledge of energy transfer processes between carotenoids and BChl in LH2 and LH1 complexes. On the other hand, these improvements brought a number of yet unanswered questions regarding origin of the involved states, routes and efficiencies of energy transfer, and effects of protein environment. An important issue is understanding the roles of the S* and 1Bu– states proposed as energy donors in carotenoid-BChl energy transfer, because the results obtained by the various research groups are often contradictory (see Polívka and Sundström, 2004, for review). The recent finding that the excited state dynamics of carotenoids in LH2 complexes depend on the intensity of excitation pulses adds a new dimension to the problem (Papagiannakis et al., 2006a). Clearly, intensity dependencies should be included in all future experiments. Although Papagiannakis et al. (2006) explored only the intensity dependence of S* state formation, it is likely that similar patterns may also be expected for the 1Bu– state. Ultrafast spectral features assigned earlier to a short-lived intermediate state (Cerullo et al., 2002) were subsequently
227 identified as being due to nonlinear effects caused by high-intensity femtosecond pulses (Kosumi et al., 2005). Another issue that deserves attention in future experiments is the role of vibrational relaxation in energy transfer. Although it was included in some data analyses, most studies attempting to disentangle the complex pattern of carotenoid-BChl energy transfer do not include vibrational levels (Koyama et al., 2004). To resolve these issues, new experimental approaches will be particularly interesting, for example the recently introduced ultrafast methods to control energy flow in LH2 complexes (Wohlleben et al., 2005). Pump-dump-probe (Papagiannakis, 2006c), femtosecond resonance Raman (McCamant et al., 2003) and two-dimensional electronic spectroscopy (Zigmantas, 2006) all hold the promise for elucidating the controlling features of energy transfer pathways in these systems. Acknowledgments This research is supported in the laboratory of HAF by the National Institutes of Health (GM-30353), the National Science Foundation (MCB-0314380), and the University of Connecticut Research Foundation. TP thanks the Czech Ministry of Education (MSM6007665808 and AV0Z50510513) for financial support. References Akahane J, Rondonuwu FS, Fiedor L, Watanabe Y and Koyama Y (2004) Dependence of singlet-energy transfer on the conjugation length of carotenoids reconstituted into the LH1 complex from Rhodospirillum rubrum G9. Chem Phys Lett 393: 184–191 Allen JP, Feher G, Yeates TO, Komiya H and Rees DC (1987) Structure of the reaction center from Rhodobacter sphaeroides R-26 — the cofactors. Proc Natl Acad Sci USA 84: 5730–5734 Andersson PO, Cogdell RJ and Gillbro T. (1996) Femtosecond dynamics of carotenoid-to-bacteriochlorophyll a energy transfer in the light-harvesting antenna complexes from the purple bacterium Chromatium purpuratum. Chem Phys 210: 195–217 Angerhofer A, Cogdell RJ and Hipkins M. (1986) A spectral characterization of the light-harvesting pigment-protein complexes from Rhodopseudomonas acidophila. Biochim Biophys Acta 848: 333–341 Angerhofer A, Bornhäuser F, Gall A and Cogdell RJ (1995) Optical and optically detected magnetic-resonance investigation on purple photosynthetic bacterial antenna complexes. Chem Phys 194: 259–274 Billsten HH, Herek JL, Garcia-Asua G, Hashøj L, Polívka T,
228 Hunter CN and Sundström V (2002a) Dynamics of energy transfer from lycopene to bacteriochlorophyll in geneticallymodified LH2 complexes of Rhodobacter sphaeroides. Biochemistry 41: 4127–4136 Billsten HH, Zigmantas D, Sundström V and Polívka T (2002b) Dynamics of vibrational relaxation in the S1 state of carotenoids having 11 conjugated C=C bonds. Chem Phys Lett 355: 465–470 Billsten HH, Pan JX, Sinha S, Pascher T, Sundström V and Polívka T (2005) Excited-state processes in the carotenoid zeaxanthin after excess energy excitation. J Phys Chem A 109: 6852–6859 Cerullo G, Polli D, Lanzani G, De Silvestri S, Hashimoto H and Cogdell RJ (2002) Photosynthetic light harveting by carotenoids: Detection of an intermediate excited state. Science 298: 2395–2398 Chadwick B, Zhang C, Cogdell RJ and Frank H (1987) The effects of lithium dodecyl-sulfate and sodium-borohydride on the absorption-spectrum of the B800-850 light-harvesting complex from Rhodopseudomonas acidophila 7750. Biochim Biophys Acta 893: 444–451 Cogdell RJ and Frank HA (1987) How carotenoids function in photosynthetic bacteria. Biochim Biophys Acta 895: 63–79 Cogdell RJ, Hipkins MF, MacDonald W and Truscott TG (1981) Energy transfer between the carotenoid and the bacteriochlorophyll within the B800-850 light-harvesting pigment-protein complex of Rhodopseudomonas sphaeroides. Biochim Biophys Acta 634: 191–202 Damjanović A, Ritz T and Schulten K (1999) Energy transfer between carotenoids and bacteriochlorophylls in light-harvesting complex II of purple bacteria. Phys Rev E 59: 3293–3311 Desamero RZB, Chynwat V, van der Hoef I, Jansen FJ, Lugtenburg J, Gosztola D, Wasielewski MR, Cua A, Bocian DF and Frank HA (1998) Mechanism of energy transfer from carotenoids to bacteriochlorophyll: Light-harvesting by carotenoids having different extents of π-electron conjugation incorporated into the B850 antenna complex from the carotenoidless bacterium Rhodobacter sphaeroides R-26.1. J Phys Chem B 102: 8151–8162 de Weerd FL, van Stokkum IHM and van Grondelle R (2002) Subpicosecond dynamics in the excited state absorption of all-trans β-carotene. Chem Phys Lett 354: 38–43 Frank HA (2001) Spectroscopic studies of the low-lying singlet excited electronic states and photochemical properties of carotenoids. Arch Biochem Biophys 385: 53–60 Frank HA and Cogdell RJ (1996) Carotenoids in photosynthesis. Photochem Photobiol 63: 257–264 Frank HA, Chynwat V, Hartwich G, Meyer M, Katheder I and Scheer H (1993) Carotenoid triplet-state formation in Rhodobacter sphaeroides R-26 reaction centers exchanged with modified bacteriochlorophyll pigments and reconstituted with spheroidene. Photosynth Res 37: 193–203 Frank HA, Bautista JA, Josue JS and Young AJ (2000a) Mechanism of nonphotochemical quenching in green plants: Energies of the lowest excited singlet states of violaxanthin and zeaxanthin. Biochemistry 39: 2831–2837 Frank HA, Bautista JA, Josue J, Pendon Z, Hiller RG, Sharples FP, Gosztola D and Wasielewski MR (2000b) Effect of the solvent environment on the spectroscopic properties and dynamics of the lowest excited states of carotenoids. J Phys Chem B 104: 4569–4577
Harry Frank and Tomáš Polívka Fraser NJ, Hashimoto H and Cogdell RJ (2001) Carotenoids and bacterial photosynthesis: The story so far. Photosynth Res 70: 249–256 Fujii R, Onaka K, Kuki M, Koyama Y and Watanabe Y (1998) The 2Ag– energies of all-trans-neurosporene and spheroidene as determined by fluorescence spectroscopy. Chem Phys Lett 288: 847–853 Furuichi K, Sashima T and Koyama Y (2002) The first detection of the 3Ag– state in carotenoids using resonance-Raman excitation profiles. Chem Phys Lett 356: 547–555 Gall A, Gardiner AT, Cogdell RJ and Robert B (2006) Carotenoid stoichiometry in the LH2 crystal: No spectral evidence for the presence of the second molecule in the α/β-apoprotein dimer. FEBS Lett 580: 3841–3844 Georgakopoulou S, van Grondelle R and van der Zwan G (2004) Circular dichroism of carotenoids in bacterial light-harvesting complexes: Experiments and modeling. Biophys J 87: 3010–3022 Gradinaru CC, Kennis JTM, Papagiannakis E, van Stokkum IHM, Cogdell RJ, Fleming GR, Niederman RA and van Grondelle R (2001) An unusual pathway of excitation energy deactivation in carotenoids: Singlet-to-triplet conversion on an ultrafast timescale in a photosynthetic antenna. Proc Natl Acad Sci USA 98: 2364–2369 Hashimoto H, Yanagi K, Yoshizawa M, Polli D, Cerullo G, Lanzani G, De Silvestri S, Gardiner AT and Cogdell RJ (2004) The very early events following photoexcitation of carotenoids. Arch Biochem Biophys 430: 61–69 Herek JL, Wohlleben W, Cogdell RJ, Zeidler D and Motzkus M (2002) Quantum control of energy flow in light harvesting. Nature 417: 533–535 Hsu CP, Walla PJ, Head-Gordon M and Fleming GR (2001) The role of the S1 state of carotenoids in photosynthetic energy transfer: The light-harvesting complex II of purple bacteria. J Phys Chem B 105:11016–11025 Koepke J, Hu X, Muenke C, Schulten K and Michel H (1996) The crystal structure of the light-harvesting complex II (B800-850) from Rhodospirillum molischianum. Structure 4: 581–597 Kosumi D, Komukai M, Hashimoto H and Yoshizawa M (2005) Ultrafast dynamics of all-trans-β-carotene explored by resonant and nonresonant photoexcitations. Phys Rev Lett 95: 213601–213604 Koyama Y, Rondonuwu FS, Fujii R and Watanabe Y (2004) Lightharvesting function of carotenoids in photosynthesis: The roles of the newly found 1Bu– state. Biopolymers 74: 2–18 Krueger BP, Scholes GD, Jimenez R and Fleming GR (1998a) Electronic excitation transfer from carotenoid to bacteriochlorophyll in the purple bacterium Rhodopseudomonas acidophila. J Phys Chem B 102: 2284–2292 Krueger BP, Scholes GD and Fleming GR (1998b) Calculation of couplings and energy-transfer pathways between the pigments of LH2 by the ab initio transition density cube method. J Phys Chem B 102: 5378–5386 Krueger BP, Yom J, Walla PJ and Fleming GR (1999) Observation of the S1 state of spheroidene in LH2 by two-photon fluorescence excitation. Chem Phys Lett 310: 57–64 Larsen DS, Papagiannakis E, van Stokkum IHM, Vengris M, Kennis JTM and van Grondelle R (2003) Excited state dynamics of β-carotene explored with dispersed multi-pulse transient absorption. Chem Phys Lett 381: 733–742 Lin S, Katilius E, Taguchi AKW and Woodbury NW (2003) Ex-
Chapter 12
Carotenoid-to-BChl Energy Transfer
citation energy transfer from carotenoid to bacteriochlorophyll in the photosynthetic purple bacterial reaction center of Rhodobacter sphaeroides. J Phys Chem B 107: 14103–14108 Lin S, Katilius E, Ilagan RP, Gibson GN, Frank HA and Woodbury NW (2006) Mechanism of carotenoid singlet excited state energy transfer in modified bacterial reaction centers. J Phys Chem B 110: 15556–15563 Macpherson AN, Arellano JB, Fraser NJ, Cogdell RJ and Gillbro T (2001) Efficient energy transfer from the carotenoid S2 state in a photosynthetic light-harvesting complex. Biophys J 80: 923–930 McCamant DW, Kukura P and Mathies RA (2003) Femtosecond time-resolved stimulated Raman spectroscopy: Application to the ultrafast internal conversion in beta-carotene. J Phys Chem A 107: 8208–8214 McDermott G, Prince SM, Freer AA, Hawthornwaite-Lawless AM, Papiz MZ, Cogdell RJ and Isaacs NW (1995) Crystal structure of an integral membrane light-harvesting complex from photosynthetic bacteria. Nature 374: 517–521 McLuskey K, Prince SM, Cogdell RJ and Isaacs NW (2001) The crystallographic structure of the B800-820 LH3 light-harvesting complex from the purple bacteria Rhodopseudomonas Acidophila strain 7050. Biochemistry 40: 8783–8789 Nagae H, Kakitani T, Katoh T and Mimuro M (1993) Calculation of the excitation transfer-matrix elements between the S2 or S1 state of carotenoid and the S2 or S1 state of bacteriochlorophyll. J Chem Phys 98: 8012–8023 Naqvi KR (1980) The mechanism of singlet-singlet excitation energy transfer from carotenoids to chlorophyll. Photochem Photobiol 31: 523–524 Niedzwiedzki DM, Sullivan JO, Polívka T, Birge RR and Frank HA (2006) Femtosecond time-resolved transient absorption spectroscopy of xanthophylls. J Phys Chem B 110: 22872–22885 Papagiannakis E, Kennis JTM, van Stokkum IHM, Cogdell RJ and van Grondelle R (2002) An alternative carotenoid-to-bacteriochlorophyll energy transfer pathway in photosynthetic light harvesting. Proc Natl Acad Sci USA 99: 6017–6022 Papagiannakis E, Das SK, Gall A, Stokkum IHM, Robert B, van Grondelle R, Frank HA and Kennis JTM (2003a) Light harvesting by carotenoids incorporated into the B850 light-harvesting complex from Rhodobacter sphaeroides R-26.1: Excited-state relaxation, ultrafast triplet formation, and energy transfer to bacteriochlorophyll. J Phys Chem B 107: 5642–5649 Papagiannakis E, van Stokkum IHM, van Grondelle R, Niederman RA, Zigmantas D, Sundström V and Polívka T (2003b) A near-infrared transient absorption study of the excited-state dynamics of the carotenoid spirilloxanthin in solution and in the LH1 complex of Rhodospirillum rubrum. J Phys Chem B 107: 11216–11223 Papagiannakis E, van Stokkum IHM, Vengris M, Cogdell RJ, van Grondelle R and Larsen DS (2006a) Excited-state dynamics of carotenoids in light-harvesting complexes. 1. Exploring the relationship between the S1 and S* states. J Phys Chem B 110: 5727–5736 Papagiannakis E, Vengris M, Valkunas L, Cogdell RJ, van Grondelle R and Larsen DS (2006b) Excited-state dynamics of carotenoids in light-harvesting complexes. 2. Dissecting pulse structures from optimal control experiment. J Phys Chem B 110: 5737–5746 Papagiannakis E, Vengris M, Larsen DS, van Stokkum IHM, Hiller RG and van Grondelle R (2006c) Use of ultrafast dispersed
229 pump-dump-probe and pump-repump-probe spectroscopies to explore the light-induced dynamics of peridinin in solution. J Phys Chem B 110: 512–521 Papiz MZ, Prince SM, Howard TD, Cogdell RJ and Isaacs NW (2003) The structure and thermal motion of the B800–850 LH2 complex from Rps. acidophila at 2.0 Å resolution and 100 K: New structural features and functionally relevant motions. J Mol Biol 326: 1523–1538. Polívka T and Sundström V (2004) Ultrafast dynamics of carotenoid excited states — From solution to natural and artificial systems. Chem Rev 104: 2021–2071 Polívka T, Herek JL, Zigmantas D, Åkerlund HE and Sundström V (1999) Direct observation of the (forbidden) S1 state in carotenoids. Proc Natl Acad Sci USA 96: 4914–4917 Polívka T, Zigmantas D, Frank HA, Bautista JA, Herek JL, Koyama Y, Fujii R and Sundström V (2001) Near-infrared time-resolved study of the S1 state dynamics of the carotenoid spheroidene. J Phys Chem B 105: 1072–1080 Polívka T, Zigmantas D, Herek JL, He Z, Pascher T, Pullerits T, Cogdell RJ, Frank HA and Sundström V (2002) The carotenoid S1 state in LH2 complexes of purple bacteria Rhodobacter sphaeroides and Rhodopseudomonas acidophila: S1 energies, dynamics, and carotenoid radical formation. J Phys Chem B 106: 11016–11025 Polívka T, Pullerits T, Frank HA Cogdell RJ and Sundström V (2004) Ultrafast formation of a carotenoid radical in LH2 antenna complexes of purple bacteria. J Phys Chem B 108: 15398–15407 Polívka T, Niedzwiedski D,Fuciman M, Sundström V and Frank HA (2007) Role of B800 in carotenoid-bacteriochlorophyll energy and electron transfer in LH2 complexes from the purple bacterium Rhodobacter sphaeroides. J Phys Chem B 111: 7422–7431 Polli D, Cerullo G, Lanzani G, De Silvestri S, Hashimoto H and Cogdell RJ (2006) Carotenoid-bacteriochlorophyll energy transfer in LH2 complexes studied with 10-fs time resolution. Biophys J 90: 2486–2497 Rademaker H, Hoff AJ, van Grondelle R and Duysens LNM (1980) Carotenoid triplet yields in normal and deuterated Rhodospirillum rubrum. Biochim Biophys Acta 592: 240–257 Ricci M, Bradforth SE, Jimenez R and Fleming GR (1996) Internal conversion and energy transfer dynamics of spheroidene in solution and in the LH1 and LH2 light-harvesting complexes. Chem Phys Lett 259: 381–390 Ritz T, Damjanovic A, Schulten K, Zhang JP and Koyama Y (2000) Efficient light harvesting through carotenoids. Photosynth Res 66: 125–144 Rondonuwu FS, Watanabe Y, Fujii R and Koyama Y (2003) A first detection of singlet to triplet conversion from the 1Bu– to the 3Ag state and triplet internal conversion from the 3Ag to the 3 Bu state in carotenoids: dependence on the conjugation length. Chem Phys Lett 376: 292–301 Rondonuwu FS, Yokoyama K, Fujii R, Koyama Y, Cogdell RJ and Watanabe Y (2004) The role of the 1Bu– state in carotenoid-tobacteriochlorophyll singlet-energy transfer in the LH2 antenna complexes from Rhodobacter sphaeroides G1C, Rhodobacter sphaeroides 2.4.1, Rhodospirillum molischianum and Rhodopseudomonas acidophila. Chem Phys Lett 390: 314–322 Roszak AW, Howard TD, Southall J, Gardiner AT, Law CJ, Isaacs NW and Cogdell RJ (2003) Crystal structure of the RC-LH1 core complex from Rhodopseudomonas palustris. Science
230 302: 1969–1972 Sashima T, Nagae H, Kuki M and Koyama Y (1999) A new singlet-excited state of all-trans-spheroidene as detected by resonance-Raman excitation profiles. Chem Phys Lett 299: 187–194 Scholes GD, Harcourt RD and Fleming GR (1997) Electronic interactions in photosynthetic light-harvesting complexes: The role of carotenoids. J Phys Chem B 101: 7302–7312 Shreve AP, Trautman JK, Frank HA, Owens TG and Albrecht AC (1991). Femtosecond energy transfer processes in the B800-850 light-harvesting complex of Rhodobacter sphaeroides 2.4.1. Biochim Biophys Acta 1058: 280–288 Tavan P and Schulten K (1987) Electronic Excitations in Finite and Infinite Polyenes. Phys Rev B 36: 4337–4358 Trautman JK, Shreve AP, Violette CA, Frank HA, Owens TG and Albrecht AC (1990) Femtosecond dynamics of energy transfer in B800-850 light-harvesting complexes of Rhodobacter sphaeroides. Proc Natl Acad Sci USA 87: 215–219 Tretiak S, Middleton C, Chernyak V and Mukamel S (2000) Bacteriochlorophyll and carotenoid excitonic couplings in the LH2 system of purple bacteria. J Phys Chem B 104: 9540–9553 van Grondelle R (1985) Excitation energy transfer, trapping and annihilation in photosynthetic systems. Biochim Biophys Acta 811: 147–195 van Grondelle R, Kramer H and Rijgersberg C (1982) Energy transfer in the B800-850-carotenoid light-harvesting complex of various mutants of Rhodopseudomonas sphaeroides and of Rhodopseudomonas capsulatus. Biochim Biophys Acta 682: 208–215 Walla PJ, Linden PA, Hsu CP, Scholes GD and Fleming GR (2000). Femtosecond dynamics of the forbidden carotenoid
Harry Frank and Tomáš Polívka S1 state in light-harvesting complexes of purple bacteria observed after two-photon excitation. Proc Natl Acad Sci USA 97: 10808–10813 Wohlleben W, Buckup T, Herek JL, Cogdell RJ, and Motzkus M (2003) Multichannel carotenoid deactivation in photosynthetic light harvesting as identified by an evolutionary target analysis. Biophys J 85: 442–450 Wohlleben W, Buckup T, Hashimoto H, Cogdell RJ, Herek JL and Motzkus M (2004) Pump-deplete-probe spectroscopy and the puzzle of carotenoid dark states. J Phys Chem 108: 3320–3325 Wohlleben W, Buckup T, Herek JL and Motzkus M (2005) Coherent control for spectroscopy and manipulation of biological dynamics. Chem Phys Chem 6: 850–857 Zhang JP, Fujii R, Qian P, Inaba T, Mizoguchi T, Koyama Y, Onaka K and Watanabe Y (2000) Mechanism of the carotenoid-to-bacteriochlorophyll energy transfer via the S-1 state in the LH2 complexes from purple bacteria. J Phys Chem B 104: 3683–3691 Zhang JP, Inaba T, Watanabe Y and Koyama Y (2001) Partition of carotenoid-to-bacteriochlorophyll singlet-energy transfer through two channels in the LH2 complex from Rhodobacter sphaeroides G1C. Chem Phys Lett 340: 484–492 Zigmantas D, Hiller RG, Sharples FP, Frank HA, Sundström V and Polívka T (2004) Effect of a conjugated carbonyl group on the photophysical properties of carotenoids. Phys Chem Chem Phys 6: 3009–3016 Zigmantas D, Read EL, Mančal T, Brixner T, Gardiner AT, Cogdell RJ and Fleming GR (2006) Two-dimensional electronic spectroscopy of the B800–B820 light-harvesting complex. Proc Natl Acad Sci USA 103: 12672–12677
Chapter 13 Spectroscopy and Dynamics of Excitation Transfer and Trapping in Purple Bacteria Rienk van Grondelle* Department of Biophysics, Faculty of Sciences, Vrije Universiteit, De Boelelaan 1081, 1081 HV Amsterdam, The Netherlands
Vladimir I. Novoderezhkin A. N. Belozersky Institute of Physico-Chemical Biology, Moscow State University, Leninskie Gory, 119992, Moscow, Russia
Summary ............................................................................................................................................................... 232 I. Introduction..................................................................................................................................................... 232 A. Collective Electronic Excitations in Photosynthetic Complexes ....................................................... 232 B. Modified Redfield and Generalized Förster Theories ....................................................................... 233 II. Structure and Exciton Spectra of Light-Harvesting 1 and 2 Bacterial Antenna Complexes .......................... 235 A. Structure and the Excited-state Properties....................................................................................... 235 B. Disordered Exciton Model ................................................................................................................ 235 C. Steady-state Spectra........................................................................................................................ 236 D. Multiple Exciton Delocalization Sizes ............................................................................................... 237 E. Superradiance and Anomalously High Nonlinear Responses.......................................................... 238 F. Exciton Transitions Viewed by Polarized Single Molecule Spectroscopy ........................................ 238 III. Equilibration Dynamics ................................................................................................................................... 239 A. Exciton Relaxation and Dynamic Red-Shift...................................................................................... 239 B. Interplay of Excitonic and Vibrational Coherences ........................................................................... 240 IV. Competition of Intraband B800-800 and Interband B800-850 Energy Transfer in the Light-Harvesting 2 Complex ........................................................................................................................................................ 241 A. Excitation-wavelength Dependent Decay of the B800 Band ............................................................ 241 B. Multiple Pathways of B800 to B850 Transfer ................................................................................... 243 C. Origin of the B800 Anisotropy Decay ............................................................................................... 243 V. Energy Trapping in the Core Reaction Center-Light-Harvesting 1 Complex ................................................. 243 A. Trapping Kinetics.............................................................................................................................. 243 B. Exciton Control of Antenna-Reaction Center Transfer Rates........................................................... 243 C. Detrapping Efficiency ....................................................................................................................... 244 VI. Slow Conformational Motions and Excitation Dynamics in the B850-Light-Harvesting 2 Complex ............... 244 A. Dynamics of Conformational Changes ............................................................................................. 244 B. Fluctuations of Spectral Shapes in a Single Complex...................................................................... 245 C. Exciton Dynamics in the Different Conformational States................................................................ 246 VII. Concluding Remarks ...................................................................................................................................... 247 Acknowledgments ................................................................................................................................................. 248 References ............................................................................................................................................................ 248
*Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 231–252. © 2009 Springer Science + Business Media B.V.
232
Rienk van Grondelle and Vladimir I. Novoderezhkin
Summary In this chapter we discuss the transfer and trapping of excitation energy in the light-harvesting antenna of purple bacteria. The determination of high-resolution X-ray crystal structures of a variety of bacterial light-harvesting complexes has allowed an interpretation of the steady-state and time-resolved picosecond/femtosecond spectral responses of these complexes at a quantitative level using the disordered exciton model in combination with the generalized Förster and modified Redfield theories. Thus a consistent physical picture of energy transfer and trapping has been obtained including a precise assignment of the energy transfer pathways together with a direct visualization of the full excitation dynamics where various regimes ranging from coherent motion of the delocalized exciton to the hopping of localized excitations are superimposed. In a single complex it is possible to observe the switching between these regimes driven by slow conformational motions. I. Introduction In photosynthetic purple bacteria solar photons are absorbed by a set of membrane-associated pigmentproteins (the light-harvesting antenna complexes LH1 and LH2) and the electronic excited state is efficiently transferred to a reaction center (RC), where it is converted into a charge separation eventually leading to the build-up of a transmembrane electro-chemical potential difference (Van Grondelle et al., 1994; Van Amerongen et al., 2001). Antenna complexes consist of ordered arrays of bacteriochlorophylls (BChls) as main light-harvesting pigments, which together with other cofactors are non-covalently bound to proteins (Chapter 8, Gabrielsen et al.; Chapter 9, Bullough et al.). The peripheral LH2 complex consists of a highly symmetric circular aggregate of 9 (or 8) pigment-protein subunits, each subunit containing 2 trans-membrane polypeptide helixes that bind 3 BChls, forming a tightly packed ring of 18 (or 16) BChls with a second ring of 9 (or 8) weakly interacting BChls (McDermott et al., 1995; Koepke et al., 1996; Papiz et al., 2003). These rings determine the well-known absorption features of LH2 with intense absorption bands at 850 and 800 nm, respectively (also called B850 and B800). The LH1 core complex has 16 or 15 subunits with 2 BChls each, forming ‘circular’ structures with 32 BChls (Karrasch et al., 1995; Scheuring et al., 2003) or an open circle with 30 BChls (Rozak et al., 2003) surrounding the RC. Abbreviations: B850, B800 – spectral bands with the 850 or 800 nm absorption; BChl – bacteriochlorophyll; Blc. – Blastochloris; CD – circular dichroism; FL – fluorescence; LH1 – core lightharvesting complex of purple bacteria; LH2 – peripheral lightharvesting complex of purple bacteria; PR – participation ratio; Qy, Qx, By, Bx – four electronic transitions of BChl; Rba. – Rhodobacter; RC – reaction center; Rps. – Rhodopseudomonas; Rsp. – Rhodospirillum; TA – transient absorption
A. Collective Electronic Excitations in Photosynthetic Complexes The distance between nearest-neighbor BChls in the tightly packed rings of LH1/LH2 is about 9 Å, thus giving rise to strong electrostatic pigment-pigment interactions. These interactions produce a coherent superposition of the excited states of the individual BChls giving rise to collective excitations (excitons) delocalized over a certain number of pigments. The weakly coupled BChls in the second (B800) ring of LH2 produce more localized excitations. In order to obtain the structure of the excited states of the whole antenna it is necessary to account for all the pigment-pigment interactions. The resulting exciton Hamiltonian can be expressed in the basis of the local excited-state wavefunctions |n〉, which correspond to the excitation of the n-th pigment: N
N
n=1
n≠ m
H = ∑ En n n + ∑ M nm n m
(1)
where N is the number of BChls in the antenna, En is the electronic transition energy of the n-th molecule, Mnm is the interaction energy between the n-th and m-th molecules. We only consider excitation of the lowest excited state of BChl corresponding to the Qy transition. Generalization of Eq. (1) taking into account the Qy, Qx, By, and Bx transitions of BChls is straightforward (Scherz and Parson, 1984; Alden et al., 1997; Koolhaas et al., 1997a). Excitations of individual pigments with the energies En are no longer eigenstates of the system due to the presence of the off-diagonal energies Mnm that reflect the excitonic coupling between the pigments. The energies ωk and wavefunctions |k〉 of the exciton eigenstates can be obtained by diagonalization of the Hamiltonian (Eq. 1):
Chapter 13
Spectroscopy and Dynamics of Excitation Transfer
N
H = ∑ ωk k k
N
;
k =1
k = ∑ cnk n n=1
(2)
where the collective exciton states |k〉 contain a coherent superposition of the individual molecular excitations |n〉. In Eq. (2) the wavefunction amplitudes cnk reflect the participation of the n-th site in the k-th exciton state. B. Modified Redfield and Generalized Förster Theories Absorption of a solar photon by LH2 populates the exciton states of the antenna. Since this population is generally not in thermal equilibrium, the system will relax. The ensuing dynamics include energy transfer between the excitonic states |k〉, where the rate of population transfer from state k´ to state k is given by Zhang et al., (1998) and Yang and Fleming (2002): ∞
Rkkk ' k ' = 2 Re ∫ dt Ak (t ) Fk*' (t ) Vkk ' (t ) 0
Ak (t ) = exp { − iω k t − gkkkk (t )
{
Fk ' (t ) = exp − iω k ' t − g
* k 'k 'k 'k '
} (t ) + 2iλ k ' k ' k ' k ' t
} (3)
where F(t) and A(t) are line-shape functions corresponding to fluorescence (FL) of the donor state and absorption of the acceptor, respectively, while V describes the electrostatic interaction between donor and acceptor. The function gkkkk(t) determines the line-broadening of the k-th exciton state due to exciton-phonon coupling, λkkkk is the corresponding reorganization energy. These quantities are (Mukamel, 1995; Zhang et al., 1998): gkkkk (t ) = ∑ ( cnk )4 gn (t ) ; λ kkkk = ∑ ( cnk )4 λ n ; n
n
∞
dω Cn (ω ) ⎡⎣ coth 2 kωBT (cos ωt − 1) − i ( sin ωt − ωt ) ⎤⎦ g n (t ) = − ∫ 2 2 −∞ πω λn =
∞
dω Cn (ω ) 2 −∞ πω
∫
(4)
where gn(t) and λn correspond to the line-broadening function and reorganization energy of the n-th molecule, which are related to the spectral density Cn(ω) of the exciton-phonon coupling, T is the temperature, kB is the Boltzmann constant. The spectral
233
density in Eq. (4) reflects a fast modulation of the transition energies of the pigments En due to phonons (intramolecular vibrations, pigment-pigment and pigment-protein modes). The phonon-induced modulation of the couplings Mnm can be included into Eq. (4) as well (Meier et al., 1997b). If the donor and acceptor states are localized at the m-th and n-th sites (i.e., cmk´ = 1 and cnk = 1) then V is given by 2
Vkk ' (t ) = M nm
(5)
where Mnm is the interaction energy corresponding to a weak coupling between the localized sites n and m. Switching to the Fourier-transforms of F(t) and A(t) we can rewrite the integral in a form of donor-acceptor spectral overlap (Yang and Fleming, 2002). Thus, we obtain the Förster formula (Förster, 1965). The standard Förster formula can be generalized to the case of energy transfer between two weakly connected clusters (Novoderezhkin and Razjivin, 1994; 1996; Sumi, 1999; Scholes and Fleming 2000; Jang et al., 2004). The rate of energy transfer from the k´-th exciton state of one cluster to the k-th state of the other cluster is 2
Vkk ' (t ) =
∑c
k n
k' m
M nm c
(6)
n ,m
where n and m designate molecules belonging to different clusters. In this generalized Förster formula, the donor and acceptor states k´ and k can have an arbitrary degree of delocalization (corresponding to arbitrarily strong excitonic interactions within each cluster). But it is important to note that the intercluster interactions Mnm are weak, thus producing only a small spatial overlap between the k´ and k wavefunctions. In the case of significant spatial overlap of the wavefunctions cmk´ and ckn the transfer rate cannot be calculated by treating Mnm as a perturbation. In this case the energy transfer should be calculated in the basis of the exciton states of the whole system. The rates of relaxation between these states (with arbitrary wavefunction overlap) are given by their interaction with phonons. The modified Redfield approach gives (Zhang et al., 1998; Yang and Fleming, 2002):
Rienk van Grondelle and Vladimir I. Novoderezhkin
234 Vkk ' (t ) = exp(2 g k ' k ' kk (t ) + 2iλ k ' k ' kk t ) × × ⎡⎣ g&&kk ' k ' k (t ) − { g& k ' kk ' k ' (t ) − g& k ' kkk (t ) + 2iλ k ' kk ' k ' } × × { g& k ' k ' kk ' (t ) − g& kkkk ' (t ) + 2iλ k ' k ' kk ' } ⎤⎦
(7) N
gkk ' k '' k ''' (t ) = ∑ cnk cnk ' cnk '' cnk ''' gn (t ) ; n=
N
λ kk ' k '' k ''' = ∑ cnk cnk ' cnk '' cnk ''' λ n n=
The exponential term in Eq. (7) contains the linebroadening function gk´k´kk(t) which is proportional to the wavefunction overlap ∑n(ck´n ckn)2. In the localized limit ck´n = 1, whereas ckn is close to zero. Considering the coupling energy Mnm between the two sites as a weak perturbation we have ckn = Mnm/(En – Em), where En and Em are the site energies. Thus, in the limit of weak interaction Eq. (7) reduces to Eq. (5) and we recover the Förster formula. As an example we calculate the energy transfer rates within a dimer as a function of the energy gap (E1–E2) between the two molecules (Fig. 1). We use the Förster (Eq. 5) and modified Redfield (Eq. 7) expressions. In the calculation we use the spectral density Cn(ω) containing one overdamped Brownian oscillator and 48 high-frequency modes with the parameters taken from the experiment (Peterman et al., 1997). The difference between energy transfer rates as predicted by the two theories is compared with the delocalization (del) length calculated as the inverse participation ratio of the exciton wavefunctions, i.e., Ndel = PR−1, where PR = (c1k)4+(ck2)4. For a dimer the inverse participation ratio (PR) varies from 1 (in the localized limit) to 2, where the excitation is uniformly delocalized over the two molecules. Figure 1 shows that in the case where the energy difference is large (E1–E2 > 5M) the excitation is almost localized (1 < Ndel < 1.1). The energy transfer has the form of hopping between the two molecules. In this case the Förster and Redfield theories give approximately the same rate. Decreasing the energy gap (E1–E2 < 5M) results in formation of delocalized states (with Ndel increasing from 1.1 up to 2). In this case the excitonic interactions create a coherent mixing of the two sites. Instead of an excitation jumping from one site to the other, we now have a relaxation between two delocalized states. Increase of the spatial overlap between the
Fig. 1. The energy transfer rate as a function of the energy gap between two molecules calculated with the Förster and modified Redfield theories (only downhill rates are shown). The delocalization length Ndel is calculated as the inverse participation ratio of the exciton wavefunctions. The Ndel = 1 line is shown to highlight the deviations from the localized limit. The two frames correspond to Mnm = 255 and 55 cm−1. We use the exciton-phonon spectral density measured for the plant light-harvesting complex (Peterman et al., 1997). The relaxation rates have been calculated for 77 K. The specific non-monotonous dependence of the rates on the energy gap is determined by the shape of the phonon wing.
delocalized wavefunctions corresponds to having much faster transfer. The deviation of the Redfield rates compared with those predicted by the Förster equation increases in proportion to the deviation of delocalization length from the localized limit. This example suggests that interband transfers in purple bacteria (for example the energy transfer from B800 to B850, which corresponds to a large energy gap and small coupling) can be calculated using the generalized Förster approach. On the contrary, intraband dynamics (for instance the B850-B850 equilibration corresponding to a small energy gap and large exciton couplings) should be modeled by the Redfield theory.
Chapter 13
Spectroscopy and Dynamics of Excitation Transfer
II. Structure and Exciton Spectra of LightHarvesting 1 and 2 Bacterial Antenna Complexes A. Structure and the Excited-state Properties Since the determination of the crystal structure of the peripheral B800-B850 LH2 antenna of purple bacteria Rhodopseudomonas (Rps.) acidophila (McDermott et al., 1995) and Rhodospirillum (Rsp.) molischianum (Koepke et al., 1996) a direct calculation of the exciton Hamiltonian (Eq. 1) from first principles has become possible (Alden et al., 1997; Scholes et al., 1999; Tretiak et al., 2000a,b; Damjanović et al., 2002; Hu et al., 2002). These calculations revealed strong exciton couplings of the Qy transitions within the tightly packed B850 ring. The corresponding values of nearest-neighbor interaction energy vary around 250–300 cm−1, depending on the method of computation; see Tretiak et al. (2000b) for a review. Such strong interactions produce a broad exciton band with the lowest states near 850 nm and higher exciton states spreading up to the 800 nm region where they are superimposed with the degenerate levels of the B800 ring (Dracheva et al., 1996; Sauer et al., 1996; Alden et al., 1997; Hu et al., 1997; Koolhaas et al., 1997b; Wu et al., 1997b; Georgakopoulou et al., 2002; Novoderezhkin et al., 2003). The B800 ring is characterized by weak pigment-pigment couplings of about 30 cm−1, corresponding to localized excited states. Nevertheless, the excitonic interactions should be taken into account to correctly reproduce the B800 line shapes (Cheng and Silbey, 2006) and the excitation dynamics within the B800 ring (Novoderezhkin et al., 2003; Zigmantas et al., 2006). Interaction of the exciton states with the phonons induces a homogeneous line broadening (given by the function gkkkk), a red-shift of the pure electronic (zero-phonon) transitions (due to reorganization effects determined by λkkkk), and the transfer of excitation energy in the complex (relaxation or hopping between different states or sites). All these features are determined by the exciton-phonon spectral density Cn(ω) that can be extracted from spectroscopic data (for example, from the low-temperature FL spectra (Kühn et al., 2002)), or obtained numerically using molecular dynamics and quantum chemical calculations (Damjanović et al., 2002).
235
B. Disordered Exciton Model Besides fast nuclear motion (phonons and vibrations) the excitonic transitions are affected by slow conformational motion of the proteins. Such motion induces random shifts (quasi-static disorder) in the transition energies En and couplings Mnm. The exciton energies and wavefunctions become perturbed correspondingly. In conventional bulk spectroscopies (with averaging over many complexes) the conformational disorder produces inhomogeneous line broadening. In single molecule techniques a slow conformational motion can be viewed as fluctuations of the spectral shapes on a time scale from milliseconds to minutes. In the homogeneous limit (without disorder) all the dipole strength of the circular aggregate (made up of N molecules having almost in-plane orientation of the Qy transition dipoles) is concentrated in a degenerate pair of levels (k = ±1). The participation ratio is PR = 1/N and 3/(2N) (i.e., 0.056, and 0.083 for N = 18) for the lowest and higher states, respectively. This homogenous exciton structure is changed dramatically in the presence of disorder depending on the ratio between the disorder width σ and the nearest neighbor interaction energy M (Monshouwer et al., 1997; Alden et al., 1997; Novoderezhkin et al., 1999a). An example of how weak, moderate, and strong disorder (σ/M = 0.52, 1.29, and 2.23) affect the absorption spectrum of the B850 antenna (with N = 18) is shown in Fig. 2. In the case of weak disorder (σ/M = 0.52) the disorder value is less than the nearest neighbor interaction energy and the exciton structure is close to that of the homogeneous limit. The k = ±1 levels are almost degenerate and each have a dipole strength of about 8.5 (in the units of monomeric dipole strength). The lowest k = 0 state becomes weakly allowed borrowing a dipole strength of about 1 from the k = ±1 levels. The PR values are 0.6–1.2 and 0.9–1.1 for k = 0 and higher levels, respectively. In the second case the disorder value exceeds the exciton coupling (σ/M = 1.29), producing significant mixing of the exciton wavefunctions. The splitting between the k = ±1 states increases to 70 cm−1. The lowest state becomes superradiant (with dipole strength of 2) and more localized (with PR values increasing to 0.4–0.5 for the red-shifted realizations). Higher states become allowed and more localized (PR is spread between 0.09 and 0.2).
236
Rienk van Grondelle and Vladimir I. Novoderezhkin
Fig. 2. Top frames: Measured (open circles) and calculated (solid lines) absorption spectra for the LH2-B850 antenna of Rps. acidophila at room temperature. The calculated absorption is shown together with the individual exciton components (thin lines). Bars show dipole strengths of the exciton components (in units of monomeric dipole strength) averaged over disorder. Line shapes are obtained with the spectral density modeled by an overdamped Brownian oscillator with coupling value of λ0 = 200 cm−1 and width (inverse bath relaxation time) of γ0 = 100 cm−1. The energies of the 1α1β, 1β2α, 1α2α, 1β2β, and 1α2β interactions were taken to be 291, 273, −50, −36, and 12 cm−1.The absorption spectrum is averaged over disorder, which was modeled by uncorrelated shifts of the site energies (diagonal disorder) taken from a Gaussian distribution with FWHM of σ = 150, 375, and 650 cm−1. Bottom frames: Participation ratio PR(k) = ∑n(cnk)4 for the 7 lowest exciton levels (i.e., k = 0, −1, 1, −2, 2, −3 and 3) calculated for 2000 realizations of the disorder. Each point shows the PR value for one exciton state for one realization as a function of the wavelength corresponding to the zero-phonon line (ZPL) of this state.
A further increase of the disorder (σ/M = 2.23 in the third case) produces an even more uniform distribution of the dipole strength. The lowest state has (on average) become more superradiant, with an average dipole strength of 3. On the other hand, there are red-shifted realizations with an almost localized lowest state (with PR value up to 0.6–0.8). The higher states are more localized as well (with PR values distributed between 0.09 and 0.35). The LH1 antenna has a structure similar to that of the B850 ring of LH2. The larger size (N = 30–32 instead of 16–18 in LH2) results in less spacing between the zero-order exciton levels, which in turn produces stronger disorder-induced mixing of the
exciton states. In particular, the lowest k = 0 state becomes more superradiant and the higher k = ±2 states (which in LH2 stay almost forbidden even with moderate disorder) now give a significant contribution to the absorption profile (Novoderezhkin et al., 1999a; 2000; 2002). C. Steady-state Spectra The steady-state spectroscopic properties of LH1/ LH2 were studied with polarized light spectroscopy (Van Mourik et al., 1992; Visschers et al., 1995), nonlinear polarized absorption (Leupold et al., 1996), hole-burning (Reddy et al., 1991, 1992, 1993; Wu
Chapter 13
Spectroscopy and Dynamics of Excitation Transfer
et al., 1997a), low-temperature FL (Freiberg et al., 2003), superradiance (Monshouwer et al., 1997), circular dichroism (CD) (Koolhaas et al., 1997b, 1998; Georgakopoulou et al., 2002), Stark spectroscopy (Beekman et al., 1997), and single-molecule techniques (Bopp et al., 1997, 1999; Van Oijen et al., 1998, 1999a,b; Tietz et al., 1999; Ketelaars et al., 2001; Gerken et al., 2003a,b; Hoffman et al., 2003; Rutkauskas et al., 2004, 2005, 2006).The data revealed important details of the exciton structure such as the exciton splitting value, disorder values (for different types of disorder), dipole strengths of the exciton transitions, and the delocalization length. The majority of the steady-state spectroscopic results can be explained within the framework of the exciton theory taking into account the influence of the static disorder. Low temperature polarized FL (Van Mourik et al., 1992) and hole-burning (Reddy et al., 1992) data revealed the presence of the allowed lowest k = 0 level together with intense higher levels broadened due to relaxation. A specific shape of the CD spectra with a red-shift of the zero-crossing (Koolhaas et al., 1997b) reflects a collective (coherent) contribution of the pigments within a large part of a complete ring. The disorder manifests itself through a mixing of the k = ±1 exciton states with the k = 0 state producing a strongly allowed superradiant lowest state observed in FL experiments (Monshouwer et al., 1997). On the other hand, it appeared that including uncorrelated disorder into the exciton model is not sufficient to explain the large splitting between the k = 0 and k = ±1 states observed in hole-burning spectra and consequently correlated disorder was introduced (Wu et al., 1997b). Also, the large splitting between the orthogonally polarized transitions observed in single molecule FL excitation spectra was explained in terms of correlated disorder produced by an elliptical deformation of the ring (Van Oijen et al., 1999b; Ketelaars et al., 2001). The simple exciton model with static disorder (uncorrelated or correlated) also fails to explain the anomalously broad FL profiles observed for LH1/LH2 at low temperature (Freiberg et al., 2003; Timpmann et al., 2004). The data suggested the presence of strong exciton-phonon coupling producing polaron features in the low temperature emission spectra (Freiberg et al., 2003; Timpmann et al., 2004). Including strong exciton-phonon coupling into the modified Redfield theory (see Eqs. 3, 4, and 7) yields more realistic line shapes with reorganization shifts
237
and phonon-induced broadening of the exciton levels in addition to the exciton splitting. These phonon-induced effects, proportional to the PR value according to Eq. (4), are more pronounced for the lowest k = 0 state (especially for strongly disordered realizations giving rise to a localized and red-shifted k = 0 state as seen in Fig. 2). This produces a much broader low-temperature FL profile as compared to the simple exciton model with just static disorder. On the other hand, the disorder of the reorganization shift values produces additional splitting between the exciton states and additional inhomogeneous broadening. Application of the modified Redfield approach allowed a quantitative explanation of the shapes of FL profiles and fluctuations of the FL peak positions observed in a single LH2 complex (Novoderezhkin et al., 2006; Rutkauskas et al., 2006; see Section VI). D. Multiple Exciton Delocalization Sizes The precise character of the excitation within the circular aggregates of the B850-LH2 and LH1 antennae, such as the degree of delocalization, multiple coherence sizes, dynamic localization, etc., has been the subject of intense debate (Reddy et al., 1992; Novoderezhkin and Razjivin, 1993, 1994, 1995a; Bradforth et al., 1995; Dracheva et al., 1995, 1997; Pullerits et al., 1996; Alden et al., 1997; Monshouwer et al., 1997; Nagarajan et al., 1999; Freiberg et al., 2003) and was critically reviewed and summarized in Novoderezhkin et al. (1999b); Meier et al. (1997c); Dahlbom et al. (2001). It should be noted that there exist many different definitions of the degree of delocalization which implies that even for the same system one may find different sizes of an exciton. Delocalization of individual exciton states can be characterized by the inverse participation ratio PR(k), which is different for different states (as illustrated in Fig.2). The effective (averaged over k) delocalization of the exciton wavefunctions Neff can be defined as PR−1(k) weighed with the populations of the exciton states at thermal equilibrium (Meier et al., 1997a). Typically the excited state in the antenna is given by a superposition of exciton states, i.e., an exciton wavepacket is formed whose evolution in time is described by the timedependent density matrix (Mukamel, 1995; Kühn and Sundström, 1997b; Meier et al., 1997a,b,c). The coherence length (or delocalization length) of the wavepacket Ncoh can be defined as the width of antidiagonal distribution of the density matrix in the site
238 representation (Kühn and Sundström, 1997). Modeling of the LH1 and LH2 antenna proteins in the presence of static disorder gave the values Neff = 5–11 (Jimenez et al., 1996; Alden et al., 1997; Meier et al., 1997c; Novoderezhkin et al., 1999) and Ncoh = 4–6 (Kühn and Sundström, 1997b; Novoderezhkin et al., 1999a,b) at room temperature. The dynamic disorder induced by exciton-phonon coupling may give rise to a further reduction of these exciton sizes due to polaron formation (Meier et al., 1997a). Notice that both Neff and Ncoh values reflect some kind of nonuniformity of the shape of the exciton wavefunction, but not a true ‘physical size.’ Generally, spectral responses of a strongly coupled aggregate reflect the cooperative behavior of some number of pigments which can be very different from the effective localization sizes, Neff , Ncoh (Koolhaas et al., 1997b; Meier et al., 1997c; Koolhaas et al., 2000). The number of molecules that contribute to the spectral response can also be different for different spectroscopic techniques. For example, the shape of the CD spectra of the B850 band of the LH2 antenna is determined by cooperativity within at least 10–12 molecules (Koolhaas et al., 1997b; 2000). The shape and amplitude of the transient absorption measured for B850 can only be reproduced by taking into account the exciton coupling between more than 12 molecules of the antenna (Novoderezhkin et al., 1999b). One could define the spectroscopic subunit as the minimal fragment of the ring which is big enough to reproduce all spectral features of the whole antenna (Koolhaas et al., 2000). Due to disorder, the size of such a subunit is less than the aggregate size N, but larger than both effective localization sizes, Neff and Ncoh. E. Superradiance and Anomalously High Nonlinear Responses One of the manifestations of the collective character of excitations in the antenna is an increase in dipole strength of the lowest exciton states (i.e., k = 0 and k = ±1 states as shown in Fig. 2). An anomalously high dipole strength of the main k = ±1 transitions gives rise to an increase of the nonlinear response amplitudes in proportion to N (Novoderezhkin and Razjivin, 1993; 1995b; Leupold et al., 1996; Novoderezhkin et al., 1999b). Pump-probe studies on the core RC-LH1 complex revealed bleaching amplitudes in the antenna that were several times bigger than those corresponding
Rienk van Grondelle and Vladimir I. Novoderezhkin to the bleaching of the special pair due to oxidation of the RC (Novoderezhkin and Razjivin, 1993, 1995b; Kennis et al., 1994; Xiao et al., 1994). In the LH2 complex the bleaching amplitude of the B850 ring was found to be much higher than for the B800 monomeric band (Kennis et al., 1996). Quantitative modeling of the shapes and relative amplitudes of the TA spectra for the LH2 antenna and isolated B820 dimeric subunit (Novoderezhkin et al., 1999b) provided direct evidence for a delocalization of the exciton over many BChls in the B850 ring; corresponding to effective localization sizes for the B850 antenna of Neff = 7.87 and Ncoh = 5.0. The dipole strength of the lowest exciton state (sensitive to the amount of the static disorder) can be determined from the low-temperature transient absorption (TA) (Kennis et al., 1997b) and superradiance (Monshouwer et al., 1997) data. Thus, the integrated intensity of the stimulated emission component in B850 corresponds to the dipole strength of the lowest exciton level of 2.3–3.4 (Kennis et al., 1997b). From the low-temperature superradiance data this value was estimated as 2.8 and 3.8 for the LH2 and LH1 antenna, respectively (Monshouwer et al., 1997). Taken together, these nonlinear spectroscopic results have lead to the disordered exciton model for the LH2-B850 and LH1 antennae (Novoderezhkin et al., 1999a). F. Exciton Transitions Viewed by Polarized Single Molecule Spectroscopy One of the most prominent manifestations of excitonic effects in the LH2-B850 and LH1 antennae is the polarization of the two main transitions (k = ±1) of a circular aggregate. Even in the presence of disorder the largest part of the dipole strength is still concentrated in the original two k = ±1 transitions that have perpendicular polarization in the plane of the ring (Novoderezhkin et al., 1999a). In contrast, localization of the excitation at one site (or on one dimeric BChl subunit) would produce N (or N/2) transitions with different (almost in-plane) polarizations having angles uniformly distributed between 0 and 2π. Such localized features were indeed observed in a single molecule study of the monomeric B800 band of LH2 (Van Oijen et al., 1999b). On the other hand, FL excitation spectra in the 850 nm region (Bopp et al., 1999; van Oijen et al., 1999a) demonstrated the existence of just two perpendicularly polarized transitions in a single LH2 complex, thus giving
Chapter 13
Spectroscopy and Dynamics of Excitation Transfer
direct evidence for pronounced excitonic features in the B850 antenna. Such a confirmation of one of the main predictions of the exciton model was not evident a priori, since in a real pigment-protein complex the ideal exciton picture may easily be destroyed by slow conformational dynamics of the complex inducing quasi-static energetic disorder, and fast phonons producing polaron self-trapping. The anomalously large splitting of the two major orthogonal excitonic transitions observed in the low temperature FL spectra of LH2 was attributed to a modulation of the coupling strength in the B850 ring that was asserted to be associated with an elliptical deformation of the LH2 ring (Bopp et al., 1999; van Oijen et al., 1999a; Mostovoy and Knoester, 2000; Ketelaars et al., 2001; Matsushita et al., 2001) (see also Section VI). III. Equilibration Dynamics The energy transfer dynamics within LH2 includes migration of localized excitations around the BChl 800 outer ring, superimposed on excitation transfer to the exciton states of the BChl 850 inner ring with subsequent equilibration in the B850 exciton manifold and the motion of the quasi-steady-state wavepacket (delocalized over 4–6 BChls) around the B850 inner ring (see Sundström et al., 1999; Renger et al., 2001; Van Grondelle and Novoderezhkin, 2001, 2006 for reviews.). The dynamics within the B850 band was studied by fs FL up-conversion (Jimenez et al., 1996), relative measurements of the induced absorption changes (Novoderezhkin et al., 1999b, Kennis et al., 1996), polarized pump-probe (Nagarajan et al., 1996, 1999; Chachisvilis et al., 1997; Vulto et al., 1999), and photon-echo (Joo et al., 1996; Jimenez et al., 1997). Modeling has included the quantitative fitting of linear and nonlinear responses (Kühn and Sundström, 1997b; Novoderezhkin et al., 1999b, 2003; Brüggemann and May, 2004). The dynamics within the LH1 antenna are very similar to the equilibration dynamics within the B850 band of LH2. But due to its larger size the LH1 antenna is characterized by a more pronounced contribution to the absorption from higher exciton levels (k = ±2 and ±3) and this feature is responsible for the increased amplitude of the dynamic Stokes shift observed in LH1 as compared with LH2 (Monshouwer et al., 1998; Novoderezhkin and Van Gron-
239
delle, 2002). Another specific property of LH1 is the stronger coupling to some low-frequency vibrations, manifested as more pronounced oscillatory features in nonlinear responses due to vibrational coherences (Chachisvilis and Sundström, 1996; Monshouwer et al., 1998; Novoderezhkin et al., 2000). A. Exciton Relaxation and Dynamic Red-Shift The direct observation of excitation dynamics in the tightly packed ring-like aggregates of the LH1 and B850-LH2 antennae (including relaxation of the exciton states and migration of the quasi-steady-state exciton wavepacket) is possible by transient absorption studies via the ps/fs pump-probe technique. Such measurements have revealed the initial ultrafast (sub100 fs) relaxation in LH1/LH2 complexes followed by slower (ps) dynamics (Xiao et al., 1994; Visser et al., 1995, 1996; Nagarajan et al., 1996, 1999; Savikhin and Struve, 1996; Chachisvilis et al., 1997; Kennis et al., 1997a,b; Monshouwer et al., 1998; Vulto et al., 1999; Polívka et al., 2000; Book et al., 2001). Fast relaxation in such a system gives rise to specific TA dynamics typical of assemblies of excitonically coupled antenna pigments, but absent from isolated dimeric subunits and monomers. Thus, Visser et al. (1995) found pronounced TA dynamics for LH1 of Rsp. rubrum with an 8–12 nm red-shift reflecting fast exciton equilibration, whereas no dynamic red-shift was found in the lowest-state absorption band of either the B820 dimeric subunit or for monomeric BChl. Short-pulse (35–50 fs) pump-probe studies of LH1 and LH2 of Rhodobacter (Rba.) sphaeroides (Nagarajan et al., 1996, 1999; Chachisvilis et al., 1997; Vulto et al., 1999) showed relaxation components in the 10–100 fs range; short-lived <20 fs components appear when tuning the wavelength to the blue (Chachisvilis et al., 1997). Selective excitation with longer pulses at the blue side of the LH2 absorption resulted in a dynamic red-shift of the TA spectrum by 3–5 nm with a time constant of 80–110 fs (Nagarajan et al., 1999; Vulto et al., 1999). This shift (attributed to relaxation from higher to lower exciton levels) was almost absent upon red excitation. Upon blue-side excitation of the LH1 of Blastochloris (Blc.) viridis (formerly Rhodopseudomonas viridis) a dynamic red-shift was observed with a time constant of 130–150 fs accompanied by a decay of the TA anisotropy with a similar time constant of about 150 fs (Monshouwer et al., 1998). The amplitude of this red-shift (15–20 nm) is several times larger than
240
Rienk van Grondelle and Vladimir I. Novoderezhkin
Fig. 3. (A). Dynamics of the TA spectrum measured upon 1017 nm excitation of the LH1 of Blc. viridis at 77K (Monshouwer et al., 1998). (B). Measured (points) and calculated (solid lines) red-shift of the TA spectrum between 0 and 400 fs upon 1017 nm excitation of the LH1 of Blc. viridis at 77K (Novoderezhkin and van Grondelle, 2002). (C). Measured (points) and modeled (solid line) TA kinetics at 1050 nm traces upon 1055 nm excitation of the LH1 of Blc. viridis at 77K (Novoderezhkin et al., 2000). See also Color Plate 4, Fig. 5.
in LH2 and is similar to that obtained for other LH1 antennae in earlier experiments (Xiao et al., 1994; Visser et al., 1995). Upon tuning the excitation wavelength to the middle of the band the amplitude of the dynamic red-shift was significantly reduced. At delays shorter than 400 fs the TA dynamics are determined mostly by a large dynamic red-shift due to fast exciton relaxation (Fig. 3A). These relaxation dynamics were explained quantitatively (Fig. 3B) using the Redfield relaxation theory and disordered exciton model, with uncorrelated diagonal disorder (Novoderezhkin and Van Grondelle, 2002). The relaxation-induced lifetimes of the higher exciton states have been calculated as 75, 41, 23, 17, and 12 fs for k = –1, 1, –2, 2, and –3, respectively. In the disordered exciton model the red-shift dynamics upon blue-side excitation are determined by relaxation from strongly allowed higher k = ±2 levels. Upon middleband excitation, corresponding to excitation of the k = ±1 levels, the red-shift is much less pronounced. Notice that for the models with correlated disorder (for instance due to elliptical deformations, correlated shift of the site energies, etc.) the absorption spectrum is determined mostly by the k = ±1 states, whereas the k = ±2 states remain forbidden (Mostovoy and Knoester, 2000; Matsushita et al., 2001; Rutkauskas
et al., 2005; Novoderezhkin et al., 2006). As a consequence these models predict that the shift towards the red of the TA upon blue excitation is almost the same as for middle-band excitation (Novoderezhkin and Van Grondelle, 2002), in contradiction of the experiment. B. Interplay of Excitonic and Vibrational Coherences Excitation dynamics in the ring-shaped antenna complexes LH1 and B850-LH2 studied in the subps to ps time range using fs pump-probe, transient grating, FL upconversion and photon-echo spectroscopies revealed pronounced oscillatory features due to vibrational coherences (Chachisvilis et al., 1994, 1995; Bradforth et al., 1995; Chachisvilis and Sundström, 1996; Joo et al., 1996; Kumble et al., 1996, Jimenez et al., 1997, Monshouwer et al., 1998). These coherences were found to be more intense and more long-lived in LH1. Thus, experimental studies of the TA dynamics for the LH1 antenna of Blc. viridis (Monshouwer et al., 1998) revealed the coupling to two vibrational modes with frequencies of about 60 and 105 cm−1, which produce oscillations in the TA signals that persist for 1–1.5 ps with a surprisingly
Chapter 13
Spectroscopy and Dynamics of Excitation Transfer
large damping time constant (600-800 fs). This time significantly exceeds the time of exciton relaxation or hopping as determined by the ultrafast redshift of the difference absorption and by anisotropy decay. At short delays (<200 fs) the oscillatory features due to vibrational coherences have a smaller amplitude than the red-shift TA dynamics, while at longer delays (after exciton relaxation) the oscillations are clearly distinguishable. The wavelength-dependent oscillatory pattern was modeled using the Redfield theory in the basis of exciton-vibrational eigenstates (Novoderezhkin et al., 2000). The impulsive excitation of electronic levels coupled to specific nuclear modes creates a wavepacket in nuclear space (both in the ground state and in the one-exciton manifold). Coherent motion of the ground state wavepacket yields oscillations in the photobleaching (PB), whereas the excited state wavepacket produces similar oscillations (but with different phase) in the one-exciton stimulated emission and the one-to-two exciton absorption. Superposition of these three contributions results in a complex wavelength-dependent pattern of oscillations in the total TA signal that can be quantitatively reproduced by the model (Fig. 3C). The simultaneous fit of the TA traces upon different excitation/detection wavelengths yields precise estimates of the frequencies, coupling parameters, and damping constants of the nuclear modes. For details, see Novoderezhkin et al. (2000). It is important to note that the relaxation of the initially created exciton wavepacket within 150 fs (shown in Fig 3) does not shorten the long-lived (within 1–1.5 ps) coherent oscillations. However, further migration of the wavepacket along the circular antenna induces an additional decay of the vibrational coherences. The time constant of this decay estimated from a fit of the oscillatory patterns is τhop = 0.9–1.5 ps (Novoderezhkin et al., 2000) and this was interpreted as the effective time of migration over a distance comparable with the delocalization length. The latter is equal to 8 BChl molecules at 77 K (Novoderezhkin et al., 1999a; Novoderezhkin and Van Grondelle, 2002). Recall that at room temperature the delocalization length in LH1/LH2 is 4–6 BChls. The hopping time of the wavepacket in this case must be shorter than τhop = 0.9–1.5 ps as suggested by the 77 K data. For instance, modeling of the room-temperature single-molecule LH2 data (see Section IV) gave τhop = 350 fs.
241
IV. Competition of Intraband B800-800 and Interband B800-850 Energy Transfer in the Light-Harvesting 2 Complex Following the determination of the structure of the LH2 complex of Rps. acidophila considerable efforts have been made to build up a consistent picture of excitation energy transfer within the whole B800B850 complex. One of the most puzzling and intriguing questions is the interplay of intraband B800 → B800 and interband B800 → B850 energy transfers. This problem has been the subject of intense studies by hole burning spectroscopy (Van der Laan et al., 1990; Reddy et al., 1991; de Caro et al., 1994; Wu et al., 1996), pump-probe techniques (Hess et al., 1993, 1995; Joo et al., 1996; Monshouwer et al., 1995; Ma et al., 1997, 1998; Kennis et al., 1996, 1997a; Pullerits et al., 1997; Ihalainen et al., 2001; Wendling et al., 2003), three-pulse photon echo techniques (Salverda et al., 2000; Agarval et al., 2001) together with quantitative modeling of nonlinear responses (Novoderezhkin et al., 2003). New insight into the problem has been reached in a recent 2-D-photon echo study of the B800-820 complex (Zigmantas et al., 2006). A. Excitation-wavelength Dependent Decay of the B800 Band Typically pump-probe kinetics measured in the B800 band show a biexponential decay of the isotropic TA (Hess et al., 1993, 1995; Monshouwer et al., 1995; Kennis et al., 1996, 1997a; Ma et al., 1997, 1998; Pullerits et al., 1997; Ihalainen et al., 2001; Wendling et al., 2003). A slow component of 1.2–1.9 ps is taken to reflect the B800 → B850 transfer, whereas a fast phase of 0.3–0.8 ps has been assigned to the B800→ B800 hopping. The anisotropy decays with approximately the same (about 0.3–0.5 ps) time, thus lending support to the assignment of the fast component to intraband B800 → B800 transfer. Modeling of energy transfer within B800-B850 complexes was done using generalized Förster theory (Sumi, 1999; Mukai et al., 1999; Scholes and Fleming 2000; Jang et al., 2004) and Redfield theory (Kühn and Sundström, 1997a; Novoderezhkin et al., 2003). The generalized Förster approach allows the correct modeling of the transfer from B800 to the lower exciton states of the B850 band (due to the big energy gap between these bands — see Section I).
242 But intraband dynamics within the B800 and B850 bands should be described by Redfield theory. At first glance the B800 dynamics corresponds to the case of non-coherent migration of localized excitations. But due to the almost iso-energetic character of the excited states even weak coupling between them creates sizable excitonic coherences that should be taken into account in order to reproduce the spectra (Cheng and Silbey, 2006), pump-probe (Novoderezhkin et al., 2003) and 2-D photon echo kinetics (Zigmantas et al., 2006) observed within the B800 band. In the combined Förster-Redfield approach (Zigmantas et al., 2006) intra-B800 and intra-B850(B820) equilibration is described by modified Redfield, whereas B800-B850(B820) transfer is calculated using generalized Förster. However, in such an approach, the resonant mixing between the B800 states and the upper exciton levels of the B850 band (B850* states) is neglected. Modeling of the whole B800-850 dynamics with the Redfield theory (Novoderezhkin et al., 2003) allows the description of all types of transfers including dynamics of the coherences between one-exciton states. The latter is important for a correct interpretation of the assignment of the B800 → B800 dynamics (see below).
Rienk van Grondelle and Vladimir I. Novoderezhkin An example of one-color polarized TA kinetics measured for LH2 of Rsp. molischianum at 77K (Wendling et al., 2003) and modeled with the Redfield theory (Novoderezhkin et al., 2003) is shown in Fig. 4. The isotropic kinetics (Fig 4A) slow down when tuning the excitation to the red side of the 800 nm band, but they become faster again in the red-most edge of the band. The anisotropy kinetics (Fig 4B) show a fast decay (from 0.42 to 0.2 with about 0.5 ps time constant) which is followed by a slower ps decay (from 0.2 to 0.1 between 1 and 8 ps delay). All these features are satisfactorily reproduced by a model taking into account all the excitonic interactions (including BChl 800-800 and BChl 850-800 couplings), site inhomogeneity (which mix the forbidden B850* states near 800 nm with the B800 states), and the full Redfield relaxation tensor (including relaxation of one-exciton populations and coherences between one-exciton states as well as non-secular terms, like coherence transfer). From this model a detailed picture of the energy transfer pathways in LH2 complex has emerged as shown in Fig. 4C; for more details, see Novoderezhkin et al. (2003).
Fig. 4. (A). One-color isotropic (magic angle) pump-probe kinetics at 793, 797, 800, and 806 nm measured (points) and modeled (solid lines) for the LH2 antenna of Rsp. molischianum at 77K (Wendling et al., 2003; Novoderezhkin et al., 2003). The difference absorption values are normalized to the maximal bleaching amplitude (reached near zero delay) and inverted. (B). Fitting of the anisotropy kinetics measured at 797 nm for the LH2 antenna of Rsp. molischianum at 77 K (Wendling et al., 2003). Experimental data (points) are compared with the kinetics calculated using the non-secular Redfield theory (solid lines), in the secular approximation (dashed curve) and in the non-coherent limit (dotted curve) (Novoderezhkin et al., 2003). (C). The exciton energy level scheme and the averaged (over disorder and over levels within the corresponding subbands) relaxation time constants for LH2 at 77K (van Grondelle and Novoderezhkin, 2006). For simplicity each of the B800, B850* and B850 manifolds is represented by two levels corresponding to the blue and red side of the band. At the blue side of the B800 band two coherently excited levels are indicated.
Chapter 13
Spectroscopy and Dynamics of Excitation Transfer
B. Multiple Pathways of B800 to B850 Transfer According to this model excitation of the B800 band creates coherence between pairs of B800 states, which decays with a time constant of 0.3–0.5 ps. This is followed by a slow migration of the localized exciton around the BChl 800 ring with a time constant of 1–3 ps (depending on the relative energies of neighboring BChls). In the energy level diagram such a migration corresponds to the population of the lowest state of the B800 excited state manifold (as shown in Fig 4C). Besides direct B800 → B850 transfer there is an additional pathway for energy transfer from the outer to inner ring, i.e., transfer through the B850* band. The effectiveness of this channel is determined by fast (600–800 fs) B800 → B850* transfer, followed by even faster (60–200 fs) B850* → B850* and B850* → B850 relaxation. The relative contribution of the B800 → B850* → B850 pathway is equal (or at least comparable) to the contribution of the direct B800→ B850 transfer. Notice that fast back B850* → B800 transfer could in principle contribute to the fast anisotropy decay observed at 800 nm, but such a ‘detour’ pathway B800 → B850* → B800 (suggested in Wu et al., 1996; Kühn and Sundström, 1997a; Renger et al., 2001) is not significant due to the fast B850* → B850* and B850* → B850 relaxations. C. Origin of the B800 Anisotropy Decay The fast decay of one-exciton coherences in the B800 band is mirrored by the sub-ps anisotropy decay from 0.4 to 0.2 in the B800 pump-probe signals, whereas the slow migration produces a further ps depolarization from 0.2 to 0.1. Note that a calculation without including coherences (taking into account the population dynamics only) gives an initial TA anisotropy of about 0.2 which decays to 0.1 with a time constant of a few ps, thus reproducing the experimental anisotropy dynamics for delays larger than 1ps (Fig. 4B). Therefore we conclude that the fast anisotropy decay is not connected with any energy migration around the ring, as was thought for the past 10 years (Hess et al., 1993, 1995; Monshouwer et al., 1995; Kennis et al., 1996, 1997a; Ma et al., 1997, 1998; Pullerits et al., 1997; Ihalainen et al., 2001; Wendling et al., 2003). Actually, the real B800 → B800 population transfer is relatively slow, even slower than the B800 → B850 transfers (having a time constant of about 1 ps — see
243
Fig. 4), in agreement with the three-pulse-photonecho-peak-shift results (Salverda et al., 2000). V. Energy Trapping in the Core Reaction Center-Light-Harvesting 1 Complex A. Trapping Kinetics Energy trapping in RC-LH1 core complexes occurs with a time constant of τ = 60-80 ps for Rba. sphaeroides (Sebban et al., 1984; Borisov et al., 1985) and τ = 50–70 ps for Rsp. rubrum as revealed by FL (Borisov et al., 1985) and absorbance kinetics (Razjivin et al., 1982; Nuijs et al., 1985; Sundström et al., 1986) at room temperature. The probability of back transfer from the RC to the antenna (upon selective excitation of the RC’s absorption band at 800 nm) was estimated as η<10–20% for Chromatium minutissimum (Abdourakhmanov et al. 1989), η < 10–15% (Danielius and Razjivin, 1987) and η = 25% (Timpmann et al., 1993) for Rsp. rubrum, η = 15% for Rba. capsulatus (Xiao et al., 1994), and η<10% for Blc. viridis (Otte et al., 1993). Analysis of the kinetics in the 77–300 K range (Visscher et al., 1989; Bergström et al., 1989; Beekman et al., 1994) showed that the trapping is limited by slow antenna-RC transfer and can be considered as almost single-step hopping from the antenna with a time constant of about 50 ps and with a small probability of back transfer, the ‘transfer-to-the-trap-limited’ model (Somsen et al, 1996). This hopping occurs to the special pair of the RC, whose lowest exciton level is in resonance with the long-wavelength absorption band of LH1 (peaking at 870 nm for Rsp. rubrum and Rba. sphaeroides, and 1015 nm for Blc. viridis). B. Exciton Control of Antenna-Reaction Center Transfer Rates Large distances between the antenna BChls and the special pair, i.e., 4–5 nm, correspond to a very slow hopping time (about 300 ps) if the excitation is assumed to be localized on one BChl of the antenna (Novoderezhkin and Razjivin, 1996). The experimental value of 50 ps can only be reproduced using the generalized Förster theory taking into account the delocalized character of excitations in the antenna (Novoderezhkin and Razjivin, 1994, 1996). It can be shown that the matrix element connecting the k-th exciton level of the antenna and the RC’s special pair
244 (given by Eq. (6)) is proportional to the dipole strength of the exciton level, d2k if the special pair is located in the center of the ring (Novoderezhkin and Razjivin, 1994). This yields an enhancement of the trapping rate of the same magnitude as the superradiance of the lowest exciton levels. It is noteworthy that the temperature dependence of the antenna-RC trapping rate and the antenna superradiance are similar, i.e., both are almost temperature independent (Visscher et al., 1989; Monshouwer et al., 1997). The trapping length (i.e., the increase in trapping rate relative to the completely localized limit) should be close to the thermally averaged superradiance length estimated as 3.8 BChls for LH1 at room temperature (Monshouwer et al., 1997). In fact the trapping length is slightly larger (about 5–6 BChls) due to a better overlap of the special pair lowest excitonic level with the emission from the higher k = ±1 states of the antenna. Upon cooling to ultralow temperatures the trapping becomes extremely heterogeneous and very dependent on the particular realization of the disorder for each RC-LH1 core (Somsen et al., 1994). This is connected with a big spread of delocalization lengths of the lowest exciton level induced by the disorder (Fig. 2). C. Detrapping Efficiency In the exciton model the interaction of the RC with the antenna is mostly determined by its interaction with the three lowest exciton levels, i.e., k = 0 and k = ±1. This can be interpreted as an interaction with three ‘supermolecules’ having anomalously large dipole moments. The effectiveness of back transfer is then roughly three times faster than the direct 50 ps hopping, i.e., about 17 ps (Novoderezhkin and Razjivin, 1994). Bearing in mind that the rate of charge separation in the RC is 3 ps (Martin et al., 1986) we obtain a detrapping efficiency of about 3/17, i.e., a bit less than the experimentally measured value of 0.2. Notice that assuming localization of the excitation at one of the N = 30–32 sites would give approximately N-times faster detrapping, corresponding to a much higher probability of escape. VI. Slow Conformational Motions and Excitation Dynamics in the B850-LightHarvesting 2 Complex Studies of excitation dynamics with conventional
Rienk van Grondelle and Vladimir I. Novoderezhkin bulk spectroscopies, although sensitive to details of the dynamics on the fs time scale, are restricted to averaging over disorder (i.e., over a large number of complexes in different conformational states). Contemporary single molecule techniques do not have fs/ps time resolution, but allow a direct visualization of the conformational changes of individual complexes (through their spectral signatures) on a ms/s time scale. Based on low- and room-temperature single molecule experiments it has been proposed that the LH2 ring can deviate from the ideal circular structure (Bopp et al., 1999; Van Oijen et al., 1999a, Ketelaars et al., 2001). The anomalously large splitting of the two major orthogonal excitonic transitions of the B850 band observed in low-temperature polarized FL excitation spectra was accounted for by introducing correlated shifts of the site energies and/or interaction energies (induced by elliptical deformation) and reducing the amount of uncorrelated disorder (Van Oijen et al., 1999a; Mostovoy and Knoester, 2000; Matsushita et al., 2001). On the other hand, the changes in the FL profile related to the abrupt movements of the FL peak observed at ambient temperatures (Rutkauskas et al., 2004, 2005, 2006) strongly suggest a large value of the uncorrelated disorder in LH2. Modeling based on the modified Redfield theory and the disordered exciton model (Novoderezhkin et al., 2006) allowed a quantitative explanation of the observed spectral fluctuations. The model suggested an evolution of an individual LH2 complex through a number of conformational sub-states with specific disorder patterns producing different types of emission profiles, each corresponding to a different regime of the excitation dynamics. A. Dynamics of Conformational Changes Figure 5A shows time-dependent fluctuations of the FL peak position measured at room temperature for a single LH2 complex (Rutkauskas et al., 2005). On a time scale of seconds the system undergoes a lot of FL peak fluctuations within the 865-875 nm range superimposed with rare large jumps to the blue (855860 nm) and also rare and even larger jumps to the red (up to 900–910 nm). Typically these large jumps produce relatively long-lived states perturbed by fast 2–3 nm fluctuations. The observed dynamics of these spectral shifts can be quantitatively reproduced by a multistate conformational model (Novoderezhkin et al., 2007),
Chapter 13
Spectroscopy and Dynamics of Excitation Transfer
245
Fig. 5. (A). Fluctuations of the FL peak position measured for a single LH2 complex of Rps. acidophila at room temperature (Rutkauskas et al., 2005). (B) Fluctuation of the FL peak calculated using a four-state conformational model (see text). To facilitate the comparison between experiment and model we selected a calculated fragment with a similar history as in the measured fragment. (C) Measured (points) and calculated (solid) fluorescence profiles corresponding to realizations with the peak positions in the range 859–861 nm (top), 869-871 nm (middle), and 889-891 nm (bottom). Calculated FL profiles are shown together with contributions of the three lowest exciton states (thin solid).
where the disorder of the site energies is modeled by assuming a finite number of conformational sub-states for each pigment and the conformational dynamics is then described by introducing phenomenological transfer rates between these sub-states. The relative values of rate constants are adjusted in order to approximate the continuous Gaussian distribution of the site energies used in the disordered exciton model. The simplest discrete model includes just two states i.e., two conformations for each pigment of the complex with different electronic transition energies (Valkunas et al., 2007). The two-state model explains the fast small-size (±10 nm) fluctuations within 860880 nm. Larger-sized jumps (especially 15–35 nm jumps to the red) clearly present in the experimental data can only be reproduced by adding one more pair of conformational states, with one state to the red and one state to the blue. Dynamics calculated with this simple four-state model, shown in Fig. 5B, (Novoderezhkin et al., 2007) display the same character of fluctuations as in the measured trace, suggesting that indeed fluctuations in only a few structural coordinates (i.e., two coordinates responsible for the small and large energy shifts, respectively) in large part determine the experimentally observed disorder.
B. Fluctuations of Spectral Shapes in a Single Complex It is important to note that large spectral shifts (Fig. 5A) are accompanied by changes in spectral shapes. Fig. 5C shows averages of FL profiles with a blue-shifted (859–861 nm), intermediate (869–871 nm), and red-shifted (889–891 nm) peak position. Red-shifted single molecule emission profiles on the average are significantly broadened and have a specific shape with a broader short-wavelength wing. Spectra with intermediate and blue-shifted peak positions feature regular FL asymmetry, i.e., a broader long-wavelength tail. Notice that blue-shifted spectra are also broader than those having an intermediate peak position. The observed spectral changes can be reproduced by calculating the line shapes with the modified Redfield theory (Fig. 5C). Remarkably, the data can be quantitatively fitted by using a discrete (four-state) distribution of the site energies or the continuous Gaussian distribution. The modeling revealed that the spectra of the three classes of realizations (shown in Fig. 5) correspond to three different types of exciton structure characterized by different relative positions, line shapes, and delocalization of the lowest exciton components k = 0, k = −1, and k = 1. The exciton states in the blue-shifted realizations are not significantly destroyed by the
246 disorder. As a result in this case the exciton structure is also not too different from that of the homogeneous ring, i.e., most of the dipole strength is concentrated in two degenerate levels (k = ±1), whereas the lowest state (k = 0) is almost forbidden due to the symmetry of the ring. The room-temperature emission originates mostly from the degenerate k = ±1 pair giving rise to a blue-shifted and broadened FL. FL profiles with intermediate peak positions result from realizations with stronger static disorder. This situation gives rise to an increase in splitting between the exciton levels and a more localized character of the exciton states. The red-shifted and more localized k = 0 state also becomes more radiant, borrowing some of the dipole strength from higher levels, and as a result its contribution to the FL profile becomes more significant. For the red-shifted FL profiles the exciton structure is very strongly affected by the disorder, which causes a large splitting between the energy levels, in particular, the splitting between the k = ±1 and the lowest k = 0 state has increased significantly (as compared with the blue-shifted and intermediate spectral profiles). Due to this large splitting between excitonic energy levels most of the FL originates from the lowest state, which is now strongly allowed and red-shifted. Its localized character gives rise to an increase in the effective phonon coupling. Such increased coupling results in a broadening of the FL profile together with an additional red-shift. It is important to note that the shape of the FL profile with intermediate and red-shifted peak position is largely determined by the shift and broadening of the k = 0 level, a feature that is the key property of the disordered ring and which is absent in the case of correlated disorder (where the k = 0 state is almost forbidden). For instance, the model with elliptical deformation and only a limited amount of uncorrelated disorder yields only a small shift of the spectrum and no significant change of its shape (Novoderezhkin et al., 2006). The increased splitting between the levels accompanied by enhanced phonon coupling is the origin of the increasingly larger broadening of the FL spectra in red-shifted realizations. The increased width of the blue-shifted FL spectra is due to another line broadening factor namely exciton relaxation. For the blue-shifted realizations a significant part of the emission originates from the higher k = ±1 levels, but due to predominant downhill relaxation the inverse lifetime increases for these higher levels. Thus the
Rienk van Grondelle and Vladimir I. Novoderezhkin more pronounced relaxation broadening of the k = ±1 levels causes the larger width of the blue-shifted FL spectra. C. Exciton Dynamics in the Different Conformational States The dynamics of the exciton wavepacket given by the time evolution of the density matrix in the site representation shows three different regimes corresponding to the three types of realizations (Fig. 6). Dynamics calculated for the blue realization shows a wavelike motion of the wavepacket delocalized over 4–5 molecules around the ring. (Notice that the delocalization length of the wavepacket is less than Ndel = 11 for the individual wavefunctions). The initially created wavepacket has its maximum at n = 15–16 (see Fig. 6) and in addition a smaller maximum at n = 2. The passage of the main maximum over half the ring occurs in 100 fs. At this point the main maximum collides with the second maximum moving in the opposite direction. For realizations with intermediate FL position the excitation is typically delocalized over a small group of pigments (in our example over sites n = 2–4 or n = 10–13). The dynamics are strongly reminiscent of hopping from one group of pigments to another with a time constant of about 350 fs. The wavelike motion is almost absent, but the signature of some wavelike flow of excitation density from one group to another can be observed to occur through the intermediate site n = 7. Due to strong disorder the wavelike motion is destroyed by scattering on impurities, producing many secondary waves moving in both directions. This results in a complicated non-uniform standing-wave pattern with oscillating populations at some sites (whereas other sites are almost not populated). Of course in an ensemble experiment a mixture of the wavelike motion and hopping-type dynamics will occur with typical time constants of 100 and 350 fs. These time constants will determine the experimentally observed dynamics for instance as reflected by the FL anisotropy decay. This estimation is in remarkable agreement with the observed bi-exponential FL anisotropy decay with 100 and 400 fs components in FL up-conversion experiments for LH1 (Bradforth et al., 1995) and the B850 ring of LH2 (Jimenez et al., 1996). The red-shifted realizations are created by a conformational change that causes a large red-shift (exceeding the interpigment coupling) at one site.
Chapter 13
Spectroscopy and Dynamics of Excitation Transfer
247
We conclude that the realizations of static disorder observed in single molecule experiments correspond to physically different limits of the excitation dynamics, i.e., coherent wavelike motion of a delocalized exciton (with a 100 fs pass over half the ring), hopping-type motion of the wavepacket (with 350 fs jumps between separated groups of 3–4 molecules), and self-trapping of the excitation that does not move from its localization site. VII. Concluding Remarks
Fig. 6. Coherent dynamics of the density matrix for a single LH2 complex, i.e., calculated without relaxation of populations and coherences for one realization of the disorder corresponding to a blue-shifted (top), an intermediate (middle), and a red-shifted (bottom) FL spectrum (Novoderezhkin et al., 2006). The initial population corresponds to the thermal equilibrium at room temperature (shown in the inserts); initial coherences have arbitrary fixed phases. The figure shows the diagonal elements of the density matrix in the site representation ρ(n,n) as a function of time. See also Color Plate 4, Fig. 6.
This produces the localization of excitation at this site (in our example at the n = 4 site) without any hopping to one of the other sites. But there is still some oscillatory modulation of the n = 4 population due to small coherences between the n = 4 and neighboring sites.
We show that nonlinear spectroscopy in combination with modeling based on high-resolution structures have unveiled the pathways of energy transfer in the bacterial antenna complexes LH1/LH2. In these complexes strongly coupled clusters of light-harvesting pigments are present together with weakly coupled aggregates, giving rise to a complicated interplay of interband conversion and intraband dynamics including fast exciton relaxation and slow migration. Obviously, the traditional Förster theory for excitation energy transfer among weakly coupled chromophores does not work in this case. We demonstrate that a self-consistent description of the dynamics and the spectroscopy can be obtained with the modified Redfield approach, where arbitrary delocalization and arbitrary strong coupling to phonons (at least for diagonal exciton-phonon coupling) are included explicitly. A combination of increasingly realistic physics with new experimental results provides a unique opportunity to reveal the exact nature of elementary events and their interplay in the total excitation dynamics using global systematic modeling. We show examples of simultaneous quantitative fits of the steady-state spectra and nonlinear kinetics for the complexes under study. To make the analysis more critical to the parameter choice it is important to use a simultaneous fit of the data obtained by several complementary nonlinear spectroscopic techniques reflecting different aspects of the dynamic behavior of the system (time- and frequency-domain techniques, pump-probe and photon echo, measurements of isotropic kinetics and anisotropy decays, etc.). For these complex systems such an approach reveals limitations (or even complete failure) of qualitative pictures based on fragmentary fittings. Besides establishing the unperturbed site energies, it is important to know the statistics of their random
248 shifts induced by slow conformations. These shifts can have dramatic effects on the excitation dynamics. To obtain an insight into the origins of the energetic disorder it is useful to combine the bulk (ensembleaveraged) spectroscopies with the single-molecule studies. The parameters of the antenna determined by the modeling allow us to determine the pathways and timescales of energy transfer and a visualization of the excitation dynamics, thus leading to a consistent physical picture of how photosynthetic pigment-proteins perform their function in storing solar energy. In this way nature has taught us how to construct a successful artificial light-harvesting sytem for future application in storing solar energy. Acknowledgments V. I. N. was supported by Russian Foundation for Basic Research (Grant 06-04-48917) and Dutch Foundation for Basic Research (NWO). References Abdurakhmanov IA, Danielius RV and Razjivin AP (1989) Efficiency of excitation trapping by reaction centers of complex B890 from Chromatium minutissimum. FEBS Letters 245: 47–50 Agarwal R, Yang M, Xu QH and Fleming GR (2001) Three pulse photon echo peak shift study of the B800 band of the LH2 complex of Rps. acidophila at room temperature: A coupled master equation and nonlinear optical response function approach. J Phys Chem B 105: 1887–1894 Alden RG, Johnson E, Nagarajan V, Parson WW, Law CJ and Cogdell RJ (1997) Calculation of spectroscopic properties of the LH2 bacteriochlorophyll-protein antenna complex of Rhodopseudomonas acidophila. J Phys Chem B 101: 4667–4680 Beekman LMP, Van Mourik F, Jones MR, Visser HM, Hunter CN and Van Grondelle R (1994) Trapping kinetics in mutants of the photosynthetic purple bacterium Rhodobacter sphaeroides: Influence of the charge separation rate and consequences for the rate-limiting step in the light-harvesting process. Biochemistry 33: 3143–3147 Beekman LMP, Frese RN, Fowler GJS, Picorel R, Cogdell RJ, Van Stokkum IHM, Hunter CN and Van Grondelle R (1997) Characterization of the light-harvesting antennas of photosynthetic purple bacteria by Stark spectroscopy. 2. LH2 complexes: Influence of the protein environment. J Phys Chem B 101: 7293–7301 Bergström H, Van Grondelle R and Sundström V (1989) Characterization of excitation energy trapping in photosynthetic purple bacteria at 77 K. FEBS Lett 250: 503–508 Book LD, Ostafin AE, Ponomarenko N, Norris JR and Scherer NF
Rienk van Grondelle and Vladimir I. Novoderezhkin (2000) Exciton delocalization and initial dephasing dynamics of purple bacterial LH2. J Phys Chem B 104: 8295–8307 Bopp MA, Jia Y, Li L, Cogdell RJ and Hochstrasser RM (1997) Fluorescence and photobleaching dynamics of single light-harvesting complexes. Proc Natl Acad Sci USA 94: 10630–10635 Bopp MA, Sytnik A, Howard TD, Cogdell RJ and Hochstrasser RM (1999) The dynamics of structural deformations of immobilized single light-harvesting complexes. Proc Natl Acad Sci USA 96: 11271–11276 Borisov AY, Freiberg AM, Godik VI, Rebane KK, and Timpmann KN (1985) Kinetics and picosecond bacteriochlorophyll fluorescence in vivo as a function of reaction center state. Biochim Biophys Acta 809: 221–229 Bradforth SE, Jimenez R, Van Mourik F, Van Grondelle R and Fleming GR (1995) Excitation transfer in the core light-harvesting complex (LH-1) of Rhodobacter sphaeroides: An ultrafast fluorescence depolarization and annihilation study. J Phys Chem 99: 16179–16191 Brüggemann B and May V (2004) Exciton exciton annihilation dynamics in chromophore complexes. II. Intensity dependent transient absorption of the LH2 antenna system. J Chem Phys 120: 2325–2336 Chachisvilis M and Sundström V (1996) Femtosecond vibrational dynamics and relaxation in the core light-harvesting complex of photosynthetic purple bacteria. Chem Phys Lett 261: 165–174 Chachisvilis M, Pullerits T, Jones MR, Hunter CN and Sundström V (1994) Vibrational dynamics in the light-harvesting complexes of the photosynthetic bacterium Rhodobacter sphaeroides. Chem Phys Lett 224: 345–351 Chachisvilis M, Fidder H, Pullerits T and Sundström V (1995) Coherent nuclear motions in light-harvesting pigments and dye molecules, probed by ultrafast spectroscopy. J Raman Spectroscopy 26: 513–522 Chachisvilis M, Kühn O, Pullerits T and Sundström V (1997) Excitons in photosynthetic purple bacteria: Wavelike motion or incoherent hopping? J Phys Chem B 101: 7275–7283 Cheng YC and Silbey RJ (2006) Coherence in the B800 ring of purple bacteria LH2. Phys Rev Lett 96: 028103(1–4) Dahlbom M, Pullerits T, Mukamel S and Sundström V (2001) Exciton delocalization in the B850 light-harvesting complex: comparison of different measures. J Phys Chem B 105: 5515–5524 Damjanovic A, Kosztin I, Kleinekathoefer U and Schulten K (2002) Excitons in a photosynthetic light-harvesting system: A combined molecular dynamics, quantum chemistry, and polaron model study. Phys Rev E 65: 031919(1–24) Danielius RV and Razjivin AP (1987) New data on the state of excitations in the light-harvesting antenna and their trapping by reaction centers. In: Rudzikas Z, Piscarskas AS and Baltramiejunas R (eds) Ultrafast Phenomena in Spectroscopy, pp 231–239. Word Scientific, Singapore Davydov AS (1971) Theory of Molecular Excitons. Plenum Press, New York De Caro CD, Visschers RW, Van Grondelle R and Völker S (1994) Inter- and intraband energy transfer in LH2-antenna complexes of purple bacteria. A fluorescence line-narrowing and holeburning study. J Phys Chem 98: 10584–10590 Dracheva TV, Novoderezhkin VI and Razjivin AP (1995) Exciton theory of spectra and energy transfer in photosynthesis:
Chapter 13
Spectroscopy and Dynamics of Excitation Transfer
Spectral hole burning in the antenna of purple bacteria. Chem Phys 194: 223–235 Dracheva TV, Novoderezhkin VI and Razjivin AP (1996) Exciton delocalization in the antenna of purple bacteria: Exciton spectra calculation using X-ray data and experimental site inhomogeneity. FEBS Lett 378: 81–84 Dracheva TV, Novoderezhkin VI and Razjivin AP (1997) Exciton delocalization in the light-harvesting LH2 complex of photosynthetic purple bacteria. Photochem Photobiol 66: 141–146 Förster T (1965) Delocalized Excitation and Excitation Transfer. In Sinanoğlu (ed) Modern Quantum Chemistry, Part III. B. Action of Light and Organic Crystals, pp 93–137. Academic Press, New York. Freiberg A, Rätsep M, Timpmann K, Trinkunas G and Woodbury NW (2003) Self-trapped excitons in LH2 antenna complexes between 5 K and ambient temperature. J Phys Chem B 107: 11510–11519 Georgakopoulou S, Frese RN, Johnson E, Koolhaas MHC, Cogdell RJ, Van Grondelle R and Van der Zwan G (2002) Absorption and CD spectroscopy and modeling of various LH2 complexes from purple bacteria. Biophys J 82: 2184–2197 Gerken U, Jelezko F, Götze B, Branschädel M, Tietz C, Ghosh R and Wrachtrup J (2003a) Membrane environment reduces the accessible conformational space available to an integral membrane protein. J Phys Chem B 107: 338–343 Gerken U, Lupo D, Tietz C, Wrachtrup J and Ghosh R (2003b) Circular symmetry of the light-harvesting 1 complex from Rhodospirillum rubrum is not perturbed by interaction with the reaction center. Biochemistry 42: 10354–10360 Hess S, Feldchtein F, Babin A, Nurgaleev I, Pullerits T, Sergeev A and Sundström V (1993) Femtosecond energy transfer within the LH2 peripheral antenna of the photosynthetic purple bacteria Rhodobacter sphaeroides and Rhodopseudomonas palustris LL. Chem Phys Lett 216: 247–257 Hess S, Åkesson E, Cogdell RJ, Pullerits T and Sundström V (1995) Energy transfer in spectrally inhomogeneous lightharvesting pigment protein complexes of purple bacteria. Biophys J 69: 2211–2225 Hofmann C, Francia F, Venturoli G, Oesterhelt D and Köhler J (2003) Energy transfer in a single self-aggregated photosynthetic unit. FEBS Letters 546: 345–348 Hu X, Ritz T, Damjanović A and Shulten K (1997) Pigment organization and transfer of electronic excitation in the photosynthetic unit of purple bacteria. J Phys Chem B 101: 3854–3871 Hu X, Ritz T, Damjanović A, Autenrieth F and Schulten K (2002) Photosynthetic apparatus of purple bacteria. Q Rev Biophys 35: 1–62 Ihalainen JA, Linnanto J, Myllyperkiö P, Van Stokkum IHM, Ücker B, Scheer H and Korppi-Tommola JEI (2001) Energy transfer in LH2 of Rhodospirillum molischianum, studied by subpicosecond spectroscopy and configuration interaction exciton calculations. J Phys Chem B 105: 9849–9856 Jang S, Newton MD and Silbey RJ (2004) Multichromophoric Förster resonance energy transfer. Phys Rev Lett 92: 218301(1–4) Jimenez R, Dikshit SN, Bradforth SE and Fleming GR (1996) Electronic excitation transfer in the LH2 complex of Rhodobacter sphaeroides. J Phys Chem 100: 6825–6834 Jimenez R, Van Mourik F, Yu J-Y and Fleming GR (1997) Threepulse photon echo measurements on LH1 and LH2 complexes of Rhodobacter sphaeroides: A nonlinear spectroscopic probe
249
of energy transfer. J Phys Chem B 101: 7350–7359 Joo T, Jia Y, Yu J-Y, Jonas DM and Fleming GR (1996) Dynamics in isolated bacterial light harvesting antenna (LH2) of Rhodobacter sphaeroides at room temperature. J Phys Chem 100: 2399–2409 Karrasch S, Bullough PA and Ghosh R (1995) The 8.5 A projection map of the light-harvesting complex I from Rhodospirillum rubrum reveals a ring composed of 16 subunits. EMBO J 14: 631–638 Kennis JTM, Aartsma TJ and Amesz J (1994) Energy trapping in the purple bacteria Chromatium vinosum and Chromatium tepidum. Biochim Biophys Acta 1188: 278–286 Kennis JTM, Streltsov AM, Aartsma TJ, Nozava T and Amesz J (1996) Energy transfer and exciton coupling in isolated B800-850 complexes of the photosynthetic purple sulfur bacterium Chromatium tepidum. The effect of structural symmetry on bacteriochlorophyll excited states. J Phys Chem 100: 2438–2442 Kennis JTM, Streltsov AM, Vulto SIE, Aartsma TJ, Nozava T and Amesz J (1997a) Femtosecond dynamics in isolated LH2 complexes of various species of purple bacteria. J Phys Chem B 101: 7827–7834 Kennis JTM, Streltsov AM, Permentier H, Aartsma TJ and Amesz J (1997b) Exciton coherence and energy transfer in the LH2 complex of Rhodopseudomonas acidophila at low temperature. J Phys Chem B 101: 8369–8374 Ketelaars M, Van Oijen AM, Matsushita M, Köhler J, Schmidt J and Aartsma TJ (2001) Spectroscopy on the B850 band of individual light-harvesting 2 complexes of Rhodopseudomonas acidophila I. Experiments and Monte Carlo simulations. Biophys J 80: 1591–1603 Koepke J, Hu X, Muenke C, Schulten K and Michel H (1996) The crystal structure of the light-harvesting complex II (B800-850) from Rhodospirillum molischianum. Structure 4: 581–597 Koolhaas MHC, Van der Zwan G, Van Mourik F and Van Grondelle R (1997a) Spectroscopy and structure of bacteriochlorophyll dimers. I. Structural consequences of non-conservative CD spectra. Biophys J 72: 1828–1841 Koolhaas MHC, Van der Zwan G, Frese RN and Van Grondelle R (1997b) Red shift of the zero crossing in the CD spectra of the LH2 antenna complex of Rhodopseudomonas acidophila: A structure-based study. J Phys Chem B 101: 7262–7270 Koolhaas MHC, Frese RN, Fowler GJS, Bibby TS, Georgakopoulou S, Van der Zwan G, Hunter CN and Van Grondelle R (1998) Identification of the upper exciton component of the B850 bacteriochlorophylls of the LH2 antenna complex, using a B800-free mutant of Rhodobacter sphaeroides. Biochemistry 37: 4693–4698 Koolhaas MHC, Van der Zwan G and Van Grondelle R (2000) Local and nonlocal contributions to the linear spectroscopy of light-harvesting antenna systems. J Phys Chem B 104: 4489–4502 Kühn O and Sundström V (1997a) Energy transfer and relaxation dynamics in light-harvesting antenna complexes of photosynthetic bacteria. J Phys Chem B 101: 3432–3440 Kühn O and Sundström V (1997b) Pump-probe spectroscopy of dissipative energy transfer dynamics in photosynthetic antenna complexes: A density matrix approach. J Chem Phys 107: 4154–4164 Kühn O, Sundström V and Pullerits T (2002) Fluorescence depolarization dynamics in the B850 complex of purple bacteria.
250 Chem Phys 275: 15–30 Kumble R, Palese S, Visschers RW, Dutton PL and Hochstrasser RM (1996) Ultrafast dynamics within the B820 subunit from the core (LH-1) antenna complex of Rs. rubrum. Chem Phys Lett 261: 396–404 Leupold D, Stiel H, Teuchner K, Novak F, Sandner W, Ucker B and Scheer H (1996) Size enhancement of transition dipoles to one- and two-exciton bands in a photosynthetic antenna. Phys Rev Lett 77: 4675–4678 Ma YZ, Cogdell RJ and Gillbro T (1997) Energy transfer and exciton annihilation in the B800–850 antenna complex of the photosynthetic purple bacterium Rhodopseudomonas acidophila (strain 10050). A femtosecond transient absorption study. J Phys Chem B 101: 1087–1095 Ma YZ, Cogdell RJ and Gillbro T (1998) Femtosecond energytransfer dynamics between bacteriochlorophylls in the B800820 antenna complex of the photosynthetic purple bacterium Rhodopseudomonas acidophila (Strain 7750). J Phys Chem B 102: 881–887 Martin J-L, Breton J, Hoff A, Migus A and Antonetti A (1986) Femtosecond spectroscopy of electron transfer in the reaction center of the photosynthetic bacterium Rhodopseudomonas sphaeroides R-26: Direct electron transfer from the dimeric bacteriochlorophyll primary donor to the bacteriopheophytin acceptor with a time constant of 2.8 ± 0.2 psec. Proc Natl Acad Sci USA 83: 957–961 Matsushita M, Ketelaars M, Van Oijen AM, Köhler J, Aartsma TJ and Schmidt J (2001) Spectroscopy on the B850 band of individual light-harvesting 2 complexes of Rhodopseudomonas acidophila II. Exciton states of an elliptically deformed ring aggregate. Biophys J 80: 1604–1614 McDermott G, Prince SM, Freer AA, Hawthornthwaite-Lawless AM, Papiz MZ, Cogdell RJ and Isaacs NW (1995) Crystal structure of an integral membrane light-harvesting complex from photosynthetic bacteria. Nature 374: 517–521 Meier T, Zhao Y, Chernyak V and Mukamel S (1997a) Polarons, localization, and excitonic coherence in superradiance of biological antenna complexes. J Chem Phys 107: 3876–3893 Meier T, Chernyak V and Mukamel S (1997b) Femtosecond photon echoes in molecular aggregates. J Chem Phys 107: 8759–8774 Meier T, Chernyak V and Mukamel S (1997c) Multiple exciton coherence sizes in photosynthetic antenna complexes viewed by pump-probe spectroscopy. J Phys Chem B 101: 7332–7342 Monshouwer R, De Zarate IO, Van Mourik F and Van Grondelle R (1995) Low-intensity pump-probe spectroscopy on the B800 to B850 transfer in the light harvesting 2 complex of Rhodobacter sphaeroides. Chem Phys Lett 246: 341–346 Monshouwer R, Abrahamsson M, Van Mourik F and Van Grondelle R (1997) Superradiance and exciton delocalization in bacterial photosynthetic light-harvesting systems. J Phys Chem B 101: 7241–7248 Monshouwer R, Baltuška A, Van Mourik F and Van Grondelle R (1998) Time-resolved absorption difference spectroscopy of the LH1 antenna of Rhodopseudomonas viridis. J Phys Chem A 102: 4360–4371 Mostovoy MV and Knoester J (2000) Statistics of optical spectra from single-ring aggregates and its application to LH2. J Phys Chem B 104: 12355–12364 Mukamel S (1995) Principles of Nonlinear Optical Spectroscopy. Oxford University Press, Oxford
Rienk van Grondelle and Vladimir I. Novoderezhkin Mukai K, Abe S and Sumi H (1999) Theory of rapid excitationenergy transfer from B800 to optically-forbidden exciton states of B850 in the antenna system LH2 of photosynthetic purple bacteria. J Phys Chem B 103: 6069–6102 Nagarajan V, Alden RG, Williams JC and Parson WW (1996) Ultrafast exciton relaxation in the B850 antenna complex of Rhodobacter sphaeroides. Proc Natl Acad Sci USA 93: 13774–13779 Nagarajan V, Johnson ET, Williams JC and Parson WW (1999) Femtosecond pump-probe spectroscopy of the B850 antenna complex of Rhodobacter sphaeroides at room temperature. J Phys Chem B 103: 2297–2309 Novoderezhkin VI and Razjivin AP (1993) Excitonic interactions in the light-harvesting antenna of photosynthetic purple bacteria and their influence on picosecond absorbance difference spectra. FEBS Lett 330: 5–7 Novoderezhkin VI and Razjivin AP (1994) Exciton states of the antenna and energy trapping by the reaction center. Photosynthesis Research 42: 9–15 Novoderezhkin VI and Razjivin AP (1995a) Exciton dynamics in circular aggregates: Application to antenna of photosynthetic purple bacteria. Biophys J 68: 1089–1100 Novoderezhkin VI and Razjivin AP (1995b) Excitation delocalization over the whole core antenna of photosynthetic purple bacteria evidenced by non-linear pump-probe spectroscopy. FEBS Lett 368: 370–372 Novoderezhkin VI and Razjivin AP (1996) The theory of Förstertype migration between clusters of strongly interacting molecules: application to light-harvesting complexes of purple bacteria. Chem Phys 211: 203–214 Novoderezhkin V and Van Grondelle R (2002) Exciton-vibrational relaxation and transient absorption dynamics in LH1 of Rhodopseudomonas viridis: A Redfield theory approach. J Phys Chem B 106: 6025–6037 Novoderezhkin V, Monshouwer R and Van Grondelle R (1999a) Disordered exciton model for the core light-harvesting antenna of Rhodopseudomonas viridis. Biophys J 77: 666–681 Novoderezhkin V, Monshouwer R and Van Grondelle R (1999b) Exciton (de)localization in the LH2 antenna of Rhodobacter sphaeroides as revealed by relative difference absorption measurements of the LH2 antenna and the B820 subunit. J Phys Chem B 103: 10540–10548 Novoderezhkin V, Monshouwer R and Van Grondelle R (2000) Electronic and vibrational coherence in the core light-harvesting antenna of Rhodopseudomonas viridis. J Phys Chem B 104: 12056–12071 Novoderezhkin V, Wendling M and Van Grondelle R (2003) Intraand interband transfers in the B800-B850 antenna of Rhodospirillum molischianum: Redfield theory modeling of polarized pump-probe kinetics. J Phys Chem B 107: 11534–11548 Novoderezhkin VI, Rutkauskas D and Van Grondelle R (2006) Dynamics of the emission spectrum of a single LH2 complex: Interplay of slow and fast nuclear motions. Biophys J 90: 2890–2902 Novoderezhkin VI, Rutkauskas D and Van Grondelle R (2007) Multistate conformational model of a single LH2 complex: Quantitative picture of time-dependent spectral fluctuations. Chem Phys 341: 45–56 Nuijs AM, Van Grondelle R, Joppe HLP, Van Bochove AC and Duysens LNM (1985) Singlet and triplet excited carotenoid and antenna bacteriochlorophyll of the photosynthetic purple
Chapter 13
Spectroscopy and Dynamics of Excitation Transfer
bacterium Rhodospirillum rubrum as studied by picosecond absorbance difference spectroscopy. Biochim Biophys Acta 810: 94–105 Otte SCM, Kleinherenbrink FAM and Amesz J (1993) Energy transfer between the reaction center and the antenna in purple bacteria Biochim Biophys Acta 1143: 84–90 Papiz MZ, Prince SM, Howard T, Cogdell RJ and Isaacs NW (2003) The structure and thermal motion of the B800-850 LH2 complex from Rps. acidophila at 2.0 Å resolution and 100 K: New structural features and functionally relevant motions. J Mol Biol 326: 1523–1538 Peterman EJG, Pullerits T, Van Grondelle R and Van Amerongen H (1997) Electron-phonon coupling and vibronic fine structure of light-harvesting complex II of green plants: Temperature dependent absorption and high-resolution fluorescence spectroscopy. J Phys Chem B 101: 4448–4457 Polívka T, Pullerits T, Herek JL and Sundström V (2000) Exciton relaxation and polaron formation in LH2 at low temperature. J Phys Chem B 104: 1088–1096 Pullerits T, Chachisvilis M and Sundström V (1996) Exciton delocalization length in the B850 antenna of Rhodobacter sphaeroides. J Phys Chem 100: 10787–10792 Razjivin AP, Danielius RV, Gadonas RA, Borisov AYu and Piskarskas AS (1982) The study of excitation transfer between light-harvesting antenna and reaction center in chromatophores from purple bacterium Rhodospirillum rubrum by selective picosecond spectroscopy. FEBS Lett 143: 40–44. Reddy NRS, Small GJ, Seibert M and Picorel R (1991) Energytransfer dynamics of the B800-B850 antenna complex of Rhodobacter sphaeroides — a hole burning study. Chem Phys Lett 181: 391–399 Reddy NRS, Picorel R and Small GJ (1992) B896 and B870 components of the Rhodobacter sphaeroides antenna — a hole burning study. J Phys Chem 96: 6458–6464 Reddy NRS, Cogdell RJ, Zhao L and Small GJ (1993) Nonphotochemical hole burning of the B800-B850 antenna complex of Rhodopseudomonas acidophila. Photochem Photobiol 57: 35–39 Redfield AG (1965) The theory of relaxation processes. Adv Mag Res 1: 1–32 Renger Th, May V and Kühn O (2001) Ultrafast excitation energy transfer dynamics in photosynthetic pigment-protein complexes. Phys Rep 343: 137–254 Roszak AW, Howard TD, Southall J, Gardiner AT, Law KJ, Isaacs NW and Cogdell RJ (2003) Crystal structure of the RC-LH1 core complex from Rhodopseudomonas palustris. Science 302: 1969–1972 Rutkauskas D, Novoderezhkin V, Cogdell RJ and Van Grondelle R (2004) Fluorescence spectral fluctuations of single LH2 complexes from Rhodopseudomonas acidophila strain 10050. Biochemistry 43: 4431–4438 Rutkauskas D, Novoderezhkin V, Cogdell RJ and Van Grondelle R (2005) Fluorescence spectroscopy of conformational changes of single LH2 complexes. Biophys J 88: 422–435 Rutkauskas D, Novoderezhkin V, Gall A, Olsen J, Cogdell RJ, Hunter CN, Van Grondelle R (2006) Spectral trends in the fluorescence of single bacterial light-harvesting complexes: experiments and modified Redfield simulations. Biophys J 90: 2475–2485 Salverda JM, Van Mourik F, Van der Zwan G, and Van Grondelle
251
R (2000) Energy transfer in the B800 rings of the peripheral bacterial light-harvesting complexes of Rhodopseudomonas acidophila and Rhodospirillum molischianum studied with photon echo techniques. J Phys Chem B 104: 11395–11408 Sauer K, Cogdell RJ, Prince SM, Freer A, Isaacs NW and Scheer H (1996) Structure-based calculations of the optical spectra of the LH2 bacteriochlorophyll-protein complex from Rhodopseudomonas acidophila. Photochem Photobiol 64: 564–576 Savikhin S and Struve WS (1996) Temperature dependence of electronic energy transfer within B850 antennae of the NF57 mutant of purple bacterium Rhodobacter sphaeroides. Chem Phys 210: 91–100 Sebban P, Jolchine G and Moya I (1984) Spectra of fluorescence lifetime and intensity of Rhodopseudomonas sphaeroides at room and low temperature. Comparison between the wild type, the C71 reaction center-less mutant and the B800–850 pigmentprotein complex. Photochem Photobiol 39: 247–253 Scherz A and Parson WW (1984) Exciton interactions in dimers of bacteriochlorophyll and related molecules. Biochim Biophys Acta 766: 666–678 Scheuring S, Seguin J, Marco S, Levy D, Robert B and Rigaud JL (2003) Nanodissection and high-resolution imaging of the Rhodopseudomonas viridis photosynthetic core complex in native membranes by AFM. Proc Natl Acad Sci USA 100: 1690–1693 Scholes GD and Fleming GR (2000) On the mechanism of light harvesting in purple bacteria: B800 to B850 energy transfer. J Phys Chem B 104: 1854–1868 Scholes GD, Gould IR, Cogdell RJ and Fleming GR (1999) Ab initio molecular orbital calculations of electronic couplings in the LH2 bacterial light-harvesting complex of Rps. acidophila. J Phys Chem B 103: 2543–2553 Somsen OJG, Van Mourik F, Van Grondelle R and Valkunas L (1994) Energy migration and trapping in a spectrally and spatially inhomogeneous light-harvesting antenna. Biophys J 66: 1580–1596 Somsen OJG, Van Grondelle R and Valkunas L (1996) A perturbed two-level model for exciton trapping in small photosynthetic systems. Biophys J 70: 669–683 Sumi H (1999) Theory on rates of excitation-energy transfer between molecular aggregates through distributed transition dipoles with application to the antenna system in bacterial photosynthesis. J Phys Chem B 103: 252–260 Sundström V, Van Grondelle R, Bergstrem H, Akesson E and Gilbro T (1986) Excitation energy transport in the BChl antenna systems of Rhodospirillum rubrum and Rhodobacter sphaeroides studied by low-intensity picosecond spectroscopy. Biochim Biophys Acta 851: 431–446 Sundström V, Pullerits T and Van Grondelle R (1999) Photosynthetic light-harvesting: Reconciling dynamics and structure of purple bacterial LH2 reveals function of photosynthetic unit. J Phys Chem B 103: 2327–2346 Tietz C, Chekhlov O, Dräbenstedt A, Schuster J and Wrachtrup J (1999) Spectroscopy on single light-harvesting complexes at low temperature. J Phys Chem B 103: 6328–6333 Timpmann KN, Zhang FG, Freiberg AM and Sundström V (1993) Detrapping of excitation energy from the reaction centre in the photosynthetic bacterium Rhodospirillum rubrum. Biochim Biophys Acta 1183: 185–193 Timpmann K, Rätsep M, Hunter CN and Freiberg A (2004)
252 Emitting excitonic polaron states in core LH1 and peripheral LH2 bacterial light-harvesting complexes. J Phys Chem B 108: 10581–10588 Tretiak S, Middleton C, Chernyak V and Mukamel S (2000a) Exciton Hamiltonian for the bacteriochlorophyll system in the LH2 antenna complex of purple bacteria. J Phys Chem B 104: 4519–4528 Tretiak S, Middleton C, Chernyak V and Mukamel S (2000b) Bacteriochlorophyll and carotenoid excitonic couplings in the LH2 system of purple bacteria. J Phys Chem B 104: 9540–9553 Valkunas L, Janušonis J, Rutkauskas D and Van Grondelle R (2007) Protein dynamics revealed in the excitonic spectra of single LH2 complexes. J Lumines 127: 269–275 Van Amerongen H, Valkunas L and Van Grondelle R (2000) Photosynthetic Excitons. World Scientific Publishers, Singapore Van der Laan H, Schmidt Th, Visschers RW, Visscher KJ, Van Grondelle R and Völker S (1990) Energy transfer in the B800–850 antenna complex of purple bacteria Rhodobacter sphaeroides: A study by spectral hole-burning. Chem Phys Lett 170: 231–238 Van Grondelle R and Novoderezhkin V (2001) Dynamics of excitation energy transfer in the LH1 and LH2 light-harvesting complexes of photosynthetic bacteria. Biochemistry 40: 15057–15068 Van Grondelle R and Novoderezhkin V (2006) Energy transfer in photosynthesis: Experimental insights and quantitative models. Phys Chem Chem Phys 8: 793–807 Van Grondelle R, Dekker JP, Gillbro T and Sundström V (1994) Energy transfer and trapping in photosynthesis. Biochim Biophys Acta 1187: 1–65 Van Mourik F, Visschers RW and Van Grondelle R (1992) Energy transfer and aggregate size effects in the inhomogeneously broadened core light-harvesting complex of Rhodobacter sphaeroides. Chem Phys Lett 193: 1–7 Van Oijen AM, Ketelaars M, Köhler J, Aartsma TJ and Schmidt J (1998) Spectroscopy of single light-harvesting complexes from purple photosynthetic bacteria at 1.2 K. J Phys Chem B 102: 9363–9366 Van Oijen AM, Ketelaars M, Köhler J, Aartsma TJ and Schmidt J (1999a) Unraveling the electronic structure of individual photosynthetic pigment-protein complexes. Science 285: 400–402 Van Oijen AM, Ketelaars M, Köhler J, Aartsma TJ and Schmidt J (1999b) Spectroscopy of individual LH2 complexes of Rhodopseudomonas acidophila: Localized excitations in the B800 band. Chem Phys 247: 53–60 Visscher KJ, Bergström H, Sundström V, Hunter CN and Van Grondelle R (1989) Temperature dependence of energy transfer from the long wavelength antenna BChl-896 to the reaction center in Rhodospirillum rubrum, Rhodobacter sphaeroides (w.t. and M21 mutant) from 77 to 177 K, studied by picosecond absorption spectroscopy. Photosynth Res 22: 211–217
Rienk van Grondelle and Vladimir I. Novoderezhkin Visschers RW, Germeroth L, Michel H, Monshouwer R and Van Grondelle R (1995) Spectroscopic properties of the lightharvesting complexes from Rhodospirillum molischianum. Biochim Biophys Acta 1230: 147–154 Visser HM, Somsen OJG, Van Mourik F, Lin S, Van Stokkum IHM and Van Grondelle R (1995) Direct observation of sub-picosecond equilibration of excitation energy in the light-harvesting antenna of Rhodospirillum rubrum. Biophys J 69: 1083–1099 Visser HM, Somsen OJG, Van Mourik F, Lin S, Van Stokkum IHM and Van Grondelle R (1996) Excited-state energy equilibration via subpicosecond energy transfer within the inhomogeneously broadened light-harvesting antenna of LH-1-only Rhodobacter sphaeroides mutants M2192 at room temperature and 4.2 K. J Phys Chem 100: 18859–18867 Vulto SIE, Kennis JTM, Streltsov AM, Amesz J and Aartsma TJ (1999) Energy relaxation within the B850 absorption band of the isolated light-harvesting complex LH2 from Rhodopseudomonas acidophila at low temperature. J Phys Chem B 103: 878–883 Wendling M, Van Mourik F, Van Stokkum IHM, Salverda JM, Michel H and Van Grondelle R (2003) Low-intensity pumpprobe measurements on the B800 band of Rhodospirillum molischianum. Biophys J 84: 440–449 Wu H-M, Savikhin S, Reddy NRS, Jankowiak R, Cogdell RJ and Small GJ (1996) Femtosecond and hole-burning studies of B800’s excitation energy relaxation dynamics in the LH2 antenna complex of Rhodopseudomonas acidophila (strain 10050). J Phys Chem 100: 12022–12033 Wu H-M, Reddy NRS and Small GJ (1997a) Direct observation and hole burning of the lowest exciton level (B870) of the LH2 antenna complex of Rhodopseudomonas acidophila (Strain 10050). J Phys Chem B 101: 651–656 Wu H-M, Ratsep M, Lee I-J, Cogdell RJ and Small GJ (1997b) Exciton level structure and energy disorder of the B850 ring of the LH2 antenna complex. J Phys Chem B 101: 7654–7663 Xiao W, Lin S, Taguchi AKW and Woodbury NW (1994) Femtosecond pump-probe analysis of energy and electron transfer in photosynthetic membranes of Rhodobacter capsulatus. Biochemistry 33: 8313–8322 Yang M and Fleming GR (2002) Influence of phonons on exciton transfer dynamics: Comparison of the Redfield, Förster, and modified Redfield equations. Chem Phys 275: 355–372 Zhang WM, Meier T, Chernyak V and Mukamel S (1998) Exciton-migration and three-pulse femtosecond optical spectroscopies of photosynthetic antenna complexes. J Chem Phys 108: 7763–7774 Zigmantas D, Read EL, Mančal T, Brixner T, Gardiner AT, Cogdell RJ and Fleming GR (2006) Two-dimensional electronic spectroscopy of the B800-820 light-harvesting complex. Proc Natl Acad Sci USA 103: 12672–12677
Chapter 14 Organization and Assembly of Light-Harvesting Complexes in the Purple Bacterial Membrane James N. Sturgis Laboratoire d’Ingénierie des Systèmes Macromoléculaires, UPR 9027, Institute de Biologie Structurale et Microbiologie, CNRS, 31 Chemin Joseph Aiguier, 13402 Marseille, France
Robert A. Niederman* Department of Molecular Biology and Biochemistry, Rutgers University, 604 Allison Road, Nelson Biological Laboratories, Piscataway, NJ 08854-8082, U.S.A.
Summary .............................................................................................................................................................. 254 I. Introduction..................................................................................................................................................... 254 II. Themes and Variations — Structural Variability of Complexes in Native Membranes ................................... 255 A. Size Heterogeneity of the Light-Harvesting 2 Complex.................................................................... 255 B. Structural Variability in Reaction Center-Light-Harvesting 1 Core Complexes ................................ 255 III. Principles of Photosynthetic Unit Organization .............................................................................................. 257 A. Models for Organization Based Upon Spectroscopic Probes .......................................................... 257 B. Multiple Patterns of Organization as Revealed by Atomic Force Microscopy of Intact Intracytoplasmic Membranes ........................................................................................................... 258 1. Blastochloris viridis and Rhodospirillum rubrum ................................................................... 258 2. Rhodospirillum photometricum and Phaeospirillum molischianum ....................................... 259 3. Rhodobacter sphaeroides ..................................................................................................... 260 4. Rhodobacter blasticus ........................................................................................................... 261 5. Rhodopseudomonas palustris ............................................................................................... 261 6. Common Points Revealed by Atomic Force Microscopy Studies ........................................... 262 IV. Proposals for the Functional Organization of Photosynthetic Units ............................................................... 263 A. Supercomplexes............................................................................................................................... 263 B. Quinone Pools .................................................................................................................................. 264 C. Functional Importance of the Observed Organization...................................................................... 264 V. In Vivo Assembly of Light-Harvesting Complexes.......................................................................................... 267 A. Complex-Specific Assembly Factors ................................................................................................ 268 B. General Membrane Assembly Factors ............................................................................................. 268 VI. Perspectives for the Next Ten Years.............................................................................................................. 269 References ........................................................................................................................................................... 270
*Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 253–273. © 2009 Springer Science + Business Media B.V.
254
James N. Sturgis and Robert A. Niederman
Summary Recent in situ surface images of the intracytoplasmic membrane (ICM) of purple photosynthetic bacteria obtained at submolecular resolution by atomic force microscopy (AFM), have revealed multiple, species-dependent patterns of supramolecular organizations for their light-harvesting (LH) antennas, suggested earlier from spectroscopic studies. These have varied from highly ordered linear arrays of dimeric reaction center-lightharvesting 1-PufX (RC-LH1-PufX) core complexes in Rhodobacter (Rba.) sphaeroides, with the peripheral LH2 antenna either interspersed between them or arranged in larger clusters, to a less orderly arrangement found in several other purple bacteria, in which randomly organized, monomeric RC-LH1-enriched domains co-exist with large, paracrystalline hexagonally packed LH2 domains. In addition, regions with apparently crystalline core complexes were observed in Rhodopseudomonas (Rps.) palustris, along with a protein-free lipid bilayer. Likely bases for the absence of the ATP synthase and cytochrome bc1 complex from the AFM topographs are discussed, including the possibility that they localize at the poles of ICM vesicles either out of view of flat regions imaged by AFM, or are removed during membrane preparation. We also discuss how the observed arrangements of LH2 and core complexes may specifically control quinol escape from the RC, emphasizing the importance of short-range diffusion within the disordered regions of the membrane in promoting the passage of quinone, and how this process may be augmented by quinone exclusion from large, ordered fields of the LH2 antenna. Possible forces that drive in vitro autoassembly of LH complexes are assessed along with the roles that complex-specific and general membrane assembly factors may play in driving the assembly and organization of photosynthetic units within the cell. Finally, we address likely perspectives for further studies on the organization and assembly of bacterial antenna complexes over the next decade. I. Introduction It was just over ten years ago that the last volume on the anoxygenic photosynthetic bacteria appeared (Blankenship et al., 1995), and since that time, an explosive growth in research on the purple bacterial LH complexes has occurred. Much of this progress stems directly from the X-ray structure of the peripheral LH2 complex of Rps. acidophila by McDermott et al. (1995), reported during the same year. As the first atomic-resolution structure to emerge for any integral membrane light-harvesting complex, this represented a major landmark in photosynthesis research. In addition, the cryo-electron microscopy studies of Karrasch et al. (1995), which appeared earlier that year, gave rise to an 8.5 Å projection map of the reconstituted LH1 core complex of Rhodospirillum (Rsp.) rubrum. As a consequence of these seminal contributions, Abbreviations: 2-D – 2-dimensional; AFM – atomic force microscopy; bc1 complex – ubiquinol-cytochrome c2 oxidoreductase; BChl – bacteriochlorophyll a; Blc. – Blastochloris; BP – bacteriophytochrome; CM – cytoplasmic membrane; E. – Eschericha; ICM – intracytoplasmic membrane; LD – linear dichroism; LH – light-harvesting; LH1 – core light-harvesting complex; LH2 – peripheral light-harvesting complex; Phs. – Phaeospirillum; PS II – Photosystem II; PSU – photosynthetic unit consisting of LH1-reaction center core structures with LH2 arranged at their peripheries; Rba. – Rhodobacter; RC – reaction center; Rps. – Rhodopseudomonas; Rsp. – Rhodospirillum
extensive efforts have focused on such aspects as the revisiting of LH2 and LH1 structure, the dynamics of intra- and inter-complex excitation energy transfer, structure-function relationships as assessed by spectroscopic and reconstitution techniques, and the modeling of photosynthetic unit (PSU) behavior (see Chapter 8, Gabrielsen et al.; Chapter 9, Bullough et al.; Chapter 10, Loach and Parkes-Loach; Chapter 11, Robert; Chapter 12, Frank and Polívka; Chapter 13, van Grondelle and Novoderezhkin; Chapter 15, Şener and Schulten). Such pioneering structural studies also laid the groundwork for the elucidation of the supramolecular organization of light-harvesting units in the native intracytoplasmic membrane (ICM), which is the subject of this chapter. Accordingly, we will consider the structural variability of the LH2 and LH1 complexes in native photosynthetic membranes, and the multiple patterns of organization of the PSU, consisting of LH2 and RC-LH1 complexes, that have been suggested from spectroscopic studies and visualized more recently by AFM. The basic physical principles that ultimately govern the observed organizations of these various pigment-protein complexes will be discussed, along with considerations on how they accommodate exciton trapping and excitation energy transfer, as well as the flow of reducing equivalents to and from the RC and to other components that catalyze light-driven
Chapter 14
In Situ Organization of Light-Harvesting Complexes
cyclic flow. In particular, we will discuss how the observed PSU arrangements may specifically control quinone diffusion and escape from the RC. Additional topics that will be addressed include the forces that drive the in vitro autoassembly of LH complexes, and how the PSU is assembled and organized in the cell. We conclude with perspectives for research on LH complexes over the next 10 years. II. Themes and Variations — Structural Variability of Complexes in Native Membranes The availability of crystallographic structures for the various LH components of the photosynthetic apparatus initially suggested that these integral membrane proteins exist in a fixed stoichiometry of heterodimeric subunits that do not undergo structural variability. However, it is becoming increasingly clear that considerable structural heterogeneity exists in the bacterial LH apparatus. This heterogeneity operates at various levels; thus we have now observed heterogeneity between species, between distinct products of gene families in the same species, and between individual complexes. The realization of the extent of this diversity is relatively new and covers both the peripheral LH2 complexes and the core complexes. A. Size Heterogeneity of the LightHarvesting 2 Complex The high-resolution structure of the LH2 antenna complex of Rps. acidophila, as resolved by X-ray crystallography, showed that it exists in a cyclic nonameric arrangement (McDermott et al., 1995), while a subsequent X-ray crystallographic analysis revealed that LH2 in Phaeospirillum (Phs.) molischianum (Koepke et al., 1996) exists as a cyclic octamer. Since these initial studies, many different peripheral antenna complexes have been studied in 2- and 3-D crystals and found to form either nonameric or octameric rings (Chapter 8, Gabrielsen et al.). There is thus a large body of evidence to suggest that while most LH2 complexes form nonameric rings, certain species are able to form smaller octameric complexes. Several studies have revealed a variability in the size of the different LH2 complexes produced by a single species. Thus, in Rps. palustris, where the peripheral antenna complexes that are synthesized depend on growth conditions, complexes isolated
255
from cells grown at very low light intensity were able to form crystals of an octameric complex designated as high-800 type LH2 (Hartigan et al., 2002), while recent AFM images of membranes from cells grown at higher light intensities showed predominantly nonameric LH2 (B800-850) complexes (Scheuring et al., 2006). It was thus a surprise to observe a variety of antenna ring sizes in an AFM analysis of native membranes of Rsp. photometricum (Fig. 1)(Scheuring et al., 2004a). The majority of LH2 antenna rings in the membrane were observed to be nonameric, with a diameter of 5.2 nm; however, considerable heterogeneity was shown to exist both in complex size (diameter) which varied between 3.8 nm and 8.8 nm, and in the number of subunits which appeared to vary from 7 to 14. While it was not possible to observe individual subunits in the rare complexes of most extreme size, individual subunits were observed in complexes made up of 8, 9 and 10 αβ pairs. As would be expected, this size heterogeneity appeared to be associated with spectral heterogeneity (Fig. 1), as first described for LH2 by Westerhuis et al. (1992), in which the absorption spectra of the smaller complexes are gradually blue shifted relative to those of the larger complexes. Thus, the above experimental results would seem to indicate that in some photosynthetic bacteria, the peripheral antenna forms rings of somewhat heterogeneous size, ranging from small rings of seven or even fewer αβ pairs up to large rings possibly containing a dozen or more αβ pairs. As also noted above, these rings appear to have slightly different spectral characteristics, as would be expected from the modified interchromophore coupling. Recently, free energy calculations (Janosi et al., 2006) suggest that although the fixed number of oligomers observed in the crystal structure are close to the lowest energy structures, the angular potential is relatively soft and thus little energy is required to modify the oligomeric size. These theoretical calculations would seem entirely consistent with the observed heterogeneity. B. Structural Variability in Reaction CenterLight-Harvesting 1 Core Complexes Over the past decade, as structural information began to emerge, it became apparent that considerable variability exists in the structures of RC-LH1 core complexes. The 8.5 Å projection map obtained from two-dimensional (2-D) crystals of the LH1 complex
256
James N. Sturgis and Robert A. Niederman
Fig. 1. Structural variability of LH2 complexes in the membrane. (a, b) high resolution AFM images of Rsp. photometricum photosynthetic membranes, showing large as indicated by arrow, (a) and mostly small (b) LH2 complexes in situ (scale bar 5 nm). (c) Class averages of LH2 complexes with 8, 9 and 10 heterodimeric αβ subunits. (d) Spectral and size heterogeneity of complexes solubilized from membranes by β-dodecylmaltoside and separated by gel filtration. The chromatogram traces, left hand scale, show that rechromatography of different fractions from the upper trace confirms a size difference, lower traces. The LH2 complexes also change the position of their main absorption band (right hand scale, and near straight line), as a function of their position in the chromatograph. See Scheuring et al. (2004a) for further details.
of Rsp. rubrum by Karrasch et al. (1995) revealed a closed ring of 16 apparent αβ heterodimers, sufficiently large to enclose the RC (Chapter 9, Bullough et al.). This was subsequently confirmed when these cryo-EM and image processing procedures were applied to the Rsp. rubrum RC-LH1 core complex (Jamieson et al., 2002). In contrast to this monomeric core structure, an EM analysis of negatively-stained tubular membranes from an LH2- strain of Rhodobacter (Rba.) sphaeroides by Jungas et al. (1999) suggested that this core complex is dimeric and consists of an open S-shaped assembly in which two RCs are surrounded by two C-shaped arcs of LH1 lacking several of the heterodimeric building blocks of the 16-membered ring structure. This basic proposal has been confirmed by the recent 8.5 Å projection structure, derived from 2-D crystals of the core RCLH1-PufX dimer of Rba. sphaeroides (Qian et al., 2005), in which 28 LH1 αβ heterodimers form an ellipsoidal, double lobed structure (14 αβ heterodimers per RC-LH1 monomer) that encloses the α-helices
of two RCs. The QB site of the RC is adjacent to the putative position of the PufX polypeptide, near the openings in each of the LH1 helical arrays. PufX, encoded by the pufX gene within the pufQBALMX operon (also encoding the β- and α-subunits of LH1 and the L- and M-subunits of the RC) has been established as an RC-LH1 core organizing protein (Frese et al., 2000, 2004), thought to facilitate ubiquinoneubiquinol exchange (Barz et al., 1995) through the gates in the dimeric core structure. A discussion of how these quinone redox species may be exchanged across the seemingly impenetrable barrier of α-helices, confronted in the closed LH1 ring-like structures, is presented below. Recent AFM topographs of photosynthetic membranes from a variety of purple bacteria (see Section III.B below for details) have shown that a dimeric RC-LH1 core structure also exists in Rba. blasticus (Scheuring et al., 2005a), while in Blastochloris (Blc.) viridis (Scheuring et al., 2003), Rsp. photometricum (Scheuring et al., 2004b), and Phs. molischianum
Chapter 14
In Situ Organization of Light-Harvesting Complexes
(Gonçalves et al., 2005), monomeric, closed elliptical assemblies of 16 LH1 αβ-heterodimers have been observed. In addition, X-ray crystallography has revealed that the Rps. palustris RC-LH1 core complex crystallizes as a monomeric, elliptical assembly, modeled with 15 LH1 transmembrane αβ-heterodimers, interrupted by the α-helix of polypeptide W, an apparent PufX homolog (Roszak et al., 2003). This monomeric core complex structure has been confirmed recently in an AFM study of photosynthetic membranes from Rps. palustris (Scheuring et al., 2006). III. Principles of Photosynthetic Unit Organization The organization of the photosynthetic apparatus in the membrane is governed by two different kinds of constraints. On the one hand, for efficient light harvesting, the RC must be functionally connected to both the LH1 and LH2 complexes, and they must be as close as possible to the RCs to ensure efficient light collection. On the other hand, the electron transfer chain requires an efficient quinol transfer from the RC to the cytochrome bc1 complex, capable of regenerating the quinone necessary for continued photosynthesis. These two different constraints both depend on migration within the plane of the photosynthetic membrane, either of excitation energy between LH complexes or of quinones/quinols between the RC and cytochrome bc1 complex. Various different solutions to these competing constraints, including the clustering of RCs, seem to have been adopted by different species. A. Models for Organization Based Upon Spectroscopic Probes Although in past years, the architecture of the ICM, as well as the localization of photosynthetic complexes, was extensively investigated by freeze-fracture electron microscopy (reviewed by Drews and Golecki, 1995), the different particles appearing in membrane fracture planes could not always be identified with certainty. In contrast, a variety of spectroscopic probes has provided sensitive techniques for assessing the distribution of functionally identified components of the PSU, from which it is possible to deduce their supramolecular organization. The reader is referred to the article by Parson and Nagarajan (2003) for a
257
more thorough description of the bases of these different techniques. Among the earliest application of laser-based probes to the elucidation of the structural organization of the photosynthetic apparatus was the examination of singlet-triplet quenching efficiencies in various strains of Rba. sphaeroides by Monger and Parson (1977). The results suggested that the LH1 complex is arranged in clusters that surround and interconnect several RCs, with LH2 positioned peripherally in large ‘lakes.’ Subsequently, singlet-singlet annihilation measurements have provided a procedure for quantitative refinement of this proposal, as well as estimates of the size and organization of antenna domains over which excitations can migrate in ICM vesicles (chromatophores) from a variety of photosynthetic bacteria (Bakker et al., 1983; Hunter et al., 1985; Vos et al., 1986; Deinum et al., 1989, 1991; Kramer et al., 1995). The results obtained at 4 K for Rba. sphaeroides by Vos et al. (1986) were explained by an arrangement in which ~4 RCs are embedded in core assemblies of ~100 LH1 BChl molecules, with LH2 BChls interspersed between them, similar to the arrangement now visualized directly by AFM in Rba. sphaeroides membranes (see Section III.B.3 below). At room temperature, thermal separation of the core clusters no longer exists, and many RCs become connected as a result of energy equilibration among >1000 BChls, again in accord with AFM images. It is also noteworthy that the rows of dimeric RC-LH1 complexes observed in AFM represent domains of LH1 BChls of a size similar to those calculated from singlet-singlet annihilation measurements (Vos et al., 1988) on membranes from a mutant lacking the LH2 complex, while the number of LH2 BChl molecules that were visualized between these rows of core complexes was also of the same order of magnitude as the connected LH2 BChl molecules deduced in excitation annihilation studies of LH2-only membranes. More recently, singlet-singlet annihilation measurements have been applied to the determination of the number of BChls remaining connected after dilution of membrane bilayer phospholipid by fusion of Rba. sphaeroides chromatophores with unilamellar liposomes (Westerhuis et al., 1998). Core assemblies, containing ~50 LH1 BChls, were dissociated into units of about half their original size, while LH2 clusters were apparently detached as single LH2 rings. These results suggest that efficient energy transfer in the membrane ultimately relies upon dense packing of the complexes, rather than on strong, specific
258 intermolecular associations, shown by these studies to be insufficiently robust to withstand bilayer dilution (discussed further in Section IV.C). Linear dichroism (LD) provides a method for direct determination of electronic transition-dipole moment orientations of selected chromophores, relative to the orientation axis of the interrogated sample. LD has thus proved to be a valuable spectroscopic probe for assessing the long-range pattern of structural organization existing within the ICM bilayer (Frese et al., 2000, 2004). Interestingly, this technique has been successfully employed to demonstrate that in Rba. sphaeroides, the PufX protein, known to be essential for dimerization of RC-LH1 core particles, and the exchange of ubiquinone redox species at the RC QB site, also induces a non-random orientation of the RC within the LH1 ring, and the alignment of all RC-LH1 cores in regular arrays. This effect was first demonstrated in membrane tubules from an RCLH1 core only strain (Frese et al., 2000), but has been extended more recently to wild-type membranes in which core arrays coexist with large domains of LH2 in the native ICM (Frese et al., 2004). B. Multiple Patterns of Organization as Revealed by Atomic Force Microscopy of Intact Intracytoplasmic Membranes In recent years, AFM has evolved into a powerful tool for obtaining surface images of membrane proteins at subnanometer resolution. Recent applications of this highly revealing structural probe to the native ICM of purple photosynthetic bacteria have provided the first in situ surface views of the supramolecular organization of any multicomponent biological membrane (reviewed by Scheuring et al., 2005b; Chapter 47, Scheuring). The earliest AFM topographs of bacterial LH complexes were obtained with the LH2 protein of Rubrivivax gelatinosus reconstituted into lipid bilayers as densely packed 2-D crystals (Scheuring et al., 2001). Although this permitted the identification of a nonameric LH2 ring structure along with membrane protruding domains, as well as the first subatomic glimpse of the LH1 rings which were present as a contaminant, recent direct AFM images of a variety of bacterial photosynthetic complexes in their native ICM environments (Fig. 2), have provided valuable structural information on their in situ oligomerization states and the intactness and flexibility of their ring-like structures, without the
James N. Sturgis and Robert A. Niederman need to resort to complicated purification, isolation and reconstitution procedures and the use of highly ordered 2-D crystals. 1. Blastochloris viridis and Rhodospirillum rubrum Blc. viridis and Rsp. rubrum represent purple photosynthetic bacteria with simplified PSUs that lack the LH2 complex. After membrane disruption, only relatively small chromatophores can be isolated from Blc. viridis, which necessitated their fusion by freezethaw cycles in order to obtain membrane preparations suitable for AFM imaging (Scheuring et al., 2003). Single RCs surrounded by a closed ellipsoid of 16 LH1 αβ-heterodimers were observed (Fig. 2a), from which the tetraheme cytochrome subunit could be nanodissected by applying a loading force to the AFM tip. This revealed that the underlying RC-L and -M subunits adopt a preferred asymmetric topography, arranged with a fixed distance distribution from the short ellipsis axis of the LH1 complex, which may represent an important constraint for energy transfer. Upon removal of the RC, the LH1 complex rearranges into a circular structure, and it was suggested that this flexible behavior could promote breathing motions that would serve to facilitate quinone/quinol passage through the normally closed LH1 structure (see Section IV.C). Although an AFM analysis of the RC-LH1 complex of Rsp. rubrum has thus far been confined to 2-D crystal of the purified core complex reconstituted into lipid bilayers, a closed ring of 16 αβ-heterodimers positioned around a central RC density has been consistently observed (Jamieson et al., 2002; Fotiadis et al., 2004). The latter study demonstrated that depending on packing constraints within the crystals, the LH1 ring is able assume a wide variety of architectures; however, their physiological importance awaits their direct AFM analysis in the Rsp. rubrum ICM. Such an in situ AFM analysis would also be of importance for understanding the arrangement of connected LH1 BChl domains in Rsp. rubrum as deduced from the singlet-singlet annihilation measurement of Vos et al. (1986). At 4 K, they were estimated to comprise ~150 BChls connecting 4 RCs, while at room temperature, an energy barrier is overcome, and the number of connected BChls was near 1,000, approaching the total number in a single chromatophore.
Chapter 14
In Situ Organization of Light-Harvesting Complexes
259
Fig. 2. Membrane organization observed in several species of purple photosynthetic bacteria by high-resolution AFM, as discussed in the text. a Blastochloris viridis, b Rhodospirillum photometricum, c Rhodobacter sphaeroides, d Rhodobacter blasticus, e and f Rhodopseudomonas palustris grown under low (e) and high (f) light intensities. In the different panels the scale bar represents 10 nm. In each panel typical monomeric (M) or dimeric (D) core complexes are indicated, and the smaller LH2 rings (2) in panels b – e. Note that in panel f, the RC-H subunit hides the details of the core complex structure.
2. Rhodospirillum photometricum and Phaeospirillum molischianum Imaging of complex membrane architectures at submolecular resolution using AFM has recently allowed the description of several new types of PSU organizations, in which the protein mixture is more complex and the organization less regular than in Blc. viridis. Among the first native multicomponent photosynthetic systems to be observed by AFM was that of Rsp. photometricum (Scheuring et al., 2004b), which unlike the closely related bacterium Rsp. rubrum, forms PSUs with an LH2 complex. Although the photosynthetic apparatus of Rsp. photometricum is localized in stacked, disk-like vesicular ICMs, sufficiently large for AFM analysis, in this study, they gave rise to chromatophore vesicles that did not collapse onto mica supports, requiring that the weakly attached upper membrane layer be removed by nanodissection with the AFM tip. This exposes the periplasmic ICM surface for direct AFM imaging. The initial observations made in this manner, as well as several subsequent ones made directly on native
membranes (Scheuring et al., 2004a; Scheuring and Sturgis 2005), showed an intriguing in situ organization of the different photosynthetic complexes (Fig. 2b). First, the main complexes observed were a monomeric oval core complex and a circular nonameric LH2 antenna complex, with no clear images of the cytochrome bc1 complex, despite the demonstration that this protein complex was present in the membrane preparations. Second, the proteins were very densely packed with an overall surface occupancy approaching 80% (Scheuring and Sturgis, 2006), posing serious questions about how such membranes might function. Third, and most unexpectedly, the proteins were not organized homogeneously, but rather they appeared to be distributed between RC-LH1-rich domains and nearly crystalline, hexagonally packed LH2 domains, again posing functional questions. The RC-LH1-rich and randomly organized domains were found to contain a constant proportion of about 3.5 LH2 complexes per core complex. A very similar organization was observed more recently in the closely related and more completely characterized bacterium, Phs. molischianum (Gonçalves et al., 2005). In particular, the membrane was
260 observed to contain a similar separation into two structurally different types of domains; paracrystalline, hexagonally packed LH2 domains, this time containing the octameric Phs. molischianum LH2, and mixed domains containing core complexes and a relatively small amount of LH2 (as in Rsp. photometricum, about 3.5 LH2 rings per core complex). Using the observed AFM images together with the known LH2 structure (Koepke et al., 1996), it was possible to precisely calculate the pigment distances between the neighboring complexes. The modulation of this organization in response to varying composition was examined in Rsp. photometricum (Scheuring and Sturgis, 2005); these observations have been extended more recently to Phs. molischianum (Gonçalves et al., 2005). In common with many photosynthetic bacteria, these species show a chromatic adaptation, varying their pigment composition in response to the environmental conditions (Cohen-Bazire et al., 1956). Surprisingly, modification of the membrane composition by addition of large quantities of LH2 does not simply dilute the core complexes, forcing them further apart. Instead the additional LH2 is added exclusively to the paracrystalline LH2 domains maintaining the original core connectivity with the disordered domains of mixed composition (Scheuring and Sturgis, 2005). This eutectic-type phase organization of the membrane is suggested to arise from the hierarchy of interactions between the different components and these essential features could be mimicked with a simple model. It has been suggested that this eutectic phase behavior is important for maintaining core complex-cytochrome bc1 connectivity (see below and Scheuring and Sturgis, 2006). 3. Rhodobacter sphaeroides Rba. sphaeroides is among the best characterized species of purple photosynthetic bacteria and has provided a unique combination of accessible molecular genetics with an ICM that is amenable to an unparalleled variety of biochemical, spectroscopic and ultrastructural probes. It was thus of considerable importance to study the native surface architecture of the otherwise well characterized Rba. sphaeroides ICM. In order to produce membrane fragments for AFM that adhered as a single layer to mica supports, it was necessary to pretreat isolated chromatophore vesicles with subsolubilizing amounts of the detergent
James N. Sturgis and Robert A. Niederman β-dodecyl maltoside. The resulting AFM topographs of these ICM fragments (Bahatyrova et al., 2004b) revealed a much more organized arrangement for the LH and RC components than that observed in other LH2-containing species (Fig. 2c). In these images, the central protruding feature within the RC-LH1 core complexes corresponded to the RC-H subunit, indicating that the cytoplasmic face of the ICM was uniformly exposed. In addition to the distinctive organization of ICM domains, crucial new features relating to the relative positions and functional associations of LH2, together with the core RC-LH1 complex, were also observed. Two types of organization were exhibited by the LH2 complex; in the first type, 10–20 molecules of LH2 formed light capture domains that interconnect linear arrays of mainly dimeric RC-LH1-PufX core complexes. In the second, LH2 existed as larger clusters in separate domains; these apparently represent the light-responsive LH2 complement as observed for Rsp. photometricum (Scheuring and Sturgis, 2005), which forms during chromatic adaptation to lowered light intensity. The close packing also observed in Rba. sphaeroides for the linear RC-LH1-PufX arrays, often separated by only narrow LH2 energy channels, demonstrates an elegant economy in photosynthetic membrane arrangement. The physical continuity between individual LH2 complexes and between LH2 and RC-LH1-PufX core dimers, assures that excitations are transferred from LH2 to LH1, and subsequently to the RC, to initiate the primary photochemical events. Moreover, the LH1 complex is ideally positioned to serve as a collection hub for excitations obtained from LH2, which are then rapidly transferred to the RC. The LH2 complexes situated between rows of RC-LH1-PufX dimers form a relatively invariant complement of pigment-protein complexes that optimally fulfill the basic requirement for efficient trapping and transmission of light energy (see also Chapter 15, Şener and Schulten). The RC-LH1-PufX core complexes are usually found in dimeric associations, arranged in rows of up to six dimers, such that an LH1 excitation can migrate along a series of dimers until an ‘open’ RC is found. These AFM images also explain why mutant cells lacking LH2 form elongated tubular membranes, which in freeze-fracture replicas showed a striking pattern of closely packed, highly ordered rows of dimeric particles arrayed helically along their entire length (Westerhuis et al., 2002); these particles were identified as open S-shaped assemblies of dimeric
Chapter 14
In Situ Organization of Light-Harvesting Complexes
261
RC-LH1-PufX complexes in an image analysis of isolated tubules after negative staining (Siebert et al., 2004). Because of the tilt of the RC-LH1-PufX complexes, it was not possible to visualize the PufX polypeptide in the AFM analysis of native ICM fragments, or to assess whether the LH1 ring is complete; however, as discussed above in Section II.B, definitive information on these structural parameters has been revealed by the recent 8.5 Å projection map of 2-D crystals of the Rba. sphaeroides core RC-LH1 dimer (Qian et al., 2005).
the forced flattening of normally curved membranes onto a flat support might modify the organization of components; however, for Rba. sphaeroides, LD measurements of oriented, untreated chromatophore preparations were consistent with the same type of organization for the LH2 and core complexes as that observed in AFM of the flat membrane patches (see Sections III.A and III.B.3 above), and further showed that this arrangement was extended over the entire chromatophore vesicle surface.
4. Rhodobacter blasticus
Recent AFM images obtained with native ICM preparations from Rps. palustris grown under different levels of illumination (Scheuring et al., 2006) clearly reflect the very versatile photosynthetic system formed by this species. Several unique organizations for the PSU components were revealed, as well as a number of similarities to arrangements described above, albeit in a novel context. Areas with mixed domains were most common under low illumination (Fig. 2e), along with crystalline peripheral antenna, while regions with apparently crystalline LH1-core complexes were observed under high illumination (Fig. 2f); a protein-free lipid bilayer was found under both light intensities. Interestingly, chromatic adaptation resulted in the modification of both the absorption and ring sizes of LH2 complexes. Thus, the unique 800 nm absorbing low-light (LL)-LH2 complex, with a ring size of eight αβ-heterodimers, was present under high intensity illumination at approximately the same level as the standard LH2 complex, with absorption bands at 800 and 850 nm and a ring size of nine αβ-heterodimers (ratio of ~1.0:0.5:0.5 for core complex/LH2 rings/ LL-LH2 rings). In contrast, the LL-LH2 became the predominant complex at low light intensity, where this ratio changed to ~1.0:0.2:2.0. As noted above in Section II.B, these AFM views were of sufficient resolution to confirm the X-ray crystallography derived model for the monomeric Rps. palustris RC-LH1 core complex (Roszak et al., 2003). It was also shown in the AFM study that the inner RC assembly induces the observed ellipticity upon the core complex (Scheuring et al., 2006), An important point to stress in interpreting the AFM studies in Rps. palustris and other purple bacteria is that a mosaic composition is expected to exist within the cytoplasmic membrane (CM)ICM continuum, representing the different stages of
For the AFM analysis on native membranes of the closely related LH2-containing species Rba. blasticus, it was necessary, as in Blc. viridis, to fuse the small isolated chromatophore vesicles by repeated cycles of freeze-thaw sonication (Scheuring et al., 2005a). This resulted in a mixture of complexes with alternating transverse orientations; however, in sufficiently large ICM domains of uniform orientation, AFM demonstrated components structured similarly to those in Rba. sphaeroides (Fig. 2d). These consist of S-shaped dimeric core complexes together with nonameric LH2 rings, but the mixing of the LH2 and core components in the Rba. blasticus membranes is much greater than that observed for Rba. sphaeroides (or indeed in Rsp. photometricum or Phs. molischianum). Thus, no lines of RCs or clearly separated antenna domains were observed, but rather a well mixed arrangement of core complexes and LH2, with no other visible components, or apparent space for additional machinery of the photosynthetic apparatus. These divergent observations merit comment concerning sample preparation; unlike the other species studied by AFM, both Rba. sphaeroides and Rba. blasticus form an ICM consisting of small, connected vesicular structures that upon cell disruption give rise to chromatophore vesicles. These isolated vesicles are inappropriate for high resolution imaging by scanning probe microscopy, as they do not adhere to the support to give sufficiently flat surfaces. As noted above, to overcome this difficulty, either freeze-thaw-induced membrane fusion (Scheuring et al 2003, 2005a), or treatment with sub-solubilizing levels of detergent (Bahatyrova et al., 2004b), were used to obtain membrane fragments suitable for imaging from Rba. blasticus and Rba. sphaeroides, respectively. It is possible that such treatments, or
5. Rhodopseudomonas palustris
262 membrane development and the distinct domains in which they occur. This ranges between unpigmented respiratory membrane, partially pigmented membrane invagination sites and fully pigmented ICM (see below). Representative membrane fractions arising form each of these regions can be isolated from Rba. sphaeroides (Parks and Niederman, 1978; Niederman et al., 1979), but only fragments of the mature ICM have thus far been examined by AFM (Bahatyrova et al., 2004b). Like Rsp. photometricum and Phs. molischianum, Rps. palustris forms infoldings of the CM that take the form of regular bundles of flattened thylakoid-like sacks. These ICM structures are differentiated into stacked and unstacked lamellae, and in freeze-fracture electron micrographs (EMs), a higher particle density was observed in the stacked regions (Varga and Staehelin, 1983). These regions may represent the membranes that were densely packed with core and LH2 complexes when visualized by AFM (Scheuring et al., 2006); however, the purely lipid membranes seen in the AFM analysis were not found in the freeze-fracture study. It is noteworthy that in freeze-fracture electron micrographs of Rba. sphaeroides cells undergoing differentiation to form the ICM, large lipid-enriched tubules attached to the CM were observed, representing an apparent developmental stage in ICM formation (Chory et al., 1984). 6. Common Points Revealed by Atomic Force Microscopy Studies A number of recurrent points seem to be worth reiterating concerning the observed membrane organizations. First, in none of the species studied have recognizable components other than peripheral light-harvesting complexes and core complexes been observed. Second, in all the reported cases the organization is not entirely random, with in most cases evidence for both core clustering and the formation of peripheral antenna domains. These effects seem to manifest themselves somewhat differently between species and with changes in the light intensity, but are nevertheless consistent factors. The lack of other components is particularly puzzling as in at least one of the species studied (Rba. sphaeroides), it has been suggested that the cytochrome bc1 complex is in close association with the RCs (Jungas et al., 1999, Joliot et al., 2005; Chapter 26, Lavergne et al; see Section IV. A). It should be noted, however, that the model of Jungas et al. (1999) for localization of the bc1 complex was proposed for
James N. Sturgis and Robert A. Niederman tubules in a strain lacking LH2, but that no bc1 complexes could be found in the tubules isolated from this strain by Siebert et al. (2004). Although all four subunits of the Rba. sphaeroides bc1 complex were demonstrated in Western blots of the ICM patches analyzed by AFM, including subunit IV which is thought to be localized largely on the cytoplasmic surface that was imaged in these preparations, it should be emphasized that in many of the images unattributed structures or gaps where the bc1 complex could be located are absent over large areas, as noted for Rsp. photometricum (Scheuring and Sturgis, 2005). Nevertheless, estimates of membrane composition in Rba. sphaeroides suggest that each 45 nm chromatophore vesicle should contain about 10 dimeric core complexes, 60 peripheral antenna complexes, 5 dimeric bc1 complexes and an ATP synthase (Geyer and Helms, 2006b), the precise numbers depending on the LH complex composition, and the related chromatophore size (Sturgis and Niederman, 1996). In this connection, Bahatyrova et al. (2004b) have suggested that a 70 nm chromatophore vesicle would comprise 15 RC-LH1-PufX dimers and approximately 100 LH2 complexes, in line with earlier estimates based on singlet-singlet annihilation data (Hunter et al., 1985). In some species and systems, it appears that a separation of the bc1 complexes and ATP synthase to other regions may exist (Geyer and Helms, 2006a; C. Mascle-Allemand, J. Lavergne and J. N. Sturgis (unpublished). The formation of separate core complex and LH2-rich regions is particularly interesting from a mechanistic point of view. First, it implies that the LH system might be sub-optimal, in light collection terms, as the LH-RC distances are not minimized. Second, it means that a complex hierarchy of interactions between the different components exists that serve to assure not only the differentiation of a stable photosynthetic membrane, but also the organization of different complexes in an evolutionarily optimal manner within these membranes. Several critical points have been raised concerning the remarkable images that have been published, or are reproduced here. These all center around the extent to which sample preparation and imaging conditions might have perturbed the surface organization of the membrane. In the earlier work, several different techniques were used to assure the adherence of sufficiently large flat membrane regions to the mica support, notably freeze-thaw fusion (Scheuring et al., 2003, 2005a), or sub-solubilizing detergent treatments (Bahatyrova et al., 2004b). As discussed above,
Chapter 14
In Situ Organization of Light-Harvesting Complexes
such treatments could well modify the organization of proteins in the membrane by altering membrane mechanical properties, assisting equilibration of kinetically trapped systems, or undermining the intermolecular interactions that determine the phase behavior of proteins in membranes. In the more recent work, such pre-treatments have proved to be unnecessary, and the membranes are deposited on the support directly after isolation (Gonçalves et al., 2005, Scheuring and Sturgis, 2005; Scheuring et al., 2006). Under these circumstances it is still possible that changes in membrane curvature or dynamics might modify the intra-molecular interactions. However, in several cases the supported structures observed by AFM strongly resemble those deduced from electron crystallography studies in terms of size, shape and curvature (reviewed by Scheuring et al., 2005b). A further argument against support driven modification of the imaged surface is that in several cases, they were not attached directly but are resting on another membrane adhering to the support (Scheuring and Sturgis, 2005). In view of the coherence of the picture that appears to be emerging, and despite differences in preparation techniques, it seems probable that most of these concerns are unjustified. Finally, it is useful to point out that all of the experimental observations described here have arisen from poorly ordered systems and therefore their interpretations do not easily lend themselves to any coherent theoretical description. While there is an increasing desire to study integrated systems in biology, we are currently somewhat deprived of the vocabulary necessary to describe such systems. As long as the structures described are homogeneous and well ordered, the idea of a supercomplex provides the necessary conceptual framework (see below). However, in the majority of environments within photosynthetic membranes, supercomplexes provide an overly structured view of an organization that is actually more diverse. Unfortunately at present, we cannot describe these systems effectively and thus fully appreciate their partially organized, as well as disorganized nature. IV. Proposals for the Functional Organization of Photosynthetic Units In this section, we will attempt to place the diversity of observed structural organizations into a coherent functional framework. First, the status of functional models involving supercomplexes, including ordered
263
quinone arrangements, will be evaluated in light of observed supramolecular arrangements. Because little structural support is found for these proposals, alternative views for the mechanism of quinone diffusion in the bilayer are presented that permit electron flow between the RC and remotely located cytochrome bc1 complexes. Finally, explanations for how paracrystalline antenna arrays may efficiently function in excitation energy transfer, as well as some possible physical bases for the self-assembly of the LH complexes, are also offered. A. Supercomplexes Thermodynamic studies of light-driven cyclic electron flow in whole Rba sphaeroides cells (most recently reviewed by Joliot et al., 2005; see also Chapter 26, Lavergne et al.), have formed the basis for a proposal in which the photosynthetic electron transfer chain is organized into supercomplexes consisting of a dimeric cytochrome bc1 complex sandwiched between two dimeric RC-LH1-PufX complexes, which interact with soluble cytochrome c2. With all the electron carriers arranged in this manner, a thermodynamic equilibrium between the primary RC donor and the high-potential Rieske FeS protein and cytochrome c1 components of the bc1 complex is thought to be mediated in <1 ms by the cytochrome c2 trapped in a single supercomplex. In contrast, equilibration between the different supercomplexes would occur 103 times more slowly. The highly ordered particles first seen in freeze-fracture electron micrographs of the tubular membranes of the LH2– strain M21 (Westerhuis et al., 2002), together with the requirement of PufX for the formation of RC dimers surrounded by open LH1 rings (Francia et al., 1999) were thought to provide structural bases for the supercomplex model. Joliot et al. (2005) have recently suggested that most of the UQ formed at the Qo site of the bc1 complex is also apparently trapped in supercomplexes. Paradoxically, the kinetic behavior attributed to electron flow within supercomplexes in whole cells cannot be demonstrated in isolated chromatophores. This has been attributed to a heterogeneity in the distribution of redox components among chromatophore vesicle populations in which cytochrome c2 is able to diffuse freely, owing to the perturbation of supercomplex structure during chromatophore preparation (Crofts et al., 1998). As a consequence, cytochrome c2 is able to diffuse within the intravesicular space in an unimpeded manner. Any model for supercomplex organization in the
264 ICM of purple bacteria would seem to require a close association between the RC and the bc1 complex; however, this is not supported by any of the submolecular ICM surface views provided by the recent AFM topographs from a variety of purple bacteria, in which no bc1 complex was visualized. The possibility has been considered that the bc1 complex might be located close to the CM or at the edges of ICM invaginations, out of view of the flat surfaces imaged in the topographs (Fig. 3) (Scheuring et al., 2005b). This might place some of the RCs at too great a distance from the bc1 complex for rapid functional coupling. This enigma is further discussed below (Section IV.C). Before leaving the topic of supercomplexes, it should be noted that in contrast to the purple bacterial ICM, considerable support now exists for the possibility that the oxidoreductase complexes of the mitochondrial respiratory chain exist in such an arrangement (recently reviewed by Boekema and Braun, 2007). The principal experimental evidence comes from the non-ionic detergent based co-purification of associated complexes (e.g., I + III2, consisting of Complex I (NADH dehydrogenase) together with the dimeric Complex III (cytochrome bc1 complex); III2 + IV1–2 (IV, cytochrome c oxidase), and large I+III2+IV1–4 supercomplexes), which were revealed by single particle EM analysis to exist in defined arrangements and stoichiometries. Such an approach has resulted in the isolation of a I+III4+IV4 supercomplex from Paracoccus dentrificans (Stroh et al., 2004); a similar supercomplex organization may hold for the oxidoreductases of the CM respiratory chain of closely related purple bacteria when grown aerobically. Under lowered aeration, the CM is the starting point for invaginations that differentiate into an ICM in which pigment-protein complexes are packed (see Section V), which could result in the disruption of supercomplex organization. More definitive evidence for functional supercomplex associations in the mitochondrial inner membrane and bacterial plasma membrane will require non-invasive techniques such as AFM of intact membrane surfaces in aqueous solution. B. Quinone Pools In a similar vein to the supercomplex model, it has more recently been proposed that the quinone pool is a misnomer and that, on metabolically relevant time scales, the assemblage of quinones is not well
James N. Sturgis and Robert A. Niederman
Fig. 3. Model of a Rba. sphaeroides chromatophore loosely based on that proposed by Geyer and Helms (2006b) showing dimeric RCs, LH complexes, the cytochrome bc1 complex, and ATP synthase in a 65 nm diameter vesicle. The vesicle backbone is formed by a linear band of dimeric RC-LH1 core complexes, similar to those observed in AFM topographs of ICM fragments by Bahatyrova et al. (2004b), with partially ordered LH2 complexes placed above and below this core dimer band. The ATP synthase is situated at the ‘North pole’ of the lower regions of the chromatophore. The LH2 rings are nonameric, as expected in Rba. sphaeroides (Walz et al., 1998). The dimeric core complex arrangement favored by Scheuring et al. (2004c) is depicted here. Note the model is also at odds with the arrangement and number of LH2 and core complexes proposed for semispherical ICM vesicles by Frese et al. (2004), on the basis of LD and AFM measurements. Important objectives in the next years include obtaining further experimental assessment of such models, refining them to include information on the stochastic variability and dynamics of such integrated systems, and defining the processes and rules involved in their assembly both in vivo and in vitro.
mixed and each quinone molecule can only access a limited number of RCs (Comayras et al., 2005a,b). This proposal was based on the effects of RC inhibitors on the apparent size of the quinone pool. These observations are related to the idea, at least in Rba. sphaeroides, that a functional unit for the electronic circuit contains two RCs, the quinone molecules and a dimeric cytochrome bc1 complex. However, in light of the inability to observe the bc1 complex in either Rba. sphaeroides chromatophore patches (Bahatyrova et al., 2004b) or isolated tubules (Siebert et al., 2004), the structural and functional measurements are not easily reconciled. C. Functional Importance of the Observed Organization Although a variety of architectures are observed for the photosynthetic apparatus, they obviously permit these systems to function efficiently in the native
Chapter 14
In Situ Organization of Light-Harvesting Complexes
environments of the different bacterial species. It is thus not surprising that distinct ecological constraints result in different solutions for the organization of the membrane to allow efficient photosynthesis. In this section, we will discuss two different aspects of these architectures; what might be the functional reasons behind the observed organizations, and what could be the forces and interactions responsible for the formation of these remarkable structures? As noted above, at the functional level, the organization of the photosynthetic complexes has to resolve several important paradoxes. First, for maximal efficiency of the LH system, it might be expected that the RC would be completely surrounded by antennae in a homogenous arrangement that minimizes the distances from the LH apparatus to the RC. Second, in order to avoid energy losses from closed RCs, the core complexes should be connected to allow energy transfer between them under intense illumination (Bahatyrova et al., 2004b; Scheuring et al., 2005b). Third, it would seem reasonable that the bc1 complex should be located in close proximity to the RC to permit closure of the electronic circuit by quinone/quinol diffusion in the membrane. Obviously, these different constraints cannot all be satisfied simultaneously. In light of the observed architectures, the apparent location of the cytochrome bc1 complex at some distance from the RC poses a major problem for quinone diffusion in the membrane bilayer. To begin with, the mode of escape of the quinol from the RC QB site through the closed LH1 ring found in many species is not readily apparent. In species such as Rba. sphaeroides, in which the RC is dimeric, the core antenna does not form closed rings, and thus escape is probably relatively simple. Here, the strategic placement of PufX, together with evidence for adjacent, highly mobile LH1 α-helices (Qian et al., 2005), implicates this region in the optimization of quinone/quinol exchange, which may also be aided by the overall flexibility demonstrated for the LH1 ring (Bahatyrova et al., 2004a). For Rps. palustris, quinone/quinol escape and re-entry appear to be facilitated through the built-in interruption by the α-helix of polypeptide W, observed within the LH1 ring (Roszak et al., 2003). In species such as Rsp. photometricum, however, in which the ring appears to be closed, quinone transfer across the LH1 wall would be dependent upon the dynamics of the process by which the integrity of the ring is breached. In this connection, it has been suggested that the several closed elliptical forms observed in AFM topographs
265
of 2-D crystals of the RC-LH1 core complex of Rsp. rubrum, reflect a flexible structure that could facilitate transient exchange of the quinone redox species through a dynamic series of molecular motions (Jamieson et al., 2002; Fotiadis et al., 2004). Support for such a diffusion process is provided by a recent molecular dynamics simulation study of possible ubiquinone shuttling pathways through the protein environment of the LH1 ring in Rsp. rubrum (Aird et al., 2007). An upper limit of the passage time for ubiquinone diffusion of ~8 × 10–3 s was calculated, compared with a turnover rate of ~3.5 ×10–3 s determined experimentally for electron transfer in monomeric RC-LH1 complexes with closed LH1 rings from a Rba. sphaeroides pufX– strain (Comayras et al., 2005b). It is relevant to this discussion that on the basis of the recent X-ray structure of Photosystem II (PS II), Loll et al. (2005) have identified a diffusion cavity for the secondary plastoquinone (PQB) that transfers redox equivalents from PS II to the photosynthetic electron transfer chain. This cavity is thought to provide a flexible, lipophilic environment for PQ/ PQH2 diffusion through an opening between the PQB-binding site and the PQ pool. The PS II protein PsbJ, which is located in a position that flanks this opening in the membrane, has been shown to influence electron transfer between PQA, PQB and the PQ pool, reminiscent of the role played by PufX, and reflecting an apparent functional evolutionary convergence. Additional details on an analogous lipophilic QB binding pocket in the Rba. sphaeroides RC can be found in the recent review of Jones (2007); see also Chapter 16 (Jones). Finally, the low resolution 3-D structure of the RC-LH1-PufX dimer from Rba. sphaeroides also reveals a possible location for a sequestered quinone pool adjacent to each RC (Chapter 9, Bullough et al.). The next difficulty faced by the quinone is diffusion in the membrane plane to reach the cytochrome bc1 complex through a crowded membrane. This problem has been studied recently by examining the dynamics of complexes in the Rsp. photometricum membrane using time-lapse AFM and Monte-Carlo simulations (Scheuring and Sturgis, 2006). It was shown that the para-crystalline regions of LH2 complexes were essentially static on long time scales, with no evidence for long-range or short-range diffusion, and packed so closely that insufficient space exists for quinones to diffuse within these arrays. In contrast, the disordered regions showed short-range corralled diffusion
266 of sufficient amplitude that would allow quinone molecules in the membrane plane to pass between complexes. This heterogeneity of dynamics, together with the exclusion of metabolically active quinones from the crystalline regions, is suggested to optimize electron transfer if the diameter of antennae domains is about the same as the distance between the RC and the cytochrome bc1 complex. This proposal seems consistent with the dynamics of electron transfer recently observed in Rsp. photometricum and Phs. molischianum (C. Mascle-Allemand, J. Lavergne and J. N. Sturgis (unpublished). The observation that the crystalline regions of the purple bacterial membrane are static, and the realization that the packing density is so great that there is insufficient space for quinone diffusion between complexes, poses a serious problem for the metabolic activity of core complexes that form crystalline domains as seen in Blc. viridis (Scheuring et al., 2003) or Rps. palustris (Scheuring et al., 2006). A possible explanation for their functional role might be that the complexes embedded in such domains are not active in cyclic electron transfer, but rather operate as light-sensitive batteries in which the stabilized charge separation between the cytochromes and bound quinones drive a rapid change in membrane potential in response to light. One of the surprises in the recent observations of the architecture of bacterial photosynthetic membranes was the clustering of peripheral LH complexes. As mentioned above, in certain species this arrangement may aid in directing electron transfer; however, the formation of large domains of LH2 was considered unlikely because of the adverse effect such domains might have on the excitation energy transfer and lightharvesting processes. Indeed, their formation can considerably increase the distance between certain LH complexes and the nearest RC. In spite of this, photosynthesis remains remarkably efficient in lowlight membranes even when antennae-RC distances are increased, though there is a small increase in the fluorescence yield (C. Mascle-Allemand, J. Lavergne and J.N. Sturgis, unpublished). In retrospect, however, this is perhaps not so surprising as LH2-LH2 energy transfer is very fast (~5 ps; Agarwal et al., 2002) compared to the excited-state B850 lifetime of ~1 ns (Bergström et al., 1988), allowing long-range exciton diffusion in the very well connected para-crystalline domains. The highly organized nature of the various protein arrays found in photosynthetic membranes poses the
James N. Sturgis and Robert A. Niederman question as to how this is achieved. Here, it should be pointed out that different ways of describing the organizations exist, and none is entirely satisfactory. On the one hand, the observed arrangements can be considered as a series of structured antenna-RC supercomplexes with some heterogeneity; such a view allows the construction of increasingly large assemblages (Scheuring et al., 2007). Alternatively, the organizations can be regarded as a disordered system in which the phase-physics and statistical properties can be examined (Scheuring and Sturgis, 2005). Because we lack the necessary vocabulary for properly describing and studying these organizations, as noted above, neither of these approaches is wholly suitable. Several observations should be highlighted in considering the detailed organization of proteins in the membrane. First, in several species, but most clearly in Rps. palustris (Scheuring et al., 2006), membrane areas rich in photosynthetic proteins are contiguous with flat lipid-rich membrane areas, emphasizing the stability of the organization of the photosynthetic apparatus (C. Mascle-Allemand, J. Lavergne and J.N. Sturgis, unpublished). It also implies that the system is able to self-assemble and thus represents an energetically favored organization. The second point worth noting is that a very simple model of interaction forces (Scheuring and Sturgis, 2005) was able to mimic the self-assembly of the two-phase organization as is observed in Rsp. photometricum or Phs. molischianum. This emphasizes the fact that the complexes can self-assemble with the final organization reflecting the interaction potentials that drive assembly. Of particular importance in this respect, are not only the interaction forces, but also the ranges of interaction (Noro and Frenkel, 2001), such that the formation of dense disordered systems appears to rely, at least in sufficiently simple systems, on the presence of long-range interactions. Because of this, it is possible that the observed membrane organization could be driven by long-range attractive forces between the components, with a species-specific hierarchy of interaction strengths defining the final arrangement (J.N. Sturgis, unpublished). However, the physical nature of these proposed long-range interactions remains obscure. The observation of very similar membrane organization in Rsp. photometricum and Phs. molischianum (Scheuring et al., 2004b; Gonçalves et al., 2005) gives some insight into the type of interactions that might be involved in organizing the different complexes
Chapter 14
In Situ Organization of Light-Harvesting Complexes
in the membrane. While these two organisms share a common supramolecular organization, the symmetry of the underlying complexes is different. In Rsp. photometricum, the LH2 complex is nonameric, while in the case of Phs. molischianum the complex is octameric; nonetheless, both complexes can form densely packed hexagonal arrays. This situation cannot be due to strong specific interactions between the LH2 complexes because of the change in symmetry, which implies, in contrast, that the association of LH2 complexes, and possibly the whole photosynthetic apparatus, is driven by non-specific interactions. Such non-specific interactions could be considered alternatively as an insolubility of the photosynthetic complexes in the lipid bilayer. A solubility problem of this magnitude could arise from several different origins, e.g., a hydrophobic mismatch as proposed in the membrane mattress model (Mouritsen and Bloom, 1984). In certain species, a difference in membrane curvature between photosynthetic membranes and cytoplasmic membranes, as previously proposed (Sturgis and Niederman, 1996) can be imagined, as well as several other elastic or chemical differences between the surfaces of the proteins and the membrane phospholipids. V. In Vivo Assembly of Light-Harvesting Complexes Owing to the numerous practical advantages pointed out above in Section III.B.3, Rba. sphaeroides and Rba. capsulatus have, over the years, provided an ideal model systems for examining the biogenesis and assembly of photosynthetic membranes. When grown photoheterotrophically (anaerobically in the light), the levels of LH2 relative to RC-LH1 core complexes are related inversely to light intensity (Cohen-Bazire et al., 1956) and shifts in illumination levels have offered a valuable paradigm for examining differential biosynthesis of photosynthetic complexes during chromatic adaptation. Although ICM formation is repressed by high oxygen tension under chemoheterotrophic growth conditions, lowering oxygen partial pressure initiates a gratuitous induction of ICM assembly in the dark (greening) by invagination of the CM, together with the synthesis and assembly of the LH and RC complexes (Takemoto and Lascelles, 1973). In Rba. sphaeroides, it is possible to isolate these CM invagination sites in the form of an upperpigmented band that sediments more slowly than the
267
ICM-derived chromatophore vesicle fraction during rate-zone centrifugation on sucrose density gradients (Niederman et al., 1979; Niederman, 2006). This membrane fraction provides a basis for the application of a variety of definitive biochemical, biophysical and ultrastructural probes to the membrane development process (Niederman et al., 1979; Bowyer et al., 1985; Hunter et al., 1988; Sturgis and Niederman, 1996; Hunter et al., 2005; Koblízek et al., 2005). These studies (reviewed recently by Niederman, 2006), demonstrated that photosynthetic units are assembled sequentially, with RC-LH1 core structures inserted initially into the CM in a form that is largely inactive in forward electron transfer. This is followed by the activation of functional electron flow, together with the addition of LH2, resulting in membrane invagination and vesicularization to form the ICM. LH2 is thought to initially pack between the linear arrays of dimeric core complexes, and to ultimately form the clusters of LH2-only domains that represent the light-responsive peripheral antenna complement. Interestingly, a rather similar mechanism of sequential assembly has been established for both PS I and PS II in cyanobacteria, and there are suggestions that the initial synthesis of thylakoid membranes in plant proplastids may involve invagination of the inner envelope, into which thylakoid proteins are first inserted (Cline, 2003). The developmental changes occurring within the CM-ICM continuum of purple bacteria are under the control of a number of interrelated two-component regulatory circuits that act at the transcriptional level to regulate the formation of both the pigment and apoprotein components of the LH and RC complexes in response to oxygen and light (see contributions in Chapter 35, Bauer et al.; Chapter 36, Klug and Masuda, for detailed descriptions of these elegant systems). Light-mediated regulation in anoxygenic phototrophs is also controlled by newly discovered bacteriophytochromes (BPs) (Giraud et al., 2002) that act as far-red(fr)/red(r) light sensors through reversible BPr to BPfr structural transformations to control expression of photosynthetic complex genes. One of the BPs from Rps. palustris has recently been shown by Evans et al. (2006) to possess a plant-like phytochrome structure consisting of N-terminal biliverdin binding (phytochrome) and C-terminal histidine kinase domains, and to function in maintaining high levels of the unique 800 nm absorbing LL-LH2 complex under low light intensity via a phosphorelay response regulator system (Evans et
268 al., 2005; Giraud et al., 2005; Chapter 40, Evans et al). Another BP of Rps. palustris is involved in the regulation of RC-LH1 core complex levels; further description of roles played by the multiple BPs of purple bacteria can be found in the recent review of Cogdell et al. (2006). The above mechanisms assure that the correct transcripts of the LH apoprotein genes are made. Subsequent stages in the formation of functional antenna complexes include: (i) the folding of the resulting nascent polypeptides into membrane insertion-competent conformations; (ii) their insertion and integration into the membrane; (iii) the binding of their BChl and carotenoid chromophores; (iv) the oligomerization of the pigment-protein protomers into their final oligomeric annular structures, and (v) their supramolecular organization into functional PSUs. With the advent of accessible molecular genetic systems, a number of site-directed mutagenesis studies have identified apoprotein regions and individual residues that are important for the assembly, chromophore attachment, and spectral tuning of both the LH1 and LH2 complexes (recently reviewed by Cogdell et al., 2006; see also Chapter 11, Robert; Chapter 46, Braun and Fiedor). These authors have pointed out that this in vivo approach exploits the cell’s normal assembly processes and the ability to provide the required pigments. Very little is known about these processes, and in the remainder of this section, we summarize the current state of knowledge in these areas and provide some speculations on how the many missing details may ultimately play out. A. Complex-Specific Assembly Factors Considerable effort in recent years has focused upon complex-specific assembly factors that are encoded by genes within photosynthesis gene clusters, and serve to assist either directly or indirectly, in LH and RC protein assembly in developing membranes (reviewed by Young and Beatty, 2003). These include the LH1-specific assembly factor LhaA, and the related LH2-specific factor PucC, that are both members of the putative ‘chlorophyll delivery branch’ of the major facilitator superfamily. A possible role for LhaA in BChl cofactor assembly is suggested by its co-enrichment with the BChl synthase protein, BchG, in the Rba. sphaeroides upper pigmented band (C.N. Hunter, personal communication; Hunter et al., 2005). These findings, together with radiolabeling studies using the heme precursor δ-aminolevulinate, suggest
James N. Sturgis and Robert A. Niederman that the upper pigmented fraction contains membrane sites for both the preferential biosynthesis of BChl cofactors and their attachment to the LH1 and RC apoproteins, which are also assembled preferentially at these sites (Niederman et al., 1979). This raises the question of whether a targeting process delivers the nascent apoproteins to the membrane at the precise location of BChl availability through the coordination of protein integration with pigment synthesis (see below). In addition to PufX, whose role in dimerization of RC-LH1 cores and their supramolecular organization is discussed above, complex-specific assembly factors encoded within the photosynthesis gene cluster in Rba. capsulatus and Rba. sphaeroides include PuhB, PuhC and PuhE, encoded by open reading frames located immediately 3´ of the puhA gene encoding the RC-H subunit (Young and Beatty 2003; Chapter 9, Bullough et al.); homologs of these open reading frames have also been identified in the Rps. palustris genome. These factors have roles in the biogenesis of RC-LH1 core complexes: PuhB has been shown to serve as an apparent dimeric RC assembly factor with secondary effects on LH1 assembly; PuhC is required for optimal RC-LH1 levels, and is thought to be involved in the reorganization of core complex (Aklujkar et al., 2006); and PuhE is a negative modulator of BChl synthesis (Aklujkar et al., 2005). PufQ, located at the 5´ end of the puf operon (Hunter et al., 1991), stimulates BChl production, acts as a putative carrier of intermediates of the BChl biosynthetic pathway, and is thought to assist in the assembly of LH1 and LH2. With regard to carotenoid attachment to the LH apoproteins, pigment composition studies have demonstrated preferential associations of all-trans species with LH apoproteins, in contrast to the 15-cis species found in the RC (Koyama and Fujii 1999), consistent with their respective lightharvesting and photoprotective functions in these protein environments. No information, however, is available on factors that may assist in the attachment of carotenoids to LH complexes. B. General Membrane Assembly Factors In contrast to the complex-specific assembly factors, very little is known about the general membrane assembly factors of purple bacteria, which undoubtedly represent more ancient homologs of the well-studied factors from Escherichia (E.) coli, since anoxygenic phototrophs are generally believed to be among the
Chapter 14
In Situ Organization of Light-Harvesting Complexes
oldest life forms. Factors studied thus far include soluble chaperonins DnaK and GroEL, which in a cell-free, transcription-translation system of Rba. capsulatus, have been shown to play roles in the synthesis and stable assembly of the LH1 apoprotein (Drews, 1996). The translocation ATPase, SecA, has also been examined in Rba. capsulatus, and is apparently involved in the movement of c-type apocytochromes to the periplasm for heme attachment (Helde et al., 1997). Other factors detected in a search of available Rba. sphaeroides genome sequences include homologs of the SecY translocon, the YidC membrane chaperonin, the SecDF membrane insertion components, and the prokaryotic signal recognition particle homologs, Fth/FtsY. In E. coli, YidC functions both independently as a membrane-protein insertase, or in conjunction with the Sec membraneprotein insertion pathway, where YidC is thought to serve as an integrase which folds transmembrane segments of nascent polypeptides to ensure their lateral integration (Dalbey and Kuhn, 2004). SecY, a homolog of the Sec61 protein-conducting channel of the endoplasmic reticulum (ER), is the major pore component of the SecYEG translocon (Van den Berg et al., 2004). The SecDF/YajC membrane insertion components have recently been shown to be involved in the Sec translocon independent pathway (Chen et al., 2005), and alternatively to link SecYEG to YidC. Targeting to the Sec translocon is mediated by the Ffh/FtsY homolog of the ER-associated signal recognition particle that recognizes the membrane targeting sequences of integral membrane proteins (Luirink and Sinning 2004). It is noteworthy that in proteomic analyses of Rps. palustris photosynthetic membranes (Fejes et al., 2003; VerBerkmoes et al., 2006), PuhB, PuhC, DnaK, GroEL, SecA, SecF, SecD, and YajC were identified. How such factors may be involved in the assembly of the PSUs of purple bacteria is unknown and remains an important topic for future investigation. Structural considerations (Hu et al., 2002) suggest that, in addition to roles for the DnaK and GroEL chaperonins, the putative Chl delivery factors and other complex-specific factors, at least some of these additional soluble and membrane-bound general assembly factors are needed to support LH1 and LH2 polypeptide membrane targeting and insertion, as well as their assembly into functional annular complexes. These factors may well include SecA, YidC, SecDF, Ffh/FtsY and the Sec translocon. While it is possible that the LH proteins in an appropriately folded state,
269
are inserted spontaneously into the lipid bilayer, the balance of recent evidence in E. coli suggests that the membrane insertion process, even for such small proteins, is enzymatically controlled by interacting membrane assembly factors, ensuring that a defined topology is assumed within the membrane (Facey and Kuhn 2004; Chen et al., 2005). Structural models for the Rba. sphaeroides LH1 and LH2 αβ-heterodimers (Olsen and Hunter, 1994; Hu et al., 2002) suggest that the apoprotein subunits each consist of single transmembrane α-helices of 23–28 residues, with charged C-terminal periplasmic extensions of 17–22 residues in the case of the α-polypeptides. Translocation of such hydrophilic domains across the membrane would be expected to require participation of an electrochemical proton gradient (Dierstein and Drews, 1986), and an enzymatically-assisted passage. Moreover, during LH protein folding within the lipid bilayer, YidC may also facilitate associations with membrane-bound proteins that function in the later stages of pigment biosynthesis and in the attachment of these cofactors, a role that has been suggested in the assembly of LH chlorophyll proteins (LHCPs) for the chloroplast YidC homolog, Oxa1p (Dalbey and Kuhn 2004). VI. Perspectives for the Next Ten Years Research over the next decade should see a continued growth in studies of ICM structure, and in particular, on the assembly processes that direct ICM formation. Among other areas that are likely to progress significantly are the development of high-resolution experimentally based models along the lines of that presented by Geyer and Helms (2006a,b)(Fig. 3), as well as the relationship between ICM architecture and ecological niches. Several aspects of the assembly process seem ripe for significant developments especially in view of the recent explosion of structural detail on the molecular organization of the ICM. These include the roles of assembly factors, the energetics of this process, and the role of protein sequence in controlling ICM organization. In this connection, it has recently been shown that PufX, which is responsible for the supramolecular dimeric organization of RC-LH1-PufX complexes in the Rba sphaeroides ICM (Siebert et al., 2004; Frese et al., 2004; Chapter 9, Bullough et al.), is present in monomeric RC-LH1-PufX complexes in the related
270 but independently evolved species Rba. veldkampii (Gubellini et al., 2006). While the kinetics of flashinduced cytochrome b561 reduction suggested that PufX apparently facilitates the quinone-mediated interaction between the RC and bc1 complexes in the same manner in both Rba. veldkampi and Rba sphaeroides, it is thought that a difference in the sequence of the cytoplasmic N-terminal domain may account for the inability of the Rba. veldkampii PufX protein to facilitate core complex dimerization. With the recent determination of the solution NMR structure of the PufX protein (Tunnicliffe et al., 2006; Wang et al., 2007), further clarification of the various assembly (Aklujkar and Beatty, 2005, 2006) and functional (Comayras et al., 2005b) roles proposed for this important core complex component should now become possible. The new knowledge of the molecular organization of the ICM, largely obtained by AFM, will make advances possible in many different unresolved aspects of this process. This can be exemplified in the need for improvement in models such as that of Geyer and Helms (2006b), depicted in Fig. 3, and this is to be expected over the course of the next decade at the level of refining the structures of the individual components and improving our vision of the organization of the elusive cytochrome bc1 and ATP synthase complexes. Other issues to be resolved include a better understanding of the stochastic variations between chromatophores, both in their nature and origin, and the further examination of the dynamics of the ICM at this level of structural resolution under different time scales (femtoseconds to milliseconds to hours). While we visualize that these essentially descriptive advances are likely to take place over the next 10 years, our understanding of the fundamental processes involved may well advance more slowly. However, we would expect a number of advances, in particular on the molecular determinants of ICM organization, a domain that was inaccessible prior to our current picture of the surface arrangement of photosynthetic complexes at submolecular resolution. It is further expected that the newer methods, such as proteomics and AFM will find application in all these types of studies, since the employment of these techniques together in the ICM of purple bacteria offers unique structural proteomic approaches currently unavailable with any other complex energy transducing membrane. AFM used in combination with confocal fluorescence microscopy (Kassies et al., 2005) also holds considerable promise. This atomic
James N. Sturgis and Robert A. Niederman force fluorescence microscopy (AFFM) technique is capable of simultaneous topographic and optical imaging with single molecule resolution in both the visible and near-IR spectral regions. The resulting molecular topography data can be integrated with specific fluorescence emission spectra for the identification of individual developing complexes, and it has been shown that these spectral imaging capabilities are sufficiently sensitive to distinguish LH1 and LH2 within the native PSU. The AFM-based nanomanipulation capabilities are also retained in the AFFM, such as the application of loading forces and measurements of forces of interaction and adhesion. References Agarwal R, Rizvi AH, Prall BS, Olsen JD, Hunter CN and Fleming GR (2002) The nature of disorder and inter-complex energy transfer in LH2 at room temperature: A three pulse photon echo peak shift study. J Phys Chem A 106: 7573–7578 Aird A Wrachtrup J, Schulten K and Tietz C (2007) Possible pathway for ubiquinone shuttling in Rhodospirillum rubrum revealed by molecular dynamics simulation. Biophys J 92: 23–33 Aklujkar M, Prince RC and Beatty JT (2005) The puhE gene of Rhodobacter capsulatus is needed for optimal transition from aerobic to photosynthetic growth and encodes a putative negative modulator of bacteriochlorophyll production. Arch Biochem Biophys 437: 186–198 Aklujkar M, Prince RC and Beatty JT (2006) The photosynthetic deficiency due to puhC gene deletion in Rhodobacter capsulatus suggests a PuhC protein-dependent process of RC/LH1/PufX complex reorganization. Arch Biochem Biophys 454: 59–71 Bakker JGC, van Grondelle R and den Hollander WTF (1983) Trapping, loss and annihilation of excitations in a photosynthetic system. II. Experiments with the purple bacteria Rhodospirillum rubrum and Rhodopseudomonas capsulata. Biochim. Biophys Acta 725: 508–518 Bahatyrova S, Frese RN, van der Werf KO, Otto C, Hunter CN and Olsen JD (2004a) Flexibility and size heterogeneity of the LH1 light harvesting complex revealed by atomic force microscopy: Functional significance for bacterial photosynthesis. J Biol Chem 279: 21327–21333 Bahatyrova S, Frese RN, Siebert, CA, van der Werf KO, van Grondelle R, Niederman RA, Bullough PA, Otto C, Olsen JD and Hunter CN (2004b) The native architecture of a photosynthetic membrane. Nature 430: 1058–1062 Barz WP, Verméglio A, Francia F, Venturoli G, Melandri BA and Oesterhelt D (1995) Role of the PufX protein in photosynthetic growth of Rhodobacter sphaeroides. 2. PufX is required for efficient ubiquinone/ubiquinol exchange between the reaction center QB site and the cytochrome bc1 complex. Biochemistry 34: 15248–15258 Bergström H, Sundström V, van Grondelle R, Gillbro T and Cogdell R (1988) Energy transfer dynamics of isolated B800–850 and B800–820 pigment-protein complexes of Rhodobacter sphaeroides and Rhodopseudomonas acidophila. Biochim Biophys Acta 936: 90–98
Chapter 14
In Situ Organization of Light-Harvesting Complexes
Blankenship RE, Madigan MT and Bauer CE (eds) (1995) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2). Kluwer Academic Publishers, Dordrecht Boekema EJ and Braun HP (2007) Supramolecular structure of the mitochondrial oxidative phosphorylation system. J Biol Chem 282: 1–4 Bowyer JR, Hunter CN, Ohnishi T and Niederman RA (1985) Photosynthetic membrane development in Rhodopseudomonas sphaeroides. Spectral and kinetic characterization of redox components of light-driven electron flow in apparent photosynthetic membrane growth initiation sites. J Biol Chem 260: 3295–3304 Chen M, Xie K, Yuan J, Yi L, Facey SJ, Pradel N, Wu LF, Kuhn A and Dalbey RE (2005) Involvement of SecDF and YidC in the membrane insertion of M13 procoat mutants. Biochemistry 44: 10741–10749 Chory J, Donohue TJ, Varga AR, Staehelin LA and Kaplan S (1984) Induction of the photosynthetic membranes of Rhodopseudomonas sphaeroides: Biochemical and morphological studies. J Bacteriol 159: 540–554 Cline K (2003) Biogenesis of green plant thylakoid membranes. In: Green BR and Parson WW (eds) Light-Harvesting Antennas in Photosynthesis (Advances in Photosynthesis and Respiration, Vol 13), pp 353–372, Kluwer Academic Publishers, Dordrecht Cogdell RJ, Gall A and Kohler J (2006) The architecture and function of the light-harvesting apparatus of purple bacteria: from single molecules to in vivo membranes. Q Rev Biophys 39: 227–324 Cohen-Bazire G, Sistrom WR and Stanier RY (1956) Kinetic studies of pigment synthesis by non-sulfur purple bacteria. J Cell Comp Physiol 49: 25–68 Comayras F, Jungas C and Lavergne J (2005a) Functional consequences of the organization of the photosynthetic apparatus in Rhodobacter sphaeroides. I. Quinone domains and excitation transfer in chromatophores and reaction center antenna complexes. J Biol Chem 280: 11203–11213 Comayras F, Jungas C and Lavergne J (2005b) Functional consequences of the organization of the photosynthetic apparatus in Rhodobacter sphaeroides: II. A study of PufX- membranes. J Biol Chem. 280: 11214–11223 Crofts A, Guergova-Kuras M and Hong S (1998) Chromatophore heterogeneity explains phenomena seen in Rhodobacter sphaeroides previously attributed to supercomplexes. Photosynth Res 55: 357– 362 Dalbey RE and Kuhn A (2004) YidC family members are involved in the membrane insertion, lateral integration, folding, and assembly of membrane proteins. J Cell Biol 166: 769–774 Deinum G, Aartsma TJ, van Grondelle R and Amesz J (1989) Singlet-singlet annihilation measurements on the antenna of Rhodospirillum rubrum between 300 and 4 K. Biochim Biophys Acta 976: 63–69 Deinum G, Otte SCM, Gardiner AT, Aartsma TJ, Cogdell RJ and Amesz J (1991) Antenna organization of Rhodopseudomonas acidophila: a study of the excitation migration. Biochim Biophys Acta 1060: 125–131 Dierstein R and Drews G (1986) Effect of uncoupler on assembly pathway for pigment-binding protein of bacterial photosynthetic membranes. J Bacteriol 168: 167–172 Drews G (1996) Formation of the light-harvesting complex I (B870) of anoxygenic phototrophic purple bacteria. Arch
271
Microbiol 166: 151–159 Drews G and Golecki JR (1995) Structure, molecular organization, and biosynthesis of membranes of purple bacteria. In: Blankenship RE, Madigan MT, Bauer CE (ed) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 231–257. Kluwer Academic Publishers, Dordrecht Evans K, Fordham-Skelton AP, Mistry H, Reynolds CD, Lawless AM and Papiz MZ (2005) A bacteriophytochrome regulates the synthesis of LH4 complexes in Rhodopseudomonas palustris. Photosynth Res 85: 169–180 Evans K, Grossmann JG, Fordham-Skelton AP and Papiz MZ. (2006) Small-angle X-ray scattering reveals the solution structure of a bacteriophytochrome in the catalytically active Pr state. J Mol Biol 364: 655–666 Facey SJ and Kuhn A (2004) Membrane integration of E. coli model membrane proteins. Biochim Biophys Acta 1694: 55–66 Fejes AP, Yi EC, Goodlett DR and Beatty JT (2003) Shotgun proteomic analysis of a chromatophore-enriched preparation from the purple phototrophic bacterium Rhodopseudomonas palustris. Photosynth Res 78: 195–203 Fotiadis D, Qian P, Philippsen A, Bullough PA, Engel A and Hunter CN (2004) Structural analysis of the reaction center light-harvesting complex I photosynthetic core complex of Rhodospirillum rubrum using atomic force microscopy. J Biol Chem 279: 2063–2068 Francia F, Wang J, Venturoli G, Melandri BA, Barz WP and Oesterhelt D. (1999) The reaction center-LH1 antenna complex of Rhodobacter sphaeroides contains one PufX molecule which is involved in dimerization of this complex. Biochemistry 38: 6834–6845 Frese RN, Olsen JD, Branvall R, Westerhuis WH, Hunter CN and van Grondelle R (2000) The long-range supraorganization of the bacterial photosynthetic unit: A key role for PufX. Proc Natl Acad Sci USA 97: 5197–5202 Frese RN, Siebert CA, Niederman RA, Hunter CN, Otto C and van Grondelle R (2004) The long-range organization of a native photosynthetic membrane. Proc Natl Acad Sci USA 101: 17994–17999 Geyer T and Helms V (2006a) Reconstruction of a kinetic model of the chromatophore vesicles from Rhodobacter sphaeroides. Biophys J 91: 927–937 Geyer T and Helms V (2006b) A spatial model of the chromatophore vesicles of Rhodobacter sphaeroides and the position of the cytochrome bc1 complex. Biophys J 91: 921–926 Giraud E, Fardoux J, Fourrier N, Hannibal L, Genty B, Bouyer P, Dreyfus B and Verméglio A (2002) Bacteriophytochrome controls photosystem synthesis in anoxygenic bacteria. Nature 417: 202–205 Giraud E, Zappa S, Vuillet L, Adriano JM, Hannibal L, Fardoux J, Berthomieu C, Bouyer P, Pignol D and Verméglio A (2005) A new type of bacteriophytochrome acts in tandem with a classical bacteriophytochrome to control the antennae synthesis in Rhodopseudomonas palustris. J Biol Chem 280: 32389–32397 Gonçalves RP, Bernadac A, Sturgis JN and Scheuring S (2005) Architecture of the native photosynthetic apparatus of Phaeospirillum molischianum. J Struct Biol 152: 221–228 Gubellini F, Francia F, Busselez J, Venturoli G and Levy D (2006) Functional and structural analysis of the photosyn-
272 thetic apparatus of Rhodobacter veldkampii. Biochemistry 45: 10512–10520 Hartigan N, Tharia HA, Sweeney F, Lawless AM and Papiz MZ (2002) The 7.5-Å electron density and spectroscopic properties of a novel low-light B800 LH2 from Rhodopseudomonas palustris. Biophys J 82: 963–977 Helde R, Wiesler B, Wachter E, Neubuser A, Hoffschulte HK, Hengelage T, Schimz KL, Stuart RA and Müller M. (1997) Comparative characterization of SecA from the alpha-subclass purple bacterium Rhodobacter capsulatus and Escherichia coli reveals differences in membrane and precursor specificity. J Bacteriol 179: 4003–4012 Hu X, Ritz T, Damjanovic A, Autenrieth F and Schulten K (2002) Photosynthetic apparatus of purple bacteria Q Rev Biophys 35: 1–62 Hunter CN, Kramer HJM and van Grondelle R (1985) Linear dichroism and fluorescence emission of antenna complexes during photosynthetic unit assembly in Rhodopseudomonas sphaeroides. Biochim Biophys Acta 807: 44–51 Hunter CN, Pennoyer JD, Sturgis JN, Farrelly D and Niederman RA (1988) Oligomerization states and associations of lightharvesting pigment-protein complexes of Rhodobacter sphaeroides as analyzed by lithium dodecyl sulfate-polyacrylamide gel electrophoresis. Biochemistry 27: 3459–3467 Hunter CN, McGlynn P, Ashby MK, Burgess JG and Olsen JD (1991) DNA sequencing and complementation/deletion analysis of the bchA-puf operon region of Rhodobacter sphaeroides: In vivo mapping of the oxygen-regulated puf promoter. Molecular Microbiology 5: 2649–2661 Hunter CN, Tucker JD and Niederman RA (2005) The assembly and organisation of photosynthetic membranes in Rhodobacter sphaeroides. Photochem Photobiol Sci 4: 1023–1027 Jamieson SJ, Wang P, Qian P, Kirkland JY, Conroy MJ, Hunter CN and Bullough PA. (2002) Projection structure of the photosynthetic reaction centre-antenna complex of Rhodospirillum rubrum at 8.5 Å resolution. EMBO J 21: 3927–3935 Janosi L, Kosztin I and Damjanovic A (2006) Theoretical prediction of spectral and optical properties of bacteriochlorophylls in thermally disordered LH2 antenna complexes. J Chem Phys 125: 014903 Joliot P, Joliot A and Verméglio (2005) Fast oxidation of the primary electron acceptor under anaerobic conditions requires the organization of the photosynthetic chain of Rhodobacter sphaeroides in supercomplexes. Biochim Biophys Acta 1706: 204–214 Jones MR (2007) Lipids in photosynthetic reaction centres: Structural roles and functional holes. Prog Lipid Res 46: 56–87 Jungas C, Ranck JL, Rigaud JL Joliot P and Verméglio A (1999) Supramolecular organization of the photosynthetic apparatus of Rhodobacter sphaeroides. EMBO J 18: 534–542 Karrasch S, Bullough PA and Ghosh R (1995) The 8.5 Å projection map of the light-harvesting complex I from Rhodospirillum rubrum reveals a ring composed of 16 subunits. EMBO J 14: 631–638 Kassies R, van der Werf KO, Lenferink A, Hunter CN, Olsen JD, Subramaniam V and Otto C (2005) Combined AFM and confocal fluorescence microscope for applications in bionanotechnology. J Microsc 217: 109–116 Koblízek M, Shih JD, Breitbart SI, Ratcliffe EC, Kolber ZS, Hunter CN and Niederman RA. (2005) Sequential assembly of
James N. Sturgis and Robert A. Niederman photosynthetic units in Rhodobacter sphaeroides as revealed by fast repetition rate analysis of variable bacteriochlorophyll a fluorescence. Biochim Biophys Acta 1706: 220–231 Koepke J, Hu X, Muenke C, Schulten K and Michel H (1996) The crystal structure of the light-harvesting complex II (B800-850) from Rhodospirillum molischianum. Structure 4: 581–597 Koyama, Y and Fujii R (1999) Cis–trans carotenoids in photosynthesis: Configurations, excited state properties and physiological functions. In: Frank HA, Young AJ, Britton G and Cogdell RJ (eds) The Photochemistry of Carotenoids (Advances in Photosynthesis and Respiration, Vol 8), pp 161–188. Kluwer Academic Publishers, Dordrecht Kramer H, Deinum G, Gardiner AT, Cogdell RJ, Francke C, Aartsma TJ and Amesz J (1995) Energy transfer in the photosynthetic antenna system of the purple-non-sulfur bacterium Rhodopseudomonas cryptolactis. Biochim Biophys Acta 1231: 33–40 Loll B, Kern J, Saenger W, Zouni A and Biesiadka J (2005) Towards complete cofactor arrangement in the 3.0 Å resolution structure of photosystem II. Nature 438: 1040–1044 Luirink J and Sinning I (2004) SRP-mediated protein targeting: structure and function revisited. Biochim Biophys Acta 1694: 17–35 McDermott G, Prince SM, Freer AA, Hawthornthwaite-Lawless AM, Papiz MZ, Cogdell RJ and Isaacs NW (1995) Crystal structure of an integral membrane light-harvesting complex from photosynthetic bacteria. Nature 374: 517–521 Monger TG and Parson WW (1977) Singlet-triplet fusion in Rhodopseudomonas spheroides chromatophores. A probe of the organization of the photosynthetic apparatus. Biochim Biophys Acta 460: 393–407 Mouritsen OG and Bloom M (1984) Mattress model of lipidprotein interactions in membranes. Biophys J 46: 141–153 Niederman RA (2006) Structure, function and formation of bacterial intracytoplasmic membranes. In: Shively JM (ed) Complex Intracellular Structures in Prokaryotes, Microbiology Monographs, Vol 2, pp 193–227. Springer-Verlag, Berlin, Heidelberg Niederman RA, Mallon DE and Parks LC (1979) Membranes of Rhodopseudomonas sphaeroides. VI. Isolation of a fraction enriched in newly synthesized bacteriochlorophyll a-protein complexes. Biochim Biophys Acta 555: 210–220 Noro MG and Frenkel D (2001) Phase behavior of a simple model for membrane proteins. J Chem Phys 114: 2477–2483 Olsen JD and Hunter CN (1994) Protein structure modelling of the bacterial light harvesting complex. Photochem Photobiol 60: 521–535 Parks LC and Niederman RA (1978) Membranes of Rhodopseudomonas sphaeroides. V. Identification of bacteriochlorophyll a-depleted cytoplasmic membrane in phototrophically grown cells. Biochim Biophys Acta 511: 70–80 Parson WW and Nagarajan V (2003) Optical spectroscopy in photosynthetic antennas. In: Green BR and Parson WW (eds) Light-Harvesting Antennas in Photosynthesis (Advances in Photosynthesis and Respiration, Vol 13), pp 83–127. Kluwer Academic Publishers, Dordrecht Qian P, Hunter CN and Bullough PA (2005) The 8.5Å projection structure of the core RC-LH1-PufX dimer of Rhodobacter sphaeroides. J Mol Biol 349: 948–960 Roszak AW, Howard TD, Southall J, Gardiner AT, Law CJ, Isaacs,
Chapter 14
In Situ Organization of Light-Harvesting Complexes
NW and Cogdell RJ (2003) Crystal structure of the RC-LH1 core complex from Rhodopseudomonas palustris. Science 302: 1969–1972 Scheuring S and Sturgis JN (2005) Chromatic adaptation of photosynthetic membranes. Science 309: 484–487 Scheuring S and Sturgis JN (2006) Dynamics and diffusion in photosynthetic membranes from Rhodospirillum photometricum. Biophys J 91: 3707–3717 Scheuring S, Reiss-Husson F, Engel A, Rigaud JL and Ranck JL (2001) High-resolution AFM topographs of Rubrivivax gelatinosus light-harvesting complex LH2. EMBO J 20: 3029–3035 Scheuring S, Seguin J, Marco S, Lévy D, Robert B and Rigaud JL (2003) Nanodissection and high-resolution imaging of the Rhodopseudomonas viridis photosynthetic core complex in native membranes by AFM. Proc Natl Acad Sci USA 100: 1690–1693 Scheuring S, Rigaud JL and Sturgis JN (2004a) Variable LH2 stoichiometry and core clustering in native membranes of Rhodospirillum photometricum. EMBO J 23: 4127–413 Scheuring S, Sturgis JN, Prima V, Bernadac A, Lévy D and Rigaud JL (2004b) Watching the photosynthetic apparatus in native membranes. Proc Natl Acad Sci USA. 101: 11293–11297 Scheuring S, Francia F, Busselez J, Melandri BA, Rigaud J-L and Lévy D (2004c) Structural role of PufX in the dimerization of the photosynthetic core complex of Rhodobacter sphaeroides. J Biol Chem 279: 3620–3626 Scheuring S, Busselez J and Lévy D (2005a) Structure of the dimeric PufX-containing core complex of Rhodobacter blasticus by in situ atomic force microscopy. J Biol Chem 280: 1426–1431 Scheuring S, Lévy D and Rigaud JL (2005b) Watching the components of photosynthetic bacterial membranes and their in situ organisation by atomic force microscopy. Biochim Biophys Acta 1712: 109–127 Scheuring S, Gonçalves RP, Prima V and Sturgis JN (2006) The photosynthetic apparatus of Rhodopseudomonas palustris: Structures and organization. J Mol Biol 358: 83–96 Scheuring S, Boudier T and Sturgis JN (2007) From high-resolution AFM topographs to atomic models of supramolecular assemblies. J Struct Biol 159: 268–276 Siebert CA, Qian P, Fotiadis D, Engel A, Hunter CN and Bullough P (2004) The role of PufX in the molecular architecture of photosynthetic membranes in Rhodobacter sphaeroides. EMBO J 23: 690–700 Stroh A, Anderka O, Pfeiffer K, Yagi T, Finel M, Ludwig B and Schägger H (2004) Assembly of respiratory complexes I, III, and IV into NADH oxidase supercomplex stabilizes complex I in Paracoccus denitrificans. J Biol Chem 279: 5000–5007 Sturgis JN and Niederman RA (1996) The effect of different levels of the B800-850 light-harvesting complex on intracytoplasmic membrane development in Rhodobacter sphaeroides. Arch Microbiol 165: 235–242 Takemoto J and Lascelles J (1973) Coupling between bacteriochlorophyll and membrane protein synthesis in Rhodopseudomonas
273
sphaeroides. Proc Natl Acad Sci USA 70: 799–803 Tunnicliffe RB, Ratcliffe EC, Hunter CN and Williamson MP (2006) The solution structure of the PufX polypeptide from Rhodobacter sphaeroides. FEBS Lett. 580: 6967–6971 Van den Berg B, Clemons WM, Collinson I, Modis Y, Hartmann E, Harrison SC and Rapoport TA (2004) X-ray structure of a protein-conducting channel. Nature 427: 36–44 Varga AR and Staehelin LA. (1983) Spatial differentiation in photosynthetic and non-photosynthetic membranes of Rhodopseudomonas palustris. J Bacteriol 154: 1414–1430 VerBerkmoes NC, Shah MB, Lankford PK, Pelletier DA, Strader MB, Tabb DL, McDonald WH, Barton JW, Hurst GB, Hauser L, Davison BH, Beatty JT, Harwood CS, Tabita FR, Hettich RL and Larimer FW (2006) Determination and comparison of the baseline proteomes of the versatile microbe Rhodopseudomonas palustris under its major metabolic states. J. Proteome Res. 5: 287–298 Vos M, van Grondelle R, van der Kooij FW, van de Poll D, Amesz J and Duysens LMN (1986) Singlet-singlet annihilation at low temperatures in the antenna of purple bacteria. Biochim Biophys Acta 850: 501–512 Vos MH, van Dorssen RJ, Amesz J, van Grondelle R and Hunter CN (1988) The organisation of the photosynthetic apparatus of Rhodobacter sphaeroides: Studies of antenna mutants using singlet-singlet quenching. Biochim Biophys Acta 933: 132–140 Walz T, Jamieson SJ, Bowers CM, Bullough PA and Hunter CN (1998) Projection structures of three photosynthetic complexes from Rhodobacter sphaeroides: LH2 at 6Å, LH1 and LH1-RC at 25Å. J Mol Biol 282: 833–845 Wang ZY, Suzuki H, Kobayashi M and Nozawa T (2007) Solution structure of the Rhodobacter sphaeroides PufX membrane protein: Implications for the quinone exchange and proteinprotein Interactions. Biochemistry 46: 3635–3642 Westerhuis WHJ, Xiao Z and Niederman RA (1992) Oligomerization-state dependent spectroscopic properties of the B850 light-harvesting complex of Rhodobacter sphaeroides R-26.1. In Murata N (ed) Research in Photosynthesis, Vol. 1, pp 93–96. Kluwer Academic Publishers, Dordrecht Westerhuis WHJ, Vos M, van Grondelle R, Amesz J and Niederman RA (1998) Altered organization of light-harvesting complexes in phospholipid-enriched Rhodobacter sphaeroides chromatophores as determined by fluorescence yield and singlet-singlet annihilation measurements. Biochim Biophys Acta 1366: 317–329 Westerhuis WHJ, Sturgis JN, Ratcliffe EC, Hunter CN and Niederman RA (2002) Isolation, size estimates, and spectral heterogeneity of an oligomeric series of light-harvesting 1 complexes from Rhodobacter sphaeroides. Biochemistry 41: 8698–8707 Young CS and Beatty JT (2003) Multi-level regulation of purple bacterial light-harvesting complexes. In: Green BR and Parson WW (ed) Light-harvesting antennas in photosynthesis (Advances in Photosynthesis and Respiration, Vol 13), pp 449–470. Kluwer Academic Publishers, Dordrecht
Chapter 15 From Atomic-Level Structure to Supramolecular Organization in the Photosynthetic Unit of Purple Bacteria Melih K. ùener* and Klaus Schulten* Beckman Institute, University of Illinois at Urbana-Champaign, Urbana, IL 61801, U.S.A.
Summary ............................................................................................................................................................... 275 I. Introduction..................................................................................................................................................... 276 II. Components of the Photosynthetic Unit ......................................................................................................... 278 III. Quantum Physics of Light-Harvesting and Excitation Energy Transfer.......................................................... 279 A. Effective Hamiltonian Formulation of Electronic Excitations ............................................................ 279 B. Excitation Energy Transfer in a Pigment Assembly ......................................................................... 280 C. Detrapping and Sharing of Excitations and the Sojourn Expansion................................................. 281 IV. Effect of Thermal Disorder on the Spectra of Light-Harvesting Complexes ................................................... 282 V. Physical Constraints Shaping the Structure of Individual Light-Harvesting Complexes................................. 284 VI. Supramolecular Organization of the Photosynthetic Unit ............................................................................... 286 VII. An Atomic-level Structural Model for a Photosynthetic Membrane Vesicle.................................................... 287 VIII. Light-Harvesting and Excitation Transfer across a Photosynthetic Membrane Vesicle ................................. 287 IX. Conclusions .................................................................................................................................................... 289 Acknowledgments ................................................................................................................................................. 290 References ............................................................................................................................................................ 290
Summary The purple bacterial photosynthetic unit (PSU) is a macromolecular assembly of remarkable simplicity that harvests sunlight with the cooperation of only half a dozen different kinds of proteins. This chapter provides a summary of recent research on the architectural and biophysical aspects of the PSU and its constituents. First, a brief overview is provided of the structure of light-harvesting components. Then the effects of thermal disorder and spectral universality on the light-harvesting function of the pigment-protein complexes is discussed, followed by an account of the physical constraints that shape the evolution of light-harvesting complexes in general. Finally, a summary is provided of recent research on the in silico assembly of an entire PSU in atomic detail. This supramolecular reconstruction of the PSU is made possible by the recent availability of not only the structural data on the individual constituent proteins but also on their global arrangement. The reconstruction is performed by combining data from X-ray crystallography, nuclear magnetic resonance, cryo-electron microscopy, and atomic force microscopy using computational modeling. The architecture of the PSU vesicle that emerges constitutes nearly two hundred light-harvesting proteins, containing around four thousand chlorophylls, which act cooperatively to maintain a very high quantum yield in a pigment array distributed over a pseudo-spherical intracytoplasmic membrane domain with an inner diameter of 60 nm.
*Authors for correspondence, email: [email protected], [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 275–294. © 2009 Springer Science + Business Media B.V.
276 I. Introduction Photosynthetic organisms, and in turn almost all terrestrial life, are powered by the fixation of light energy in the form of chemical bonds. This stabilization and storage of energy is rarely performed by individual chlorophylls (Chls). As suggested already by the work of Emerson and Arnold (1932), up to thousands of Chls distributed over hundreds of proteins cooperate by forming a PSU which converts light energy into increasingly more stable forms usable by the cell. This is described in several reviews (Xu and Schulten, 1992; Hu et al., 1997; Ritz et al., 1998; Schulten, 1999; van Amerongen et al., 2000; Govindjee, 2000; Ritz and Schulten, 2001; Blankenship, 2002; Hu et al., 2002; Ritz et al., 2002; Şener and Schulten, 2005; Cogdell et al., 2006; Chapter 13, van Grondelle and Novoderezkhin; Chapter 14, Sturgis and Niederman). In a PSU, light energy initially migrates via radiationless excitation transfer (Oppenheimer, 1941; Förster, 1948; Arnold and Oppenheimer, 1950) from light-harvesting complexes to a reaction center (RC) initiating a charge separation (Marcus, 1956a,b; Onuchic et al., 1986) across the membrane, which is later utilized for ATP synthesis (Junge et al., 1997; Fillingame, 2000; Fillingame et al., 2000; Chapter 24, Feniouk and Junge). The structure and function of the biomolecules that constitute the PSU have been studied over many decades. Elucidation of individual aspects of light-harvesting in the PSU has brought Nobel prizes to Wilstätter (1915) and Fischer (1930) for the purification and the subsequent study of the chemical properties of Chls; to Karrer (1937) and Kuhn (1938) for determining the structure and function of carotenoids; to Deisenhofer, Huber, and Michel (1988) (Deisenhofer et al., 1985) for crystallographic determination of the RC structure; and to Marcus (1992) (Marcus, 1956a,b) for his theory of electron transfer reactions. The availability of high resolution atomic structures of the constituent light-harvesting proteins is essential for our current molecular level understanding of photosynthesis. In purple bacteria, the PSU is built from a small number of different protein types. In the order of Abbreviations: AFM – atomic force microscopy; bc1 complex – ubiquinol-cytochrome c2 oxidoreductase; BChl – bacteriochlorophyll; Chl – chlorophyll; cryo-EM – cryo-electron microscopy; LD – linear dichroism; LH1 – light-harvesting complex 1; LH2 – light-harvesting complex 2; PS I – Photosystem I; PSU – photosynthetic unit; RC – reaction center
Melih K. ùener and Klaus Schulten energy flow, these are the peripheral light-harvesting complex, LH2, the core light-harvesting complex, LH1 (McDermott et al., 1995; Karrasch et al., 1995; Hu et al., 1995; Koepke et al., 1996; Papiz et al., 2003; Roszak et al., 2003; Chapter 8, Gabrielsen et al.; Chapter 9, Bullough et al.), and the RC (Deisenhofer et al., 1985; Allen et al., 1987; Ermler et al., 1994; Camara-Artigas et al., 2002). The RC reduces quinone to hydroquinone as a result of electron transfer; subsequently the ubiquinol-cytochrome c2 oxidoreductase (bc1 complex) (Crofts et al., 1983; Velasco and Crofts, 1991; Gennis et al., 1993; Xia et al., 1997; Chapter 22, Berry et al., Chapter 23, Kramer et al.) oxidizes hydroquinone and reduces cytochrome c2, which completes the cycle by shuttling electrons back to the RC (Chapter 17, Axelrod et al.). The ATP-synthase (Junge et al., 1997; Fillingame et al., 2000; Fillingame, 2000; Chapter 24, Feniouk and Junge) makes use of the resulting proton gradient. The purple bacterial PSU displays a certain simplicity in contrast to its counterpart in plants and cyanobacteria. The components of the oxygenic light-harvesting apparatus (Krauss et al., 1993; Krauß et al., 1996; Jordan et al., 2001; Zouni et al., 2001; Ben-Shem et al., 2003; Bibby et al., 2003; Loll et al., 2005; Nelson and Yocum, 2006; Amunts et al., 2007; Murray and Barber, 2007) are notably more complex than their bacterial analogs as they are located higher on the evolutionary tree (Xiong et al., 2000). The structural simplicity displayed by the modular, cylindrically symmetric purple bacterial light-harvesting proteins also brings about a lower packing efficiency in terms of pigment density (see Fig. 1). The RC-LH1 complex contains 1 bacteriochlorophyll (BChl) per 72 amino acids and LH2 contains 1 BChl per 33 amino acids, whereas the cyanobacterial Photosystem I (PS I) contains 1 Chl per 27 amino acids (PS I is better compared to a RC-LH1 complex with two accompanying LH2 complexes, which together have approximately 1 BChl per 40 amino acids). However, the quantum efficiency itself is not strongly affected by the packing density and is nearly unity for both systems due to the Förster radius, defined as the distance over which energy transfer efficiency reduces by half, being much larger than the typical inter-(B)Chl separation. Quantum mechanics is fundamentally important for the description of photosynthesis as it is necessary to portray the interaction of light with living matter and the dynamics of the electronically excited quantum states of the pigments (van Grondelle et
Chapter 15
Supramolecular Organization of the Photosynthetic Unit
277
Fig. 1. Structural hierarchy and comparison of light-harvesting components. The modular architectures of the PSU light-harvesting complexes are readily visible from the corresponding structures. Three BChl molecules (illustrated in (a) in terms of the porphyrin ring) and two rhodopin glucoside molecules (rendered in the following in licorice representation in black and gray, respectively) accompany each α-β subunit (b), nine of which join to form the LH2 complex of Rps. acidophila, shown in side view (from the plane of the cytoplasmic membrane) in (c) and in top view in (d) (Papiz et al., 2003) (PDIB: 1NKZ). The same modularity is seen in the (pseudo-)cylindrical LH1 complex surrounding the RC shown in (e). A structural model for the S-shaped RC-LH1 dimer of Rba. sphaeroides based on cryo-EM images (Qian et al., 2005) was constructed in Şener et al. (2007), shown in (f) (BChls only). Finally, comparison with the cyanobacterial PS I (g) shows that evolutionarily more advanced photosynthetic species adapted a higher packing density at the expense of modularity.
al., 1994; Mukamel, 1995; Hu et al., 1997; Treutlein et al., 1988a,b; Schulten and Tesch, 1991; Nonella and Schulten, 1991; Xu and Schulten, 1994; Ritz et al., 1998; Hu et al., 1998; Hu and Schulten, 1998; Krueger et al., 1998; Sundström et al., 1999; Scholes et al., 1999; Tretiak et al., 2000; Damjanović et al., 2000a,b; Ritz et al., 2000; Şener and Schulten, 2002; Park et al., 2003; Şener et al., 2002, 2004,
2005, 2007; Damjanović et al., 2002a,b; Kosztin and Schulten, 2008; Chapter 13, van Grondelle and Novoderezhkin). This is further complicated by the presence of thermal disorder. Ordinary zero-temperature quantum theory is not sufficient to account for the spectral properties of the pigment complexes at a finite temperature. Thermal effects must be accounted for in terms of dynamic (Damjanović et al.,
278 2002a; Kosztin and Schulten, 2008) or static (Şener and Schulten, 2002) disorder models. The physics that governs the light-harvesting process, including the effects of finite temperature disorder, provides integral design constraints that appear to have shaped the evolution of pigment-protein complexes (Şener and Schulten, 2005). A combination of structural data with physicsbased computational models can accurately portray the light-harvesting function of a pigment-protein complex as discussed in the subsequent sections. However, a greater challenge lies in the description of hundreds of proteins that cooperate to form a functional unit. The supramolecular organization of the constituent proteins is difficult to determine and must be ascertained by combining atomic force microscopy (AFM) (Bahatyrova et al., 2004; Scheuring et al., 2004, 2005; Fotiadis et al., 2004), cryo-electron microscopy (cryo-EM) (Jungas et al., 1999; Jamieson et al., 2002; Siebert et al., 2004; Qian et al., 2005), and linear dichroism (LD) (Frese et al., 2004) studies. Furthermore, the sheer size of the resulting molecular aggregate provides entirely new challenges for computational models. The intracytoplasmic vesicle which forms the PSU is a pseudo-spherical region of an inner diameter of approximately 60 nm. The energy transfer that follows light absorption must be described as a partially coherent and partially incoherent quantum mechanical process over this large domain containing thousands of BChls (Şener et al., 2007). The organization of this review is as follows: in the next section we discuss the structures of the proteins that constitute the PSU with emphasis on the pigment-protein complexes; the quantum mechanical aspects of light-harvesting and excitation transfer processes across a pigment-assembly, including one distributed over multiple proteins, are outlined in Section III. The effects of thermal disorder on the spectral properties of a pigment array are discussed in Section IV followed, in Section V, by a summary of the physical constraints that shape the structure of a light-harvesting complex; Section VI discusses the experimental data and theoretical models on the supramolecular organization of the PSU which is brought together in an atomistic model of an entire vesicle in Section VII. Finally, an account of the quantum mechanical process of light-harvesting and energy transfer across the resulting photosynthetic vesicle is given in Section VIII.
Melih K. ùener and Klaus Schulten II. Components of the Photosynthetic Unit Studies of light-harvesting components of the purple bacterial PSU currently benefit from the availability of a large body of structural information (McDermott et al., 1995; Koepke et al., 1996; Conroy et al., 2000; Camara-Artigas et al., 2002; Papiz et al., 2003; Fotiadis et al., 2004). Since it is beyond the scope of this manuscript to provide a comparison of all related structural data, the reader is instead referred to a recent review (Cogdell et al., 2006) and other chapters in this volume (Chapter 8, Gabrielsen et al.; Chapter 9, Bullough et al.). In the following, we present the structures and structural models for LH2, LH1 and RC complexes of the Rhodobacter (Rba.) sphaeroides PSU. These structures are of integral importance for the atomic-detail reconstruction of an entire intracytoplasmic membrane described in Section VII. A notable common feature of all PSU light-harvesting complexes is the modularity of their structure. A (pseudo-)cylindrical arrangement of α−β subunits (a pair of transmembrane helices) which non-covalently bind BChls and carotenoids forms the building blocks of both LH1 and LH2 (Chapter 10, Loach and ParkesLoach; Chapter 46, Braun and Fiedor). For instance, in Rhodospirillum (Rsp.) molischianum the pigmentprotein complexes display an eight-fold symmetry (Koepke et al., 1996) whereas in Rhodopseudomonas (Rps.) acidophila (McDermott et al., 1995; Papiz et al., 2003) and Rba. sphaeroides (Walz et al., 1998) the symmetry is nine-fold. A collection of the relevant light-harvesting complexes is displayed in Fig. 1. This modular construction is in great contrast to the analogous cyanobacterial and plant systems (Jordan et al., 2001; Bibby et al., 2003; Ben-Shem et al., 2003; Loll et al., 2005; Nelson and Yocum, 2006; Amunts et al., 2007; Murray and Barber, 2007) which show a greater degree of complexity (See Fig. 1g). The symmetry of the pigment-protein complexes is subtly broken in many cases. Notably, a small polypeptide, PufX, that likely plays a role in quinone traffic in and out of the RC, constitutes a discontinuity on the LH1 ring surrounding the RC. In Rba. sphaeroides RC, LH1, and PufX form S-shaped dimeric supercomplexes as observed in negative stain EM of intact membranes and cryo-EM projection maps of purified complexes (Jungas et al., 1999; Frese et al., 2000, 2004; Siebert et al., 2004; Qian et al., 2005; Scheuring and Sturgis, 2005). There are currently no atomic structures available for the RC-
Chapter 15
Supramolecular Organization of the Photosynthetic Unit
LH1-PufX supercomplex. A structural model of the pigments based on cryo-EM data (Qian et al., 2005) was constructed and described in Şener et al. (2007) as shown in Fig. 1(f). Recent data indicate that the RC-LH1-PufX supercomplex is actually bent out of the plane in a manner consistent with a curvature which corresponds to that of the spherical invaginations of the intracytoplasmic membrane (Chapter 9, Bullough et al.). For every RC there are about five to ten LH2 units in a typical PSU. In Rba. sphaeroides, the LH2 contains a total of 27 BChls, 18 of which form the strongly coupled B850 band whereas the remaining 9 form the weakly coupled B800 band. The stronger coupling of the B850 BChls is a direct reflection of the corresponding inter-pigment distances (around 10 Å vs. 20 Å). The stronger coupling results in delocalization of B850 excitonic states over the entire ring, which subsequently is broken by the effects of thermal disorder (Şener and Schulten, 2002; Damjanović et al., 2002a; Kosztin and Schulten, 2008). In LH1, excitonically coupled BChls absorb at 875 nm. In addition to the light-harvesting complexes, LH2, LH1, and RC, the purple bacterial PSU contains biomolecules which convert electronic excitation energy to a charge gradient and finally to stable chemical bonds. Upon receiving an electronic excitation the RC initiates a transmembrane electron transfer reducing the membrane-diffusible electron carrier quinone to hydroquinone. Subsequently, hydroquinones diffuse through the cell membrane to the cytochrome bc1 complex (Crofts et al., 1983; Velasco and Crofts, 1991; Gennis et al., 1993; Xia et al., 1997), which in turn oxidizes hydroquinone by transferring the electrons to the cytochrome c2 complex while pumping protons across the membrane. The proton gradient, thus created, is used for ATP synthesis by ATP synthase (Junge et al., 1997; Fillingame et al., 2000; Fillingame, 2000). The circuit is completed by the cytochrome c2 which shuttles the electrons back to the RC resetting the system. With the exception of cytochrome c2 all aforementioned components are embedded in the membrane. The components responsible for the electron transfer reactions or the process of ATP synthesis are not considered further in this review as the vesicle construction in Section VII concentrates explicitly on light-harvesting complexes only. This focus also reflects the absence of the bc1 complex and ATP synthase from AFM data of the membranes (Bahatyrova et al., 2004; Scheuring et al., 2004, 2005; Frese et al., 2004; Fotiadis et al.,
279
2004) which renders it impossible to unambiguously assign locations for these complexes in the membrane (see Section VI). III. Quantum Physics of Light-Harvesting and Excitation Energy Transfer A. Effective Hamiltonian Formulation of Electronic Excitations Interaction of light with biomolecules is necessarily described by quantum theory. Past research has shown that the biophysical processes involved in photosynthesis can be accurately modeled in atomic detail at the quantum mechanical level (Hu et al., 1997, 1998; Ritz et al., 1998; Hu and Schulten, 1998; Sundström et al., 1999; Scholes et al., 1999; Tretiak et al., 2000; Şener and Schulten, 2002; Şener et al., 2002, 2004, 2005, 2007; Damjanović et al., 2002a; Kosztin and Schulten, 2008). Central to these formulations is the construction of an effective Hamiltonian for the light-harvesting pigment assembly in terms of electronic excited states. In the simplest case, such a Hamiltonian is given in terms of the lowest excited state of the BChl molecule, the Qy state (Scheer, 1991). Therefore, a set of basis states for an effective Hamiltonian is given by | i 〉 = |φ1 : φ2 : … φ*i : … : φN〉 in which the ith BChl is Qy-excited and all other BChls are in their ground states. In this basis the effective Hamiltonian is given by (Hu et al., 1997, 1998; Ritz et al., 1998; Damjanović et al., 1999; Tretiak et al., 2000; Damjanović et al., 2000a; Ritz et al., 2001; Şener and Schulten, 2002; Şener et al., 2002) ⎛ ε1 V12 ⎜V ε2 H = ⎜ 21 ⎜L L ⎜ ⎝ VN 1 VN 2
V13 V23 L VN 3
L V1N ⎞ L V2 N ⎟ ⎟ L L⎟ ⎟ L εN ⎠
(1)
where N is the total number of BChls, εi denotes the site energy of excited BChl i, and Vij are the interBChl electronic couplings. For most typical interBChl distances the so-called excitonic couplings Vij are dominated by the direct Coulomb contribution (Förster, 1948) rather than the electron exchange term (Dexter, 1953). For BChls that are sufficiently far apart the couplings Vij can be approximated by the induced dipole-induced dipole interactions of the BChl Qy
Melih K. ùener and Klaus Schulten
280 states, which are given by
(
)(
⎛ d ⋅d 3 r ij ⋅ d i r ij ⋅ d j i j Vij = C ⎜ − 3 ⎜⎝ rij rij5
) ⎞⎟ ⎟⎠
between BChls i and j is given by (Förster, 1948; Ritz et al., 2001; Şener et al., 2002) (2)
where di is the unit vector along the transition dipole moment of the Qy state of BChl i, and C is a constant determined empirically. The value of the constant C has been taken to be 348000 Å3 cm–1 in order to reproduce LH2 exciton spectra (Knox and Spring, 2003; Şener and Schulten, 2002; Şener et al., 2007). As the inter-pigment distances approach 10 Å or less, the dipole-dipole approximation becomes increasingly inaccurate, requiring the evaluation of the couplings by quantum chemistry methods (Cory et al., 1998; Tretiak et al., 2000; Şener et al., 2002). Specifically, the nearest neighbor couplings of the LH2 B850 ring are taken to be 363 cm–1 and 320 cm–1 (Tretiak et al., 2000; Şener and Schulten, 2002); the nearest neighbor couplings of the LH1 B875 ‘S-band’ are taken to be 300 cm–1 and 233 cm–1 (Koolhaas et al., 1998; Damjanović et al., 2000a). The intra- and inter-subunit couplings alternate in both proteins because of the modular structure. The special pair BChl coupling is taken to be 500 cm–1 (Eccles et al., 1988; Damjanović et al., 2000a). All other couplings are computed according to the dipole-dipole approximation (Eq. 2) regardless of whether the pigments belong to different proteins. BChl site energies are determined by spectroscopic data as well as taken from earlier models (Eccles et al., 1988; Koolhaas et al., 1998; Groot et al., 1998; Damjanović et al., 2000a; Ritz et al., 2001; Şener and Schulten, 2002). B. Excitation Energy Transfer in a Pigment Assembly The effective Hamiltonian given by Eq. (1) determines on the one hand the spectral properties and on the other hand the energy transfer dynamics of the assembly. The description of spectral properties at finite temperature is discussed in the next section whereas this subsection provides an overview of excitation transfer dynamics according to Förster theory (Förster, 1948; Ritz et al., 2001). In a pigment-protein complex that is under the influence of thermal disorder electronic excitations travel in the form of excitons that are delocalized over only a few pigments. The rate of excitation transfer
Tij =
2 2π Vij J ij ,
J ij = ∫ SiD ( E )S jA ( E ) dE
(3)
where Vij are the electronic couplings introduced in Eq. (1) and Jij is the spectral overlap between the donor emission spectrum SiD(E) and the acceptor absorption spectrum SjA(E), which are in general temperature dependent. Transfer times between nearby BChls are typically in the sub-picosecond range. The probability pi(t) that BChl i is electronically excited at a given time t is governed by a master equation d p (t ) = ∑ K ij p j (t ) dt i j ⎞ ⎛ K ij = T ji − δ ij ⎜ ∑ Tik + k diss + δ i,RC k ET ⎟ ⎠ ⎝ k
(4)
where δi,RC is one if the BChl i belongs to the reaction center and zero otherwise. The dissipation rate of kdiss=(1 ns)–1 and the electron transfer rate of kET = (3 ps)–1 are determined empirically. The average excitation lifetime τ, the quantum yield q, and the dissipation probability d can be solved from the master equation Eq. (4) to yield (Ritz et al., 2001; Şener et al., 2002) τ=−
1 1 K −1 1 N
(5)
q=−
1 k RC K −1 1 N ET
(6)
d=−
1 k diss 1 K −1 1 N
(7)
where |RC 〉 ≡ Σi δi,RC |i〉 and |1〉 ≡ Σi|i〉. The quantum yield q is related to the dissipation probability d by the relation q = 1 – d = 1 – kdissτ. The excitation migration model thus introduced in terms of single-exciton hopping events is insufficient to adequately describe transfer events involving strongly coupled BChl groups such as the B850 BChls of LH2, B875 BChls of LH1, or the special pair BChls in the RC. This model must be superseded
Chapter 15
Supramolecular Organization of the Photosynthetic Unit
by one that involves coherent energy transfer between resonant states of neighboring BChl clusters. Thus, the rate of excitation transfer TDA between a donor BChl cluster D and an acceptor BChl cluster A follows from the assumption that the donor excited states are populated according to the Boltzmann distribution (Ritz et al., 2001; Şener et al., 2004, 2005) − E D /k T
TDA
2 2π e m B = V DA dE S mD ( E )S nA ( E ) ∑ ∑ D m∈D n∈A ∑ e − El /k BT mn ∫ l∈D
(8) where EDm are the donor energy levels; spectral overlaps are computed similar to Eq. (3). In terms of the cluster-transfer rates TDA (Eq. 8), the excitation lifetime τ and the quantum yield q of the system is given by (Ritz et al., 2001; Şener et al., 2002, 2004, 2007) τ=−
1 1 K −1 0 M M cl
q=−
1 k RC K cl −1 0 M ET
(9)
Section VII contains nearly 4000 BChls and therefore is studied within the framework of Förster theory. C. Detrapping and Sharing of Excitations and the Sojourn Expansion An electronic excitation that arrives at a RC does not always initiate a charge transfer event and may instead escape the RC to be recaptured at the same or another RC. This nonzero detrapping probability, which is around 20–30%, results from the excitation back-transfer rate of (8 ps)–1 at the RC being not small compared to the charge separation rate of (3 ps)–1 (Timpmann et al., 1995; Bernhardt and Trissl, 2000; Amesz and Neerken, 2002; Ritz et al., 1998; Şener et al., 2007). Subsequently, the excitation migration process in the PSU can be described in terms of an expansion involving repeated detrapping and retrapping (sojourn) events. This sojourn expansion of the average excitation lifetime (Eq. 9) is based on separating from Kcl the operator ∆cl that describes detrapping events from any of the RCs (Şener et al., 2002, 2004, 2007) K cl ≡ κ cl + ∆ cl M
where |0〉 denotes the initial state and Kcl is the M×M matrix given by ⎞ ⎛ ( K cl )ij = T ji − δ ij ⎜ k ET δ i,RC + k diss + ∑ Tik ⎟ ⎠ ⎝
(10)
k
Here M is the number of BChl clusters. All indices run over clusters and not individual pigments. For the PSU, separate BChl clusters include the B850 and B800 bands for the LH2, B875 band for the LH1, and the RC BChls. Of these groups, the LH2 B800 band is actually not very strongly coupled and is treated as a cluster only for computational simplicity. The Förster theory formulation outlined above becomes increasingly less accurate for strong coupling, fast timescale processes in which case it can be supplemented by the more accurate modified Redfield theory (Yang and Fleming, 2002; Rutkauskas et al., 2005; van Grondelle and Novoderezhkin, 2006; Chapter 13, van Grondelle and Novoderezkhin). However, the computational complexity of the latter method is considerable and its application does not typically go beyond systems containing several dozen pigments. The intracytoplasmic membrane vesicle introduced in
281
∆ cl ≡ ∑
∑T
k =1 j∈RC
jk
k
j
(11)
The subscript ‘cl’ indicates that the matrix indices refer to the BChl clusters rather than individual BChls. The matrix Kcl can be expanded in terms of repeated detrapping events K cl −1 = (κ cl + ∆ cl ) −1 = κ cl −1 − κ cl −1 ∆ cl κ cl −1 + κ cl −1 ∆ cl κ cl −1 ∆ cl κ cl −1 − L
(12) Subsequently, the average excitation lifetime τ is given in terms of the expansion τ = τ 0 + τ1 + τ 2 L
(
∆ cl κ −1 τ k = − 1 κ −1 cl cl
)
k −1
0 ,k >0
(13)
The term τ0 is the first usage time corresponding to an excitation lifetime with no detrapping events. The expansion in Eq. (13) can be expressed in terms of detrapping and cross-transfer probabilities to yield
Melih K. ùener and Klaus Schulten
282 (Şener et al., 2002, 2004, 2007)
(
)
τ = τ 0 + τ soj ⋅ 1 M − Q X ⋅ Q D
(14)
where (τsoj)j is the sojourn time defined as the average excitation lifetime subsequent to an immediate detrapping from RC j, (QD)j is the detrapping probability from the same reaction center, and 1M denotes the identity matrix of dimension M. The term (QX)jk is the conditional cross-transfer probability between the RCs j and k corresponding to a retrapping of excitation at RCj subsequent to a detrapping at RCk: Q jkX = − k ET RC j K cl−1 Tk
(15)
where |Tk〉 is the quantum state resulting from a detrapping event at RCk. In Section VIII these crosstrapping probabilities are presented as a measure of excitation sharing between up to 36 RCs of a PSU vesicle (Şener et al., 2007). IV. Effect of Thermal Disorder on the Spectra of Light-Harvesting Complexes Zero temperature quantum theory is insufficient to describe light-harvesting processes accurately at physiological temperatures. Finite temperature effects are more than a mere modeling challenge for studies of pigment-protein complexes since photosynthetic systems depend on thermal disorder for their function. Thermal broadening of spectral line-shapes enables efficient energy transfer across a greater range of frequencies. For example, the quantum yield in Photosystem I in cyanobacteria and plants (Kennis et al., 2001; Melkozernov, 2001; Melkozernov et al., 2001; Gobets et al., 2001; Şener et al., 2002, 2004; Şener and Schulten, 2005) drops with decreasing temperature from nearly unity at room temperature for a wide range of wavelengths to about 50% or less and becomes wavelength dependent as the excitation runs the risk of being trapped at low energy pigments without reaching the RC. Effects of thermal disorder on spectral properties can be introduced either as temporal averages of the dynamics of one light-harvesting complex or as ensemble averages of multiple complexes. The first approach is referred to as dynamic disorder (Damjanović et al., 2002a; Kosztin and Schulten,
2008) whereas the second is called static disorder (Şener and Schulten, 2002). These two different models of thermal disorder are, in principle, related to one another by virtue of the ergodic theorem. The effect of dynamic disorder on the electronic properties of light-harvesting complexes was studied in Damjanović et al. (2002a) combining molecular dynamics (MD) with quantum chemistry and a polaron model analysis. The exciton-phonon coupling that enters into the polaron model, and the corresponding phonon spectral function were derived from MD and quantum chemistry simulations. It was predicted that excitons in the B850 BChl ring are delocalized over five pigments at room temperature. The polaron model permits the calculation of the absorption and circular dichroism spectra of the B850 excitons from the sole knowledge of the autocorrelation function of the excitation energies of individual BChls (Damjanović et al., 2002a; Kosztin and Schulten, 2008). Static disorder in a thermal ensemble of BChl rings was described in Şener and Schulten (2002) for the B850 band of LH2 in Rsp. molischianum using analytical methods of random matrix theory (Efetov, 1995; Guhr et al., 1998; Şener, 1999). Random matrix models introduce disorder at the level of the effective Hamiltonian formalism by adding a random disorder term to the noise-free, zero-temperature Hamiltonian. The physics of such deterministic-plus-random systems has traditionally been studied analytically in the large N-limit (N being the dimension of the Hamiltonian). However, typical light-harvesting systems are mesoscopic in size (for example, N=16 for the B850 band of LH2 in Rsp. molischianum), resulting in considerable challenges in adopting readily available analytical methods. Thermal disorder models of light-harvesting complexes aim to compute spectral quantities such as the absorption spectrum αT (ω) and the density of states ρT (ω) at a finite temperature T as a function of the zero-temperature spectral properties of the Hamiltonian (Eq. 1). In terms of the effective Hamiltonian in Eq. (1), the density of states and the directionally averaged absorption spectrum are defined by N
ρT (ω) =
∑ δ(ω − E )
(16)
i
i=1
α T (ω) =
4π 2 ωn 3c
∑D i
2 i
δ(ω − Ei )
(17)
Chapter 15
Supramolecular Organization of the Photosynthetic Unit
where Di is the transition dipole moment for the eigenstate i and the ensemble averages are defined with respect to a temperature dependent weight function. A simple approach to introducing thermal disorder is the Gaussians-on-sticks picture of spectral quantities which ‘dress’ the discrete spectrum N
ρ0 (ω) = ∑ δ(ω − Ek ) k =1
of the effective Hamiltonian H of the pigment complex with Gaussians: N
ρT (ω) : ∑ k =1
(ω − E ) exp(− k
2υ 2
2
)
283
describing the overall width of disorder and is mostly independent of the shape and the total number of the disorder terms Ri (Efetov, 1995; Guhr et al., 1998; Şener, 1999; Şener and Schulten, 2002) (see Fig. 2). Spectral universality becomes more accurate as the system size N increases and becomes exact for N→∞. This result is conceptually similar to the ubiquity of the central limit theorem which establishes that summing up random variables will generally yield a Gaussian distribution regardless of the shape of the original random function. Spectral universality is the reason why many diverse and disjoint disorder models succeed in modeling bulk thermal effects in light-harvesting systems whereas they are bound to fail to distinguish various forms of disorder from one another. Since spectral universality theorems of random
where the temperature dependent width υ is determined heuristically. Unfortunately, this approach becomes highly inaccurate as multiple energy levels become near degenerate (as compared to the disorder parameter υ). This is because the energy levels Ek(T) of a disordered quantum system actually begin to separate as a result of the avoided level crossings during the onset of disorder (Efetov, 1995; Andreev et al., 1996; Guhr et al., 1998). Another approach for treating static disorder is to numerically diagonalize an ensemble of effective Hamiltonians containing noise terms H T = H 0 + R1 + R2 + L
(18)
where the random terms Ri (in the simplest case only a single term R) are chosen from ensembles of matrices with given weight functions. For example, site energy fluctuations may be assigned as diagonal disorder terms whereas thermal fluctuations of interpigment couplings are introduced as off-diagonal disorder terms (Cogdell et al., 2006). The shapes of the disorder terms Ri are typically taken as Gaussians, and their widths are determined empirically. However, as a result of the spectral universality theorems of random matrix theory (Guhr et al., 1998), different ‘shapes’ of disorder (such as diagonal versus off-diagonal disorder) are largely indistinguishable from one another in terms of their effects on spectra. Thus, the average spectrum of the random ensemble in Eq. (18) is dependent largely on a single parameter
Fig. 2. Spectral universality under different classes of disorder. The density of states (a) (Eq. 16) and the absorption spectrum (b) (Eq. 17) for the LH2 B850 band of Rsp. molischianum are shown for three different forms of disorder (Şener and Schulten, 2002). Real-symmetric (orthogonal) disorder: solid line, complex-hermitian (unitary) disorder: dashed line, diagonal disorder: dotted line. Only the zero-temperature spectrum and the width of each disorder distribution determines the bulk shape of the finite temperature spectral properties, not the class of disorder. The approximate agreement between different forms of disorder shown here becomes more pronounced as the system size N increases.
Melih K. ùener and Klaus Schulten
284 matrix theory (Guhr et al., 1998) imply that various forms of disorder result in the same spectral shape, it is convenient to compute spectral quantities for the mathematically simplest of disorder models. Analytical results are most easily obtained for unitary disorder (Guhr et al., 1998). In Şener and Schulten (2002) the finite temperature density of states of a system of arbitrary size N was analytically evaluated. The result can be expressed in a closed and compact, if combinatorially complex, form in terms of the system size N and the eigenvalues Ei (Şener and Schulten, 2002): ρT ( ω ) =
∞
1 dα A0 A1 π ∫0 ∞
A0 =
N 1 Im ∫ dξ e -ξ Π i=1 π -∞
∞
A1 =
1
2
∫ dη e
−∞
− η2
N ⎛ Π⎜ η − i=1 ⎝
ξ − iε − 1 2υ T
(ω + υ α − E )
1
2 T
2υ T
i
⎞
( ω + υ α − E )⎟ 2 T
i
⎠
(19) Here the only free parameter is the disorder term υT corresponding to the overall width of disorder.
An effective field theory approximation to Eq. (19) can be introduced for systems of larger N where the combinatorial cost to evaluate this analytical expression becomes high. None of the aforementioned computational methods for modeling thermal disorder effects has so far been applied to systems containing more than a handful of pigment-protein complexes. It remains a challenge to properly account for thermal effects for a system as large as the PSU vesicle consisting of up to 200 light-harvesting proteins and four thousand pigments as presented in Section VII. V. Physical Constraints Shaping the Structure of Individual Light-Harvesting Complexes A comparison of pigment-protein complexes from both oxygenic and anoxygenic species reveals a number of common design principles dictated by the physics of the light-harvesting process shaping the evolution of the related proteins. These principles, which are also relevant for engineering artificial
light-harvesting systems, are briefly summarized below; see also the reviews of Şener and Schulten (2005); Cogdell et al. (2006); Chapter 10, Loach and Parkes-Loach; Chapter 46, Braun and Fiedor. A comparison is also provided between the purple bacterial photosynthetic apparatus and the oxygenic systems in terms of their architectural contrast. The overall efficiency of a light-harvesting system is constrained by any bottlenecks in the sequence of processes leading to the storage of energy in the form of chemical bonds. Therefore, photosynthesis needs to be efficient at all the steps of photon absorption, excitation migration, and charge transfer. As discussed in the previous section thermal broadening of line shapes increases both the range of frequencies over which photons are absorbed as well as the efficiency of resonant energy transfer between neighboring pigments. It is not a coincidence that the spread of site energies in the PSU of around 230 cm–1 (Sumi, 2000, 2001; Cogdell et al., 2006) is of the same magnitude as kBT (for T=300 K). A funnel-like arrangement of BChl energies (B800 → B850 → B875) facilitates a downhill transfer of energy in the PSU. However, a comparison with Photosystem I (Damjanović et al., 2002b; Şener et al., 2002) reveals that an energy funnel toward the RC is not necessary for a high quantum yield at room temperature. In fact, the socalled red Chl pool of Photosystem I (Pålsson et al., 1998; Gobets et al., 2001; Zazubovich et al., 2002), which constitutes a trap for excitation at cryogenic temperatures, extends the functionality of light-harvesting into longer wavelengths without a noticeable loss of efficiency. It follows from Eq. (3) that efficient transfer of energy between pigments requires, in addition to a strong spectral overlap Jij, a strong inter-coupling coupling Vij. This is achieved by inter-pigment separations that are much smaller than the Förster radius. Thus, a high pigment density facilitates more efficient excitation transfer. As mentioned in the Introduction, the pigment density increases in the evolutionarily more advanced plant and cyanobacterial light-harvesting complexes. The modular structure of the purple bacterial light-harvesting complexes and the implication of their symmetry on the nature of excitonic quantum states appear to be largely irrelevant when thermal disorder effects are taken into account. Exciton bands that would be delocalized over the entire ring at zerotemperature are localized over a handful of pigments (Damjanović et al., 2002a) at physiological tempera-
Chapter 15
Supramolecular Organization of the Photosynthetic Unit
tures. Thus modularity was likely a structural artifact at the early stages of photosynthesis evolution that was superseded later. A similar scenario of modularity based on gene-duplication was also proposed for the early evolution of Photosystem I (Nelson and BenShem, 2004; Kosztin and Schulten, 2008). A notable exception to the aforementioned demand on high pigment density is the immediate vicinity of the RC. Both bacterial and plant systems display a no-fly zone (also called a cordon sanitaire) referring to an absence of pigments within about 20 Å of the electron transfer chain (see Fig. 3). This is to avoid the loss of transported electrons along the electron transfer chain. Electron transfer requires a direct overlap of electronic wavefunctions and therefore direct spatial proximity between pigments. The electron transfer rate decays exponentially with distance. In contrast, excitation transfer decays as 1/r6 where r is the pigment-pigment distance. Thus, with a ‘no-fly zone’-architecture it is possible to maintain efficient excitation transfer into the RC without having a substantial loss of electrons away from the electron transfer chain. Not surprisingly, the RC is the most demanding structural component of the light-harvesting assembly. For example, photosystems I and II in cyanobacteria have developed special mechanisms to deal with iron deficiency which poses a challenge on the synthesis and assembly of new RCs (Bibby et al., 2001a,b, 2003). Reflecting this prominence of the RC, photosynthetic systems including the PSU have evolved to regulate the number of pigments per RC in a way to avoid keeping the RCs idle. Thus, the total number of BChls in the PSU times the light intensity (in terms of photons/BChl/second) roughly equals the charge separation rate times the number of RCs. Therefore, growth conditions, especially the light intensity, affect the stoichiometry of the PSU (Şener et al., 2007). Oxygenic and anoxygenic light-harvesting systems display differences in terms of the optimization of their pigment architectures. A robustness and optimality of Chl geometries and site energies was reported for cyanobacterial and plant photosystems I and II (Şener et al., 2002, 2004; Damjanović et al., 2002b; Yang et al., 2003; Vasil’ev and Bruce, 2004; Vasil’ev et al., 2004). Robustness is manifested in the form of a parameter insensitivity, enabling a high quantum yield irrespective of site energy fluctuations resulting from thermal disorder or modifications to the geometry of the pigment network. Another aspect of robustness is
285
Fig. 3. Exclusion of chlorophylls from the immediate vicinity of the RC. Both in anoxygenic (a; RC-LH1) and oxygenic (b; PS I) light-harvesting systems a ‘no-fly-zone’ surrounds the RC, where the presence of pigments that are not part of the electron transfer chain is excluded to prevent the escape of electrons subsequent to the trapping of excitation.
graceful degradation, referring to a tolerance for the loss of individual components. Removing a pigment at random from the network usually does not have a substantial effect on the overall efficiency. Graceful degradation provides a protective buffer against photodamage and constitutes part of the mechanism that photosynthetic systems have evolved to handle harmful effects of excess light energy. Optimality, on the other hand, indicates that within the narrow range (near unity) of quantum yields corresponding to different pigment geometries and site energies, the one corresponding to the actual pigment architecture
286 found in nature is nearly the highest. This is likely a result of the selective pressure during the evolution projecting itself on to one of the dimensions of the fitness landscape (the quantum yield). It is important to note that while the aforementioned robustness, which follows largely from the high pigment density as compared with the Förster radius, is universally present in oxygenic and anoxygenic systems alike, the signs of optimality of the pigment geometry are only prominent for the oxygenic photosynthetic systems. No comparable optimization of BChl geometries was reported among the modular architectures of the evolutionarily simpler purple bacterial PSU. It is possible that the circular, modular arrangement of pigments in LH2 and LH1 is simply an artifact of the simplest feasible assembly scheme, namely one based on identical subunits. Thermal disorder at physiological temperatures likely overrides any quantum mechanical concerns which would be valid at zero temperature. The regular arrangement of pigments is apparently traded for higher pigment density at higher tiers of evolution. VI. Supramolecular Organization of the Photosynthetic Unit Structure and function of individual light-harvesting proteins discussed above must be reconciled with the global architecture of the PSU before system-level functional models can be constructed that connect multiple steps of the light-harvesting process. In this section and the next a computational model for a PSU vesicle (Şener et al., 2007) is presented that contains hundreds of light-harvesting proteins based on AFM, cryo-EM, LD, and X-ray crystallography data. EM of negatively stained thin sections of photosynthetically grown cells shows that the PSU of Rba. sphaeroides forms pseudo-spherical intracytoplasmic protrusions with an inner diameter of approximately 60 nm. Radiolabeling, biochemical, and spectroscopic techniques reveal that it is the clustering of RC-LH1-PufX and LH2 pigment-protein complexes combined with lipid biosynthesis that is responsible for the membrane curvature causing the PSU to invaginate (Niederman et al., 1979; Hunter et al., 1985; Pugh et al., 1998; Hunter et al., 2005). This curvature-inducing assembly process, likely akin to the clustering of heavy beads on an elastic surface, has been modeled recently using an application of colloid theory (Frese et al., 2008) as well as molecular
Melih K. ùener and Klaus Schulten dynamics simulations (Chandler et al., 2008). See Chapter 9, Bullough et al. for further discussion of membrane curvature. In a PSU vesicle up to two hundred light-harvesting complexes containing around three to four thousand chlorophylls surround approximately 30 RCs (Hunter et al., 1985). Recent AFM data (Bahatyrova et al., 2004; Scheuring et al., 2004, 2005) reveal that the RC-LH1-PufX dimers form linear arrays that are surrounded by a ‘lake’ of LH2s. Notably, neither the bc1 complex nor the ATP synthase has been observed directly in AFM images. A likely explanation of the absence of the bc1 complexes from AFM data is that they may populate the neck region of the vesicle which is not imaged by AFM. In fact, noninvaginated cytoplasmic membrane regions not only lack photosynthetic complexes, but are also enriched in respiratory components (Parks and Niederman, 1978). ATP synthase is expected to be found in most, but not necessarily all, vesicles at a ratio of about one per vesicle; some vesicles possibly lack a directly associated ATP synthase complex (Feniouk et al., 2002) which argues that ATP synthase is located outside of the bulbous chromatophore. The stoichiometry of the bc1 complex was suggested earlier to be one bc1 complex per two RCs (Crofts et al., 1983) and nine bc1 complexes per vesicle (Velasco and Crofts, 1991). However, a recent model of electron transport components concludes about five bc1 complexes per vesicle (Geyer and Helms, 2006). The precise stoichiometry of the PSU depends on growth conditions, especially the light intensity. While there is not sufficient information to unambiguously model the electron transport components, a model of the light-harvesting components of the PSU vesicle can be constructed by combining a multitude of AFM images. This construction requires, firstly, the mapping of the planar domains containing LH2 and RC-LH1-PufX complexes on to a spherical region. This plane-to-sphere mapping process is outlined in the next section (Snyder, 1987). Subsequently, the individual LH-domains must be oriented with respect to one another and to the cytoplasmic membrane. Adopting a directional convention that the vesicle is connected at its ‘south pole’ to the cell membrane, LD data (Frese et al., 2004) and the observed kinetics of LH2 incorporation (Niederman et al., 1976, 1979; Hunter et al., 2005) suggest that the RC-LH1-PufX dimers are oriented in a north-south alignment. This is also consistent with an aggregation model of the vesicle where the addition of new RC-LH1-PufX
Chapter 15
Supramolecular Organization of the Photosynthetic Unit
dimers with accompanying LH2s deepens the invagination toward the cytoplasm. In fact, elongated RC-LH1-PufX dimer arrays were reported in mutants lacking LH2 (Jungas et al., 1999; Frese et al., 2000; Siebert et al., 2004). VII. An Atomic-level Structural Model for a Photosynthetic Membrane Vesicle In this section, the spatial arrangement of the lightharvesting components discussed above are combined with the atomic structures of individual complexes to provide a structural model for a PSU vesicle at atomic-detail level. There are several difficulties associated with combining planar AFM images (Bahatyrova et al., 2004) to produce a pseudo-spherical chromatophore model. First, the preparation process for AFM imaging tears and distorts the vesicle patches during their flattening. This process is not area-preserving and alters the packing density of the light-harvesting complexes. Also, the vesicles are not necessarily of an exact spherical shape and their approximate diameter prior to imaging is not known. Finally, the relative alignment of individual patches with respect to one another is not known. The patches were aligned in accord with the LD data of intact membrane vesicles (Frese et al., 2004). The conversion of planar images onto spherical patches was performed using an area-preserving transformation, the inverse-Mollweide projection (Snyder, 1987; Şener et al., 2007), that maps a point (x,y) on the plane onto a point (φ,λ) on a sphere of radius R, where φ is the latitude and λ is the longitude
πx / R
(20)
2 2 cos ( Θ )
where
(
Θ = sin −1 y / 2 R
)
on the plane. Another advantage of the transformation in Eq. (20) is that it maps an infinitesimal circle on the plane onto a circle of equal area on the target sphere. Thus, deformations of the relative geometry of the light-harvesting complexes are minimized as long as multiple small patches are used instead of a few large ones. The inverse-Mollweide transformation (Eq. 20) was applied for mapping the atomic structures of the individual complexes introduced in Section II onto a spherical vesicle. However, the transformation was used not for mapping individual atomic coordinates, but the center of mass and the relative orientation of each complex, thereby avoiding distortion artifacts within individual complexes. The resulting assembly was then manually edited to remove steric clashes as well as gaps between proteins. A lower than natural packing density would adversely affect the overall quantum efficiency of the system. Two vesicle assemblies with a 60 nm inner diameter were constructed in Şener et al. (2007) corresponding to different light intensity growth conditions. The first vesicle contains 18 RC-LH1-PufX dimers and 101 LH2 complexes with a total of 3879 BChls and an overall BChl:RC ratio of 108, corresponding to a highlight growth condition. The second vesicle contains 9 RC-LH1-PufX dimers and 144 LH2 complexes with a total of 4464 BChls and an overall BChl:RC ratio of 248, corresponding to a low-light growth condition. The overall BChl:RC ratios follow those obtained for wild type chromatophores (Aagaard and Sistrom, 1972). The high-light vesicle is depicted in Fig. 4 whereas planar projections of both vesicles are shown in Fig. 6. VIII. Light-Harvesting and Excitation Transfer across a Photosynthetic Membrane Vesicle
⎛ 2Θ + sin ( 2Θ ) ⎞ φ = sin −1 ⎜ ⎟ π ⎠ ⎝ λ=
287
(21)
The inverse-Mollweide projection is used in geography to produce the familiar rendering of the globe which maps the circles of latitude onto parallel lines
The atomic-level reconstruction of the PSU vesicles outlined in the previous section makes it possible to describe the excitation migration process over a pigment network stretched across hundreds of proteins. Following the methods presented in Section III, an effective Hamiltonian (Eq. 1) for the pigment ensemble (see Figs. 4, 5, and 6) was constructed in Şener et al. (2007) and the rates of excitation transfer between BChl groups (B800, B850, B875, and RC BChls; see Section III.B and Fig. 5) were computed according to Eq. (8). Not surprisingly, the excitation transfer rates for
288
Fig. 4. Organization of the PSU from the supramolecular architecture to individual chlorophylls and the energy transfer network. The high-light adapted vesicle model for Rba. sphaeroides (Şener et al., 2007) is shown. The three different constituent proteins and their BChls are shown: LH2, LH1, and RC. From left to right three different rendering styles are featured. On the left, the light-harvesting proteins are shown in the representation (backbone only, no pigments); in the middle, the proteins are rendered transparently, revealing the BChls, represented by their porphyrin rings; on the right, only the BChls are shown but this time together with their respective electronic couplings (Eq. 1). Only the strongest couplings are shown. The vesicle construction depicted here is based on AFM data on intracytoplasmic membranes (Bahatyrova et al., 2004). Figure made with VMD (Humphrey et al., 1996). See also Color Plate 5, Fig. 8.
B800 → B850 (<1 ps), B850 → B850, and B850 → B875 (for neighboring LH2 pairs or LH2 → LH1 transfers; approximately 10 ps) agree with results of earlier models (Ritz et al., 1998, 2001; Hu et al., 1998; Damjanović et al., 1999, 2000a). The computed B875 → RC forward transfer time of 20 ps also compares favorably with a value of 15 ps reported in earlier studies (Damjanović et al., 2000a; Ritz et al., 2001). However, when the effective Hamiltonian parameters of the monomeric circular RC-LH1 complex are substituted for the dimeric S-shaped RC-LH1-PufX structure found in Rba. sphaeroides, the computed RC → LH1 backtransfer time (1.4 ps) is too fast compared to the empirically supported value of 8 ps. This back-transfer time is especially sensitive to the Hamiltonian parameters of the RC BChls which are not known accurately for the dimeric complex. Therefore, this disagreement for the back transfer time is likely due to lack of sufficient structural data
Melih K. ùener and Klaus Schulten
Fig. 5. Excitation transfer between the bacteriochlorophyll groups of the PSU vesicle. Shown are BChl groups of the highlight adapted vesicle model for Rba. sphaeroides (Şener et al., 2007). RC and LH1 B875 BChls are depicted in black, whereas LH2 B850 and B800 BChls are depicted in dark and light gray, respectively. Also shown are the center of mass of each B850 and B875 BChl cluster, rendered as a sphere of the same cluster, and the excitation transfer rates between them. The radii of the bonds connecting two centers of mass are proportional to the logarithm of the excitation transfer between the corresponding BChl groups computed according to Eq. (8). For the sake of clarity only the strongest transfer rates between BChl groups are rendered and transfers involving the B800 and RC BChls are not shown. Figure made with VMD (Humphrey et al., 1996).
for computing more accurate Hamiltonian parameters and shortcomings of the cluster-cluster transfer picture (Eq. 8) when the assumptions of coherent transfer and Boltzmann equilibrium become suspect. In Şener et al. (2007) a heuristic RC → LH1 back-transfer time of 8 ps was adopted corresponding to a detrapping probability of 27% in approximate agreement with the observed value of 20% previously reported (Timpmann et al., 1995; Bernhardt and Trissl, 2000; Amesz and Neerken, 2002). The transfer rates thus computed between BChl clusters enable the evaluation of the average excitation lifetime τ and the overall quantum efficiency q of the pigment network according to Eqs. (5,6) (see Section III.B). For the first (high-light adapted) vesicle, an average excitation lifetime of 50 ps and a corresponding quantum efficiency of 95% were reported in Şener et al. (2007) in accord with observations (Sundström et al., 1999). The second (low-light adapted) vesicle features a longer 162 ps lifetime with
Chapter 15
Supramolecular Organization of the Photosynthetic Unit
289
Fig. 6. Excitation transfer dynamics of two PSU vesicles. The distributions of average excitation lifetime as a function of the initially excited BChl are shown for (a) the high-light adapted and (b) the low-light adapted vesicles modeled in Şener et al. (2007). The second vesicle displays an increased average lifetime due to the increased mean-free path for excitations to travel to a RC. Excitation sharing between nearby RCs is depicted in (c) and (d) for the two respective vesicles in terms of the cross-trapping probability of excitation at a given RC after it has detrapped from the RC shown at the center of the coordinate system. The probability of excitation detrapped at the central RC to be retrapped at the same RC is only 13% denoting significant excitation sharing between the RCs of the same dimer cluster.
only an 84% efficiency due to a greater mean-free path for excitations to travel from further LH2 complexes. For both vesicles, a continuous distribution of excitation lifetimes was reported as a function of the initial BChl-RC distance indicating a seamlessly connected, efficient energy transfer network. Notably, a substantial degree of excitation sharing is reported among neighboring RCs. In fact, the average probability that a detrapped excitation is recaptured at the same RC is only about 13% (Şener et al., 2007) (see Fig. 6). The cross-trapping probabilities Q Xjk among RCs, computed according to Eq. (15) correspond to the coefficients of the sojourn expansion describing repeated detrapping-retrapping events (see Section III.C). It is not yet clear how much excitation sharing is exploited by nature, for example at high light intensities, as a measure of maintaining high efficiency and preventing photodamage. It is entirely possible that the rationale for the linear clustered arrangement of RC-LH1 dimers is to simplify chromatophore assembly and facilitate quinone diffusion rather than to optimize excitation migration.
IX. Conclusions For the first time, modeling techniques permit us to describe the fundamental biochemical processes underlying photosynthesis all the way down from the atomic detail level up to the system level, bridging timescales ranging from femtoseconds to milliseconds. This accomplishment of decades-long collaboration between experimental and theoretical studies provides an example of constructing working structural models for entire cellular organelles by combining imaging techniques such as AFM, cryoEM, X-ray crystallography, and spectroscopy. Future system-level PSU studies must put greater emphasis on the processes subsequent to excitation migration such as charge transfer, quinone diffusion, and ATP synthesis. This is only possible by a closer cooperation between what have long been disjointed fields of study. Specifically, there is now a great need for more refined data on the architecture and development of the PSU, especially in reference to the location of the missing components such as the bc1 complex or the ATP synthase.
290 We expect that over the next decade all-atom dynamical models combining the structural and functional aspects of entire vesicles will become commonplace. Such models can describe the steps of photosynthesis from light-harvesting all the way to the storage of energy in chemical bonds, also accounting fully for the effects of thermal disorder, thus paving the way for the design of efficient man-made light-harvesting devices. Acknowledgments The authors would like to thank Neil Hunter and Marion Thurnauer for editorial assistance. K. S. has been supported by NSF grant MCB 0234938, NIH grant PHS 2 P41 RR05969, and by a Humboldt Foundation award. M. K. Ş. acknowledges the support of the Department of Physiology and Biophysics and Institute for Computational Biomedicine of Weill Cornell Medical College. References Aagaard J and Sistrom W (1972) Control of synthesis of reaction center bacteriochlorophyll in photosynthetic bacteria. Photochem Photobiol 15: 209–225 Allen J, Yeates T, Komiya H and Rees D (1987) Structure of the reaction center from Rhodobacter sphaeroides R-26: The protein subunits. Proc Natl Acad Sci USA 84: 6162–6166 Amesz J and Neerken S (2002) Excitation energy trapping in anoxygenic photosynthetic bacteria. Photosynth Res 73: 73–81 Amunts A, Drory O and Nelson N (2007) The structure of a plant Photosystem I supercomplex at 3.4 Å resolution. Nature 447: 58–63 Andreev AV, Agam O, Altshuler BL and Simons BD (1996) Quantum chaos, irreversible classical dynamics, and random matrix theory. Phys Rev Lett 76: 3947–3950 Arnold W and Oppenheimer JR (1950) Internal conversion in the photosynthetic mechanism of blue-green algae. J Gen Physiol 33: 423–435 Bahatyrova S, Frese RN, Siebert CA, Olsen JD, vander Werf KO, van Grondelle R, Niederman RA, Bullough PA, Otto C and Hunter CN (2004) The native architecture of a photosynthetic membrane. Nature 430: 1058–1062 Beekman LMP, van Mourik F, Jones MR, Visser HM, Hunter CN and van Grondelle R (1994) Trapping kinetics in mutants of the photosynthetic purple bacterium Rhodobacter sphaeroides: the influence of the charge separation rate. Biochemistry 33: 3143–3147 Ben-Shem A, Frolow F and Nelson N (2003) Crystal structure of plant Photosystem I. Nature 426: 630–635 Bernhardt K and Trissl H-W (2000) Escape probability and trapping mechanism in purple bacteria: Revisited. Biochim Biophys Acta 1457: 1–17
Melih K. ùener and Klaus Schulten Bibby TS, Nield J and Barber J (2001a) Iron deficiency induces the formation of an antenna ring around trimeric Photosystem I in cyanobacteria. Nature 412: 743–745 Bibby TS, Nield J and Barber J (2001b) Three-dimensional model and characterization of the iron stress-induced CP43´Photosystem I supercomplex isolated from the cyanobacterium Synechocystis PCC 6803. J Biol Chem 276: 43246–43252 Bibby TS, Nield J, Chen M, Larkum AWD and Barber J (2003) Structure of a Photosystem II supercomplex isolated from Prochloron didemni retaining its chlorophyll a/b light-harvesting system. Proc Natl Acad Sci USA 100: 9050–9054 Blankenship RE (2002) Molecular Mechanisms of Photosynthesis. Blackwell Science, Malden Camara-Artigas A, Brune D and Allen JP (2002) Interactions between lipids and bacterial reaction centers determined by protein crystallography. Proc Nat Acad Sci USA 99: 11055–11060 Chandler D, Hsin J, Harrison CB, Gumbart J and Schulten K (2008) Intrinsic curvature properties of photosynthetic proteins in chromatophores. Biophys J, in press Cogdell RJ, Gall A and Köhler J (2006) The architecture and function of the light-harvesting apparatus of purple bacteria: From single molecules to in vivo membranes. Quart Rev Biophys 39: 227–324 Conroy MJ, Westerhuis W, Parkes-Loach PS, Loach PA, Hunter CN and Williamson MP (2000) The solution structure of the Rhodobacter sphaeroides LH1 polypeptide reveals two helical domains separated by a flexible region: Structural consequences for the LH1 complex. J Mol Biol 298: 83–94 Cory MG, Zerner MC, Hu X and Schulten K (1998) Electronic excitations in aggregates of bacteriochlorophylls. J Phys Chem B 102: 7640–7650 Crofts AR, Meinhardt SW, Jones KR and Snozzi M (1983) The role of the quinone pool in the cyclic electron transfer chain of Rps sphaeroides: A modified Q-cycle mechanism. Biochim Biophys Acta 723: 202–218 Damjanović A, Ritz T and Schulten K (1999) Energy transfer between carotenoids and bacteriochlorophylls in a light harvesting protein. Phys Rev E 59: 3293–3311 Damjanović A, Ritz T and Schulten K (2000a) Excitation energy trapping by the reaction center of Rhodobacter sphaeroides. Int J Quantum Chem 77: 139–151 Damjanović A, Ritz T and Schulten K (2000b) Excitation transfer in the peridinin-chlorophyll-protein of Amphidinium carterae. Biophys J 79: 1695–1705 Damjanović A, Kosztin I, Kleinekathoefer U and Schulten K (2002a) Excitons in a photosynthetic light-harvesting system: A combined molecular dynamics, quantum chemistry and polaron model study. Phys Rev E 65: 031919 Damjanović A, Vaswani HM, Fromme P and Fleming GR (2002b) Chlorophyll excitations in Photosystem I of Synechococcus elongates. J Phys Chem B 106: 10251–10262 Deisenhofer J, Epp O, Mikki K, Huber R and Michel H (1985) Structure of the protein subunits in the photosynthetic reaction centre of Rhodopseudomonas viridis at 3 Å resolution. Nature 318: 618–624 Dexter D (1953) A theory of sensitized luminescence in solids. J Chem Phys 21: 836–850 Eccles J, Honig B and Schulten K (1988) Spectroscopic determinants in the reaction center of Rhodopseudomonas viridis. Biophys J 53: 137–144 Efetov K (1995) Supersymmetry in quantum chaos and meso-
Chapter 15
Supramolecular Organization of the Photosynthetic Unit
scopic physics. Physica D83: 151–162 Emerson R and Arnold A (1932) The photochemical reaction in photosynthesis. J Gen Physiol 16: 191–205 Ermler U, Fritzsch G, Buchanan SK and Michel H (1994) Structure of the photosynthetic reaction center from Rhodobacter sphaeroides at 2.65Å resolution: Cofactors and protein-cofactor interactions. Structure 2: 925–936 Feniouk BA, Cherepanov DA, Voskoboynikova NE, Mulkidjanian AY and Junge W (2002) Chromatophore vesicles of Rhodobacter capsulatus contain on average one F0F1-ATP synthase each. Biophys J 82: 1115–1122 Fillingame RH (2000) Getting to the bottom of the F1-ATPase. Nature Struct Biol 7: 1002–1004 Fillingame RH, Jiang W and Dmitriev OY (2000) Coupling H+ transport to rotary catalysis in F-type ATP synthases: Structure and organization of the transmembrane rotary motor. J Exp Biol 203: 9–17 Förster T (1948) Zwischenmolekulare Energiewanderung und Fluoreszenz. Ann Phys (Leipzig) 2: 55–75 Fotiadis D, Qian P, Philippsen A, Bullough PA, Engel A and Hunter CN (2004) Structural analysis of the reaction center light-harvesting complex I photosynthetic core complex of Rhodospirillum rubrum using atomic force microscopy. J Biol Chem 279: 2063–2068 Frese RN, Olsen J, Branvall R, Westerhuis W, Hunter CN and van Grondelle R (2000) The long-range supraorganization of the bacterial photosynthetic unit: A key role for PufX. Proc Natl Acad Sci USA 97: 5197–5202 Frese RN, Siebert CA, Niederman RA, Hunter CN, Otto C and van Grondelle R (2004) The long-range organization of a native photosynthetic membrane. Proc Natl Acad Sci USA 101: 17994–17999 Frese RN, Pàmies, JC, Olsen, JD, Bahatyrova S, van der Weij-de Wit CD, Aartsma TJ., Otto C, Hunter CN, Frenkel D and van Grondelle R (2008) Protein shape and crowding drive domain formation and curvature in biological membranes. Biophys J 19: 640–647 Gennis RB, Barquera B, Hacker B, Van Doren SR, Arnaud S, Crofts AR, Davidson E, Gray KA and Daldal F (1993) The bc1 complexes of Rhodobacter sphaeroides and Rhodobacter capsulatus. J Bioener Biomemb 25: 195–209 Geyer T and Helms V (2006) Reconstruction of a kinetic model of the chromatophore vesicles from Rhodobacter sphaeroides. Biophys J 91: 927–937 Gobets B, van Stokkum IH, Rögner M, Kruip J, Schlodder E, Karapetyan NV, Dekker JP and van Grondelle R (2001) Timeresolved fluorescence emission measurements of Photosystem I particles of various cyanobacteria: A unified compartmental model. Biophys J 81: 407–424 Govindjee (2000) Milestones in photosynthesis research. In: Yunus M, Pathre U and Mohanty P (eds) Probing Photosynthesis: Mechanisms, Regulation, and Adaptation, pp 9–39. Taylor and Francis, New York Groot M, Yu J, Agarwal R, Norris JR and Fleming GR (1998) Three-pulse photon echo measurements on the accessory pigments in the reaction center of Rhodobacter sphaeroides. J Phys Chem B 102: 5923–5931 Guhr T, Muller-Groeling A and Weidenmuller HA (1998) Random matrix theories in quantum physics: Common concepts. Phys Rep 299: 189–425 Hu X and Schulten K (1998) A model for the light-harvesting
291
complex I (B875) of Rhodobacter sphaeroides. Biophys J 75: 683–694 Hu X, Xu D, Hamer K, Schulten K, Koepke J and Michel H (1995) Predicting the structure of the light-harvesting complex II of Rhodospirillum molischianum. Prot Sci 4: 1670–1682 Hu X, Ritz T, Damjanović A and Schulten K (1997) Pigment organization and transfer of electronic excitation in the purple bacteria. J Phys Chem B 101: 3854–3871 Hu X, Damjanović A, Ritz T and Schulten K (1998) Architecture and function of the light harvesting apparatus of purple bacteria. Proc Natl Acad Sci USA 95: 5935–5941 Hu X, Ritz T, Damjanović A, Autenrieth F and Schulten K (2002) Photosynthetic apparatus of purple bacteria. Quart Rev Biophys 35: 1–62 Humphrey W, Dalke A and Schulten K (1996) VMD — Visual Molecular Dynamics. J Mol Graphics 14: 33–38 Hunter CN, Kramer HJM and van Grondelle R (1985) Linear dichroism and fluorescence emission of antenna complexes during photosynthetic unit assembly in Rhodopseudomonas sphaeroides. Biochim Biophys Acta 807: 44–51 Hunter CN, Tucker JD, Niederman RA (2005) Perspective on the assembly and organisation of photosynthetic membranes in Rhodobacter sphaeroides. Photochem Photobiol Sci 4: 1023–1027 Jamieson SJ, Wang P, Qian P, Kirkland JY, Conroy MJ, Hunter CN and Bullough PA (2002) Projection structure of the photosynthetic reaction centre-antenna complex of Rhodospirillum rubrum at 8.5 Å resolution. J Mol Biol 21: 3927–3935 Jordan P, Fromme P, Witt HT, Klukas O, Saenger W and Krauß N (2001) Three-dimensional structure of cyanobacterial Photosystem I at 2.5 Å resolution. Nature 411: 909–917 Jungas C, Ranck J, Rigaud J, Joliot P and Verméglio A (1999) Supramolecular organization of the photosynthetic apparatus of Rhodobacter sphaeroides. EMBO J 18: 534–542 Junge W, Lill H and Engelbrecht S (1997) ATP synthase: An electrochemical transducer with rotatory mechanics. Trends Biochem Sci 22: 420–423 Karrasch S, Bullough P and Ghosh R (1995) 8.5 Å projection map of the light-harvesting complex I from Rhodospirillum rubrum reveals a ring composed of 16 subunits. EMBO J 14: 631–638 Kennis JTM, Gobets B, van Stokkum IHM, Dekker JP, van Grondelle R and Fleming GR (2001) Light harvesting by chlorophylls and carotenoids in the Photosystem I core complex of Synechococcus elongatus: A fluorescence upconversion study. J Phys Chem B 105: 4485–4494 Knox RS and Spring BQ (2003) Dipole strengths in the chlorophylls. Photochem Photobiol 77: 497–501 Koepke J, Hu X, Muenke C, Schulten K and Michel H (1996) The crystal structure of the light harvesting complex II (B800-850) from Rhodospirillum molischianum. Structure 4: 581–597 Koolhaas MHC, Frese RN, Fowler GJS, Bibby TS, Georgakopoulou S, van der Zwan G, Hunter CN and van Grondelle R (1998) Identification of the upper exciton component of the B850 bacteriochlorophylls of the LH2 antenna complex, using a B800-free mutant of Rhodobacter sphaeroides. Biochemistry 37: 4693–4698 Kosztin I and Schulten K (2008) Molecular dynamics methods for bioeletronic systems in photosynthesis. In: Aartsma T and Matysik J (eds) Biophysical Techniques in Photosynthesis II (Advances in Photosynthesis and Respiration, Vol 26), pp
292 445–464. Springer, Dordrecht Krauss N, Hinrichs W, Witt I, Fromme P, Pritzkow W, Dauter Z, Betzel C, Wilson KS, Witt HT and Saenger W (1993) 3-dimensional structure of system-I of photosynthesis at 6 Ångström resolution. Nature 361: 326–331 Krauß N, Schubert W-D, Klukas O, Fromme P, Witt HT and Saenger W (1996) Photosystem I at 4 Å resolution represents the first structural model of a joint photosynthetic reaction centre and core antenna system. Nature Struct Biol 3: 965–973 Krueger BP, Scholes GD and Fleming GR (1998) Calculation of couplings and energy-transfer pathways between the pigments of LH2 by the ab initio transition density cube method. J Phys Chem B 102: 5378–5386 Loll B, Kern J, Saenger W, Zouni A and Biesiadka J (2005) Towards complete cofactor arrangement in the 3.0 Å resolution structure of Photosystem II. Nature 438: 1040–1044 Marcus RA (1956a) Electrostatic free energy and other properties of states having nonequilibrium polarization II. J Chem Phys 24: 979–989 Marcus RA (1956b) On the energy of oxidation-reduction reactions involving electron transfer I. J Chem Phys 24: 966–978 McDermott G, Prince SM, Freer AA, Hawthornthwaite-Lawless AM, Papiz MZ, Cogdell RJ and Isaacs NW (1995) Crystal structure of an integral membrane light-harvesting complex from photosynthetic bacteria. Nature 374: 517–521 Melkozernov AN (2001) Excitation energy transfer in Photosystem I from oxygenic organisms. Photosynth Res 70: 129–153 Melkozernov AN, Lin S, Blankenship RE and Valkunas L (2001) Spectral inhomogeneity of Photosystem I and its influence on excitation equilibration and trapping in the cyanobacterium Synechocystis sp PCC6803 at 77 K. Biophys J 81: 1144–1154 Mukamel S (1995) Principles of Nonlinear Optical Spectroscopy. Oxford University Press, New York Murray JW and Barber J (2007) Structural characteristics of channels and pathways in Photosystem II including the identification of an oxygen channel. J Struct Biol 159: 228–237 Nelson N and Ben-Shem A (2004) The complex architecture of oxygenic photosynthesis. Nat Rev Mol Cell Biol 5: 971–982 Nelson N and Yocum CF (2006) Structure and Function of Photosystems I and II. Annu Rev Plant Biol 57: 521–565 Niederman RA, Mallon DE and Langan JJ (1976) Membranes of Rhodopseudomonas sphaeroides. IV. Assembly of chromatophores in low-aeration cell suspensions. Biochim Biophys Acta 440: 429–447 Niederman RA, Mallon DE and Parks LC (1979) Membranes of Rhodopseudomonas sphaeroides. VI. Isolation of a fraction enriched in newly synthesized bacteriochlorophyll a-protein complexes. Biochim Biophys Acta 555: 210–220 Nonella M and Schulten K (1991) Molecular dynamics simulation of electron transfer in proteins: Theory and application to QA → QB transfer in the photosynthetic reaction center. J Phys Chem 95: 2059–2067 Onuchic JN, Beratan DN and Hopfield JJ (1986) Some aspects of electron-transfer reaction dynamics. J Phys Chem 90: 3707–3721 Oppenheimer JR (1941) Internal conversion in photosynthesis. In: Proceedings of the American Physical Society. Phys Rev 60: 158 Papiz MZ, Prince SM, Howard T, Cogdell RJ and Isaacs NW (2003) The structure and thermal motion of the B800-850 LH2
Melih K. ùener and Klaus Schulten complex from Rps. acidophila at 2.0 Å resolution and 100K: New structural features and functionally relevant motions. J Mol Biol 326: 1523–1538 Park S, Şener MK, Lu D and Schulten K (2003) Reaction paths based on mean first-passage times. J Chem Phys 119: 1313–1319 Parks LC and Niederman RA (1978) Membranes of Rhodopseudomonas sphaeroides. V. Identification of bacteriochlorophyll a-depleted cytoplasmic membrane in phototrophically grown cells. Biochim Biophys Acta 511: 70–82 Pålsson L-O, Flemming C, Gobets B, van Grondelle R, Dekker JP and Schlodder E (1998) Energy transfer and charge separation in Photosystem I: P700 oxidation upon selective excitation of the long-wavelength antenna chlorophylls of Synechococcus elongatus. Biophys J 74: 2611–2622 Pugh R, McGlynn P, Jones M and Hunter CN (1998) The LH1-RC core complex of Rhodobacter sphaeroides: Interaction between components, time-dependent assembly, and topology of the PufX protein. Biochim Biophys Acta 1366: 301–316 Qian P, Hunter CN and Bullough PA (2005) The 8.5Å, projection structure of the core RC-LH1-PufX dimer of Rhodobacter sphaeroides. J Mol Biol 349: 948–960 Ritz T and Schulten K (2001) Physik der Photosynthese. Physikalische Blaetter 57: 49–53 Ritz T, Hu X, Damjanović A and Schulten K (1998) Excitons and excitation transfer in the photosynthetic unit of purple bacteria. J Lumin 76-77: 310–321 Ritz T, Damjanović A, Schulten K, Zhang J and Koyama Y (2000) Efficient light harvesting through carotenoids. Photosynth Res 66: 125–144 Ritz T, Park S and Schulten K (2001) Kinetics of excitation migration and trapping in the photosynthetic unit of purple bacteria. J Phys Chem B 105: 8259–8267 Ritz T, Damjanović A and Schulten K (2002) The quantum physics of photosynthesis. Chem Phys Chem 3: 243–248 Roszak AW, Howard TD, Southall J, Gardiner AT, Law CJ, Isaacs NW and Cogdell RJ (2003) Crystal structure of the RC-LH1 core complex from Rhodopseudomonas palustris. Science 302: 1969–1972 Rutkauskas D, Novoderezhkin VI, Cogdell RJ and van Grondelle R (2005) Fluorescence spectroscopy of conformational changes of single LH2 complexes. Biophys J 88: 422–435 Scheer H (1991) Chlorophylls. CRC Press, Boca Raton Scheuring S, Levy D and Rigaud J-L (2005) Watching the components of photosynthetic bacterial membranes and their in situ organisation by atomic force microscopy. Biochim Biophys Acta 1712: 109–127 Scheuring S and Sturgis J (2005) Chromatic adaptation of photosynthetic membranes. Science 309: 484–487 Scheuring S, Sturgis JN, Prima V, Bernadac A, Levy D and Rigaud J-L (2004) Watching the photosynthetic apparatus in native membranes. Proc Natl Acad Sci USA 91: 11293–11297 Scholes G, Gould I, Cogdell R and Fleming G (1999) Ab initio molecular orbital calculations of electronic couplings in the LH2 bacterial light-harvesting complex of Rps Acidophila. J Phys Chem B 103: 2543–2553 Schulten K (1999) From simplicity to complexity and back: Function, architecture and mechanism of light harvesting systems in photosynthetic bacteria. In: Frauenfelder H, Deisenhofer J and Wolynes PG (eds) Simplicity and Complexity in Proteins and
Chapter 15
Supramolecular Organization of the Photosynthetic Unit
Nucleic Acids, pp 227–253. Dahlem University Press, Berlin Schulten K and Tesch M (1991) Coupling of protein motion to electron transfer: Molecular dynamics and stochastic quantum mechanics study of photosynthetic reaction centers. Chem Phys 158: 421–446 Şener MK (1999) Universality in random matrix models of quantum chromodynamics. PhD thesis, State University of New York at Stony Brook, Dept. of Physics, Stony Brook, NY Şener M and Schulten K (2002) A general random matrix approach to account for the effect of static disorder on the spectral properties of light harvesting systems. Phys Rev E 65: 031916 Şener M and Schulten K (2005) Physical principles of efficient excitation transfer in light harvesting. In: Andrews DL (ed), Energy Harvesting Materials, pp 1–26. World Scientific, Singapore Şener MK, Lu D, Ritz T, Park S, Fromme P and Schulten K (2002) Robustness and optimality of light harvesting in cyanobacterial Photosystem I. J Phys Chem B 106: 7948–7960 Şener MK, Park S, Lu D, Damjanović A, Ritz T, Fromme P and Schulten K (2004) Excitation migration in trimeric cyanobacterial Photosystem I. J Chem Phys 120: 11183–11195 Şener MK, Jolley C, Ben-Shem A, Fromme P, Nelson N, Croce R and Schulten K (2005) Comparison of the light harvesting networks of plant and cyanobacterial Photosystem I. Biophys J 89: 1630–1642 Şener MK, Olsen JD, Hunter CN and Schulten K (2007) Atomic-level structural and functional model of a bacterial photosynthetic membrane vesicle. Proc Nat Acad Sci USA 104: 15723–15728 Siebert CA, Qian P, Fotiadis D, Engel A, Hunter CN and Bullough P (2004) The role of PufX in the molecular architecture of photosynthetic membranes in Rhodobacter sphaeroides. EMBO J 23: 690–700 Snyder JP (1987) Map Projections—A working Manual. U S Geological Survey Professional Paper 1395. US Government Printing Office, Washington, DC Sumi H (2000) Structural strategies in the antenna system of photosynthesis on the basis of quantum-mechanical coherence among pigments. J Luminesc 87-89: 71–76 Sumi H (2001) Bacterial photosynthesis begins with quantummechanical coherence. Chem Record 1: 480–493 Sundström V, Pullerits T and van Grondelle R (1999) Photosynthetic light-harvesting: Reconciling dynamics and structure of purple bacterial LH2 reveals function of photosynthetic unit. J Phys Chem B 103: 2327–2346 Timpmann K, Freiberg A and Sundström V (1995) Energy trapping and detrapping in the photosynthetic bacterium Rhodopseudomonas viridis: Transfer-to-trap-limited dynamics. Chem Phys 194: 275–283 Tretiak S, Middleton C, Chernyak V and Mukamel S (2000) Bacteriochlorophyll and carotenoid excitonic couplings in the LH2 system of purple bacteria. J Phys Chem B 104: 9540–9553 Treutlein H, Schulten K, Deisenhofer J, Michel H, Brünger A and Karplus M (1988a) Molecular dynamics simulation of the primary processes in the photosynthetic reaction center of Rhodopseudomonas viridis. In: Breton J, Verméglio A (eds) The Photosynthetic Bacterial Reaction Center: Structure and Dynamics, Vol 149, pp 139–150, NATO Sci Ser A. Plenum, New York Treutlein H, Schulten K, Niedermeier C, Deisenhofer J, Michel H
293
and Devault D (1988b) Electrostatic control of electron transfer in the photosynthetic reaction center of Rhodopseudomonas viridis. In: Breton J, Verméglio A (eds), The Photosynthetic Bacterial Reaction Center: Structure and Dynamics, Vol 149, pp 369–377, NATO Sci Ser A. Plenum, New York van Amerongen H, Valkunas L and van Grondelle R (2000) Photosynthetic Excitons. World Scientific, Singapore van Grondelle R and Novoderezhkin VI (2006) Energy transfer in photosynthesis: Experimental insights and quantitative models. Phys Chem Chem Phys 8: 793–807 van Grondelle R, Dekker JP, Gillbro T and Sundström V (1994) Energy transfer and trapping in photosynthesis. Biochim Biophys Acta 1187: 1–65 Vasil’ev S and Bruce D (2004) Optimization and evolution of light harvesting in photosynthesis: The role of antenna chlorophyll conserved between Photosystem II and Photosystem I. Plant Cell 16: 3059–3068 Vasil’ev S, Shen J-R, Kamiya N and Bruce D (2004) The orientations of core antenna chlorophylls in Photosystem II are optimized to maximize the quantum yield of photosynthesis. FEBS Lett 561: 111–116 Velasco F and Crofts AR (1991) Complexes or super complexes: Inhibitor titrations show that electron transfer in chromatophores from Rb. sphaeroides involves a dimeric ubiquinol: cytochrome c oxidoreductase, and is delocalized. Biochem Soc Trans 19: 588–593 Visscher KJ, Bergstrom H, Sundström V, Hunter CN and van Grondelle R (1989) Temperature dependence of energy transfer from the long wavelength antenna BChl-896 to the reaction center in Rhodospirillum rubrum, Rhodobacter sphaeroides (wt and M21 mutant) from 77 to 177 K, studied by picosecond absorption spectroscopy. Photosynth Res 22: 211–217 Walz T, Jamieson SJ, Bowers CM, Bullough PA and Hunter CN (1998) Projection structures of three photosynthetic complexes from Rhodobacter sphaeroides: LH2 at 6 Å, LH1 and LH1-RC at 25 Å. J Mol Biol 282: 833–845 Xia D, Yu C-A, Kim H, Xia J-Z, Kachurin AM, Zhang L, Yu L and Deisenhofer J (1997) Crystal structure of the cytochrome bc1 complex from bovine heart mitochondria. Science 277: 60–66 Xiong J, Fischer WM, Inoue K, Nakahara M and Bauer CE (2000) Molecular evidence for the early evolution of photosynthesis. Science 289: 1724–1730 Xu D and Schulten K (1992) Multi-mode coupling of protein motion to electron transfer in the photosynthetic reaction center: Spin-boson theory based on a classical molecular dynamics simulation. In: Breton J and Verméglio A (eds), The Photosynthetic Bacterial Reaction Center: II Structure, Spectroscopy and Dynamics, pp 301–312. NATO Sci Ser A. Plenum Press, New York Xu D and Schulten K (1994) Coupling of protein motion to electron transfer in a photosynthetic reaction center: Investigating the low temperature behaviour in the framework of the spin-boson model. Chem Phys 182: 91–117 Yang M and Fleming GR (2002) Influence of phonons on exciton transfer dynamics: comparison of the Redfield, Förster, and modified Redfield equations. Chem Phys 282: 163–180 Yang M, Damjanović A, Vaswani HM and Fleming GR (2003) Energy transfer in Photosystem I of cyanobacteria Synechococcus elongatus: Model study with structure-based semi-empiri-
294 cal Hamiltonian and experimental spectral density. Biophys J 85: 140–158 Zazubovich V, Matsuzaki S, Johnson TW, Hayes JM, Chitnis PR and Small GJ (2002) Red antenna states of Photosystem I from cyanobacterium Synechococcus elongatus: A spectral
Melih K. ùener and Klaus Schulten hole burning study. Chem Phys 275: 47–59 Zouni A, Witt H-T, Kern J, Fromme P, Krauss N, Saenger W and Orth P (2001) Crystal structure of Photosystem II from Synechococcus elongatus at 3.8 Å resolution. Nature 409: 739–743
Chapter 16 Structural Plasticity of Reaction Centers from Purple Bacteria Michael R. Jones* Department of Biochemistry, School of Medical Sciences, University of Bristol, University Walk, Bristol, BS8 1TD, U.K.
Summary ............................................................................................................................................................... 295 I. The Plastic Purple Reaction Center ............................................................................................................... 296 II. Biochemical and Genetic Alteration of Polypeptide Composition................................................................... 297 A. Depletion of the H-Polypeptide......................................................................................................... 297 B. Depletion of the Cytochrome Subunit............................................................................................... 298 C. Chimeric RCs ................................................................................................................................... 298 III. Cofactor Exclusion ......................................................................................................................................... 299 A. Carotenoid Exclusion ....................................................................................................................... 299 B. Quinone Exclusion ........................................................................................................................... 301 C. Bacteriochlorin Exclusion ................................................................................................................. 301 IV. Cofactor Replacement.................................................................................................................................... 302 A. Bacteriochlorin Replacement ........................................................................................................... 302 B. Quinone Replacement...................................................................................................................... 303 C. Carotenoid Replacement ................................................................................................................. 304 D. Iron Removal and Replacement....................................................................................................... 305 E. Genetic Approaches to Cofactor Replacement ................................................................................ 305 V. Helix Symmetrization...................................................................................................................................... 307 A. DLL, Cousins and Offspring ............................................................................................................... 307 B. Sym-1 and Siblings .......................................................................................................................... 308 C. QAQA and A6D1 ................................................................................................................................ 309 VI. Water and Other Unexpected Things in Electron Density Maps .................................................................... 310 VII. Building New Functionality ............................................................................................................................. 312 VIII. Conclusions .................................................................................................................................................... 313 Acknowledgments ................................................................................................................................................. 314 References ............................................................................................................................................................ 314
Summary The purple bacterial reaction center has made major contributions to our understanding of photosynthetic energy transduction and biological electron transfer. A major element in this work has been alteration of the structure and/or composition of the reaction center through site-directed mutagenesis or biochemical treatments, and much of the usefulness of the reaction center as an experimental system has come from its (sometimes surprising) tolerance of significant changes in cofactor composition or amino acid sequence. The purpose of this chapter is to illustrate this structural plasticity, by documenting the extent to which the structure and composition of the reaction center has been altered in order to investigate details of its mechanism, or general *Email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 295–321. © 2009 Springer Science + Business Media B.V.
296
Michael R. Jones
principles pertaining to biological electron transfer. For the cofactors, the article looks at how and why individual cofactors have been replaced and the main outcomes of such studies. For the protein, the emphasis is placed on mutations involving large numbers of amino acids or single mutations that alter the cofactor composition of the complex. The chapter also considers the significant influence that changes in solvent (water) structure can have on reaction center structure and mechanism, attempts to construct chimeric reaction centers, and protein engineering aimed at introducing new functionality into the reaction center. I. The Plastic Purple Reaction Center According to the Encyclopaedia Britannica Online (2006), ‘plasticity’ is defined as ‘the ability of certain solids to flow or to change shape permanently when subjected to stresses of intermediate magnitude between those producing temporary deformation, or elastic behavior, and those causing failure of the material, or rupture.’ Even more concisely, Wikipedia defines ‘plasticity’ in a physics and engineering context as ‘the propensity of a material to undergo permanent deformation under load’ (Wikipedia-The Free Encyclopedia, 2006). The purpose of this chapter is to document the extent to which the purple bacterial reaction center (RC) has been remolded in terms of structure and composition in order to investigate its innate function as a transducer of light energy, and during use as a tool to investigate broader subjects such as biological electron transfer. Such remolding can either be achieved genetically, through site-directed mutagenesis and the like, or through biochemical treatments that affect either the protein or the cofactors. On the genetic side, it is obvious that a detailed account of the structural and functional consequences of every single mutation ever engineered into the RC would be a mammoth task, and so the chapter focuses mainly on larger scale alterations involving multiple amino acids, or point mutations that change the cofactor composition. On the biochemical side changes to both protein and cofactors are considered, but emphasis is placed mainly on work carried out during the last ten years or so. All purple bacterial RCs have a core domain made Abbreviations: BA – A-side monomeric BChl; BB – B-side monomeric BChl; BChl – bacteriochlorophyll; Blc. – Blastochloris; BPhe – bacteriopheophytin; Chl – chlorophyll; Em P/P+ – mid-point redox potential of P/P+; HA – A-side BPhe; HB – B-side BPhe; LDAO – lauryldimethylamine oxide; LH1 – core light harvesting complex; LH2 – peripheral light harvesting complex; P – primary electron donor; PA – A-side P BChl; PB – B-side P BChl; QA – primary acceptor quinone; QB – secondary (dissociable) acceptor quinone; Rba. – Rhodobacter; RC – reaction center
up from the H-, L- and M-polypeptides, encoded by the puhA, pufL and pufM genes, respectively. These polypeptides encase the bacteriochlorophyll (BChl), bacteriopheophytin (BPhe), quinone and carotenoid cofactors in a three-dimensional arrangement that has become very familiar since the publication of crystal structures in the mid-1980’s for the RCs from Blastochloris (Blc.) viridis (formerly Rhodopseudomonas viridis) and Rhodobacter (Rba.) sphaeroides (see Hoff and Deisenhofer,1997, for a comprehensive review that includes original references). In many species, including Blc. viridis, a fourth C-polypeptide is bound to the periplasmic faces of the L- and M-polypeptides, and binds heme groups that act as the electron donor to the photo-oxidized BChl cofactors. Gene deletion experiments have shown that the H-, L- and M-polypeptides are indispensable for assembly of the RC (Tehrani et al., 2003; Tehrani and Beatty, 2004), but in some species it is possible to delete the gene encoding the C-polypeptide and still retain photosynthetic growth (Nagashima et al., 1996; Masuda et al., 2002). Unlike the H-, L- and M-polypeptides that form their scaffold, not all of the intramembrane cofactors are essential for assembly of the RC. Figure 1 summarizes the arrangement of the polypeptides (Fig. 1a) and cofactors (Fig. 1b) of the Rba. sphaeroides RC. Two membrane-spanning branches of cofactors extend from a pair of BChls on the periplasmic side of the membrane (denoted P — formed by the PA and PB BChls) to a quinone on the cytoplasmic side (denoted QA and QB on the so-called A- and B-branches, respectively). On each branch the P BChls are connected to the quinone cofactor by a ‘monomeric’ BChl (BA and BB) and a BPhe (HA and HB) (Fig. 1b). The wild-type RC also has a carotenoid cofactor located close to the BB monomeric BChl and an iron atom between the QA and QB quinone. The structure and mechanism of the purple bacterial RC has been extensively reviewed (Parson, 1991; Deisenhofer and Norris, 1993; Woodbury and Allen, 1995; Parson, 1996; Hoff and Deisenhofer, 1997; Okamura et al., 2000; van Brederode and Jones, 2000;
Chapter 16
Reaction Center Structural Plasticity
297 such as the light-harvesting 2 (LH2) antenna found in Rba. sphaeroides. II. Biochemical and Genetic Alteration of Polypeptide Composition A. Depletion of the H-Polypeptide
Fig. 1. The structure of the Rba. sphaeroides RC. (a) The Lpolypeptide (dark-grey ribbon) and M-polypeptide (light-grey ribbon) form a heterodimeric scaffold that encases the 10 cofactors (grey spheres). The H-polypeptide (light-grey tube) has a single membrane-spanning helix and a cytoplasmic domain. Grey boxes denote the approximate position of the membrane. (b) The BChl (PA, PB, BA, BB), BPhe (HA, HB) and ubiquinone (QA, QB) cofactors form two membrane-spanning branches around an axis of pseudo-twofold symmetry (dotted line). The monomeric BChls are highlighted in dark grey and the hydrocarbon side-chains of the bacteriochlorins and quinones have been removed for clarity. Also shown are the carotenoid (Crt), the non-heme iron (Fe) and the route of light-driven electron transfer (arrows).
Wraight, 2004), and various aspects of structure and mechanism are addressed in other chapters in this volume. In all purple photosynthetic bacteria the RC is closely associated with a light-harvesting-1 (LH1) antenna pigment-protein, forming the socalled RC-LH1 core complex, and in many species these RC-LH1 complexes are surrounded by one or more peripheral light-harvesting antenna complexes
As outlined above, the L- and M-polypeptides of the RC bind the electron transfer cofactors and are indispensable components of the complex (Tehrani and Beatty, 2004). Gene deletion experiments have shown that the H-polypeptide is also needed for assembly of the Rba. sphaeroides RC (Wong et al., 1996; Chen et al., 1998), but once assembled and solubilized from the membrane using detergent it is possible to remove the H-polypeptide from this complex by incubation for 1 hour at 25 °C with a mixture of LiC1O4, CaCl2, ethanol, Tris-HC1, sodium cholate and EDTA (Debus et al., 1985) The remaining ‘LM’ complex retains the non-heme iron and both quinones, shows only very minor changes in bacteriochlorin absorbance, and carries out charge separation to form P+QA– (Debus et al., 1985). The H-polypeptide can also be reconstituted to yield RCs with essentially wild-type properties (Debus et al., 1985; Utschig et al., 1997). Figure 2a shows a view of the cytoplasmic surface of the Rba. sphaeroides RC with the H-polypeptide removed. The QA, QB and Fe binding sites are located just below the now-exposed cytoplasmic surfaces of the L- and M-polypeptides, consistent with the observation that these cofactors are not released when the H-polypeptide is removed. However the QA quinone is partially visible from the outside of the LM complex in this model (highlighted in darkgrey in Fig. 2a). As shown in Fig. 2b the head-groups of the QA and QB quinones (dark-grey spheres) are overlaid by just a few amino acids from the M- and L-polypeptides, respectively (shown as sticks in Fig. 2b), and so will be much closer to the aqueous phase on the cytoplasmic side of the membrane than is the case when the H-polypeptide is present. The main effects of H-polypeptide depletion are a two-fold acceleration in the rate of P+QA– recombination and a 102–103-fold slowing in the rate of electron transfer from QA– to QB (Debus et al., 1985), effects consistent with a significant change in the environment of these quinones. There is also a decrease in the affinity of the QB site for quinone (Debus et al., 1985). LM preps have been used to study the influence of the
Michael R. Jones
298
polyacrylamide gel electrophoresis. This protocol produces a preparation containing the L- and M-polypeptides and associated cofactors but lacking both the H-polypeptide and the tetra-heme cytochrome subunit (Fukushima et al., 1988; Matsuura et al., 1988). In the same report it was commented that the cytochrome subunit was rather easily separated from the Rvi. gelatinosus RC using detergents such as sodium dodecyl sulfate (SDS) or lauryldimethylamine oxide (LDAO), or chaotropes such as NaSCN or KClO4. In a subsequent report (Matsuura et al., 1988), it was established that water-soluble Rvi. gelatinosus cytochrome c2 or horse-heart cytochrome c can directly reduce the photo-oxidized primary donor in such LM preparations of the Rvi. gelatinosus RC, demonstrating that the tetra-heme cytochrome subunit is dispensable in this species in vitro. Studies involving disruption of the gene for the cytochrome subunit in Rvi. gelatinosus and Rhodovulum sulfidophilum have also led to the conclusion that this component of the RC is dispensable in these species, with strains of bacteria lacking the cytochrome subunit being capable of photosynthetic growth, albeit at a reduced rate (Nagashima et al., 1996; Masuda et al., 2002). C. Chimeric RCs
Fig. 2. The cytoplasmic surface of a model LM preparation of the Rba. sphaeroides RC. (a) The LM heterodimer is shown as a solid object, with the quinones shown as dark-grey spheres. The view is along the symmetry axis, and parts of the QA quinone are visible from the outside of the complex. (b) View as in (a) but with amino acids overlaying the quinone sites shown in stick format, revealing the underlying quinones. In both panels the dotted ovals show the approximate positions of the QA and QB quinones. See also Color Plate 7, Fig. 10.
H-polypeptide on electron transfer from HA– to QA (Liu et al., 1991a; Schelvis et al., 1992; Utschig et al., 1997; Tang et al., 1999; Knox et al., 2001) and carotenoid triplet states (Sebban and Lindqvist, 1987). A protocol for the depletion of the H-polypeptide from the Rps. viridis RC has also been reported (Hara et al., 1998). B. Depletion of the Cytochrome Subunit Fukushima and co-workers have described a protocol involving treatment of RC-LH1 core complexes from Rubrivivax (Rvi.) gelatinosus with octyl thioglucoside, urea and Triton X-100, followed by preparative
The L-, M- and H-polypeptides are all required for RC assembly, but there have been some reports of the construction of chimeric RCs that contain polypeptides from more than one species. Expression of the puf operon from Rba. sphaeroides in a puf operon deletion strain of Rba. capsulatus has been found to produce a RC comprising the Rba. sphaeroides L- and Mpolypeptides and the Rba. capsulatus H-polypeptide (Zilsel et al., 1989). Fluorescence measurements on the resulting strain were interpreted in terms of some RC misassembly, but nevertheless the resulting RCs were capable of supporting photosynthetic growth at near-to-normal rates at moderate light intensity. Along similar lines, Tehrani and co-workers have examined whether parts of the H-polypeptide of the Rba. sphaeroides RC can be replaced with equivalent sequences from the Blc. viridis H-polypeptide (Tehrani et al., 2003). The large cytoplasmic domain could not be exchanged without abolishing RC assembly, but a construct containing a Blc. viridis sequence between residues 1 and 34, encompassing the small (11 amino acid) periplasmic domain and the membrane-spanning helix of the H-polypeptide, yielded RCs capable of supporting photosynthetic growth, as did a construct in which just the membrane-span-
Chapter 16
Reaction Center Structural Plasticity
ning helix was substituted. Findings with these and related constructs have been used to develop a model for assembly of the RC (Tehrani et al., 2003; Tehrani and Beatty, 2004). Taguchi and co-workers have also reported the construction of chimeric RCs in an intriguing experiment in which sequences of the pufM gene from Rba. sphaeroides were used to restore photosyntheticallycompetent RCs to a strain of Rba. capsulatus with a 43 bp deletion in the pufM gene (Taguchi et al., 1993). This deletion removed a loop of amino acids that form part of the binding pocket of the QA quinone at the cytoplasmic end of the membrane-spanning D-helix. A number of photosynthetically-competent derivative strains were generated in which all or part of the C-terminal end of the defective Rba. capsulatus M-polypeptide was replaced by the equivalent sequence from Rba. sphaeroides. All of these included the whole of the membrane-spanning D-helix and the adjacent short surface helix (denoted ‘de’ as it connects the D and E membrane-spanning helices). Although most protein engineering studies have been carried out on Rba. sphaeroides or Rba. capsulatus RCs, studies of the tetra-heme cytochrome that forms an extramembrane component of many purple bacterial RCs has necessitated work with other species. The X-ray crystal structure of the cytochrome subunit in the Blc. viridis RC is known, but this organism presents experimental challenges due to its poor growth under non-photosynthetic conditions. In order to examine the cytochrome component in a more experimentally-amenable organism, Maki and co-workers have constructed a chimeric RC by expressing the pufC gene from Blc. viridis in a pufC-deletion strain of Rvi. gelatinosus (Maki et al., 2003). The first point worthy of note is that the uncomplemented pufC deletion strain of Rvi. gelatinosus was capable of photosynthetic growth at approximately half the rate exhibited by the wild-type strain, demonstrating that functional RCs can assemble in this organism in the absence of the cytochrome component (Nagashima et al., 1996) (see above). Secondly, when the Blc. viridis pufC gene was added the resulting RCs were also capable of supporting photosynthetic growth, showed evidence for rapid reduction of the RC BChls by a cytochrome and biochemical evidence for association of the cytochrome with the remainder of the RC (Maki et al., 2003). This was despite the fact that the pucC gene product shows only 45% identity between the two species, and only 8 out of 30 cytochrome residues
299 involved at the interface with the L/M polypeptides are conserved. The chimera also included exchange of 34 amino acids from the C-terminus of the Mpolypeptide, a region that shows a little over 60% identity between the two species. The chimeric RC was sufficiently robust to withstand additional point mutations of the cytochrome subunit (Alric et al., 2004, 2006), although detergent purification of this chimeric RC has not been reported. Finally, Kortluke and co-workers have described the expression of the puf operon from Roseobacter (Rsb.) denitrificans in a double puf and puc deletion mutant of Rba. capsulatus (Kortluke et al., 1997). The latter strain lacks RCs due to the absence of the pufLM genes, the LH1 antenna due to the absence of the pufBA genes, and the LH2 antenna due to the absence of the pucBA genes. Expression of the Rsb. denitrificans puf operon in this double deletion mutant, together with the Rba. capsulatus pufQ and pufX genes, produced a strain that showed spectroscopic and biochemical evidence of RC assembly at a reduced level. Photoexcitation of membranes containing these chimeric RCs produced spectroscopic evidence for the P+QA– radical pair state but not a longer lived P+QB– state, and the RCs were not capable of supporting photosynthetic growth of the transconjugant strain (Kortluke et al., 1997). III. Cofactor Exclusion A. Carotenoid Exclusion A useful feature of the RC is that it can be induced to assemble without certain cofactors, with a very limited impact on the structure of the rest of the pigment-protein complex. In fact a significant proportion of the research carried out on the Rba. sphaeroides RC has involved a particular variant isolated from a carotenoid-deficient strain, the so-called R-26 RC (Clayton and Smith, 1960; Reed and Clayton, 1968). The R-26 strain has lowered levels of the LH2 light harvesting complex, assisting purification of the RC and providing biochemists with an easier way to satisfy sample-hungry spectroscopists and crystallographers. The RC from the related strain R-26.1, which differs from R-26 in the profile of antenna complexes present (Davidson and Cogdell, 1981; Hunter, 1995), has also used been used as a carotenoid-less RC in a number of studies. The expression level of the RC is not adversely af-
300
Michael R. Jones
Fig. 3. Stereo view of structural changes in the carotenoid-deficient R-26.1 RC. The structures of the wild-type (light-grey) and R-26.1 (dark grey) RCs are overlaid. Most components are shown as stick models with oxygen or nitrogen atoms highlighted as small spheres. The alpha-carbon of Gly M71 is shown as a large sphere. In the R-26.1 RC a molecule of LDAO (thick dark-grey sticks) occupies the central part of the binding pocket of the carotenoid found in the wild-type RC (thick light-grey sticks). The phytol chains of the HB and BB bacteriochlorins are shown from the point that their conformations start to deviate in the two structures (atom nearest the label).
fected by a lack of carotenoid in the R-26 strain, and structural comparisons have shown that the absence of the carotenoid does not produce large scale changes in the structure of the complex, or changes in conformation of the majority of amino acid side chains lining the carotenoid binding pocket. The structure of the R-26 RC at a resolution of 2.8 Å shows that a molecule of the detergent LDAO occupies part of the carotenoid binding pocket (Yeates et al., 1988), and this was also found to be the case in a more recent comparison of the structures of the R-26.1 RC and a carotenoid-containing Rba. sphaeroides RC (Roszak et al., 2004). Figure 3 shows a stereo view of an overlay of the two structures from the latter study. In addition to the presence of an LDAO, the only significant structural consequences of the lack of carotenoid were shifts in the positions of the phytol tails of the BB BChl and HB BPhe, which form part of one end of the carotenoid binding pocket, and the side chain of Phe M162. This residue undergoes significant movement into the empty carotenoid pocket in the R-26.1 RC (Fig. 3), and has been proposed to act as a ‘gatekeeper,’ occluding one end of the binding pocket in the absence of carotenoid and so preventing insertion of carotenoid into the complex from the wrong end (Roszak et al., 2004). The polar end of the carotenoid engages in a bonding interaction with Trp M75 at the opposite end of the binding pocket (McAuley et al., 2000), and this is also shown in Fig. 3.
The RC from carotenoid-containing strains of Rba. sphaeroides can be engineered to be carotenoid-deficient by a single mutation. The carotenoid is exposed at the intramembrane surface of the RC at both ends of its curved binding pocket, one end passing between the A and B helices of the M-polypeptide and the other between the C helix of the M-polypeptide and the phytol tail of the BB accessory BChl. A Gly, M71, is located at the point at which the carotenoid emerges between the A and B helices of the M-polypeptide, and mutagenesis to a larger Leu (GM71L) blocks the binding pocket with the result that carotenoid is not incorporated into the GM71L RC during assembly (Ridge et al., 1999). The alpha-carbon of this Gly is shown as a large sphere in Fig. 3. An X-ray crystal structure of the GM71L RC, determined to a resolution of 2.65 Å, showed that the carotenoid binding pocket was empty (Ridge, 1998), which could indicate that the LDAO depicted in Fig. 3 inserts from the ‘Gly M71 end’ of the carotenoid binding pocket, consistent with the gatekeeper role of Phe M162 (see above). The principal function of the RC carotenoid is to quench the energy of BChl triplet states before they can react to produce singlet oxygen, and so this GM71L mutation provides a means to engineer a RC in which BChl triplet states are long-lived, as well as providing a means of expressing a carotenoid-less RC in the presence of carotenoid-containing antenna complexes.
Chapter 16
Reaction Center Structural Plasticity
B. Quinone Exclusion Occupancy of the QB site in purified RCs is variable, due to the dissociable nature of the QB quinone, but the structure of the binding pocket does not seem to be affected by the presence or absence of the quinone. In the case of the QA quinone, mutations have been reported that either completely prevent incorporation of the quinone during assembly of the RC, or loosen its binding to such an extent that the quinone is lost on purification of the complex. The best characterized of these is an Ala to Trp change at the M260 position (AM260W), that causes the RC to assemble without a QA quinone by filling up space in the binding pocket normally occupied by the quinone head-group. The absence of quinone at QA in both membrane-bound and purified RCs has been demonstrated through a variety of spectroscopic techniques (Ridge et al., 1999; Wakeham et al., 2003, 2004) and X-ray crystallography (McAuley et al., 2000). This mutation blocks electron transfer along the A-branch of RC cofactors, and leaves the QB cofactor as the sole quinone in the complex, and so has been used by a number of research groups looking into mutations that promote B-branch electron transfer (for a review, see Wakeham and Jones, 2005, and also Paddock et al., 2005). Structural changes caused by steric exclusion of the QA quinone are restricted to part of a short loop of amino acids connecting the membrane-spanning E-helix of the M subunit with the de-helix (M256M260), and residue Gln H32 (McAuley et al., 2000). A chloride ion is also incorporated into the vacant QA binding pocket (see below). Although not characterized in as much depth, the mutation Ala M248 to Trp also causes the RC to assemble without a QA quinone (Ridge et al., 1999). A second class of mutations loosen binding of the QA quinone such that, although the QA site is wholly or partially occupied in membrane-embedded RCs, the QA site is unoccupied in detergent purified complexes. This has been achieved by mutating Trp M252 to a range of alternatives in Rb sphaeroides (Stilz et al., 1994), or the equivalent Trp M250 to Val (TM250V) in Rba. capsulatus (Coleman and Youvan, 1990). The structural consequences of mutation of this Trp residue are not known, but the TM250V change has been used as a QA-removing mutation in studies of B-branch electron transfer (Laible et al., 1998; Kirmaier et al., 2003; Laible et al., 2003). In the cases of mutation of Trp M252 to Phe and Tyr it is possible to reconstitute quinone into the QA site,
301 but this does not occur if this Trp is replaced by a non-aromatic amino acid (Stilz et al., 1994). In line with this, incubation of the purified TM250V RC with 30 – 40 molar equivalents of ubiquinone-6 achieves selective reconstitution of the QB ubiquinone (Laible et al., 1998, 2003; Kirmaier et al., 2003). C. Bacteriochlorin Exclusion Although they are larger than either quinone or carotenoid, certain of the bacteriochlorin cofactors can also be excluded from the RC by single point mutations. The best characterized of these is Ala M149 to Trp (AM260W), which causes the Rba. sphaeroides RC to assemble without the B-branch BPhe (Watson et al., 2005). This absence of the HB cofactor, which is evident from the X-ray crystal structure and absorbance spectra of membrane-bound RCs, has no obvious effect on function. The reported 3.4 Å resolution crystal structure (Watson et al., 2005), and a subsequent structure at 2.2 Å resolution (Watson, 2005), show that the absence of the HB BPhe has very little impact on the structure of the binding pocket and the neighboring cofactors. This is illustrated in Fig. 4, which shows the structure of the RC in the vicinity of the HB binding site in the wild-type and AM260W RC. The only significant change in the structure of the RC is a change in the conformation of the last 25 or so carbon atoms of the phytol chain of one of the special pair BChls (PB) in a proportion of the RCs in the crystal. In the wild-type RC this phytol chain projects laterally out of the RC between the phytol chain of HB and the BB cofactor (mid-grey sticks in Fig. 4b). In the AM149W RC the phytol tail of PB adopts roughly this conformation in around 40% of the RCs, and in the remainder occupies space deep in the HB binding pocket that is vacated by loss of the HB cofactor (shown as dark-grey sticks projecting vertically in Fig. 4b). This aside, the loss of the BPhe cofactor has very little impact on the structure of the HB binding pocket and neighboring cofactors, as can be seen by comparing the protein topology in Figs. 4a and 4b. This lack of any gross structural changes suggests a rather rigid protein structure, which would be consistent with the finding that many of the RC cofactors can be extracted from the complex and replaced by alternatives (see below). As with the carotenoid binding pocket it is not clear what fills the remainder of the HB pocket in the membrane-bound AM149W RC, but presumably lipid chains can extend into the cavity that is created.
302
Michael R. Jones
Fig. 4. Region around the HB site in the wild-type and AM149W RC complexes from Rba. sphaeroides. The protein is shown as a solid light-grey object, and the BB and QB cofactors as a solid mid-grey object. In (a) the HB cofactor is shown as dark-grey sticks and the phytol chain of PB as mid-grey sticks. In (b) the phytol chain of PB in the normal conformation is shown as mid-grey sticks (white arrow) and the phytol chain of PB in the alternate conformation is shown as dark-grey sticks (black arrow). See also Color Plate 7, Fig. 11.
A second mutant RC that lacks one of the RC BPhes, termed DLL, is discussed in Section V.A below. Two ‘dimer-less’ mutants of the Rba. sphaeroides RC have been described that appear to lack the primary donor BChls. The simplest of these has a Val to Arg mutation at position L157, which is located close to ring III of the PA BChl (Jackson et al., 1997). Purified VL157R RCs have a BChl:BPhe ratio of 1.25, which is consistent with the loss of an average of 1.5 BChls per RC, and the QY absorbance band of the P BChls is completely absent from the room temperature absorbance spectrum. This suggests that most, but perhaps not all, of the P BChls are absent from purified VL157R RCs. The second dimer-less RC has a triple glycine mutation (HL168G/HL173G/ HM202G – Triple-G) that removes the three His residues that engage in bonding interactions with the P BChls (Moore and Boxer, 1998). Purified RCs have a BChl:BPhe ratio of 1.1 which could indicate the loss of up to two BChls per RC, and again absorbance spectra of the Triple-G RC lack a P QY band. IV. Cofactor Replacement A. Bacteriochlorin Replacement Much of the site-directed mutagenesis carried out on
the RC has been aimed at modulating the properties of the cofactors, particularly in the case of the bacteriochlorins responsible for primary charge separation. So for example extensive use has been made of mutations that alter the redox potential of BChl or BPhe, or which substitute BChl for BPhe or vice versa (see below). An alternative approach, applied by Scheer and co-workers in particular, has been to use biochemical treatments to replace the native BChl a or BPhe a by an alternative bacteriochlorin. This subject and the underlying methodology has been reviewed in detail by Scheer and Struck (1993) and Scheer and Hartwich (1995), and so the following will be confined to a brief résumé of what is possible with this approach. The first reports of successful BChl replacement concerned the accessory cofactors BA and BB of the Rba. sphaeroides complex and involved incubating carotenoid-less R-26 RCs (see above) with a 20fold excess of the replacement bacteriochlorin for 90 min at 42.2 °C in Tris buffer containing 0.08 % LDAO and 10 % methanol. Similar experiments with wild-type RCs achieved selective replacement of the BA BChl (Scheer and Hartwich, 1995). Scheer and Struck (1993) have commented that temperature is a crucial factor in these exchange experiments, it being necessary to heat the protein to close to its ‘melting temperature’ in order to induce sufficient flexibility in the protein structure that the native cofactors can
Chapter 16
Reaction Center Structural Plasticity
exchange with the excess of replacement cofactor in solution, but without causing irreversible structural changes or denaturation. Franken and co-workers have reported that selective exchange of BA can be achieved in R26 RCs through variation of the incubation temperature (Franken et al., 1997a,b). Exchange of the BPhe cofactors HA and HB has also been achieved using this type of approach (Meyer and Scheer, 1995; Franken et al., 1997a), and bacteriochlorin exchange has also been reported for the RCs from Rsp. rubrum and Blc. viridis (Scheer and Hartwich, 1995). A variety of bacteriochlorins has been successfully introduced at the accessory BChl and BPhe locations, including 132-hydroxy, 3-(α-hydroxyethyl), 3-vinyl and (3-vinyl)-132-hydroxy derivatives of BChl a. A range of other bacteriochlorins has also been tested but with negative results, suggesting that the cofactor binding pockets do exert some selectivity on the type of bacteriochlorin that can be bound (see Storch et al., 1996, for experiments on stereoselectivity). The native Mg bacteriochlorins have also been replaced with Zn- and Ni-BChl a analogs in order to further vary redox potential (Scheer and Hartwich, 1995). It has not proven possible to replace BChl a with chlorophyll a, perhaps due to a lack of flexibility of the macrocycle (Scheer and Struck, 1993), but the native BPhe a cofactors have been successfully replaced by pheophytin a, pheophytin b and a number of derivatives thereof with altered substituent groups, as well as derivatives of BPhe a (Meyer and Scheer, 1995; Scheer and Hartwich, 1995). RCs with replaced bacteriochlorins have been used for a variety of purposes. Complexes with replaced accessory BChls have been used to examine ultrafast energy transfer and relaxation events amongst the accessory and P BChls (Hartwich et al., 1995a), assignments of the QY absorbance bands of the BA, BB and P BChls (Hartwich et al., 1995b), the role of BA in primary charge separation (Finkele et al., 1992, Sporlein et al., 2000), and the role of BB in the transfer of triplet state energy to the carotenoid (Frank et al., 1993, 1996). This last study involved replacement of BB with 132-hydroxy-Zn-BChl a or (3-vinyl)-132-hydroxy-BChl a in carotenoid-less R-26 RCs followed by reconstitution of spheroidenone into the carotenoid binding pocket. Complexes with pheophytin a in place of the BPhe a have been used to examine the issue of whether primary charge separation between P and HA is a two-step reaction involving BA– as a discrete intermediate, the pheophytin a providing a means of
303 altering the free energy of the P+HA– state to a point approximately isoenergetic with that of P+BA– (Shkuropatov and Shuvalov, 1993; Schmidt et al., 1994; Kennis et al., 1997; Huber et al., 1998, Sporlein et al., 2000). BPhe a replacement by pheophytin a, by the same method as that applied to R-26 RCs (Meyer and Scheer, 1995; Franken et al., 1997a), has also been applied to a mutant Rba. sphaeroides RC with a relatively long-lived primary donor singlet excited state (Shkuropatov et al., 2003). To date there does not appear to have been a report of exchange of the primary donor BChls by the technique outlined above. These BChls are rather less exposed to the intramembrane phase than the accessory BChls or BPhes, and therefore it is conceivable that they are less able to detach from the RC during the ‘protein breathing’ (Scheer and Struck, 1993) brought on by moderate heating of RCs in detergent. Partial (~30%) replacement of the primary donor BChls by Zn-BChl a has been reported by Kobayashi and co-workers using a somewhat different procedure in which the bacteriochlorin cofactors were released from Rba. sphaeroides R26 RCs in buffer with 5% octylglucoside using a 3-fold volume of acetone that contained an excess of either BChl a or Zn-BChl a (Kobayashi et al., 2004). The sample was then freezedried and the dried material was resuspended in water. A solution of Rba. sphaeroides lipids in buffer with 0.5 % deoxycholate was then added, and the detergent removed by dialysis. This procedure produced RCs at a yield of around 30% with a near-to-normal QY absorbance spectrum when added BChl a was used, and produced RCs with an altered spectrum when Zn-BChl a was used. HPLC analysis of these Zn-BChl a containing RCs indicated a 6:4 ratio of BChl to Zn-BChl, and simulation of absorbance and CD spectra suggested that 30% of the primary donor BChls had been replaced by Zn-BChl. Although the exchange is not stoichiometric, and the RCs containing Zn-BChl require further analysis, this appears to be an intriguing method for cofactor exchange involving reconstitution of extracted cofactors into the protein concurrent with reconstitution of RCs into liposomes. B. Quinone Replacement One of the first procedures to be developed for significantly altering the structure of the purple bacterial RC involved removal of the QA quinone from its intraprotein binding pocket, and its replacement
Michael R. Jones
304 by alternative quinone or non-quinone molecules (Okamura et al., 1975). Such QA-removed and QAsubstituted RCs have been used for a variety of purposes, but they have made particular contributions to study of the mechanism of biological electron transfer (Moser et al., 1992), through studies of the free energy dependence of the HA– to QA and QA– to P+ reactions. Changing the quinone species at the QA site alters the mid-point reduction potential of this cofactor, and so alters the free energy of the P+QA– state relative to the ground state or other radical pairs such as P+HA– or P+QB– . Incubation of purified RCs at 25 °C with LDAO and the quinone site inhibitor o-phenanthroline causes dissociation of the QA quinone without causing irreversible denaturation of the protein (Okamura et al., 1975). Quinone removal can be achieved by binding RCs to an ion exchange column and washing with the quinone removal buffer. Either native or non-native quinone species can then be reconstituted into the site by incubation of the QA-depleted RCs with an excess of the quinone in question (Okamura et al., 1975). A number of protocols have been described for QA depletion and reconstitution in both Rba. sphaeroides (Okamura et al., 1975; Gunner at al, 1986; Woodbury et al., 1986; Kálmán and Maroti, 1994) and Blc. viridis (Liu and Hoff, 1990; Breton, 1997), including procedures also involving depletion or replacement (with Zn2+) of the non-heme Fe (Liu et al., 1991b) (see below for more details). Non-quinone species have also been substituted into the QA site and tested for QA function (Warncke and Dutton, 1993a). RCs modified in this way have been used for a variety of purposes, including examining characteristics of the reduction of QA by HA– (Gunner and Dutton, 1989; Schelvis et al., 1992; Warncke and Dutton, 1993a; Laporte et al., 1995), the relationship between the rate and free energy for P+QA– recombination (Gopher et al., 1985; Kleinfeld et al., 1985; Gunner at al, 1986; Woodbury et al., 1986; Lersch and Michel-Beyerle, 1987; Kálmán and Maroti, 1994; Turzo et al., 2000; Xu and Gunner, 2000; Hucke et al., 2002), proton binding on reduction of QA (Kálmán and Maroti, 1994), P+QB– recombination (Labahn et al., 1995; Allen et al., 1998; Schmid and Labahn, 2000), features of the P+QA– radical pair state (Sebban, 1988; McPherson et al., 1990; van den Brink et al., 1994a; Morris et al., 1995; Hulsebosch et al., 1999), electron transfer from QA to QB (Graige et al., 1996, 1998, 1999; Li et al., 1998; Paddock et al., 1999b; Li et al., 2000b), affinity of the QB site for
various quinones (Giangiacomo and Dutton, 1989), enthalpy and volume changes on P+QA– formation using photoacoustic spectroscopy (Edens et al., 2000) and detailed analysis of the quinone environment using isotopically-labeled quinones (Breton et al., 1994a,b,c; Brudler et al., 1994; van den Brink et al., 1994b; van Liemt et al., 1995). The quinone replacement approach has also been used to study the affinity of the QA site for quinone (Warncke and Dutton, 1993b; Mallardi et al., 1998), including the influence of different types of tail (Warncke et al., 1994), and various substituents on the head-group that affect size and/or charge (Warncke and Dutton, 1993a; Hucke et al., 2002; Katz et al., 1991; Madeo and Gunner, 2005). In one case, for a RC with anthraquinone substituted into the QA site, the X-ray crystal structure has been determined to a resolution of 2.4 Å. The structure showed no detectable changes to the conformation of the backbone and residues that form the QA site (Kuglstatter et al., 2000), the three ring anthraquinone being oriented approximately coplanar to the native ubiquinone. Hydrogen bonding of the anthraquinone to the surrounding protein via its carbonyl groups was similar to that seen for ubiquinone in the wildtype RC (Kuglstatter et al., 2000). Electron density in the region of the isoprenoid quinone tail, which is absent in anthraquinone, was attributed to a molecule of LDAO (Kuglstatter et al., 2000). C. Carotenoid Replacement Experiments involving the reconstitution of non-native carotenoids have mainly involved carotenoid-less strains of purple bacteria such as Rba. sphaeroides strains R-26 and R-26.1 and Rsp. rubrum strain G9. A number of protocols have been described for the reconstitution of non-native carotenoids into such carotenoid-less RCs (Boucheret al., 1977; Agalidis et al., 1980; Davidson and Cogdell, 1981), and in the case of the R-26.1 RC this is achieved by sonication of a mixture of RCs and added carotenoid in the dark and on ice, followed by column chromatography to remove excess unbound carotenoid (Roszak et al., 2004). RCs with reconstituted carotenoids have been used to examine the influence of substituent groups and the extent of the conjugated π electron system on the ability of the carotenoid to carry out its photoprotective function, quenching of primary donor triplet states (Chadwick and Frank, 1986; Frank et al., 1986; Farhoosh et al., 1997), the role
Chapter 16
Reaction Center Structural Plasticity
played by the BA BChl in triplet transfer (Frank et al., 1993, 1996) and issues relating to the 15–15´-cis stereochemistry of the carotenoid in Rba. sphaeroides (Gebhard et al., 1991; Kok et al., 1994, 1997; Bautista et al., 1998). In the most recent studies Yanagi et al. (2005) have used R-26.1 RCs reconstituted with 3,4-dihydrospheroidene to examine the effect of the carotenoid cofactor on the electrostatic environment of the primary donor BChls, and Roszak and co-workers have used R-26.1 RCs reconstituted with spheroidene or 3,4-dihydrospheroidene to examine structural factors that ensure unidirectional binding of the carotenoid cofactor into its elongated, tubular binding pocket (Roszak et al., 2004). D. Iron Removal and Replacement Perhaps the most extensively employed structural modification of the RC is the removal of the nonheme iron atom, and its replacement with alternative divalent metals such as zinc. This iron replacement is a prerequisite to studies of the detailed structure and properties of the radical anions QA– • and QB– • through electron paramagnetic resonance spectroscopy (EPR), and related techniques such as ENDOR (electron nuclear double resonance) and ESEEM (electron spin-echo envelope modulation). This is because in the wild-type RC these radicals are strongly magnetically coupled to the high-spin Fe2+, and so only very broad EPR spectra can be obtained (see Utschig and Thurnauer (2004) for a review). Removal of the iron, or its replacement with the diamagnetic Zn2+, yields more detailed spectroscopic information. A number of protocols have been developed for both iron removal and metal exchange in Rba. sphaeroides (Debus et al., 1986; Utschig et al., 1997, 2005; Poluektov et al., 2005) and Blc. viridis (Gardiner et al., 1999; Utschig et al., 2005). In general these protocols involve incubating RCs with o-phenanthroline and a high concentration of lithium- or potassium thiocyanate to prepare Fe-removed RCs, followed by addition of ZnCl2 or ZnSO4 to prepare Zn-replaced RCs. Hermes and co-workers have also recently described a procedure in which Rba. sphaeroides was grown on a malate yeast medium in which FeSO4 was replaced with MnSO4 (Hermes et al., 2006 ). This resulted in complete replacement of the Fe atom with Mn, negating the need to carry out a biochemical exchange (see also discussion in Debus et al., 1986). Although it has only modest effects on the rate of electron transfer from QA– to QB (Debus et al., 1986),
305 Fe2+ removal causes an approximately 20-fold reduction in the rate of electron transfer from HA– to QA (Kirmaier et al., 1986). If Zn2+ is substituted into the site then a normal rate can be obtained for this reaction, depending on the procedure used (see Utschig et al., 1997, for a discussion and a review of earlier data). This effect on the HA– to QA reaction has been attributed to a change in reorganization energy for this step following Fe2+-depletion (Tang et al., 1999). Fe-removed and Zn-substituted RCs have been used for a variety of purposes; see the following for reviews of the older literature (Stehlik and Mobius, 1997; Lubitz and Feher, 1999; Möbius, 2000; Lubitz et al., 2002; Utschig and Thurnauer, 2004). Most recently, these preparations have been used to study how QA– and QB– hydrogen bond to the surrounding protein (Lubitz and Feher, 1999; Schnegg et al., 2002; Flores et al., 2006), protein reorganization on photogeneration of P+QA– (Zech et al., 1997; Borovykh et al., 1998; Poluektov et al., 2001; Poluektov et al., 2005), properties of the P+HA– state (Hulsebosch et al., 1999), light-induced conformational changes in and around the QB pocket (Utschig et al., 2005), the temperature dependence of the conformational flexibility of RCs (Borovykh et al., 2005), and spin-spin interactions between the iron and quinone or BPhe radicals (Calvo et al., 2002) and between QA– and QB– (Calvo et al., 2000, 2001). Fe2+-removed and Cu2+- substituted RCs have also been used to identify a Cu2+ binding site on the surface of the RC that modulates the rate of electron transfer from QA– to QB (Utschig et al., 2000; Utschig et al., 2001), similar to the Zn2+ binding site that also exerts this effect (Utschig et al., 1998). E. Genetic Approaches to Cofactor Replacement The use of site-directed mutagenesis to change the chemical character of the RC cofactors has focused on the bacteriochlorins, involving replacement of BChl by BPhe or vice versa. Each of the primary donor BChls has been replaced by BPhe by mutating the His residue that provides the axial ligand to the central Mg to a residue that cannot provide such a ligand, producing a BChl:BPhe heterodimer. This is achieved by mutating either His M202 to Leu for PA or His M173 to Leu for PB (Bylina and Youvan, 1988; Kirmaier et al., 1988). Mutation of His M202 to Glu also produces a heterodimer phenotype (Nabedryk et al., 2000). These heterodimer mutants have been heavily characterized, and their
306 properties and uses have been covered in a number of reviews (Coleman and Youvan, 1990; Woodbury and Allen, 1995; Parson, 1996; Hoff and Deisenhofer, 1997). Most recently they have been used in studies assessing the contribution of electronic coupling to B-branch electron transfer (Kirmaier et al., 2005), examining how the protein influences electronic asymmetry in the primary donor BChls (Moore et al., 1999), investigating alternative pathways for charge separation initiated by excitation of BA (van Brederode et al., 1999) and examination of the effects of hydrogen bond mutations on the electronic structure of the heterodimer (Nabedryk et al., 1998). The heterodimer mutants have also been structurally characterized using FTIR spectroscopy (Albouy et al., 1997; Nabedryk et al., 2000) and the X-ray crystal structure of the HM202L mutant has been reported (Camara-Artigas et al., 2002). The latter shows good preservation of the structure of the bacteriochlorin dimer. The mutation does produce very small shifts in the positions of backbone and side-chain atoms for residues M196-M206, but no longer range changes in the structure of the protein component. A water molecule normally located between the side-chain of His M202 and the keto carbonyl of BA is absent in the HM202L mutant; this water has a strong influence on the rate of primary charge separation (Potter et al., 2005) and so the effects of the heterodimer mutation on this process are likely to be due to a combination of the replacement of the BChl by BPhe and loss of this water. In the case of the accessory BChls, mutagenesis of His M182 to Leu causes replacement of the BB BChl with BPhe (Katilius et al., 1999). This mutant has been used in studies of B-branch electron transfer (for a review, see Wakeham and Jones, 2005) and has also been used to demonstrate the role of BB as an intermediate in singlet energy transfer from the carotenoid to P (Lin et al., 2006). On the A-branch, mutagenesis of His L153 to Leu in Blc. viridis causes the BA BChl to be replaced by BPhe. The main effect of this is to produce a very long-lived P+BA– state (Arlt et al., 1996). In Rba. sphaeroides the equivalent mutation (or a HL153F change) does not produce a clean replacement of BChl by BPhe, but rather produces more complex effects including alterations in the properties of the P BChls (Katilius et al., 2004). His L153 has also been replaced by Asp, Glu, Gln, Gly, Ser, Tyr and Val, and purified RCs from many of these mutants also show evidence of significant changes to the properties of P
Michael R. Jones in addition to blue-shifts of the QX and QY absorbance bands of BA (Katilius et al., 2004). Thus it would appear that replacement of His L153 without causing larger scale structural disturbance is possible in Blc. viridis but not in Rba. sphaeroides, possibly because the presence of the cytochrome subunit in the former helps to stabilize the structure of the L-polypeptide on the periplasmic side of the membrane. For the BPhe cofactors, mutagenesis of Leu M214 to His in Rba. sphaeroides (Kirmaier et al., 1991) or the equivalent Leu M212 to His in Rba. capsulatus (Heller et al., 1995b) causes replacement of BPhe by BChl. The properties of this so-called beta mutant have again been characterized in depth, and have been covered in a number of reviews (Woodbury and Allen, 1995; Parson, 1996; Hoff and Deisenhofer, 1997). These mutations have been particularly useful in studies of B-branch electron transfer (see Wakeham and Jones, 2005, for a review and for more recent reports see Kirmaier et al., 2005; Paddock et al., 2005; Kee et al., 2006). In addition the X-ray crystal structure of a Rba. sphaeroides RC with multiple mutations including the LM214H change has been reported, showing a clean replacement of BPhe by BChl with no additional significant changes in the structure of the protein or neighboring cofactors (Paddock et al., 2005). The HA BPhe is also replaced by BChl following mutation of Ala L124 to His, and a double mutation of Phe L121 to His and Phe L97 to either Val or Cys (Heller et al., 1995a). The structural basis for the latter changes is less obvious than for the more heavily characterized LM214H mutant, as the Phe at the L121 position does not sit over the center of the BPhe macrocycle, and so the new His could not donate a ligand to Mg without some more complex structural changes. One possibility is that water molecules are involved in the binding of BChl (Heller et al., 1995a) (see below). On the B-branch, mutation of Leu L185 to His causes replacement of the HB BPhe by BChl (Watson et al., 2005). This residue is the symmetrical equivalent of Leu M214 on the A-branch. In addition to site-directed mutagenesis, interruption or augmentation of the biosynthetic pathway for BChl has been used to alter the cofactor composition of the RC. In particular, this approach has been used to investigate whether the chemical nature of the ‘tail’ of BChl, a C20 isoprenoid, has an effect on assembly of the RC or its function. Bollivar and co-workers have reported a strain of Rba. capsulatus that produces BChl a esterified with geranylgeraniol rather than the
Chapter 16
Reaction Center Structural Plasticity
native phytol due to a lesion in the gene bchP (Bollivar et al., 1994). This gene encodes a biosynthetic enzyme, termed BchlaGG reductase, that is responsible for hydrogenation of the geranylgeraniol tail to phytol through the removal of three (of four) C=C bonds. The larger number of double bonds in geranylgeraniol is expected to affect properties such as the flexibility of the BChl a side chain. The mutant strain, designated DB391, was capable of photosynthetic growth at a six-fold reduced rate which appeared to be due to (again six-fold) lowered levels of the photosynthetic apparatus (Bollivar et al., 1994). RCs isolated from the DB391 strain showed largely normal absorbance and functional properties, apart from a small blueshift of the absorbance band of the accessory BChls that could indicate some small changes in structure of the complex, but showed evidence of a markedly decreased stability. This combination of properties would be consistent with a RC in which the tetrapyrrole rings are correctly positioned in the structure, producing normal functional properties, but where the esterified side-chains are incorrectly packed into the structure, compromising the stability of the complex. Addlesee and Hunter (1999) have produced an equivalent Rba. sphaeroides mutant with a lesion in bchP. This mutant showed a 27 % reduction in photosynthetic growth rate and around a 30% reduction in levels of light harvesting complex, consistent with the findings of Bollivar et al., but perhaps indicating a less serious effect on levels of RC complex in Rba. sphaeroides compared to Rba. capsulatus. In Rhodospirillum (Rsp.) rubrum the BChl synthesis pathway naturally produces BChl a esterified with geranylgeraniol as the end product for incorporation into RCs. Addlesee and Hunter (2002) have shown that heterologous expression of the bchP gene from Rba. sphaeroides in Rsp. rubrum causes the photosystem to assemble with BChl a esterified with phytol. The RCs and LH1 complexes from this strain were capable of supporting photosynthetic growth and showed only modest changes in spectroscopic properties (Qian et al., 2003). V. Helix Symmetrization The majority of spectroscopic studies aimed at probing RC mechanism are precluded in wild-type strains due the presence of the BChl and/or carotenoid cofactors of the antenna complexes. RCs therefore have to be removed from the membrane and purified, a procedure which is straightforward for wild-type
307 RCs but which can provide a stumbling block for studies on less stable mutant complexes. To overcome this problem, genetic systems involving gene deletions or point mutations have been devised for both Rba. sphaeroides and Rba. capsulatus that remove the LH1 and LH2 antenna complexes, isolating the RC as the sole BChl-containing complex in the cell. The most widely used of these are the U43 strain of Rba. capsulatus, first constructed by Youvan and co-workers (Youvan et al., 1985), and the DD13 and DD13/G1 strains of Rba. sphaeroides, first constructed by Hunter and co-workers (Jones et al., 1992a,b). Strains of this type have proven particularly useful for studying RCs with large scale mutations, such as those described in the rest of this section. A. DLL, Cousins and Offspring The two-fold rotational structural symmetry displayed by the RC is in marked contrast to its strong functional asymmetry, and this intriguing aspect of RC design prompted a number of experiments aimed at ‘symmetrizing’ the RC through large scale mutations. In addition to producing a number of interesting phenotypes, this approach has illustrated the structural robustness of the purple bacterial RC, particularly when embedded in the native membrane. The best known and most extensively characterized RC of this type is the so-called DLL mutant of Rba. capsulatus, in which residues 192 to 217 of the M-polypeptide were replaced by the structurallyequivalent residues 165 to 190 of the L-polypeptide, a change equivalent to 13 single point mutations (Robles et al., 1990). The resulting RC assembles without a BPhe at the HA site, blocking A-branch electron transfer and greatly extending the lifetime of the primary donor excited state P*. Figure 5a shows the central location of the residues in question, in a model of the L- and M-polypeptides of the Rba. capsulatus RC constructed by Foloppe and coworkers (Foloppe et al., 1995). The replaced section of the M-polypeptide includes part of the D-helix and a loop connecting this to a helix exposed at the periplasmic surface of the complex (light-grey ribbon), and this crosses the symmetrically-equivalent section of the L-polypeptide (dark-grey ribbon) in the core of the RC. Figure 5b shows a more detailed view of the replaced region of the M-polypeptide and the adjacent HA BPhe (left), and an equivalent view of the D-helix of the L-polypeptide. In the region adjacent to the HA BPhe the most obvious differences between the two helices are a Trp in place of Leu at
Michael R. Jones
308
Fig. 5. Sequences involved in the DLL mutant from Rba. capsulatus. (a) The LM heterodimer in a model of the Rba. capsulatus RC (see text). In the DLL mutant residues 192-217 of the M-polypeptide (light-grey ribbons) are replaced by residues 165–190 of the Lpolypeptide (dark-grey ribbons). The location of residue L237 is indicated by the spheres. (b) Residues M192-217 and the adjacent HA BPhe (left) and an equivalent view of residues L165-190 (right). Color coding is as for (a) and residues mentioned in the text are labeled.
the position closest to the center of the HB macrocycle (locus M212), and a Thr in place of Gly at a position one turn further along the helix (locus M209) (Fig. 5b). It has been suggested that these two differences could be responsible for the lack of a HA BPhe in the DLL RC, the larger Trp and Thr preventing insertion of the BPhe into the complex (Foloppe et al., 1995). In support of this idea, the strain containing the DLL RC was able to regain the capacity to grow under photosynthetic conditions (at varying rates) after undergoing one or more spontaneous mutations that restored assembly of the RC with an HA BPhe (Robles et al., 1990). These mutations were
WM212L and WM212L/IM204T, which remove the large Trp residue adjacent to the expected position of the HA macrocycle, and MM216V and MM216T which reduce the volume of the side chain one turn of the helix away from this Trp (Fig. 5b). However, in Rba. sphaeroides at least, mutation of the Leu/Gly pair at the equivalent of the M212/M209 positions to Trp/Met does not cause the RC to assemble without this BPhe (unpublished observations). The principal use of the DLL RC has been in examining the properties of the P* state, and, because it is not stable in detergents such as LDAO, these studies have been carried out on DLL RCs embedded in native antenna-deficient membranes (Vos et al., 1991, 1993; Foloppe et al., 1995). Most recently this large scale mutation has been used in studies of B-branch electron transfer employing RCs solubilized in the detergent Deriphat 160-C, and purified using a His-tag (Chuang et al., 2006). The DLL mutation was combined with three additional changes, His M195 to Phe and His L168 to Phe to restore the mid-point redox potential of the P/P+ couple (Em P/P+) to close to the value in wild-type RCs, and a Phe L181 to Tyr change which, together with the Tyr M208 to Phe change already in the DLL multiple mutation, promotes B-branch electron transfer. The resulting mutant (named DLL-FYLFM) has a yield of 70% electron transfer along the B-branch to form P+HB–, more than twice that reported for any other mutant RC (Chuang et al., 2006). DMM helix-symmetrization and DLM helix-swap mutants have also been constructed, but these RCs are not stable, even when in the native membrane (Robles et al., 1990). This illustrates the point that although the RC can tolerate large scale mutation, whether or not the assembly or stability of the RC is compromised depends on the exact location and extent of the mutations, and symmetrical mutations do not always yield complementary results. Finally, a strain containing the DMM RC was able to regain the capacity for photosynthetic growth after undergoing the mutation FL187L or AL237P. The FL187L mutation restores the native, smaller side chain on the opposite side of the helix to the BPhe binding pocket, suggesting that incorrect helix packing could be the source of the instability of the DMM mutant. B. Sym-1 and Siblings In the so-called sym-1 mutant of Rba. capsulatus, residues 187 to 203 of the M-polypeptide were replaced with residues 160 to 176 of the L-polypeptide,
Chapter 16
Reaction Center Structural Plasticity
corresponding to 9 point mutations (Taguchi et al., 1992). The most notable effect of this large scale mutation was a large increase in Em P/P+, an effect common to the DLL mutant and attributed to replacement of Phe M195 by His (Stocker et al., 1992). This mutation causes donation of a new hydrogen bond to the acetyl carbonyl group of the PM BChl which could increase Em P/P+ (Wachtveitl et al., 1993; Mattioli et al., 1994, 1995), although Spiedel and co-workers have proposed that the strong increase in Em P/P+ is actually attributable to a charge-dipole interaction between P and the new His side chain (Spiedel et al., 2002). Subsequently, Taguchi and co-workers reported the construction of nine further symmetrizing mutations that, together with sym-1, cumulatively replaced 113 residues between positions 162 and 280 of the Mpolypeptide by the structurally-equivalent residues from the L-polypeptide (Taguchi et al., 1996). These mutations involved changes of stretches of between 5 and 24 amino acids, corresponding to between 4 and 16 point mutations in any one mutant. Five of these mutants were capable of photosynthetic growth, and of the remaining four only one did not appear to assemble RCs. Photosynthetic growth seemed particularly sensitive to mutations near the quinones, and in two of the non-photosynthetic mutants (sym0 and sym2-2) a photosynthetic-positive phenotype could be restored through a single point mutation in the M-polypeptide. Three of the mutants, sym0, sym2-1 and sym5-2 were subsequently studied by ultrafast transient absorption spectroscopy and the sym2-1 mutant in particular showed a reduced (35%) yield of A-branch charge separation and some evidence of 10–20 % electron transfer along the B-branch (Lin et al., 1996). The sym2-1 RC had a replacement of residues M205 to M210 by their L-polypeptide counterparts, and despite the altered primary photochemistry was still capable of supporting photosynthetic growth (Taguchi et al., 1996). This sym2-1 RC was subsequently used to study spectrally-distinct conformations of the RC that can be interchanged through alterations in ionic strength, detergent concentration or temperature (Eastman et al., 2000). The spectral changes involved the long-wavelength QY band of the P BChls, which also underwent a reversible change of around 30 mV in Em P/P+. C. QAQA and A6D1 Perhaps the most remarkable symmetrizing muta-
309
Fig. 6. Sequences involved in the QAQA and A6D1 mutants from Rba. capsulatus. In the QAQA mutant residues L193-227 (light grey ribbons) from around the QB quinone (light-grey spheres on right) are replaced by residues M220-261 (dark-grey ribbons) from around the QA quinone (light-grey spheres on left). In the A6D1 mutant residues Met M144 and Ala M145 (dark-grey spheres) are replaced by Ile and Ser, respectively.
tion is that seen in the so-called QAQA RC of Rba. capsulatus, in which 35 residues of the L-polypeptide (193–227) are replaced by 42 residues from the M-polypeptide (220–261) in a region of very weak sequence homology between the two (Coleman and Youvan, 1993). The location of the relevant sequences is shown in Fig. 6, which again is based on a model of the Rba. capsulatus RC by Foloppe and co-workers (Foloppe et al., 1995). In each polypeptide the sequence in question corresponds to two loops and a short helix that connect the cytoplasmic ends of the D and E membrane-spanning helices, and forms part of the contact surface between the membrane-embedded LM heterodimer and the cytoplasmic domain of the H-polypeptide. The mutation constructed by Coleman and Youvan replaced the region of the protein that comprises the QB binding pocket with an equivalent region from the QA binding pocket, and according to the alignment presented by Li et al. (2000a) this corresponds to 31 point mutations together with the introduction of an additional seven amino acids. Among the residues mutated by this large scale resculpting are groups known to be key to QB function such as Glu L212, Asp L213 and Ser L223. Perhaps not surprisingly the strain containing the QAQA RC was not capable of photosynthetic growth, but a remarkable result obtained with a derivative of the QAQA RC illustrates another aspect of the structural plasticity of the purple
310 bacterial RC. Returning to definitions of plasticity, in neuroscience the term neuroplasticity refers to the capacity of an area of the brain to assume some of the function of a damaged area. In a (slightly contrived) parallel, the Rba. capsulatus RC shows remarkable structural plasticity in response to the loss of photosynthetic function induced by the QAQA mutation, the RC regaining the capacity to support photosynthetic growth through just two spontaneous point mutations near the HB cofactor (Coleman and Youvan, 1993). These mutations, of Met M144 to Ile and Ala M145 to Ser, are located in a region of M-polypeptide well outside the symmetrized region (see Fig. 6) and arose in response to incubation of the QAQA strain under photosynthetic growth conditions. The resulting so-called A6D1 complex was stable in detergent solution and, despite the extensive mutagenic alteration, showed only modest changes in the characteristics of electron transfer reactions involving QB. The affinity of ubiquinone for the new QB site was only three-fold lower than that in wild-type RC, and the strongest effect was a ~500-fold reduction in the rate of the first electron transfer from QA– to QB (Li et al., 2000a). Although Met M144 is rather close to the head-group of the QB quinone (Fig. 6), in the absence of direct structural information it is difficult to understand how these two point mutations can partially overcome the impairment of function engendered by the large scale QAQA mutation, and produce a QB site that is sufficiently functional to support photosynthetic growth. Studies of mutants of Glu L212 and Asp L213, two key residues in the QB site involved in proton delivery to QB (for reviews, see Okamura et al., 2000; Wraight, 2004), have also shown how photosynthetic function can be restored by mutations somewhat removed from the original mutation site. Around a dozen mutations have been identified that restore photosynthetic capacity to RCs with mutations at these key acidic amino acids, some of which are a considerable distance from the QB site. A feature of many of these mutations is that they either introduce a negative charge or remove a positive charge, indicating that their primary influence is electrostatic (Okamura et al., 2000). However the distance of some of these mutations from the QB site suggests that larger scale structural changes may also be involved, altering the electrostatic contributions made by a number of residues (Sebban et al., 1995; Paddock et al., 1999a; Tandori et al., 2001).
Michael R. Jones VI. Water and Other Unexpected Things in Electron Density Maps For some of the suppressor mutations outlined in the last section it is possible to rationalize the restoration of QB function, but for others this is not easy, particularly in the case of more distant mutations. One possibility that has to be borne in mind for many of these mutations is a change in the number and position of bound water molecules near the QB site. Water, due to its polar character, propensity to form H-bonds and ability to act as a proton donor, can exert significant effects that can be difficult to predict when designing a mutagenesis experiment. As an example, mutation of Ser L223 to Ala or Asn produces a RC with a non-functional QB site, whereas mutation to Gly does not impair QB function (Paddock et al., 1995). Similarly, it has been found that a Glu L212 to Gln mutant is non-functional whereas a Glu L212 to Ala mutant retains function (Miksovska et al., 1997). In both cases it has been proposed that the smaller side chain creates a cavity that is occupied by a water that is able to substitute for the native Ser or Glu (Paddock et al., 1995; Miksovska et al., 1997). This is not possible when the Ser is replaced by a larger Ala or Asn, or when Glu is replaced by Gln. Changes in water structure have also been proposed to explain restoration of QB function by a Leu L227 to Phe mutation (Miksovska et al., 1997). Water is also thought to play a role in BChl binding in mutants where the His residues that donate the axial ligands to the P BChls are replaced by Gly. Mutation of either of these His residues (L173 and M202) to a non-polar side-chain such as Leu or Phe is documented to cause the RC to assemble with a BPhe in place of the native BChl, forming a so-called heterodimer RC with a BChl:BPhe primary electron donor. In order to explain the presence of a largely unperturbed BChl:BChl pair in a HL173G and HM202G mutant, Goldsmith et al. (1996) proposed that a water molecule occupies the cavity created by the His to Gly change and provides the essential fifth ligand to the BChl Mg (see also, Czarnecki et al., 2006). A similar explanation has been put forward by Heller et al. (1995a) to explain replacement of the active branch BPhe by BChl in a FL97V/FL124H double mutant where the new L124 His is poorly positioned to be the ligand donor (see above). In order to be certain as to the extent of any functional influence of new (or excluded) waters and other small molecules in a mutant complex it is advanta-
Chapter 16
Reaction Center Structural Plasticity
311
Fig. 7. Stereo view of the distribution of water molecules in an X-ray crystal structure of the Rba. sphaeroides RC. Waters are shown as dark-grey spheres.
geous to combine spectroscopic investigations with X-ray crystallography. However, even then it may not be easy to be certain about the role played by an individual water. The numbers of water molecules included in the various structures for the purple bacterial RC are highly variable, reflecting variations in the quality of electron density from one dataset to the next. Figure 7 shows the distribution of 368 water molecules included in a recently published structure for a mutant Rba. sphaeroides RC, at 2.4 Å resolution (Potter et al., 2005). Waters are mainly grouped at the periplasmic and cytoplasmic sides of the membrane, with relatively few in the region of the protein that occupies the hydrophobic core of the membrane. However significant numbers of waters are found in the vicinity of the P BChls and the quinones, and as the individual located half-way across the membrane demonstrates it is possible for water molecules to be found buried deep in the intramembrane region of the protein interior. The number and arrangement of waters is particularly varied in parts of the structure where the electron density map is complex, such as surface-exposed areas and regions such as the QB site. The latter is particularly challenging as the quinone can be bound at substoichiometric levels and in at least two binding positions (Lancaster and Michel, 1997; Stowell et al., 1997; Kuglstatter et al., 2001), and the number and position of water molecules
may vary depending on the presence and position of the quinone. Obtaining an accurate picture of the arrangement of water molecules in a particular RC is challenging, but is often important if the properties of a particular mutant are to be fully understood. An example of a mutant RC where the influence of an individual water molecule is revealed is the GM203L RC of Rba. sphaeroides, where a Gly residue located between the primary donor BChls and the BA BChl is changed to a larger Leu (Potter et al., 2005). One result of this mutation is an approximately eight-fold slowing in the rate of primary electron transfer, one of the strongest effects reported for a single point mutation. Immediately adjacent to the Gly side chain at the M202 position is a water molecule that is in hydrogen bond distance to the keto carbonyl group of BA and to the side-chain of His M202. The X-ray crystal structure of the GM203L mutant shows that the larger Leu side-chain takes up the volume normally occupied by this water, excluding it from the structure (Potter et al., 2005). The mutation thus removes any electrostatic or hydrogen bond interactions that exist between this water and the primary electron donor on one side, and the primary electron acceptor on the other. Although the exact reason(s) for the slowing of primary electron transfer in response to removal of this water remain to be determined, and could
312 arise from more than one source (see discussion in Yakovlev et al., 2005), it is clear the water has a strong influence on the kinetics of the reaction. This underscores the importance of treating the solvent as a significant element of the structure of the RC, alongside the protein and cofactors, and the need to be aware of the significant effects that changes in the structure of this solvent can have on the properties of mutant RCs, particularly when larger scale changes in structure are being attempted. An aspect of the X-ray crystal structure of the QAdeficient AM260W RC described above is the presence of a feature in the QA binding pocket assigned as a chloride ion (McAuley et al., 2000; Paddock et al., 2005). This ion fills part of the space usually occupied by the head-group of the QA quinone and is in bonding distance to nitrogens in the side chains of His M219, Trp M252 and the new Trp M260. The presence of this ion was only detected when the X-ray crystal structure of the AM260W RC was solved and its effects on the functional properties of the RC (if any) are unclear. One possibility that has been raised is that a negatively charged ion at this location could influence the binding position of the QB quinone, although this remains speculation. This finding highlights a further hazard of protein engineering that produces significant changes in structure – the potential for additional ions or small molecules of uncertain influence being incorporated into the altered protein structure – and underscores the need for structural data to be obtained where possible. A chloride is also present in almost exactly the same position in the X-ray crystal structure of the AM248W RC, which as mentioned above also lacks the QA quinone (M. R. Jones and P. K. Fyfe, unpublished). As with the AM260W RC, the chloride sits in a cavity that is lined by nitrogen atoms of the side-chains of a His and two Trp residues. The electronegative chloride therefore is not only an appropriate size to fit into this particular cavity, but also can solvate the partial positive charges on the surrounding side-chain nitrogens. On this theme, Takahashi and Wraight (2006) have recently proposed that small molecules can penetrate the interior of the RC protein and influence functional properties by showing that a variety of weak acids can restore proton transfer to the QB quinone in a double mutant RC (Asp L210 to Asn and Asp M17 to Asn) that has a very strong impairment of proton transfer ability. These acidic amino acids are located in the middle of a putative proton transfer pathway that leads
Michael R. Jones from the cytoplasmic surface of the H-polypeptide to the buried QB site, and it has been suggested that smaller weak acids such as azide or fluoride can penetrate into the interior of the RC and partially overcome the lesion created when these acidic amino acids are mutated to Asn. If this is correct then this represents another example of a non-native molecule being incorporated into the structure of the RC and influencing functional properties of the complex. VII. Building New Functionality The first steps have been taken towards building new functionality into the purple bacterial RC, with the engineering of Rba. sphaeroides RCs capable of the photo-oxidation of tyrosine and, more recently, the tight binding and oxidation of manganese. These experiments address a major difference between the purple bacterial RC and Photosystem II, with chlorophyll photo-oxidation in the latter producing water oxidation in the so-called manganese cluster as a result of a series of electron transfer reactions facilitated by an intervening tyrosine residue. To achieve a potential sufficiently oxidizing to extract an electron from a Tyr, Kálmán and co-workers raised the mid-point potential for single electron oxidation of the primary donor BChls by several hundred millivolts to around +0.8 V through mutation of Leu L131, Leu M160 and Phe M197 to His, and Tyr M210 to Trp (Kálmán et al., 1999). Evidence for tyrosine oxidation was then seen in RCs carrying an additional tyrosine substitution at Arg M164, Arg L135 or Phe L167, namely the spectral signature of a tyrosyl radical obtained through EPR spectroscopy (Kálmán et al., 1999, 2003b). Histidine mutations were also introduced near each of these engineered tyrosines, and near the native Tyr L162 and Tyr M193 residues, to provide an acceptor for the phenolic proton of the Tyr that is released on photo-oxidation (Narváez et al., 2004). This study showed that each of the two native or three engineered tyrosines can be photooxidized to form a neutral tyrosyl radical, with yields that were influenced by pH and the availability of a proton acceptor. RCs with the triple His mutation at residues Leu L131, Leu M160 and Phe M197 have also been shown to be capable of oxidizing weakly bound manganese (Kálmán et al., 2003a), and more recently Thielges and co-workers have reported structural modifications of the RC that result in the tight-binding of a
Chapter 16
Reaction Center Structural Plasticity
313
Fig. 8. An engineered Mn2+ binding site on the periplasmic surface of the Rba. sphaeroides RC. The stereo view shows the protein as a solid object except for amino acids (sticks) and waters (mid-grey spheres) involved in binding the Mn2+ (dark-grey sphere). For the amino acids, nitrogen or oxygen atoms are shown in mid-grey. The residues are (left to right) Glu M173, Glu M168, Tyr M164, Asp M288, His M193 and Asp M292.
photo-oxidizable manganese (Thielges et al., 2005). The design of this site was based around residue Glu M173 which occupies an approximately equivalent position to residue Asp 170 of the D1 polypeptide in Photosystem II, a key residue involved in binding of the manganese cluster (see Thielges et al., 2005, for a detailed discussion of design principles). Four mutant RCs with different combinations of Asp and Glu residues near Glu M173 were constructed, using a background that included the LL131H, LM160H, FM197H and YM210W mutations that raise the oxidation potential of P, and the RM164Y mutation that introduces a photo-oxidizable Tyr near the primary donor. Of the four mutants constructed, three were found to bind manganese and the most tightly-binding of these, termed M2, had a dissociation constant of 1 µM and transferred an electron from the manganese to the photo-oxidized primary donor with a lifetime of 12 ms (Thielges et al., 2005). The binding of Mn was found to be pH dependent, and around pH 8 binding was concomitant with the release of two protons (Kálmán et al., 2005). The M2 mutant, which contained the changes Met M168 to Glu and Gly M288 to Asp in addition to those detailed above, was also characterized by Xray crystallography, and the key structural features are depicted in Fig. 8. Although caution has to be
exercised with this structure given a limited resolution (4.5 Å), it shows a manganese ion and two water molecules bound to the periplasmic surface of the RC and engaging in interactions with four acidic residues and a His. Residue Glu M173 is shown on the left in Fig. 8, together with (left to right) the engineered Glu M168 and Asp M288, and the native residues His M193 and Asp M292. Immediately below this surface-exposed binding site is the engineered Tyr M164. Thielges and co-workers have discussed the extent to which this structure matches the original design for a Mn site involving these mutations and, returning to a theme explored above, have commented on the unanticipated role played by water in forming part of the binding site (Thielges et al., 2005). As well as providing interesting insights into the rational design of a new binding site for a redox-active metal cofactor, this work addresses evolutionary connections between the purple bacterial RC and its water oxidizing counterpart from oxygenic photosynthetic organisms. VIII. Conclusions To close, the structural plasticity exhibited by the purple bacterial RC has made a major contribution
314 to our understanding of the mechanism of photosynthetic energy transduction and biological electron transfer. Although often difficult to predict, the tolerance displayed to significant changes in structure and composition makes this protein an ideal tool for uncovering the more deeply-buried secrets of these fundamental biological processes, and opens the door to the engineering of new biological function based around light-activated redox chemistry. Acknowledgments The author thanks laboratory members and collaborators past and present for enjoyable discussions, and the Biotechnology and Biological Research Council for financial support. All Figures were produced using PyMOL (DeLano, 2002) References Addlesee HA and Hunter CN (1999) Physical mapping and functional assignment of the geranylgeranyl-bacteriochlorophyll reductase gene, bchP, of Rhodobacter sphaeroides. J Bact 181: 7248–7255 Addlesee HA and Hunter CN (2002) Rhodospirillum rubrum possesses a variant of the bchP gene, encoding geranylgeranylbacteriopheophytin reductase. J Bact 184: 1578–1586 Agalidis I, Lutz M and Reiss-Husson F (1980) Binding of carotenoids on reaction centers from Rhodopseudomonas sphaeroides R-26. Biochim Biophys Acta 589: 264–274 Albouy D, Kuhn M, Williams JC, Allen JP, Lubitz W and Mattioli TA (1997) Fourier transform Raman investigation of the electronic structure and charge localization in a bacteriochlorophyll-bacteriopheophytin dimer of reaction centers from Rhodobacter sphaeroides. Biochim Biophys Acta-Bioenerg 1321: 137–148 Allen JP, Williams JC, Graige MS, Paddock ML, Labahn A, Feher G and Okamura MY (1998) Free energy dependence of the direct charge recombination from the primary and secondary quinones in reaction centers from Rhodobacter sphaeroides. Photosynth Res 55: 227–233 Alric J, Cuni A, Maki H, Nagashima KVP, Verméglio A and Rappaport F (2004) Electrostatic interaction between redox cofactors in photosynthetic reaction centers. J Biol Chem 279: 47849–47855 Alric J, Lavergne J, Rappaport F, Verméglio A, Matsuura K, Shimada K and Nagashima KVP (2006) Kinetic performance and energy profile in a roller coaster electron transfer chain: A study of modified tetraheme-reaction center constructs. J Am Chem Soc 128: 4136–4145 Arlt T, Dohse B, Schmidt S, Wachtveitl J, Laussermair E, Zinth W and Oesterhelt D (1996) Electron transfer dynamics of Rhodopseudomonas viridis reaction centers with a modified binding site for the accessory bacteriochlorophyll. Biochemistry
Michael R. Jones 35: 9235–9244 Bautista JA, Chynwat V, Cua A, Jansen FJ, Lugtenburg J, Gosztola D, Wasielewski MR and Frank HA (1998) The spectroscopic and photochemical properties of locked-15,15´-cis-spheroidene in solution and incorporated into the reaction center of Rhodobacter sphaeroides R-26.1. Photosynth Res 55: 49–65 Bollivar DW, Wang SJ, Allen JP and Bauer CE (1994) Molecular-genetic analysis of terminal steps in bacteriochlorophyll a biosynthesis — characterization of a Rhodobacter capsulatus strain that synthesizes geranylgeraniol-esterified bacteriochlorophyll a. Biochemistry 33: 12763–12768 Borovykh IV, Dzuba SA, Proskuryakov II, Gast P and Hoff AJ (1998) Light-induced structural changes in photosynthetic reaction centers studied by ESEEM of spin-correlated D+QA– radical pairs. Biochim Biophys Acta-Bioenerg 1363: 182–186 Borovykh IV, Gast P and Dzuba SA (2005) ‘Glass transition’ near 200 K in the bacterial photosynthetic reaction center protein detected by studying the distances in the transient P+QA– radical pair. J Phys Chem B 109: 7535–7539 Boucher F, van der Rest M and Gingras G (1977) Structure and function of carotenoids in photoreaction center from Rhodospirillum rubrum. Biochim Biophys Acta 461: 339–357 Breton J (1997) Efficient exchange of the primary quinone acceptor QA in isolated reaction centers of Rhodopseudomonas viridis. Proc Natl Acad Sci USA 94: 11318–11323 Breton J, Boullais C, Burie JR, Nabedryk E and Mioskowski C (1994a) Binding-sites of quinones in photosynthetic bacterial reaction centers investigated by light-induced FTIR difference spectroscopy — assignment of the interactions of each carbonyl of QA in Rhodobacter sphaeroides using site-specific C13-labeled ubiquinone. Biochemistry 33: 14378–14386 Breton J, Burie J-R, Berthomieu C, Berger G and Nabedryk EC (1994b) The binding-sites of quinones in photosynthetic bacterial reaction centers investigated by light-induced FTIR difference spectroscopy: Assignment binding of the QA vibrations in Rhodobacter-sphaeroides using 18O- or 13C-labeled ubiquinone and vitamin K1. Biochemistry 33: 4953–4965 Breton J, Burie JR, Boullais C, Berger G and Nabedryk E (1994c) Binding-sites of quinones in photosynthetic bacterial reaction centers investigated by light-induced FTIR difference spectroscopy — binding of chainless symmetrical quinones to the QA site of Rhodobacter sphaeroides. Biochemistry 33: 12405–12415 Brudler R, De Groot HJM, van Liemt WBS, Steggerda WF, Esmeijer R, Gast P, Hoff AJ, Lugtenburg J and Gerwert K (1994) Asymmetric binding of the 1-C=0 and 4-C=0 groups of QA in Rhodobacter sphaeroides R26 reaction centers monitored by Fourier-transform infrared-spectroscopy using site-specific isotopically labeled ubiquinone-10. EMBO J 13: 5523–5530 Bylina EJ and Youvan DC (1988) Directed mutations affecting spectroscopic and electron-transfer properties of the primary donor in the photosynthetic reaction center. Proc Natl Acad Sci USA 85: 7226–7230 Calvo R, Abresch EC, Bittl R, Feher G, Hofbauer W, Isaacson RA, Lubitz W, Okamura MY and Paddock ML (2000) EPR study of the molecular and electronic structure of the semiquinone biradical QA–• QB– • in photosynthetic reaction centers from Rhodobacter sphaeroides. J Am Chem Soc 122: 7327–7341 Calvo R, Isaacson RA, Paddock ML, Abresch EC, Okamura MY, Maniero AL, Brunel LC and Feher G (2001) EPR study of the semiquinone biradical QA–• QB– • in photosynthetic reaction cen-
Chapter 16
Reaction Center Structural Plasticity
ters of Rhodobacter sphaeroides at 326 GHz: Determination of the exchange interaction Jo. J Phys Chem B 105: 4053–4057 Calvo R, Isaacson RA, Abresch EC, Okamura MY and Feher G (2002) Spin-lattice relaxation of coupled metal-radical spindimers in proteins: Application to Fe2+-cofactor (QA– •, QB– •, Φ – •) dimers in reaction centers from photosynthetic bacteria. Biophys J 83: 2440–2456 Camara-Artigas A, Magee C, Goetsch A and Allen JP (2002) The structure of the heterodimer reaction center from Rhodobacter sphaeroides at 2.55 Ångstrom resolution. Photosynth Res 74: 87–93 Chadwick BW and Frank HA (1986) Electron-spin-resonance studies of carotenoids incorporated into reaction centers of Rhodobacter sphaeroides R26.1. Biochim Biophys Acta 851: 257–266 Chen X-Y, Yurkov V, Paddock ML, Okamura MY and Beatty JT (1998) A puhA gene deletion and plasmid complementation system for facile site directed mutagenesis studies of the reaction center H protein of Rhodobacter sphaeroides. Photosynth Res 55: 369–373 Chuang JI, Boxer SG, Holten D, Kirmaier C (2006) High yield of M-side electron transfer in mutants of Rhodobacter capsulatus reaction centers lacking the L-side bacteriopheophytin. Biochemistry 45: 3845–3851 Clayton RK and Smith C (1960) Rhodopseudomonas spheroides: High catalase and blue-green double mutants. Biochem Biophys Res Commun 3: 143–145 Coleman WJ and Youvan DC (1990) Spectroscopic analysis of genetically modified photosynthetic reaction centers. Ann Rev Biophys Biophys Chem 19: 333–367 Coleman WJ and Youvan DC (1993) Atavistic reaction center. Nature 366: 517–518 Czarnecki K, Chen L, Diers JR, Frank HA, Bocian DF (2006) Low-frequency resonance Raman studies of the H(M202)G cavity mutant of bacterial photosynthetic reaction centers. Photosynth Res 88: 31–41 Davidson E and Cogdell RJ (1981) The polypeptide composition of the B850 light-harvesting pigment-protein complex from Rhodopseudomonas sphaeroides, R26.1. FEBS Letts 132: 81–84 Debus RJ, Feher G and Okamura MY (1985) LM complex of reaction centers from Rhodopseudomonas sphaeroides R26 — characterization and reconstitution with the H-subunit. Biochemistry 24: 2488–2500 Debus RJ, Feher G and Okamura MY (1986) Iron-depleted reaction centers from Rhodopseudomonas sphaeroides R26.1 — characterization and reconstitution with Fe2+, Mn2+, Co2+, Ni2+, Cu2+, and Zn2+. Biochemistry 25: 2276–2287 Deisenhofer J and Norris JR (eds) (1993) The Photosynthetic Reaction Center. Vols 1 and 2. Academic Press, San Diego DeLano WL (2002) The PyMOL Molecular Graphics System, DeLano Scientific, San Carlos. http://pymol.sourceforge.net/ (July 06, 2007) Eastman JE, Taguchi AKW, Lin S, Jackson JA and Woodbury NW (2000) Characterization of a Rhodobacter capsulatus reaction center mutant that enhances the distinction between spectral forms of the initial electron donor. Biochemistry 39: 14787–14798 Edens GJ, Gunner MR, Xu Q and Mauzerall D (2000) The enthalpy and entropy of reaction for formation of P+QA– from excited reaction centers of Rhodobacter sphaeroides. J Am
315 Chem Soc 122: 1479–1485 Encyclopaedia Britannica Online. http://www.britannica.com (July 27, 2006) Farhoosh R, Chynwat V, Gebhard R, Lugtenburg J and Frank HA (1997) Triplet energy transfer between the primary donor and carotenoids in Rhodobacter sphaeroides R-26.1 reaction centers incorporated with spheroidene analogs having different extents of π-electron conjugation. Photochem Photobiol 66: 97–104 Finkele U, Lauterwasser C, Struck A, Scheer H and Zinth W (1992) Primary electron transfer kinetics in bacterial reaction centers with modified bacteriochlorophylls at the monomeric sites BA,B. Proc Natl Acad Sci USA 89: 9514–9518 Flores M, Isaacson R, Abresch E, Calvo R, Lubitz W and Feher G (2006) Protein-cofactor interactions in bacterial reaction centers from Rhodobacter sphaeroides R-26: I. Identification of the ENDOR lines associated with the hydrogen bonds to the primary quinone QA•–. Biophys J 90: 3356–3362 Foloppe N, Breton J and Smith JC (1995) Conformational energetics of a partially symmetrized photosynthetic reaction center. Chem Phys Letts 242: 238–243 Frank HA, Chadwick BW, Taremi S, Kolaczkowski S and Bowman MK (1986) Singlet and triplet absorption-spectra of carotenoids bound in the reaction centers of Rhodopseudomonas sphaeroides R26. FEBS Letts 203: 157–163 Frank HA, Chynwat V, Hartwich G, Meyer M, Katheder I and Scheer H (1993) Carotenoid triplet-state formation in Rhodobacter sphaeroides R-26 reaction centers exchanged with modified bacteriochlorophyll pigments and reconstituted with spheroidene. Photosynth Res 37: 193–203 Frank HA, Chynwat V, Posteraro A, Hartwich G, Simonin I and Scheer H (1996) Triplet state energy transfer between the primary donor and the carotenoid in Rhodobacter sphaeroides R-26.1 reaction centers exchanged with modified bacteriochlorophyll pigments and reconstituted with spheroidene. Photochem Photobiol 64: 823–831 Franken EM, Shkuropatov AY, Francke C, Neerken S, Gast P, Shuvalov VA, Hoff AJ and Aartsma TJ (1997a) Reaction centers of Rhodobacter sphaeroides R-26 with selective replacement of bacteriopheophytin by pheophytin a. I. Characterisation of steady-state absorbance and circular dichroism, and of the P+QA– state. Biochim Biophys Acta 1319: 242–250 Franken EM, Shkuropatov AY, Francke C, Neerken S, Gast P, Shuvalov VA, Hoff AJ and Aartsma TJ (1997b) Reaction centers of Rhodobacter sphaeroides R-26 with selective replacement of bacteriopheophytin by pheophytin a. II. Temperature dependence of the quantum yield of P+QA– and 3P formation. Biochim Biophys Acta 1321: 1–9 Fukushima A, Matsuura K, Shimada K and Satoh T (1988) Reaction center-B870 pigment protein complexes with bound cytochromes c-555 and c-551 from Rhodocyclus gelatinosus. Biochim Biophys Acta 933: 399–405 Gardiner AT, Zech SG, MacMillan F, Käss H, Bittl R, Schlodder E, Lendzian F, and Lubitz W (1999) Electron paramagnetic resonance studies of zinc-substituted reaction centers from Rhodopseudomonas viridis. Biochemistry 38: 11773–11787 Gebhard R, van der Hoef K, Violette CA, de Groot HJM, Frank HA and Lugtenburg J (1991) C13 magic angle spinning NMR evidence for a 15,15´-Z configuration of the spheroidene chromophore in the Rhodobacter sphaeroides reaction center — synthesis of C13-labeled and H2-labeled spheroidenes. Pure Applied Chem 63: 115–122
316 Giangiacomo KM and Dutton PL (1989) In photosynthetic reaction centers, the free-energy difference for electron-transfer between quinones bound at the primary and secondary quinonebinding sites governs the observed secondary site specificity. Proc Natl Acad Sci USA 86: 2658–2662 Goldsmith JO, King B and Boxer SG (1996) Mg coordination by amino acid side chains is not required for assembly and function of the special pair in bacterial photosynthetic reaction centers. Biochemistry 35: 2421–2428 Gopher A, Blatt Y, Schonfeld M, Okamura MY, Feher G and Montal M (1985) The effect of an applied electric field on the charge recombination kinetics in reaction centers reconstituted in planar lipid bilayers Biophys. J. 48: 311–320 Graige MS, Paddock ML, Bruce JM, Feher G and Okamura MY (1996) Mechanism of proton-coupled electron transfer for quinone QB reduction in reaction centers of Rb sphaeroides. J Am Chem Soc 118: 9005–9016 Graige MS, Feher G and Okamura MY (1998) Conformational gating of the electron transfer reaction QA– • QB– • → QAQB– • in bacterial reaction centers of Rhodobacter sphaeroides determined by a driving force assay. Proc Natl Acad Sci USA 95: 11679–11684 Graige MS, Paddock ML, Feher G and Okamura MY (1999) Observation of the protonated semiquinone intermediate in isolated reaction centers from Rhodobacter sphaeroides: Implications for the mechanism of electron and proton transfer in proteins. Biochemistry 38: 11465–11473 Gunner MR and Dutton PL (1989) Temperature and ∆G° dependence of the electron-transfer from Bph– to QA in reaction center protein from Rhodobacter sphaeroides with different quinones as QA. J Am Chem Soc 111: 3400–3412 Gunner MR, Robertson DE and Dutton PL (1986) Kinetic-studies on the reaction center protein from Rhodopseudomonas sphaeroides — the temperature and free-energy dependence of electron-transfer between various quinones in the QA site and the oxidized bacteriochlorophyll dimer. J Phys Chem 90: 3783–3795 Hara M, Kaneko T, Nakamura C, Asada Y and Miyake J (1998) Redox properties of an H-subunit-depleted photosynthetic reaction center from Rhodopseudomonas viridis. Biochim Biophys Acta-Bioenerg 1363: 199–208 Hartwich G, Friese M, Scheer H, Ogrodnik A and Michel-Beyerle ME (1995a) Ultrafast internal conversion in 132-OH-Ni-bacteriochlorophyll in reaction centers of Rhodobacter sphaeroides R26. Chem Phys 197: 423–434 Hartwich G, Scheer H, Aust V, Angerhofer A (1995b) Absorption and ADMR studies on bacterial photosynthetic reaction centers with modified pigments. Biochim Biophys Acta-Bioenerg 1230: 97–113 Heller BA, Holten D and Kirmaier C (1995a) Characterization of bacterial reaction centers having mutations of aromatic residues in the binding-site of the bacteriopheophytin intermediary electron carrier. Biochemistry 34: 5294–5302 Heller BA, Holten D and Kirmaier C (1995b) Control of electron-transfer between the L-side and M-side of photosynthetic reaction centers. Science 269: 940–945 Hermes S, Bremm O, Garczarek F, Derrien V, Liebisch P, Loja P, Sebban P, Gerwert K and Haumann M (2006) A time-resolved iron-specific X-ray absorption experiment yields no evidence for an Fe2+ → Fe3+ transition during QA → QB electron transfer in the photosynthetic reaction center. Biochemistry 45: 353–359
Michael R. Jones Hoff AJ and Deisenhofer J (1997) Photophysics of photosynthesis: Structure and spectroscopy of reaction centres of purple bacteria. Physics Reports-Review Section of Physics Letters 287: 2–247 Huber H, Meyer M, Scheer H, Zinth W and Wachtveitl J (1998) Temperature dependence of the primary electron transfer reaction in pigment-modified bacterial reaction centers. Photosynth Res 55: 153–162 Hucke O, Schmid R and Labahn A (2002) Exploring the primary electron acceptor QA-site of the bacterial reaction center from Rhodobacter sphaeroides — binding mode of vitamin K derivatives. Eur J Biochem 269: 1096–1108 Hulsebosch RJ, Borovykh IV, Paschenko SV, Gast P and Hoff AJ (1999) Radical pair dynamics and interactions in quinone-reconstituted photosynthetic reaction centers of Rb. sphaeroides R26: A multifrequency magnetic resonance study. J Phys Chem B 103: 6815–6823 Hunter CN (1995) Genetic manipulation of the antenna complexes of purple bacteria. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 473–501. Kluwer Academic Publishers, Dordrecht Jackson JA, Lin S, Taguchi AKW, Williams JC, Allen JP and Woodbury NW (1997) Energy transfer in Rhodobacter sphaeroides reaction centers with the initial electron donor oxidized or missing. J Phys Chem B 101: 5747–5754 Jones MR, Fowler GJS, Gibson LCD, Grief GG, Olsen JD, Crielaard W and Hunter CN (1992a) Construction of mutants of Rhodobacter sphaeroides lacking one or more pigment-protein complexes and complementation with reaction centre, LH1 and LH2 genes. Mol Microbiol 6: 1173–1184 Jones MR, Visschers RW, van Grondelle R and Hunter CN (1992b) Construction and characterisation of a mutant of Rhodobacter sphaeroides with the reaction centre as the sole pigment-protein complex. Biochemistry 31: 4458–4465 Kálmán L and Maroti P (1994) Stabilization of reduced primary quinone by proton uptake in reaction centers of Rhodobactersphaeroides. Biochemistry 33: 9237–9244 Kálmán L, LoBrutto R, Allen JP and Williams JC (1999) Modified reaction centres oxidize tyrosine in reactions that mirror Photosystem II. Nature 402: 696–699 Kálmán L, LoBrutto R, Allen JP and Williams JC (2003a) Manganese oxidation by modified reaction centers from Rhodobacter sphaeroides. Biochemistry 42: 11016–11022 Kálmán L, LoBrutto R, Narváez AJ, Williams JC and Allen JP (2003b) Correlation of proton release and electrochromic shifts of the optical spectrum due to oxidation of tyrosine in reaction centers from Rhodobacter sphaeroides. Biochemistry 42: 13280–13286 Kálmán L, Thielges MC, Williams JC and Allen JP (2005) Proton release due to manganese binding and oxidation in modified bacterial reaction centers. Biochemistry 44: 13266–13273 Katilius E, Turanchik T, Lin S, Taguchi AKW and Woodbury NW (1999) B-side electron transfer in a Rhodobacter sphaeroides reaction center mutant in which the B-side monomer bacteriochlorophyll is replaced with bacteriopheophytin. J Phys Chem B 103: 7386–7389 Katilius E, Babendure JL, Lin S and Woodbury NW (2004) Electron transfer dynamics in Rhodobacter sphaeroides reaction center mutants with a modified ligand for the monomer bacteriochlorophyll on the active side. Photosynth Res 81:
Chapter 16
Reaction Center Structural Plasticity
165–180 Katz EY, Shkuropatov AY, Klevanik AV, Adanin VM and Shuvalov VA (1991) Effect of the structure of exogenous quinone on its ability to function as the primary quinone in the reaction centers from Rhodobacter sphaeroides R-26. Biol Memb 8: 468–475 Kee HL, Laible PD, Bautista JA, Hanson DK, Holten D and Kirmaier C (2006) Determination of the rate and yield of B-side quinone reduction in Rhodobacter capsulatus reaction centers. Biochemistry 45: 7314–7322 Kennis JTM, Shkuropatov AY, van Stokkum IHM, Gast P, Hoff AJ, Shuvalov VA and Aartsma TJ (1997) Formation of a long-lived P+BA– state in plant pheophytin-exchanged reaction centers of Rhodobacter sphaeroides R26 at low temperature. Biochemistry 36: 16231–16238 Kirmaier C, Holten D, Debus RJ, Feher G and Okamura MY (1986) Primary photochemistry of iron-depleted and zinc-reconstituted reaction centers from Rhodopseudomonas sphaeroides. Proc Natl Acad Sci USA 83: 6407–6411 Kirmaier C, Holten D, Bylina EJ and Youvan DC (1988) Electron-transfer in a genetically modified bacterial reaction center containing a heterodimer. Proc Natl Acad Sci USA 85: 7562–7566 Kirmaier C, Gaul D, DeBey R, Holten D and Schenck CC (1991) Charge separation in a reaction center incorporating bacteriochlorophyll for photoactive bacteriopheophytin. Science 251: 922–927 Kirmaier C, Laible PD, Hanson DK and Holten D (2003) B-side charge separation in bacterial photosynthetic reaction centers: Nanosecond time scale electron transfer from HB– to QB. Biochemistry 42: 2016–2024 Kirmaier C, Bautista JA, Laible PD, Hanson DK and Holten D (2005) Probing the contribution of electronic coupling to the directionality of electron transfer in photosynthetic reaction centers. J Phys Chem B 109: 24160–24172 Kleinfeld D, Okamura MY and Feher G (1985) Charge recombination kinetics as a probe of protonation of the primary acceptor in photosynthetic reaction centers Biophys J 48: 849–852 Knox PP, Churbanova IY, Zakharova NI, Krasil’nikov PM, Lukashev EP, Rubin AB and Shaitan KV (2001) Effects of extraction of the H-subunit from Rhodobacter sphaeroides reaction centers on relaxation processes associated with charge separation. Biochem-Moscow 66: 91–95 Kobayashi M, Takaya A, Kanai N, Ota Y, Saito T, Wang Z-Y and Nozawa T (2004) Reconstitution and replacement of bacteriochlorophyll a molecules in photosynthetic reaction centers. J Biochem 136: 363–369 Kok P, Kohler J, Groenen EJJ, Gebhard R, van der Hoef K, Lugtenburg J, Hoff AJ, Farhoosh R and Frank HA (1994) Towards a vibrational analysis of spheroidene — resonance Ramanspectroscopy of C13-labeled spheroidenes in petroleum ether and in the Rhodobacter sphaeroides reaction center. Biochim Biophys Acta 1185: 188–192 Kok P, Kohler J, Groenen EJJ, Gebhard R, vanderHoef I, Lugtenburg J, Farhoosh R and Frank HA (1997) Resonance Raman spectroscopy of H2-labelled spheroidenes in petroleum ether and in the Rhodobacter sphaeroides reaction centre. Spectrochim Acta A-Mol Biomol Spec 53: 381–392 Kortluke C, Breese K, Gadon N, Labahn A and Drews G (1997) Structure of the puf operon of the obligately aerobic, bacteriochlorophyll a-containing bacterium Roseobacter denitrificans
317 OCh114 and its expression in a Rhodobacter capsulatus puf puc deletion mutant. J Bact 179: 5247–5258 Kuglstatter A, Miksovska J, Sebban P and Fritzsch G (2000) Structure of the photosynthetic reaction centre from Rhodobacter sphaeroides reconstituted with anthraquinone as primary quinone QA. FEBS Letts 472: 114–116 Kuglstatter A, Ermler U, Michel H, Baciou L and Fritzsch G. (2001) X-ray structure analysis of photosynthetic reaction center variants from Rhodobacter sphaeroides: Structural changes induced by point mutations at position L209 modulate electron and proton transfer. Biochemistry 40: 4253–4260 Labahn A, Bruce JM, Okamura MY and Feher G (1995) Direct charge recombination from D+QAQB– to DQAQB in bacterial reaction centers from Rhodobacter sphaeroides containing low potential quinone in the QA site. Chem Phys 197: 355–366 Laible PD, Kirmaier C, Holten D, Tiede DM, Schiffer M and Hanson DK (1998) Formation of P+QB– via B-branch electron transfer in mutant reaction centers. In: Garab G (ed) Photosynthesis: Mechanisms and Effects, pp 849–852. Kluwer Academic Publishers, Dordrecht Laible PD, Kirmaier C, Udawatte CSM, Hofman SJ, Holten D and Hanson DK (2003) Quinone reduction via secondary Bbranch electron transfer in mutant bacterial reaction centers. Biochemistry 42: 1718–1730 Lancaster CRD and Michel H (1997) The coupling of lightinduced electron transfer and proton uptake as derived from crystal structures of reaction centres from Rhodopseudomonas viridis modified at the binding site of the secondary quinone, QB. Structure 5: 1339–1359 Laporte L, Kirmaier C, Schenck CC and Holten D (1995) Free-energy dependence of the rate of electron-transfer to the primary quinone in beta-type reaction centers. Chem Phys 197: 225–237 Lersch W and Michel-Beyerle ME (1987) Implications of spin dynamics for the charge recombination in iron-depleted and quinone-substituted reaction centers from Rhodobacter sphaeroides R-26. Biochim Biophys Acta 891: 265–274 Li JL, Gilroy D, Tiede DM and Gunner MR (1998) Kinetic phases in the electron transfer from P+QA– QB to P+QAQB– and the associated processes in Rhodobacter sphaeroides R-26 reaction centers. Biochemistry 37: 2818–2829 Li J, Coleman WJ, Youvan DC and Gunner MR (2000a) Characterization of a symmetrized mutant RC with 42 residues from the QA site replacing residues in the QB site. Photosynth Res 64: 41–52 Li JL, Takahashi E and Gunner MR (2000b) –∆G°AB, and pH dependence of the electron transfer from P+QA– QB to P+QAQB– in Rhodobacter sphaeroides reaction centers. Biochemistry 39: 7445–7454 Lin S, Xiao WZ, Eastman JE, Taguchi AKW and Woodbury NW (1996) Low-temperature femtosecond-resolution transient absorption spectroscopy of large-scale symmetry mutants of bacterial reaction centers. Biochemistry 35: 3187–3196 Lin S, Katilius E, Ilagan RP, Gibson GN, Frank HA and Woodbury NW (2006) Mechanism of carotenoid singlet excited state energy transfer in modified bacterial reaction centers. J Phys Chem B 110: 15556–15563 Liu BL and Hoff AJ (1990) QA-depletion and reconstitution of a reaction center preparation from the photosynthetic bacterium Rhodopseudomonas viridis. FEBS Letts 269: 354–357 Liu BL, van Kan PJM and Hoff AJ (1991a) Influence of the H-
318 subunit and Fe2+ on electron-transport from I– to QA in Fe2+-free and or H-free reaction centers from Rhodobacter sphaeroides R-26. FEBS Letts 289: 23–28 Liu BL, Yang LH and Hoff AJ (1991b) On the depletion and reconstitution of both QA and metal in reaction centers of the photosynthetic bacterium Rb sphaeroides R-26. Photosynth Res 28: 51–58 Lubitz W and Feher G (1999) The primary and secondary accepters in bacterial photosynthesis III. Characterization of the quinone radicals QA•– and QB•– by EPR and ENDOR. Appl Mag Res 17: 1–48 Lubitz W, Lendzian F and Bittl R (2002) Radicals, radical pairs and triplet states in photosynthesis. Acc Chem Res 35: 313–320 Madeo J and Gunner MR (2005) Modeling binding kinetics at the QA site in bacterial reaction centers. Biochemistry 44: 10994–11004 Maki H, Matsuura K, Shimada K and Nagashima KVP (2003) Chimeric photosynthetic reaction center complex of purple bacteria composed of the core subunits of Rubrivivax gelatinosus and the cytochrome subunit of Blastochloris viridis. J Biol Chem 278: 3921–3928 Mallardi A, Giustini M and Palazzo G (1998) Binding of ubiquinone to photosynthetic reaction centers. 2. Determination of enthalpy and entropy changes for the binding to the QA site in reverse micelles. J Phys Chem B 102: 9168–9173 Masuda S, Tsukatani Y, Kimura Y, Nagashima KVP, Shimada K and Matsuura K (2002) Mutational analyses of the photosynthetic reaction center-bound triheme cytochrome subunit and cytochrome c2 in the purple bacterium Rhodovulum sulfidophilum. Biochemistry 41: 11211–11217 Matsuura K, Fukushima A, Shimada K and Satoh T (1988) Direct and indirect electron transfer from cytochromes c and c2 to the photosynthetic reaction center in pigment-protein complexes isolated from Rhodocyclus gelatinosus. FEBS Lett 237: 21–25 Mattioli TA, Williams JC, Allen JP and Robert B (1994) Changes in primary donor hydrogen-bonding interactions in mutant reaction centers from Rhodobacter sphaeroides: Identification of the vibrational frequencies of all the conjugated carbonyl groups. Biochemistry 33: 1636–1643 Mattioli TA, Lin X, Allen JP and Williams JC (1995) Correlation between multiple hydrogen-bonding and alteration of the oxidation potential of the bacteriochlorophyll dimer of reaction centers from Rhodobacter sphaeroides. Biochemistry 34: 6142–6152 McAuley KE, Fyfe PK, Ridge JP, Cogdell RJ, Isaacs NW and Jones MR (2000) Ubiquinone binding, ubiquinone exclusion, and detailed cofactor conformation in a mutant bacterial reaction center. Biochemistry 39: 15032–15043. McPherson PH, Nagarajan V, Parson WW, Okamura MY and Feher G (1990) pH-dependence of the free-energy gap between DQA and D+QA– determined from delayed fluorescence in reaction centers from Rhodobacter sphaeroides R-26. Biochim Biophys Acta 1019: 91–94 Meyer M and Scheer H (1995) Reaction centers of Rhodobacter sphaeroides R26 containing C-3 acetyl and vinyl (bacterio)pheophytins at sites HA,B . Photosynth Res 44:55– 65 Miksovska J, Kálmán L, Schiffer M, Maroti P, Sebban P and Hanson DK (1997) In bacterial reaction centers rapid delivery of the second proton to QB can be achieved in the absence of
Michael R. Jones L212Glu. Biochemistry 36: 12216–12226 Möbius K (2000) Primary processes in photosynthesis: What do we learn from high-field EPR spectroscopy? Chem Soc Rev 29: 129–139 Moore L and Boxer SG (1998) Inter-chromophore interactions in pigment-modified and dimer-less bacterial photosynthetic reaction centers. Photosynth Res 55: 173–180 Moore LJ, Zhou HL and Boxer SG (1999) Excited-state electronic asymmetry of the special pair in photosynthetic reaction center mutants: Absorption and stark spectroscopy. Biochemistry 38: 11949–11960 Morris AL, Snyder SW, Zhang YN, Tang J, Thurnauer MC, Dutton PL, Robertson DE and Gunner MR (1995) Electron-spin polarization model applied to sequential electron-transfer in iron-containing photosynthetic bacterial reaction centers with different quinones as QA. J Phys Chem 99: 3854–3866 Moser CC, Keske JM, Warncke K, Farid R S and Dutton PL (1992) Nature of biological electron transfer. Nature 355: 796–802 Nabedryk E, Breton J, Williams JC, Allen JP, Kuhn M and Lubitz W (1998) FTIR characterization of the primary electron donor in double mutants combining the heterodimer HL(M202) with the LH(L131), HF(L168), FH(M197), or LH(M160) mutations. Spectrochim Acta A-Mol Biomol Spec 54: 1219–1230 Nabedryk E, Schulz C, Muh F, Lubitz W and Breton J (2000) Heterodimeric versus homodimeric structure of the primary electron donor in Rhodobacter sphaeroides reaction centers genetically modified at position M202. Photochem Photobiol 71: 582–588 Nagashima KVP, Shimada K and Matsuura K (1996) Shortcut of the photosynthetic electron transfer in a mutant lacking the reaction center-bound cytochrome subunit by gene disruption in a purple bacterium, Rubrivivax gelatinosus. FEBS Letts 385: 209–213 Narváez AJ, LoBrutto R, Allen JP and Williams JC (2004) Trapped tyrosyl radical populations in modified reaction centers from Rhodobacter sphaeroides. Biochemistry 43: 14379–14384 Okamura MY, Isaacson RA and Feher G. (1975) Primary acceptor in bacterial photosynthesis — obligatory role of ubiquinone in photoactive reaction centers of Rhodopseudomonas spheroides. Proc Natl Acad Sci USA 72: 3491–3495 Okamura MY, Paddock ML, Graige MS and Feher G (2000) Proton and electron transfer in bacterial reaction centers. Biochim Biophys Acta: Bioenerg 1458: 148–163 Paddock ML, Feher G and Okamura MY (1995) Pathway of proton transfer in bacterial reaction centers: Further investigations on the role of Ser-L223 studied by site-directed mutagenesis. Biochemistry 34: 15742–15750 Paddock ML, Axelrod HL, Abresch EC, Yeh AP, Rees DC, Feher G and Okamura MY (1999a) Crystal structure of a photosynthetic revertant from Rb. sphaeroides: Mechanism of action of a long distant suppressor mutation. Biophys J 76: A141-A141 Paddock ML, Graige MS, Feher G and Okamura MY (1999b) Identification of the proton pathway in bacterial reaction centers: Inhibition of proton transfer by binding of Zn2+ and Cd2+. Proc Natl Acad Sci USA 96: 6183–6188 Paddock ML, Chang C, Xu Q, Abresch EC, Axelrod HL, Feher G and Okamura MY (2005) Quinone (QB) reduction by Bbranch electron transfer in mutant bacterial reaction centers from Rhodobacter sphaeroides: Quantum efficiency and X-ray structure. Biochemistry 44: 6920–6928 Parson WW (1991) Reaction centers. In: Scheer H (ed) Chloro-
Chapter 16
Reaction Center Structural Plasticity
phylls, pp 1153–1180. CRC Press, Boca Raton Parson WW (1996) Photosynthetic bacterial reaction centers. In: Bendall DS (ed) Protein Electron Transfer, pp 125–160. BIOS Scientific Publishers, Oxford Poluektov OG, Utschig LM, Tang J, Dubinski AA, Schlesselman S, Thurnauer MC (2001) High-frequency EPR approach to the electron spin-polarization effects observed in the photosynthetic reaction centers. Appl Mag Res 21: 311–323 Poluektov OG, Utschig LM, Dubinskij AA and Thurnauer MC (2005) Electron-transfer pathways and protein response to charge separation in photosynthetic reaction centers: timeresolved high-field ENDOR of the spin-correlated radical pair P865+QA–. J Am Chem Soc 127: 4049–4059 Potter JA, Fyfe PK, Frolov D, Wakeham MC, van Grondelle R, Robert B and Jones MR (2005) Strong effects of an individual water molecule on the rate of primary charge separation in the Rhodobacter sphaeroides reaction centre. J Biol Chem 280: 27155–27164 Qian P, Addlesee HA, Ruban AV, Wang PY, Bullough PA and Hunter CN (2003) A reaction center-light-harvesting 1 complex (RC-LH1) from a Rhodospirillum rubrum mutant with altered esterifying pigments. J Biol Chem 278: 23678–23685 Reed W and Clayton RK (1968) Isolation of a reaction center fraction from Rhodopseudomonas spheroides. Biochem Biophys Res Commun 30: 471–475 Ridge JP (1998) The role of the protein during energy transduction in the bacterial reaction centre. PhD Thesis, University of Sheffield Ridge JP, van Brederode ME, Goodwin MG, van Grondelle R and Jones MR (1999) Mutations that modify or exclude binding of the QA ubiquinone and carotenoid in the reaction center from Rhodobacter sphaeroides. Photosynth Res 59: 9–26 Robles SJ, Breton J and Youvan DC (1990) Partial symmetrization of the photosynthetic reaction center. Science 248: 1402–1405 Roszak AW, McKendrick K, Gardiner AT, Mitchell IA, Isaacs NW, Cogdell RJ, Hashimoto H and Frank HA (2004) Protein regulation of carotenoid binding: Gatekeeper and locking amino acid residues in reaction centers of Rhodobacter sphaeroides. Structure 12: 765–773 Scheer H and Hartwich G (1995) Bacterial reaction centers with modified tetrapyrrole chromophores. In: Blankenship RE, Madigan MT and Bauer C (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 649–663, Kluwer Academic Publishers, Dordrecht Scheer H and Struck A (1993) Bacterial reaction centers with modified tetrapyrrole chromophores. In: Deisenhofer J and Norris JR (eds) The Photosynthetic Reaction Center, Vol. 1, pp 157–192. Academic Press, San Diego Schelvis JPM, Liu BL, Aartsma TJ and Hoff AJ (1992) The electron-transfer rate from BphA– to QA in reaction centers of Rhodobacter-sphaeroides R-26 -influence of the H-subunit, the QA and Fe2+ cofactors, and the isoprene tail of QA. Biochim Biophys Acta 1102: 229–236 Schmid R and Labahn A (2000) Temperature and free energy dependence of the direct charge recombination rate from the secondary quinone in bacterial reaction centers from Rhodobacter sphaeroides. J Phys Chem B 104: 2928–2936 Schmidt S, Arlt T, Harem P, Huber H, Nagele T, Wachtveitl J, Meyer M, Scheer H and Zinth W (1994) Energetics of the primary electron transfer reaction revealed by ultrafast spec-
319 troscopy on modified bacterial reaction centers. Chem Phys Letts 223: 116–120 Schnegg A, Fuhs M, Rohrer M, Lubitz W, Prisner TF and Möbius K (2002) Molecular dynamics of QA– • and QB– • in photosynthetic bacterial reaction centers studied by pulsed high-field EPR at 95 GHz. J Phys Chem B 106: 9454–9462 Sebban P (1988) Activation-energy of the rate-constant of P+QA– absorption decay in reaction centers from Rhodobactersphaeroides reconstituted with different anthraquinones. FEBS Letts 233: 331–334 Sebban P and Lindqvist L (1987) Kinetic-study of PF and CarT states in the LM subunit purified from the wild-type Rhodobacter sphaeroides reaction centers. Photosynth Res 13: 57–67 Sebban P, Maroti P, Schiffer M and Hanson DK (1995) Electrostatic dominoes — long-distance propagation of mutational effects in photosynthetic reaction centers of Rhodobacter capsulatus. Biochemistry 34: 8390–8397 Shkuropatov AY and Shuvalov VA (1993) Electron transfer in pheophytin a-modified reaction centers from Rhodobacter sphaeroides (R-26). FEBS Letts 322: 168–172 Shkuropatov AY, Neerken S, Permentier HP, de Wijn R, Schmidt KA, Shuvalov VA, Aartsma TSJ, Gast P and Hoff AJ (2003) The effect of exchange of bacteriopheophytin a with plant pheophytin a on charge separation in Y(M210)W mutant reaction centers of Rhodobacter sphaeroides at low temperature. Biochim Biophys Acta-Bioenerg 1557: 1–12 Spiedel D, Jones MR and Robert B (2002) Tuning of the redox potential of the primary electron donor in reaction centres of purple bacteria: Effects of amino acid polarity and position. FEBS Letts 527: 171–175 Sporlein S, Zinth W, Meyer M, Scheer H and Wachtveitl J (2000) Primary electron transfer in modified bacterial reaction centers: Optimization of the first events in photosynthesis. Chem Phys Letts 322: 454–464 Stehlik D and Möbius K (1997) New EPR methods for investigating photoprocesses with paramagnetic intermediates. Ann Rev Phys Chem 48: 745–784 Stilz HU, Finkele U, Holzapfel W, Lauterwasser C, Zinth W and Oesterhelt D (1994) Influence of M-subunit Thr222 and Trp252 on quinone binding and electron-transfer in Rhodobacter sphaeroides reaction centers. Eur J Biochem 223: 233–242 Stocker JW, Taguchi AKW, Murchison HA, Woodbury NW and Boxer SG (1992) Spectroscopic and redox properties of Sym1 and (M)F195H — Rhodobacter capsulatus reaction center symmetry mutants which affect the initial electron-donor. Biochemistry 31: 10356–10362 Storch KF, Cmiel E, Schafer W and Scheer H (1996) Stereoselectivity of pigment exchange with 132-hydroxylated tetrapyrroles in reaction centers of Rhodobacter sphaeroides R26. Eur J Biochem 238: 280–286 Stowell MHB, McPhillips TM, Rees DC, Soltis SM, Abresch E and Feher G (1997) Light-induced structural changes in photosynthetic reaction center: Implications for mechanism of electron-proton transfer. Science 276: 812–816 Taguchi AKW, Stocker JW, Alden RG, Causgrove TP, Peloquin JM, Boxer SG and Woodbury NW (1992) Biochemical characterization and electron transfer reactions of Sym1, a Rhodobacter capsulatus reaction center symmetry mutant which affects the initial electron donor. Biochemistry 31: 10345–10355 Taguchi AKW, Stocker JW, Boxer SG and Woodbury NW (1993)
320 Photosynthetic reaction center mutagenesis via chimeric rescue of a non-functional Rhodobacter capsulatus puf operon with sequences from Rhodobacter sphaeroides. Photosynth Res 36: 43–58 Taguchi AKW, Eastman JE, Gallo DM, Sheagley E, Xiao WZ and Woodbury, NW (1996) Asymmetry requirements in the photosynthetic reaction center of Rhodobacter capsulatus. Biochemistry 35: 3175–3186 Takahashi E and Wraight CA (2006) Small weak acids reactivate proton transfer in reaction centers from Rhodobacter sphaeroides mutated at Asp(L210) and Asp(M17). J Biol Chem 281: 4413–4422 Tandori J, Baciou L, Alexov E, Maroti P, Schiffer M, Hanson DK and Sebban P (2001) Revealing the involvement of extended hydrogen bond networks in the cooperative function between distant sites in bacterial reaction centers. J Biol Chem 276: 45513–45515 Tang J, Utschig LM, Poluektov O and Thurnauer MC (1999) Transient W-band EPR study of sequential electron transfer in photosynthetic bacterial reaction centers. J Phys Chem B 103: 5145–5150 Tehrani A and Beatty JT (2004) Effects of precise deletions in Rhodobacter sphaeroides reaction center genes on steady-state levels of reaction center proteins: A revised model for reaction center assembly. Photosynth Res 79: 101–108 Tehrani A, Prince RC and Beatty JT (2003) Effects of photosynthetic reaction center H protein domain mutations on photosynthetic properties and reaction center assembly in Rhodobacter sphaeroides. Biochemistry 42: 8919–8928 Thielges M, Uyeda G, Cámara-Artigas A, Kálmán L, Williams JC and Allen JP (2005) Design of a redox-linked active metal site: Manganese bound to bacterial reaction centers at a site resembling that of Photosystem II. Biochemistry 44: 7389–7394 Turzo K, Laczko G, Filus Z and Maroti P (2000) Quinonedependent delayed fluorescence from the reaction center of photosynthetic bacteria. Biophys J 79: 14–25 Utschig LM and Thurnauer MC (2004) Metal ion modulated electron transfer in photosynthetic proteins. Acc Chem Res 37: 439–447 Utschig LM, Greenfield SR, Tang J, Laible PD and Thurnauer MC (1997) Influence of iron-removal procedures on sequential electron transfer in photosynthetic bacterial reaction centers studied by transient EPR spectroscopy. Biochemistry 36: 8548–8558 Utschig LM, Ohigashi Y, Thurnauer MC and Tiede DM (1998) A new metal-binding site in photosynthetic bacterial reaction centers that modulates QA to QB electron transfer. Biochemistry 37: 8278–8281 Utschig LM, Poluektov O, Tiede DM and Thurnauer MC (2000) EPR investigation of Cu2+-substituted photosynthetic bacterial reaction centers: Evidence for histidine ligation at the surface metal site. Biochemistry 39: 2961–2969 Utschig LM, Poluektov O, Schlesselman SL, Thurnauer MC and Tiede DM (2001) Cu2+ site in photosynthetic bacterial reaction centers from Rhodobacter sphaeroides, Rhodobacter capsulatus, and Rhodopseudomonas viridis. Biochemistry 40: 6132–6141 Utschig LM, Thurnauer MC, Tiede DM and Poluektov OG (2005) Low-temperature interquinone electron transfer in photosynthetic reaction centers from Rhodobacter sphaeroides and Blastochloris viridis: Characterization of QB-states by
Michael R. Jones high-frequency electron paramagnetic resonance (EPR) and electron-nuclear double resonance (ENDOR). Biochemistry 44: 14131–14142 van Brederode ME and Jones MR (2000) Reaction centres of purple bacteria. In: Scrutton NS and Holzenburg A (eds) Enzyme-Catalysed Electron and Radical Transfer, pp 621–676. Kluwer Academic/Plenum Publishers, New York van Brederode ME, van Stokkum IHM, Katilius E, van Mourik F, Jones MR and van Grondelle R (1999) Primary charge separation routes in the BChl:BPhe heterodimer reaction centers of Rhodobacter sphaeroides. Biochemistry 38: 7545–7555 van den Brink JS, Hulsebosch RJ, Gast P, Hore PJ and Hoff AJ (1994a) QA binding in reaction centers of the photosynthetic purple bacterium Rhodobacter sphaeroides R26 investigated with electron spin polarization spectroscopy. Biochemistry 33: 13668–13677 van den Brink JS, Spoyalov AP, Gast P, van Liemt WBS, Raap J, Lugtenburg J and Hoff AJ (1994b) Asymmetric binding of the primary acceptor quinone in reaction centers of the photosynthetic bacterium Rhodobacter sphaeroides R26, probed with Q-band (35 GHz) EPR spectroscopy. FEBS Letts 353: 273–276. van Liemt WBS, Boender GJ, Gast P, Hoff AJ, Lugtenburg J and de Groot HJM (1995) C13 magic-angle-spinning NMR characterization of the functionally asymmetric QA binding in Rhodobacter sphaeroides R26 photosynthetic reaction centers using site-specific C13-labeled ubiquinone-10. Biochemistry 34: 10229–10236 Vos MH, Lambry JC, Robles SJ, Youvan DC, Breton J and Martin JL (1991) Direct observation of vibrational coherence in bacterial reaction centers using femtosecond absorptionspectroscopy. Proc Natl Acad Sci USA 88: 8885–8889 Vos MH, Rappaport F, Lambry JC, Breton J and Martin JL (1993) Visualization of coherent nuclear motion in a membrane-protein by femtosecond spectroscopy. Nature 363: 320–325 Wachtveitl J, Farchaus JW, Das R, Lutz M, Robert B and Mattioli TA (1993) Structure, spectroscopic and redox properties of Rhodobacter sphaeroides reaction centers bearing point mutations near the primary electron donor. Biochemistry 32: 12875–12886 Wakeham MC and Jones MR (2005) Rewiring photosynthesis: engineering wrong-way electron transfer in the purple bacterial reaction centre. Biochem Soc Trans 33: 851–857 Wakeham MC, Goodwin MG, McKibbin C and Jones MR (2003) Photo-accumulation of the P+QB– radical pair state in purple bacterial reaction centres that lack the QA ubiquinone. FEBS Letts 540: 234–240 Wakeham MC, Breton J, Nabedryk E and Jones MR (2004) Formation of a semiquinone at the QB site by A-branch or Bbranch electron transfer in the reaction centre from Rhodobacter sphaeroides. Biochemistry 43: 4755–4763 Warncke K and Dutton PL (1993a) Influence of QA site redox cofactor structure on equilibrium binding, in situ electrochemistry, and electron-transfer performance in the photosynthetic reaction center protein. Biochemistry 32: 4769–4779 Warncke K and Dutton PL (1993b) Experimental resolution of the free-energies of aqueous solvation contributions to ligand protein-binding — quinone-QA site interactions in the photosynthetic reaction center protein. Proc Natl Acad Sci USA 90: 2920–2924 Warncke K, Gunner MR, Braun BS, Gu LQ, Yu CA, Bruce JM
Chapter 16
Reaction Center Structural Plasticity
and Dutton PL (1994) Influence of hydrocarbon tail structure on quinone binding and electron-transfer performance at the QA and QB sites of the photosynthetic reaction-center protein. Biochemistry 33: 7830–7841 Watson AJ (2005) Stability and interactions of the purple bacterial reaction centre. PhD Thesis, University of Bristol Watson AJ, Fyfe PK, Frolov D, Wakeham MC, Nabedryk E, van Grondelle R, Breton J and Jones MR (2005) Replacement or exclusion of the B-branch bacteriopheophytin in the purple bacterial reaction centre: the HB cofactor is not required for assembly or core function of the Rhodobacter sphaeroides complex. Biochim Biophys Acta: Bioenerg 1710: 34–46 Wikipedia, The Free Encyclopedia. http://en.wikipedia.org (July 27, 2006) Wong DK-H, Collins WJ, Harmer A, Lilburn TG and Beatty JT (1996) Directed mutagenesis of the Rhodobacter capsulatus puhA gene and orf 214: Pleiotropic effects on photosynthetic reaction center and light-harvesting 1 complexes. J Bact 178: 2334–2342 Woodbury NW and Allen JP (1995) The pathway, kinetics and thermodynamics of electron transfer in wild-type and mutant bacterial reaction centers of purple nonsulfur bacteria. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 527–557. Kluwer Academic Publishers, Dordrecht Woodbury NW, Parson WW, Gunner MR, Prince RC and Dutton PL (1986) Radical-pair energetics and decay mechanisms in reaction centers containing anthraquinones, naphthoquinones or benzoquinones in place of ubiquinone. Biochim Biophys Acta 851: 6–22 Wraight CA (2004) Proton and electron transfer in the acceptor
321 quinone complex of photosynthetic reaction centers from Rhodobacter sphaeroides. Frontiers in Biosciences 9: 309–337 Xu Q and Gunner MR (2000) Temperature dependence of the free energy, enthalpy, and entropy of P+QA– charge recombination in Rhodobacter sphaeroides R-26 reaction centers. J Phys Chem B 104: 8035–8043 Yakovlev AG, Jones MR, Potter JA, Fyfe PK, Vasilieva LG, Shkuropatov AYa and Shuvalov VA (2005) Primary charge separation between P* and BA: Electron-transfer pathways in native and mutant GM203L bacterial reaction centers. Chem Phys 319: 297–307 Yanagi K, Shimizu M, Hashimoto H, Gardiner AT, Roszak AW and Cogdell RJ (2005) Local electrostatic field induced by the carotenoid bound to the reaction center of the purple photosynthetic bacterium Rhodobacter sphaeroides. J Phys Chem B 109: 992–998 Yeates TO, Komiya H, Chirino A, Rees DC, Allen JP and Feher G (1988) Structure of the reaction center from Rhodobactersphaeroides R-26 and 2.4.1-protein-cofactor (bacteriochlorophyll, bacteriopheophytin, and carotenoid) interactions. Proc Natl Acad Sci USA 85: 7993–7997 Youvan DC, Ismail S and Bylina EJ (1985) Chromosomal deletion and plasmid complementation of the photosynthetic reaction center and light-harvesting genes from Rhodopseudomonas capsulata. Gene 38: 19–30 Zech SG, Bittl R, Gardiner AT and Lubitz W (1997) Transient and pulsed EPR spectroscopy on the radical pair state P+•865QA–• to study light-induced changes in bacterial reaction centers. App Mag Res 13: 517–529 Zilsel J, Lilburn TG and Beatty JT (1989) Formation of functional inter-species hybrid photosynthetic complexes in Rhodobacter capsulatus. FEBS Letts 253: 247–252
Chapter 17 Structure and Function of the Cytochrome c2:Reaction Center Complex from Rhodobacter sphaeroides Herbert Axelrod Stanford Synchrotron Radiation Laboratory, 2575 Sand Hill Rd., Menlo Park, CA 94025, U.S.A.
Osamu Miyashita Department of Biochemistry and Molecular Biophysics, University of Arizona, 1041 East Lowell Street, Tucson, AZ 85721, U.S.A.
Melvin Okamura* Department of Physics, University of California, San Diego, 9500 Gilman Dr., La Jolla, CA 92093-0354, U.S.A.
Summary ............................................................................................................................................................... 323 I. Introduction..................................................................................................................................................... 324 II. History ............................................................................................................................................................ 324 III. Structure of the Cytochrome c2:Reaction Center Complex ............................................................................ 325 IV. Electron Transfer Reactions ........................................................................................................................... 327 V. Effects of Mutation ......................................................................................................................................... 329 A. Effects on Binding, KD ..................................................................................................................... 329 B. Effects on Electron Transfer in the Bound State, ke ........................................................................ 330 C. Effects on the Second Order Rate Constant, k2 .............................................................................. 330 VI. Mechanism of Inter-Protein Electron Transfer................................................................................................ 332 A. First Order Electron Transfer............................................................................................................ 332 B. Second Order Electron Transfer ...................................................................................................... 332 Acknowledgments ................................................................................................................................................. 333 References ............................................................................................................................................................ 333
Summary In purple non-sulfur photosynthetic bacteria, a c-type heme protein, cytochrome (Cyt) c2, serves as the electron donor to the reaction center (RC) which is the site of the initial photochemical electron transfer. The second order rate of electron transfer from Cyt c2 to the RC is diffusion limited and optimized to facilitate electron transfer through the photosynthetic apparatus. This review summarizes the X-ray crystal structure of the Cyt c2:RC complex from Rhodobacter (Rba.) sphaeroides and studies based on the structure that elucidate the molecular basis for the role of the complex in electron transfer. The structure of the complex shows the heme cofactor in van der Waals contact with the reaction center and in close proximity to the bacteriochlorophyll *Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 323–336. © 2009 Springer Science + Business Media B.V.
324
Herbert Axelrod, Osamu Miyashita and Melvin Okamura
dimer, the primary electron donor. The binding interface region includes: a) a solvent-separated region with long-range electrostatic interactions between complementary charged residues on Cyt c2 and the RC which play a role in protein docking and binding, and b) a small central region with short-range interactions, including hydrophobic, hydrogen bonding, and a cation-π interaction, which play a role in binding the Cyt in close contact to the RC surface to facilitate strong electronic coupling between cofactors for rapid inter-protein electron transfer. Both types of interactions contribute to the binding. However, the two types of interactions have markedly different effects on the electron transfer kinetics. The long-range electrostatic interactions change the second order rate constant by changing the association rate but not the electron transfer rate in the bound state. The short-range interactions do not affect the association rate but change the dissociation rate as well as the electron transfer rate in the bound state. The strength of the binding interactions is optimized to allow sufficiently fast dissociation of the oxidized Cyt c2 that does not limit the rate of cyclic electron transfer. I. Introduction Inter-protein electron transfer plays an important role in the machinery for energy conversion in photosynthesis and respiration. The primary light induced reactions and proton pumping electron transfer processes take place in the reaction center (RC), a membrane protein that is embedded in the photosynthetic membrane. In order for electron transfer to occur rapidly, Cyt c2 a mobile electron transfer protein carries electrons between the RC and other membrane protein components. The electron transfer reactions between the mobile carrier and membrane bound proteins are optimized for rapid association, electron transfer, and dissociation. A crystal structure of the Cyt c2:RC complex from Rhodobacter (Rba.) sphaeroides has been determined (Axelrod et al., 2002). In this review we discuss the interactions between Cyt c2 and the RC, focusing on structural information from the crystal structure of the complex and studies of interactions identified in this structure that elucidate the molecular basis for the inter-protein electron transfer. For more comprehensive discussion of earlier studies see Dutton and Prince (1978) and Tiede and Dutton (1993). For an earlier review of the structure and function of the Cyt c2:RC complex see Axelrod and Okamura (2005). The photosynthetic electron transfer components in membranes of purple bacteria are shown in Fig. 1 (Blankenship, 2002). The initial photochemistry occurs in the RC where light induced electron transfer proceeds from a primary donor, D, a bacteriochlorophyll (BChl) dimer (BChl2), through a series of Abbreviations: BChl – bacteriochlorophyll; BChl2 – bacteriochlorophyll dimer; Blc. – Blastochloris; Cyt c2 – cytochrome c2; Rba. – Rhodobacter; RC – reaction center; Rps. – Rhodopseudomonas; Rsp. – Rhodospirillum; Tch. – Thermochromatium
acceptors, doubly reducing a bound quinone QB. The doubly reduced QB binds two protons and dissociates from the RC molecule and is oxidized by Cyt bc1 complex. The electron transfer cycle is completed by Cyt c2 which shuttles the electron from Cyt bc1 back to the RC. In the process, protons are pumped across the membrane, creating a membrane potential that drives ATP synthesis. In photosynthetic bacteria lacking Cyt c2 other mobile electron carriers replace this Cyt. (Meyer and Donohue, 1995). Other photosynthetic bacteria such as Blastochloris (Blc.) viridis (previously Rhodopseudomonas (Rps.) viridis) and Thermochromatium (Tch.) tepidum contain a RC associated Cyt complex that serves as the electron donor to the RC (Nitschke and Dracheva, 1995). The interactions of mobile electron carriers with RC associated Cyts in Blc. viridis and Tch. tepidum have been discussed (Nogi et al., 2005). II. History The reactions between Cyt c2 and the RC are readily followed by transient optical absorption changes upon illumination and were some of the first biophysical studies of electron transfer in photosynthesis. The oxidation of BChl (Duysens, 1952) and Cyt (Vernon and Kamen, 1953; Duysens, 1954; Chance and Smith, 1955) were observed upon illumination of Rhodospirillum (Rsp.) rubrum membranes. However, the nature of the primary reaction was not fully established until the definitive time resolved measurements of Parson (1968) showed that the fast oxidation of BChl was followed by its reduction in 2 µs with the concomitant oxidation of the bound Cyt in Chromatium vinosum chromatophores. Studies using isolated RCs from Rba. sphaeroides by Ke et al. (1970) using exogenous mammalian Cyt and Prince et al. (1974) using Cyt c2
Chapter 17
Cytochrome c2 Interactions
325
Fig. 1. The photosynthetic electron transfer apparatus in purple bacterial membranes. Modified from Axelrod and Okamura (2005).
showed that the rate of reaction between RC and Cyt varied with Cyt concentration indicating a second order reaction. The second order rate constant was dependent on the ionic strength of the medium indicating an electrostatic interaction between the Cyt and the RC. At low ionic strength, the second order rate constant for the reaction between the RC and Cyt c2 was extremely fast (k2 ~109 s–1 M–1), close to the diffusion limit. At higher concentrations of Cyt c2, Overfield et al. (1979) observed that the oxidation of Cyt c2 was biphasic with fast and slow kinetic phases. The fast (τ ~1 µs) phase was found to be first order (independent of Cyt concentration) and was attributed to the electron transfer from bound Cyt c2. From the Cyt concentration dependence of the fraction of RCs with fast phase kinetics, dissociation constants (KD ≅ 1–10µM) for the Cyt have been determined (Tiede and Dutton, 1993). III. Structure of the Cytochrome c2:Reaction Center Complex The structures of the RC (Allen et al., 1986; Chang et al., 1986; Ermler et al., 1994) and Cyt c2 (Axelrod et al., 1994) from Rba. sphaeroides are known from X-ray crystallography. Experimental approaches involving cross-linking (Rosen et al., 1983; Drepper et al., 1997), chemical modification (Long et al., 1989) and site directed mutation (Caffrey et al., 1992) showed the importance of electrostatic
interactions in positioning of the Cyt on the RC surface. The periplasmic surface of the RC contains a cluster of negatively charged residues surrounding the central region directly over the BChl2. The heme edge of the Cyt c2 is solvent exposed and is surrounded by a positively charged region that can interact with a negatively charged region on the RC. Several proposals for the structure of the complex were made based on electrostatic complementarity between the two proteins as well as the structure of the Blc. viridis RC which contains a bound tetra-heme cytochrome positioned over the BChl2 (Deisenhofer and Michel, 1989). Allen et al. (1987) modeled the Cyt c2 directly above the BChl2 similar to that of the permanently bound Cyt in the RC of Blc. viridis in a position that is optimized for electron transfer. Tiede and Chang (1988) modeled the Cyt c2 tilted slightly to increase electrostatic contacts between the Cyt and RC surface consistent with optical dichroism measurements (Tiede, 1987). Adir et al. (1996) modeled the Cyt c2 off-center on the RC M-subunit side of the periplasmic surface over a cluster of negatively charged residues, based on electrostatic calculations and a low resolution electron density map of the Cyt: RC complex. The definitive model for the complex comes from X-ray crystallography. The structure of the Cyt c2:RC complex from Rba. sphaeroides was determined from co-crystals of Cyt c2 and RC obtained by crystallization of a mixture of the two proteins at low ionic strength. Two structures were obtained at resolutions
326
Herbert Axelrod, Osamu Miyashita and Melvin Okamura
of 3.25 and 2.4 Å, from two crystal forms of the complex (Axelrod et al., 2002). Under these conditions the stoichiometry of the bound Cyt was high (0.9 Cyt/RC). The co-crystals were assayed for activity by micro-spectrophotometry. They displayed electron transfer from Cyt to RC with a transfer time (τ = 0.9 ± 0.1 µs) identical to the value measured for electron transfer between Cyt and RC in solution. Since the electron transfer rate is extremely sensitive to the distance between donor and acceptor, this indicates that the structure in the crystal is very similar to the structure in solution. The overall structure of the complex is shown in Fig. 2A. The cytochrome is positioned in the center
of the periplasmic surface of the RC with the solventexposed heme edge directly over the BChl dimer, in position for fast electron transfer. This position of the heme is similar to that found in the structure of the Blc. viridis RC (Deisenhofer and Michel, 1989). An interesting feature of the structure is the kink in a polypeptide chain of the RC (shown by the arrow in Fig. 2A). This kink is far from the Cyt binding site and is due to lattice contacts with a Cyt c2 molecule bound to an adjacent RC in the crystal. Apparently, the binding of the Cyt c2 onto the RC surface alters the conformation due to packing constraints. Since the electron transfer rates in the crystal and in solution are identical and the rates are exquisitely sensitive
Fig. 2. Structure of the Cyt c2 : RC complex. A) Side view of the complex. The Cyt c2 is bound with heme positioned directly over the BChl2. A small region of close contact is indicated by the circle. A conformational change, indicated by the arrow, is due to crystal packing interactions with an adjacent bound Cyt. B) View of the structure in the region between the cofactors. The heme and BChl are separated by the planar aromatic ring of Tyr L162. The closest distance between the conjugated rings in the cofactors is 14.2 Å as shown by the long arrow. A strong tunneling pathway between the two cofactors is indicated by the short arrows. C) Open book view of the interface between the complexes. The two proteins are separated by rotating around the center line. The central region of shortrange contacts (indicated by the circle) is surrounded by charged residues involved in long-range electrostatic interactions. Interacting residues are color coded (see color plate XX). The contact between the heme of Cyt c2 and the ‘hot spot’ Tyr L162 of the RC is shown by the arrow. Hydrophobic residues Leu M191, Val M192 on the RC interact with Phe C102 on the Cyt. Hydrogen bonds are formed between atoms on the RC and the Cyt. A cation-π interaction is formed between Tyr M295 and Arg C32. Electrostatic interactions occur between negatively charged groups on the RC and positively charged groups on the Cyt. Other groups in van der Waals contact are also shown. Modified from Axelrod et al. ( 2002). See also Color Plate 6, Fig. 9.
Chapter 17
Cytochrome c2 Interactions
to distance, this result indicates that the binding of the Cyt on the RC in the crystal is at a rigidly-fixed, specific site. At the center of the binding region, the two proteins exhibit close van der Waals contacts that are important for binding and electron transfer. The small number of hydrophobic interactions between the Cyt and RC in the region of the heme edge are shown in Fig. 2B. The solvent exposed methyl group of the heme is in van der Waals contact with Tyr L162 on the RC directly above the BChldimer. In addition, there is an interaction between Phe (C102) from the cytochrome with Val M192 and Leu M191. The closest distance between the donor and acceptor molecules, (heme and BChl2) is 8.4 Å. This is the distance between the methyl group on the heme and the methyl group on the BChl which are separated only by the ring of Tyr L162. Neither of these two methyl groups is in conjugation with the aromatic ring of either cofactor but they are ends of covalent extensions of the cofactors that are proposed to be important in a tunneling pathway for electron transfer from the Cyt to the RC (Aquino et al., 1997; Miyashita et al., 2003a). The closest distance between conjugated atoms in the cofactors is 14.2 Å. The electron transfer rates calculated from this structure will be discussed later. The interface region between the Cyt and the RC can be divided into two domains—a central domain having short-range contacts, surrounded by a longrange electrostatic domain of complementary charged residues (Fig. 2C). The short-range interactions include hydrophobic interactions, hydrogen bonding interactions and a novel cation-π interaction. The hydrophobic interactions are between Tyr L162, Leu M191 and Val M192 on the RC and Phe C102 and the heme methyl group on the Cyt c2. Hydrogen bonds are observed between amide residues Gln L258, Asn M187, and Asn M188 on the RC with backbone atoms on the Cyt. A cation-π complex is formed between Tyr M295 on the RC and Arg C32 on the Cyt. The short-range interaction region is tightly packed with a low solvent accessibility. Surrounding this region of close contact is a region of solvent separated charged residues. The negatively charged residues on the RC are around Tyr L162 and are predominantly on the M side of the periplasmic surface. The positively charged residues of the Cyt are located opposite the negatively charged residues in the complex. In the space between charged residues, the protein surfaces are farther apart than in the central region and the residues exhibit a higher solvent acces-
327 sibility. The positively charged residues on the Cyt c2 and negatively charged residues on the RC are further than 4.5 Å apart and do not exhibit ion-paired close contact interactions. Thus, the interactions are due to long-range electrostatic attraction between solvated charged groups. The interactions between residues in both the hydrophobic short-range interaction region and electrostatic long-range interaction region are important for binding as shown by studies of sitedirected mutations. However, mutations of residues having electrostatic interactions have very different effects on electron transfer rates than mutations of residues having hydrophobic interactions. These effects will be discussed below. IV. Electron Transfer Reactions The inter-protein electron transfer between Cyt c2 and RC involves the processes of binding and electron transfer. The simplest scheme for the overall reaction is shown below, K A =1/ K D
⎯⎯ ⎯⎯ ⎯⎯ ⎯ → Cyt 2+: DQ Cyt 2+ + DQ ← ⎯
↓
hν Cyt
2+
hν
+D Q +
–
k on
↓
⎯⎯ ⎯⎯ → Cyt 2+: D + Q – ⎯k⎯ → Cyt 3+ + DQ – ← k e
off
(1) where D represents the donor, Q represents the bound quinone. Before illumination (top line) the RC and Cyt are in equilibrium between a bound state and unbound state with a dissociation constant KD which represents the binding of the reduced Cyt to the RC. After illumination (bottom line) the fraction of RCs in the bound state reacts with first order rate constant ke or dissociates with a dissociation rate koff. The RCs in the unbound state react with unbound Cyt with a rate that is proportional to the Cyt concentration, with a second order constant k2. The value of the second order rate constant k2 depends on the association rate (kon), dissociation rate (koff) and electron transfer rate (ke) (Bendall, 1996). k2 =
kon ke ke + koff
(2)
Two cases with different electron transfer behavior can be delineated. In the diffusion limited case, ke >>
ABSORBANCE CHANGE, $A (865nm)
0.02 0.00
A)
Native RCs (R26) 0.7 M M
0.02
0.00
B) YA (L162)
0.01
0.04
0.3 M M
0.06
0.10
7 MM
0.03
no cyt c
0
2
4
2
Time [ms]
6
8
96 MM
27 MM
0.02
0.1 M M
0.08
0.12
ABSORBANCE CHANGE, $A (865nm)
Herbert Axelrod, Osamu Miyashita and Melvin Okamura
328
no cytc2 0
200
400
TIME (Ms)
600
800
laser flash
Fig. 3. Electron transfer kinetics of the reaction between Cyt c2 and the RC. The oxidation and recovery of the BChl2 is monitored at 865 nm. A) Native RCs (diffusion-limited) – the recovery due to the reaction with Cyt c2 is biphasic. The relative amplitudes of the fast and slow phases and the rate of thes slow phase depend on the Cyt c2 concentration. B) Mutant RCs in which Tyr (L162) is changed to Ala (fast exchange limit). The recovery, due to reaction with Cyt c2, is monophasic. The rate of recovery is dependent on Cyt c2 concentration.
koff , the second order rate constant is the association rate constant, i.e., k2 = kon. The diffusion limit applies to native RCs in which ke = 106 s–1 >> koff = 103 s–1. In this case, electron transfer occurs upon each binding event. In the diffusion limited case the rate limiting step is not the electron transfer rate but is determined by the formation of the bound state. A different behavior occurs when the dissociation rate is faster than the electron transfer rate, the fast exchange limit (i.e., koff >> ke). In this case electron transfer does not occur upon each binding event and depends upon the electron transfer rate and the occupancy of the bound state. In the fast exchange limit, the second order rate constant becomes k2 ≅ ke/KD, and monophasic rather than biphasic kinetics is observed in flash kinetic experiments (i.e., no fast first order rate is observed). The different features of the kinetics for electron transfer following a single saturating laser flash in the diffusion-limited and fast exchange regimes are shown in Fig. 3. For native RCs for which the reaction with Cyt is diffusion-limited, the reduction of the oxidized primary donor monitored at 865 nm is biphasic (Fig. 3A). The fast phase is due to electron transfer from bound Cyt c2. The time constant for this reaction (unresolved in this trace) is τe = 1µs. The slow phase has decays with a rate that is dependent on the Cyt concentration: k = k2 [Cyt]. The relative amplitudes of the fast and slow phases and the rate of the slow phase depend on the cytochrome concentration. For some mutant RCs having changes to hydrophobic residues the reaction rates and binding
affinity lead to reaction dynamics in the fast exchange limit (Fig. 3B). In the fast exchange limit the donor reduction is monophasic and the rate is dependent on the Cyt c2 concentration, reaching a limit at high Cyt concentration (Gong et al., 2003). A critical parameter in the electron transfer process is the dissociation rate koff. Graige (1998) measured the dissociation rate for Cyt c2 from the RC (Rba. sphaeroides) and found koff = 1.7 × 103 s–1 by using two laser flashes in the presence of Cyt c2 where reduction of the oxidized donor after the second flash was limited by the rate of dissociation of oxidized Cyt. Gerencser et al. (1999) found a dissociation rate for horse heart Cyt c from Rba. sphaeroides RC to be koff =103 s–1 and showed that at low ionic strength the dissociation rate of horse heart Cyt became the rate limiting step for turnover at high light intensities. The dissociation rate of electron transfer found in these studies is fast enough not to be rate limiting in photosynthetic membranes under physiological light conditions (Crofts and Wraight, 1983). To a rough approximation the dissociation rate may be estimated from the dissociation constant and the association rate. At low ionic strength this gives a value of approximately koff = kon × KD = 109 × 10–6 = 103 s–1. However, this is a rough estimate since kon is measured for the association of reduced Cyt, and koff is measured for the dissociation of oxidized Cyt and the dissociation and association rates, and thus the binding constant, can depend on the redox state of the Cyt and D. However, several studies indicate that the differences of binding constants and rates between
Chapter 17
Cytochrome c2 Interactions
oxidized and reduced forms are relatively small. In early studies Rosen (1980) measured the KD values for reduced and oxidized Rba. sphaeroides Cyt c2 to the Rba. sphaeroides RC (10 mM Tris pH 8) to be 1.0 and 1.5 µM, respectively. Larson and Wraight (2000) found a preferential binding of oxidized horse Cyt to Rba. sphaeroides RC compared to the reduced form by a factor of 3. However, experiments by Devanathan et al. (2004) on membrane reconstituted RCs using plasmon-waveguide resonance spectroscopy have found larger differences in binding of oxidized and reduced forms. They find two binding sites, a high affinity and low affinity site, and preferential binding of the reduced Cyt to the high affinity site. In some kinetic studies involving Cyt c2 and RCs, a slow first order component (τ~500 µs) was observed in addition to the slow second order phase (Overfield et al., 1979; Moser and Dutton, 1988). This slow phase has been proposed to be due to the transition from a configuration in which the Cyt is bound in an inactive state to an active configuration. However, the presence of the slow first order phase is variable and not observed in all studies. The amount of the slow phase varies with different preparations (Tiede and Dutton, 1993). Two other explanations for the 500 µs phase have been advanced: 1) dissociation of a fraction of oxidized Cyt, and 2) heterogeneity of the RC preparation due to aggregation. Larson and Wraight (2000) reported a problem of double-hits to RCs excited with xenon flash lamps leading to multiphasic kinetics which may explain results from earlier experiments. The longer illumination times result in the excitation of RCs containing oxidized Cyt giving rise to an additional slow phase where donor reduction is rate limited by the dissociation of oxidized Cyt. Tiede (2000) used neutron diffraction and found that different preparations of isolated RCs are in different states of aggregation. The presence of the slow first order phase was found in aggregated RCs but not in monomeric RCs. He suggested that the slow first order phase is due to aggregation of RCs. The presence of aggregation effects agrees with recent studies which indicate that the apparent KD for Cyt c2 binding increases at higher RC concentrations which could be due to an aggregation of RC that blocks Cyt binding (E. Abresch, unpublished). More recent measurements of the dissociation constant and second order rate constant have been routinely performed at low RC concentrations (~0.1 µM) in order to reduce the effect of aggregation (Tetreault et al., 2001; Gong et al., 2003).
329 In summary, the high second order rate of electron transfer, the rapid first order rate of electron transfer, and the relatively low binding affinity (giving a dissociation rate that is not rate limiting for turnover) are well optimized for the function of the electron transfer cycle through the RC. The electron transfer rates and binding will be discussed in more detail below in relation to the structure of the Cyt:RC complex. V. Effects of Mutation A. Effects on Binding, KD Residues in the interface region between the Cyt and the RC were mutated to determine how changing the interactions between the two proteins affects binding. Tetreault et al. (2001) found that mutation of negatively charged acid residues to neutral or positively charged residues on the RC increased KD. The largest change (~1000 fold) was obtained for the mutation of Asp M184 → Lys. Asp M184 is in the center of a cluster of acid residues and close to Tyr L162 in the center of the interface region. Gong et al. (2003) found that mutation of RC hydrophobic residues Tyr L162 and Leu M191 to Ala and other residues increased KD. The largest increase in KD (~100-fold) due to Ala substitution was observed for the mutation of Tyr L162. This residue is located in the center of the short-range interaction domain (see Fig. 2C). This large change in binding shows that Tyr L162 is a ‘hot spot’ for binding. Similar, ‘hot spots’ have been observed in studies of Ala replacement mutagenesis on protein-protein association (Bogan and Thorn, 1998). Interestingly, the mutation of Val M192 to Ala resulted in an increase in binding affinity. The changes in binding affinity due to Ala mutation were found to correlate with the solvent accessibility of the mutated residues. Tyr L162, whose mutation produced a large decrease in binding affinity, was inaccessible to solvent in the Cyt:RC complex, while Val M192 whose mutation resulted in increased binding affinity was solvated in the Cyt:RC complex. These results support the idea that the hydrophobic binding interaction is due to a closely fitting association between protein residues with the exclusion of solvent. The function of the cation-S interaction between Lys C32 and Tyr M295 in the Cyt:RC complex was studied by mutation of both residues (Paddock et al., 2005). The mutation of Tyr M295 to non-aromatic
Herbert Axelrod, Osamu Miyashita and Melvin Okamura
330
A) Electrostatic
8
B) Hydrophobic
6
10
5
7
ke (s )
10
-1
-1
ke (s )
10
10
4
10
6
10
3
3
10
4
10
5
10
6
7
10
10
8
10
-1
KA=1/KD (M )
10
3
10
4
10
5
10
6
10
7
10
-1
KA=1/KD (M )
Fig. 4. Electron transfer rate ke with Cyt c2 in the bound state vs. the association constant KA for different mutant RCs. A) RCs modified by changes to charged residues (Tetreault et al., 2001). B) RCs modified by changes to hydrophobic residues (Gong et al., 2003).
residues had only a relatively small (3-fold ) effect on KD in contrast to theoretical estimates which predict a large binding interaction for the complex (Gallivan and Dougherty, 1999). The small effect on binding due to the mutation probably results from the energy needed for desolvation of the positively charged Arg C32 in order to form the cation-π complex. This desolvation energy reduces the binding energy for formation of the cation-π bond. Mutation of residues forming hydrogen bonds across the Cyt:RC interface resulted in changes in KD (up to six-fold) indicating that hydrogen bonding interactions stabilize the bound state. (Abresch et al., 2006) As with the case of hydrophobic interactions the magnitude of the change in KD was correlated with solvent accessibility. The residues exhibiting the largest changes in KD showed the lowest solvent accessibility in the Cyt:RC complex. B. Effects on Electron Transfer in the Bound State, ke Residues in the interface region were mutated to determine how structural changes affect electron transfer. The changes to ke were different for mutations of charged residues or mutations of residues having short-range inter-protein contacts. Mutation of charged residues changed the binding affinity KA=1/KD but did not change the electron transfer rate, ke, in the bound state (Fig. 4A) (Tetreault et al., 2001; Tetreault et al., 2002). In contrast, the mutation of hydrophobic residues changed both the binding affinity and electron transfer rate (Fig. 4B) (Gong et al., 2003). These results can be understood
in terms of changes in the electronic coupling interaction between donor and acceptor groups, which is responsible for electron transfer. The electronic coupling interaction decreases exponentially with increasing distance between donor and acceptor and depends on the pathway for electron transfer between donor and acceptor. The mutational results indicate that the pathway for electron transfer is through the short-range interaction domain around Tyr L162 in contact with the heme edge (Fig. 2C). The mutations of charged residues change the binding energy, but do not change the structure of the tunneling contact region in the bound state. Thus, ke is not changed by mutation of charged residues. On the other hand, mutations of residues in the short-range interaction region change the short-range contacts between the electron donor (heme) and acceptor (BChl2). Modifications of these contacts change the electronic coupling responsible for electron transfer. Thus, ke is changed by mutations in the short-range interaction region. An interesting feature of the results of mutations is the strong correlation between the changes in binding affinity KA and in the change in electron transfer rate ke. The correlation suggests that there is a relationship between the interactions responsible for electron transfer and the interactions responsible for binding (Gong et al., 2003). C. Effects on the Second Order Rate Constant, k2 The effects of mutations on the second order rate constant k2 for electron transfer between Cyt c2 and the RC were studied to determine the mechanism of
Cytochrome c2 Interactions
Chapter 17
331
10
10
9
10
k2 (s M )
9
1
1
10
1
1
k2 (s M )
A) Electrostatic
8
10
B) Hydrophobic
8
10
7
10 3
10
4
10
5
6
10
7
10
10
8
10
- -11
K =1/K (M ) A
D
3
10
4
5
10
6
10
10
K =1/K A
D
7
10
-1
(M )
Fig. 5. Second order electron transfer rate constant, k2, vs. association constant KA for reaction between Cyt c2 and mutant RCs. A) RCs modified by changes to charged residues (Tetreault et al., 2001). B) RCs modified by changes to hydrophobic residues (Gong et al., 2003). The triangles indicate reactions displaying single exponential (fast exchange) kinetics.
the reaction. Lin et al. (1994) showed that the driving force dependence of k2 is independent of the driving force for electron transfer. This finding indicates that electron transfer is not the rate limiting step for thesecond order rate. Instead, the association process for the formation of the bound state is the rate limiting step expected for a diffusion limited reaction. (i.e., k2 = kon ). The mutation of charged residues on the RC (Tetreault et al., 2001, 2002) and on the Cyt (Caffrey et al., 1992) changed the second order rate constant, indicating the role of electrostatic interactions in the association process. The changes in k2 due to mutations that modify charge were generally smaller than the changes in KD but were correlated (Fig. 5A). The effects of charge changes on k2 can be understood in terms of transition state theory for a chemical reaction. The transition state is the configuration in which the Cyt is in position to form the bound state of the Cyt: RC complex. Electrostatic interactions stabilize the Cyt in the transition state. However, since the distance between the Cyt and the RC is greater in the transition state than in the bound state, the electrostatic effect due to the Coulomb potential energy (which falls as 1/r ) is smaller for the transition state (affecting k2) than for the bound state (affecting KD). The relative changes in k2 and KD can be accounted for if the Cyt in the transition state is about 10 Å farther away from the RC than in the bound state (Tetreault et al., 2001; Miyashita et al., 2003b). The changes in k2 due to mutations of hydrophobic residues (Gong et al., 2003) showed complex behavior (Fig. 5B). Hydrophobic mutations resulting in small
decreases in KA resulted in only small changes in k2 (i.e., the slope is low on the right side of Fig. 5B). These small changes in k2 indicate that kon is almost unaffected by short-range interactions. The hydrophobic mutations change KD by changing koff. Mutations that produced large decreases in KA resulted in large decreases in k2. (i.e., the slope is higher on the left side of Fig. 5B). The onset of the large decreases in k2 was coincident with the appearance of monophasic rather than biphasic kinetics (triangles in Fig. 5B). This change in kinetics indicates a change from the diffusion limit (ke >> koff) to the fast exchange limit ke << koff. The switch from the diffusion limit to the fast exchange limit is caused by two factors; a faster dissociation rate koff resulting from disruption of short-range interactions, and a decrease in the electron transfer rate ke, resulting from disruption of electron transfer pathways. These studies show that short-range hydrophobic interactions are important for the second order rate constant. The hydrophobic interactions bind the Cyt in the active site for a time long enough for electron transfer to occur. This allows the reaction to proceed in the diffusion limited regime. The fact that some hydrophobic mutations change KA but not k2 (i.e., the mutations plotted on the right side of Fig. 5B) is not an indication that hydrophobic interactions are unimportant. These mutations shorten the lifetime in the bound state, but do not reduce the lifetime enough to allow dissociation to occur before electron transfer. The insensitivity of k2 to small changes at the short- range interaction interface is an indication of the robustness in the functional design of the photosynthetic machinery.
Herbert Axelrod, Osamu Miyashita and Melvin Okamura
332
In summary, the long-range electrostatic interactions are important in docking the Cyt on the RC surface. The short-range interactions affect the dissociation rate and the electron transfer rate and help to optimize the two rates so that the second order electron transfer reaction is in the diffusion limited regime. VI. Mechanism of Inter-Protein Electron Transfer A. First Order Electron Transfer The first order electron transfer rate can be calculated based on the X-ray crystal structure of the Cyt:RC complex using theory for electron transfer between fixed sites in a protein. The general form for the rate of electron transfer follows Fermi’s Golden Rule (Devault, 1980) ke =
2 2π TDA ( FC)
(3)
where TDA is the tunneling matrix element coupling the electron donor and acceptor (which is distance dependent) and FC is the Franck-Condon factor (which depends on the free energy difference). Moser and Dutton (1992) use a phenomenological approach, where the distance dependence of the rate is fit with an exponential decay k = koe–βr (β =1.4 Å–1) and the Franck-Condon factor is fit using a function similar to the Marcus equation (Marcus, 1964). Using the measured value of 14.2 Å for the distance between aromatic atoms in heme and BChl gives a value for the electron transfer rate ke = 6 × 105 s–1 close to the observed value of 106 s–1. In this calculation the values of reorganization energy λ = 0.5 eV and free energy difference ∆G = –0.16 eV were used (Lin et al., 1994). Alternatively, electron transfer rates can be calculated including the details of the structure of the protein using the Pathway model of Beratan et al. (1991). Calculation of the electron transfer from Cyt to RC gives a lower electron transfer rate due to the two through-space jumps between heme and Tyr L162 and Tyr L162 and BChl, greatly reducing the electronic coupling between electron donor and acceptor (Aquino et al., 1997; Miyashita et al., 2005). However, recent quantum mechanical studies (Prytkova et al., 2005; Kurlancheek and Cave, 2006)
Fig. 6. Reaction coordinate diagram for the diffusion controlled electron transfer reaction. The Cyt and RC approach to form an electrostatic Encounter Complex followed by passage over the Transition State barrier to the Active Complex in which electron transfer occurs. Mutations to charged residues change the energies of the Active Complex, Transition State and Encounter Complex (dashed line). Modified from Tetrault et al. (2001).
indicate a stronger through-space coupling than in the original Pathway model. Values for the electron transfer rate from Cyt to RC calculated using the modified Pathway model gave a value ke =5 × 106 s–1 close to the experimental results (O. Miyashita, unpublished). In this calculation the Pathway model (Beratan et al., 1991) was used with values for through-space jumps given by Kurlancheek and Cave (2006). The Marcus equation for the Franck-Condon factor and tunneling matrix element estimated by Hopfield (1974) were used. These results indicate that the fast electron transfer rate across the protein boundary can be reasonably accounted for by the structure of the Cyt:RC complex which brings the two proteins in close contact. B. Second Order Electron Transfer The mechanism of inter-protein electron transfer reactions has been discussed in terms of key encounter complex and transition state intermediates. These states are involved in docking and electron transfer, determining the second order reaction rate that is important for rapid delivery of electrons to the RC (Bendall, 1996). These are shown on the reaction coordinate diagram for the reaction between Cyt and RC in Fig. 6. As the RC and Cyt approach each other the free energy increases due to the reduced translational entropy of the complex until the electrostatic interactions between complementary charged groups lower the energy of the two proteins, thus forming a loosely-bound electrostatic encounter complex.
Chapter 17
Cytochrome c2 Interactions
This is followed by passage across the transition state barrier that leads to the active complex within which electron transfer occurs. Computational results on the Cyt:RC complex have provided a detailed picture of the association process (Miyashita et al., 2003b, 2004; Autenrieth et al., 2004). The encounter complex was found to be a diffuse ensemble of configurations in which the Cyt is generally positioned with its positively charged active surface over the negatively charged M-subunit side of the RC surface, with the cytochrome at a solvent-separated distance from the RC surface and the heme edge not oriented (Tiede et al., 1993; Adir et al., 1996; Miyashita et al., 2003b). The transition state, in contrast, was found to have the cytochrome still at a solvent separated distance (~10 Å) but oriented with its heme edge directed toward Tyr L162, the ‘hot spot’ for binding and electron transfer (Miyashita et al., 2003b, 2004). The positioning of the Cyt within proximity of the binding surface of the RC in the transition state is followed by the collapse to the active state which is assumed to have the co-crystal structure. In this process the solvent is excluded from the region between the proteins. The role of water molecules in the binding process has been studied (Autenrieth et al., 2004). Short-range hydrophobic contacts are then established and the heme edge is positioned in Van der Waals contact with Tyr L162 optimized for rapid electron transfer. The short-range hydrophobic and hydrogen bonded contacts do not contribute to the association rate kon as shown by the mutation results presented in Fig. 5B, consistent with the large separation between Cyt and RC at the transition state. However, they serve to hold the Cyt on the RC surface long enough for electron transfer to occur before dissociation (i.e., ke > koff ), thus allowing the reaction to proceed in the diffusion-limited regime. In addition, the binding interactions are weak enough to allow a dissociation rate (koff ~ 103 s–1) fast enough not to be rate limiting in turnover. This mechanism for electron transfer between Cyt and RC explains the highly optimized diffusion-limited rate constant that is important for the rapid turnover. The role of a specific complex between electron transfer partners may vary for different systems. Several stable protein complexes have been studied in which the cofactors are in position for fast electron transfer as in the case for the Cyt:RC complex. These include complexes between cytochrome c and cytochrome c peroxidase (Pelletier and Kraut, 1992) and cytochrome c and cytochrome bc1 (Hunte et al., 2002).
333 On the other hand, in other systems a stable complex between the two proteins has not been observed. For these systems the reactions most likely occur in the fast-exchange regime by a dynamic docking mechanism (Liang et al., 2004). When a permanent bound state is not observed, the electron transfer can occur from transient formation of a specific complex in which rapid electron transfer occurs. In this case the rapid electron transfer (k2 = ke/KD) would be enhanced by close tunneling contacts between donor and acceptor that increase ke to compensate for a large KD. Such a specific complex may explain the reaction between the Cyt c2 and RC from Rsp. rubrum where the kinetics are strictly monophasic (van der Wal et al., 1987). Although there is no direct evidence for complex formation, a computational study of the binding between Cyt c2 and the RC from Rsp. rubrum shows that a complex similar to that found in Rba. sphaeroides can be formed but with lower binding affinity (Pogorelov et al., 2007). Alternatively, electron transfer may occur from an ensemble of states such as in a non-specific electrostatic encounter complex. In this case, the complex between the two proteins may not be very specific or well optimized for electron transfer as suggested in the ‘Velcro model’ of McClendon (1991) and the electron transfer may proceed through longer distances through water. Studies that indicate that long distance electron transfer through water can occur (Lin et al., 2005; Miyashita et al., 2005). The actual electron transfer mechanism for these systems depends upon the relative rates and occupancies at different docking positions. Further studies of the electron transfer are needed to establish the mechanisms that form the molecular basis for these critical biological processes. Acknowledgments We wish to thank the many colleagues who contributed to the work from our group on the cytochrome: RC complex reported here. Special thanks to George Feher, Mark Paddock, Ed Abresch and Charlene Chang for contributions to ongoing research and to José Onuchic for theoretical discussions. Work supported by NIH grant GM 41637. References Abresch EC, Villalobos M, Paddock ML, Chang C and Okamura MY (2006) The importance of buried H-bonds on binding and
334
Herbert Axelrod, Osamu Miyashita and Melvin Okamura
electron transfer in the cytochrome c2:reaction center complex. Biophysical Society Meeting Abstracts, Biophys J, Supplement, 503a, Abstract, 2400 Adir N, Axelrod H, Beroza P, Isaacson R, Rongey S, Okamura M and Feher G (1996) Co-crystallization and characterization of the photosynthetic reaction center-cytochrome c2 complex from Rhodobacter sphaeroides. Biochemistry 35: 2535–2547 Allen JP, Feher G, Yeates TO, Rees DC, Deisenhofer J, Michel H and Huber R (1986). Structural homology of reaction centers from Rhodopseudomonas sphaeroides and Rhodopseudomonas viridis as determined by x-ray diffraction. Proc Natl Acad Sci USA 83: 8589–8593 Allen JP, Feher G, Yeates TO, Komiya H and Rees DC (1987) Structure of the reaction center from Rhodobacter sphaeroides R-26: The protein subunits. Proc Natl Acad Sci USA 84: 6162–6166 Aquino A, Beroza P, Reagan J and Onuchic J (1997) Estimating the effect of protein dynamics on electron transfer to the special pair in the photosynthetic reaction center. Chem Phys Lett 275: 181–187 Autenrieth F, Tajkhorshid E, Schulten K and Luthey-Schulten Z (2004) Role of water in transient cytochrome c2 docking. J Phys Chem B 108: 20376–20387 Axelrod H and Okamura M (2005) The structure and function of the cytochrome c2:reaction center electron transfer complex from Rhodobacter sphaeroides. Photosynth Res 85: 101–114 Axelrod H, Feher G, Allen J, Chirino A, Day M, Hsu B and Rees D (1994) Crystallization and X-ray structure determination of cytochrome c2 from Rhodobacter sphaeroides in three crystal forms. Acta Cryst D: Biol Crystallogr 50: 596–602 Axelrod HL, Abresch EC, Okamura MY, Yeh AP, Rees DC and Feher G (2002) X-ray structure determination of the cytochrome c2:reaction center electron transfer complex from Rhodobacter sphaeroides. J Mol Biol 319: 501–515 Bendall D (1996) Interprotein electron transfer. In: Bendall D (ed) Protein Electron Transfer, pp 43–68. Bios Scientific Publishers Ltd, Oxford Beratan D, Betts J and Onuchic J (1991) Protein electron transfer rates set by the bridging secondary and tertiary structure. Science 252: 1285–1288 Blankenship RE (2002) Molecular Mechanisms of Photosynthesis. Blackwell Science, London Bogan A and Thorn K (1998) Anatomy of hot spots in protein interfaces. J Mol Biol 280: 1–9 Caffrey MS, Bartsch RG and Cusanovich MA (1992) Study of the cytochrome c2-reaction center interaction by site-directed mutagenesis. J Biol Chem 267: 6317–6321 Chance B and Smith L (1955) Respiratory pigments of Rhodospirillum rubrum. Nature 175: 803–806 Chang C-H, Tiede D, Tang J, Smith U, Norris J and Schiffer M (1986) Structure of Rhodopseudomonas sphaeroides R-26 reaction center. FEBS Lett 205: 82–86 Crofts AR and Wraight CA (1983) The electrochemical domain of photosynthesis. Biochim Biophys Acta 726: 149–185 Deisenhofer J and Michel H (1989) The photosynthetic reaction center from the purple bacterium Rhodopseudomonas viridis. EMBO J 8: 2149–2170 Devanathan S, Salamon Z, Tollin G, Fitch J, Meyer T and Cusanovich M (2004) Binding of oxidized and reduced cytochrome c2 to photosynthetic reaction centers: Plasmon-waveguide resonance
spectroscopy. Biochemistry 43: 16405–16415 Devault D (1980) Quantum Mechanical Tunneling in Biological Systems. Q Rev Biophys 13: 387–564 Drepper F, Dorlet P and Mathis P (1997) Cross-linked electron transfer complex between cytochrome c2 and the photosynthetic reaction center of Rhodobacter sphaeroides. Biochemistry 36: 1418–1427 Dutton PL and Prince RC (1978) Reaction center driven cytochrome interactions. In: Clayton RK and Sistrom WR (eds) The Photosynthetic Bacteria, pp 525–570. Plenum Press, New York Duysens L (1952) Transfer of excitation energy in photosynthesis. Doctoral dissertation. Utrecht University, Utrecht, The Netherlands Duysens L (1954) Reversible photo-oxidation of a cytochrome pigment in photosynthesizing Rhodospirillum rubrum. Nature 173: 692–693 Ermler U, Fritsch G, Buchanan S and Michel H (1994) Structure of the photosynthetic reaction center from Rhodobacter sphaeroides at 2.65 Å resolution — Cofactors and protein cofactor interactions. Structure 2: 925–936 Hopfield JJ (1974) Electron transfer between biological molecules by thermally activated tunneling. Proc Natl Acad Sci USA 71: 3640–3645 Gallivan JP and Dougherty DA (1999) Cation-π interactions in structural biology. Proc Natl Acad Sci USA 96: 9459–9464 Gerencsér L, Laczkó G and Maróti P (1999) Unbinding of oxidized cytochrome c from photosynthetic reaction center of Rhodobacter sphaeroides is the bottleneck of fast turnover. Biochemistry 38: 16866–16875 Gong X, Paddock M and Okamura M (2003) Interactions between cytochrome c2 and photosynthetic reaction center from Rhodobacter sphaeroides: Changes in binding affinity and electron transfer rate due to mutation of interfacial hydrophobic residues are strongly correlated. Biochemistry 42: 14492–14500 Graige M, Feher G and Okamura M (1998) Conformational gating of the electron transfer reaction QA–.QB → QAQB–. in bacterial reaction centers of Rhodobacter sphaeroides determined by a driving force assay. Proc Natl Acad Sci USA 95: 11679–11684 Hunte C, Solmaz S and Lange C (2002) Electron transfer between yeast cytochrome bc1 complex and cytochrome c: A structural analysis. Biochim Biophys Acta 1555: 21–28 Ke B, Chaney TH and Reed DW (1970) The electrostatic interaction between the reaction center bacteriochlorophyll derived from Rhodopseudomonas sphaeroides and mammalian cytochrome c and its effects on light activated electron transport. Biochim Biophys Acta 216: 373–383 Kurlancheek W and Cave RJ (2006) Tunneling through weak interactions: A comparison of through-space, H-bond, and throughbond mediated tunneling. J Phys Chem 110: 14018–14028 Larson J and Wraight C (2000) Preferential binding of equine ferricytochrome c to the bacterial photosynthetic reaction center from Rhodobacter sphaeroides. Biochemistry 39: 14822–14830 Liang ZX, Kurnikov IV, Nocek JM, Mauk AG, Beratan DN and Hoffman BM (2004) Dynamic docking and electron-transfer between cytochrome b5 and a suite of myoglobin surface-charge mutants. Introduction of a functional-docking algorithm for protein-protein complexes. J Am Chem Soc 126: 2785–2798
Chapter 17
Cytochrome c2 Interactions
Lin J, Balabin I and Beratan D (2005) The nature of aqueous tunneling pathways between electron-transfer proteins. Science 310: 1311–1313 Lin X, Williams JC, Allen J and Mathis P (1994) Relationship between rate and free energy difference for electron transfer from cytochrome c2 to the reaction center in Rhodobacter sphaeroides. Biochemistry 33: 13517–13523 Long J, Durham B, Okamura M and Millett F (1989) Role of specific lysine residues in binding cytochrome c2 to the Rhodobacter sphaeroides reaction center in optimal orientation for rapid electron transfer. Biochemistry 28: 6970–6974 Marcus RA (1964) Chemical and electrochemical electron transfer theory. Ann Rev Phys Chem 15: 155–196 McLendon G (1991) Control of biological electron transport via molecular recognition and binding: The ‘Velcro’ model. Struct Bond 75: 160–174 Meyer T and Donohue T (1995) Cytochromes, Iron-sulfur, and copper proteins mediating electron transfer from the Cyt bc1 complex to photosynthetic reaction center complexes. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 725–745. Kluwer Academic Publishers, Dordrecht Miyashita O, Okamura MY and Onuchic JN (2003a) Theoretical understanding of the interprotein electron transfer between cytochrome c2 and the photosynthetic reaction center. J Phys Chem B 107: 1230–1241 Miyashita O, Onuchic JN and Okamura MY (2003b) Continuum electrostatic model for the binding of cytochrome c2 to the photosynthetic reaction center from Rhodobacter sphaeroides. Biochemistry 42: 11651–11660 Miyashita O, Onuchic J and Okamura M (2004) Transition state and encounter complex for fast association of cytochrome c2with bacterial reaction center. Proc Natl Acad Sci USA 101: 16174–16179 Miyashita O, Okamura and Onuchic J (2005) Interprotein electron transfer from cytochrome c2 to photosynthetic reaction center: Tunneling across an aqueous interface. Proc Natl Acad Sci USA 102: 3558–3563 Moser C and Dutton PL (1988). Cytochrome c and c2 binding dynamics and electron transfer with photosynthetic reaction center protein and other integral membrane redox proteins. Biochemistry 27: 2450–2461 Moser C and Dutton P (1992) Engineering protein structure for electron transfer in photosynthetic reaction centers. Biochim Biophys Acta 1101: 171–176 Nitschke W and Dracheva SM (1995) Reaction center associated cytochromes. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration), pp 775–805. Kluwer Academic Publishers, Dordrecht Nogi T, Hirano Y and Miki K (2005) Structural and functional studies on the tetraheme cytochrome subunit and its electron donor proteins: The possible docking mechanisms during the electron transfer reaction. Photosynth Res. 85: 87–99 Overfield RE, Wraight CA and Devault DC (1979) Microsecond photooxidation kinetics of cytochrome c2 from Rhodopseudomonas sphaeroides: In vivo and solution studies. FEBS Lett. 105: 137–142 Paddock M, Weber K, Chang C and Okamura M (2005) Interac-
335 tions between cytochrome c2 and the photosynthetic reaction center from Rhodobacter sphaeroides: The cation-π interaction. Biochemistry 44: 9619–9625 Parson W (1968) The role of P870 in bacterial photosynthesis. Biochim Biophys Acta 153: 248–259 Pelletier H and Kraut J (1992) Crystal structure of a complex between electron transfer partners, cytochrome c peroxidase and cytochrome c. Science 258: 1748–1755 Pogorelov T, Autenrieth F, Roberts E and Luthey-Schulten Z (2007) Cytochrome c2 exit strategy: Dissociation studies and evolutionary implications. J Phys Chem B 111: 618–634 Prince RC, Cogdell RJ and Crofts AR (1974) The photo-oxidation of horse heart cytochrome c and native cytochrome c2 by reaction centres from Rhodopseudomonas spheroides R-26. Biochim Biophys Acta 347: 1–13 Prytkova TR, Kurnikov IV and Beratan DN (2005) Ab initio based calculations of electron-transfer rates in metalloproteins. J Phys Chem B 109: 1618–1625 Rosen D, Okamura MY and G Feher (1980) Interaction of cytochrome c with reaction centers of Rhodopseudomonas sphaeroides R-26: Determination of number of binding sites and dissociation constants by equilibrium dialysis. Biochemistry 19: 5687–5692 Rosen D, Okamura MY, Abresch EC, Valkirs GE and Feher G (1983) Interaction of cytochrome c with reaction centers of Rhodopseudomonas sphaeroides R-26: Localization of the binding site by chemical cross-linking and immunochemical studies. Biochemistry 22: 335–341 Tetreault M, Rongey SH, Feher G and Okamura M (2001) Interaction between cytochrome c2 and the photosynthetic reaction center from Rhodobacter sphaeroides: Effects of charge-modifying mutations on binding and electron transfer. Biochemistry 40: 8452–8462 Tetreault M, Cusanovich M, Meyer T, Axelrod H and Okamura M (2002) Double mutant studies identify electrostatic interactions that are important for docking cytochrome c2 onto the bacterial reaction center. Biochemistry 41: 5807–5815 Tiede D (1987) Cytochrome c orientation in electron transfer complexes with photosynthetic reaction centers of Rhodobacter sphaeroides and when bound to the surface of negatively charged membranes: Characterization by optical linear dichroism. Biochemistry 26: 397–410 Tiede DM and Chang CH (1988) The cytochrome-c binding surface of reaction centers from Rhodobacter sphaeroides. Isr J Chem 28: 183–191 Tiede D and Dutton P (1993) Electron transfer between bacterial reaction centers and mobile c-type cytochromes. In: Deisenhofer J and Norris J (eds) The Photosynthetic Reaction Center, Vol 1, pp 258–288. Academic Press, San Diego Tiede D, Vashishta A and Gunner M (1993) Electron-transfer kinetics and electrostatic properties of the Rhodobacter sphaeroides reaction center and soluble c-cytochromes. Biochemistry 32: 4515–4531 Tiede DM, Littrell K, Marone PA, Zhang R and Thiyagarajan P (2000) Solution structure of a biological bimolecular electron transfer complex: characterization of the photosynthetic reaction center-cytochrome c2 protein complex by small angle neutron scattering. J Appl Cryst 33: 560–564 van der Wal H, van Grondelle R, Millett F and Knaff D (1987) Oxidation of cytochrome c2 and of cytochrome c by reaction cen-
336
Herbert Axelrod, Osamu Miyashita and Melvin Okamura
ters of Rhodospirillum rubrum and Rhodobacter sphaeroides. The effect of ionic strength and of lysine modification on oxidation rates. Biochim Biophys Acta 893: 490–498 Vernon L and Kamen M (1953) Studies on the metabolism of
photosynthetic bacteria. XV. Photoautoxidation of ferrocytochrome c in extracts of Rhodospirillum rubrum. Arch Biochem Biophys 44: 298–311
Chapter 18 Directed Modification of Reaction Centers from Purple Bacteria JoAnn C. Williams and James P. Allen* Department of Chemistry and Biochemistry and Center for Bioenergy & Photosynthesis, Arizona State University, Tempe AZ 85287-1604, U.S.A.
Summary ................................................................................................................................................................... 1 I. Introduction......................................................................................................................................................... 2 II. Properties of the Cofactors................................................................................................................................. 2 A. Identity, Substitution, and Removal ...................................................................................................... 2 B. Optical Spectra ..................................................................................................................................... 3 C. Oxidation/Reduction Midpoint Potentials ............................................................................................. 4 D. Modeling the Electronic Structure of the Bacteriochlorophyll Dimer .................................................... 6 III. Electron Transfer Concepts................................................................................................................................ 7 A. Energetics ............................................................................................................................................ 7 B. Coupling ............................................................................................................................................... 9 C. Dynamics ............................................................................................................................................. 9 IV. Pathways of Electron Transfer ......................................................................................................................... 10 A. B-side Electron Transfer..................................................................................................................... 10 B. New Electron Transfer Reactions....................................................................................................... 12 V. Conclusions ...................................................................................................................................................... 13 Acknowledgments ................................................................................................................................................... 13 References .............................................................................................................................................................. 13
Summary Reaction centers from purple bacteria form a superb test system for the manipulation of electron transfer parameters. The wealth of cofactors and electron transfer reactions provides opportunities for directed modification of specific properties. In particular, the energies of each cofactor can be selectively changed by mutations of neighboring amino acid residues. The starting point for the initial electron transfer, the bacteriochlorophyll dimer, has proven to be exceptionally malleable, allowing large changes in energetics and rates. Most of the other cofactors can be exchanged or eliminated entirely, permitting considerable alteration of pathways. By orchestrating multiple changes in the reaction center, the light-initiated electron transfer pathway can be directed towards alternate ends, for example down the B branch of cofactors rather than the naturally preferred A branch. Extensive modeling of features of electron transfer such as the energetics, the coupling, and the protein dynamics has been corroborated by observed changes in the characteristics of the reactions after modification of the cofactor properties. For example, the maximum rates for several electron transfer reactions, determined by application of Marcus theory to the rates of reactions in a range of mutants, show a correlation with the
*Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 337–353. © 2009 Springer Science + Business Media B.V.
338
JoAnn C. Williams and James P. Allen
distance between the cofactors. Other measurements revealing the intimate interaction of the protein and cofactors show that protein motion controls the rate of the initial electron transfer. Thus the reaction center provides a natural and modifiable template for understanding the factors governing electron transfer. I. Introduction What happens when the reaction center is excited by light? The light energy is converted into chemical energy through a series of electron and proton transfer reactions involving the cofactors of the reaction center. These reactions are able to proceed with essentially every light photon producing useful reactions, corresponding to a quantum efficiency of nearly 100%. The balancing act of capturing light energy while not destroying the molecules involved or producing unfavorable side reactions is achieved by fine tuning the properties of the cofactors through interactions with the protein in which they are embedded. The protein can influence several aspects of electron transfer identified by theoretical treatments, including the energetics, the coupling, and the protein dynamics. Although it is difficult to separate the contributions of each of these components, experiments have probed their effects, notably by altering the light-induced reactions in specific ways. This chapter will review the partnership between the cofactors and the protein scaffold as it relates to the parameters of electron transfer models, focusing on examples where changes in the properties have measurable effects on electron transfer. The examples will be primarily of reaction centers from Rhodobacter (Rba.) sphaeroides and Rba. capsulatus, the two commonly studied reaction center systems of purple bacteria. II. Properties of the Cofactors A. Identity, Substitution, and Removal The reaction center from Rba. sphaeroides and Rba. capsulatus is an integral membrane protein complex composed of three proteins: the L, M, and H subunits. The L and M subunits form the core of the protein and are largely composed of five transmembrane helices Abbreviations: BA – bacteriochlorophyll monomer on A branch of cofactors; BB – bacteriochlorophyll monomer on B branch of cofactors; HA – bacteriopheophytin on A branch of cofactors; HB – bacteriopheophytin on B branch of cofactors; P – bacteriochlorophyll dimer; QA – quinone on A branch of cofactors; QB – quinone on B branch of cofactors; Rba. – Rhodobacter
that are structurally related to each other by an approximate two-fold symmetry axis. The H subunit is more peripheral, containing one transmembrane helix and a large cytoplasmic domain. Embedded in the middle of the L and M subunits are ten cofactors, all of which can participate in some manner in energy or electron transfer. The ten cofactors in the reaction center are: two bacteriochlorophyll a molecules that form a dimer (P), two bacteriochlorophyll a monomers (BA and BB), two bacteriopheophytin a molecules (HA and HB), two ubiquinone molecules (QA and QB), a carotenoid molecule, and an iron (Fig. 1). These cofactors are arranged into two branches, identified as the A and B branches, which are related by the same two-fold symmetry as found for the L and M subunits. Although the cofactors are normally expressed with a well-defined composition, some of these cofactors can be substituted by molecules in the same class, for example bacteriochlorophylls for bacteriopheophytins. The monomer bacteriochlorophylls and bacteriopheophytins, the quinones, the carotenoid, and the iron can be biochemically removed and replaced. Mutagenesis can also result in biosynthetic substitutions, primarily by replacement of the amino acid residues coordinating the cofactors. For example, when the ligand to one of the central Mg atoms of P is changed, a bacteriopheophytin is incorporated rather than bacteriochlorophyll in a mutant that has been termed a heterodimer (His L173 to Leu and His M202 to Leu in Rba. sphaeroides, L173 and M200 in Rba. capsulatus) (Bylina and Youvan, 1988; Kirmaier et al., 1988; McDowell et al., 1991; Allen et al., 1996; van Brederode et al., 1999; King et al., 2001). Both halves of the dimer can be individually replaced this way, although the double bacteriopheophytin dimer appears to be unstable. Similarly the binding site for the B-side bacteriochlorophyll monomer is found to contain a bacteriopheophytin when the residue forming its Mg ligand is changed (φ mutant, His M182 to Leu in Rba. sphaeroides), although analogous mutations on the A side do not appear to have the same substitution effect (Katilius et al., 1999, 2004). Conversely, the bacteriopheophytin on the A side can be converted to bacteriochlorophyll by introduction of a residue to act as a ligand (β mutant, Leu M214
Chapter 18
Reaction Center Alteration
Fig. 1. Three-dimensional structure of the cofactors of the reaction center from Rba. sphaeroides R-26. Shown are the bacteriochlorophyll a dimer (P) (shaded dark), the two bacteriochlorophyll a monomers (BA and BB), the two bacteriopheophytin a molecules (HA and HB), the two ubiquinone molecules (QA and QB), and the iron (Fe). Although the carotenoid is present in wild type, it is not present in the R-26 strain. The view is perpendicular to the approximate two-fold symmetry axis that passes from P to Fe in the plane of the paper.
to His in Rba. sphaeroides, M212 in Rba. capsulatus) (Kirmaier et al., 1991). The addition of a histidine near the bacteriopheophytin on the B branch also results in incorporation of a bacteriochlorophyll (Leu L185 to His in Rba. sphaeroides) (Watson et al., 2005). Changing iron ligands results in a loss of metal specificity, and in one case a significant amount of zinc is incorporated (His M266 to Cys in Rba. sphaeroides) (Williams et al., 2007). In addition to altering the cofactor composition, some of the cofactors can be removed or their incorporation can be blocked. Many studies of reaction centers have been performed on a carotenoid-less strain of Rba. sphaeroides, identified as R-26, which shows properties essentially identical to the carotenoid-containing wild type except for the loss of the ability to trap excess energy. The quinones can be taken out by exposing the reaction centers to a detergent treatment, which initially results in a decrease in QB followed by loss of QA. Biosynthetic incorporation of QA can be blocked by substitution of amino acid residues forming the binding pocket,
339 for example by removing the tryptophan in van der Waals contact with QA (M252 in Rba. sphaeroides and M250 in Rba. capsulatus), or adding a tryptophan in place of a smaller residue (Ala M260 to Trp in Rba. sphaeroides) (Breton et al., 2004). Certain mutations near P result in reaction centers that lack a functional P (Val L157 to Arg, His L153 to Glu, Leu, Gln, or Tyr, His L173 to Gly, and His M202 to Gly in Rba. sphaeroides) (Jackson et al., 1997; Moore and Boxer, 1998; Katilius et al., 2004). The loss of HA is one outcome of the large-scale alterations of the DLL mutant, in which the D transmembrane sequence of the M subunit is replaced with the symmetry-related segment of the L subunit (M192 to M217 replaced with L165 to L190 in Rba. capsulatus) (Robles et al., 1990). Similarly, the B-branch bacteriopheophytin is not required for assembly of the reaction center as shown by a mutant with the change of an alanine that is adjacent to HB to tryptophan (M149 in Rba. sphaeroides)(Watson et al., 2005). A loss of bacteriochlorophyll (presumably in the dimer) was also reported to be due to structural and electrostatic changes in a residue located between BB and P (Ile L177 to His in Rba. sphaeroides) (Khatypov et al., 2005). See Chapter 16, Jones, for a summary of the effects of exclusion and replacement of reaction center cofactors. The ability to alter the cofactor composition provides the opportunity to manipulate the electron transfer reactions as discussed below. B. Optical Spectra One of the most accessible properties of the cofactors is the absorption spectrum (Fig. 2). The tetrapyrrole pigments (P, BA, BB, HA, and HB) have absorption peaks in the near-infrared region, the visible region and the UV region, and the quinones have an unresolved band in the visible region. The bacteriopheophytins, monomer bacteriochlorophylls, and dimer bacteriochlorophylls can be distinguished from each other in the near-infrared peaks at 760 nm, 800 nm, and 865 nm, respectively. In the visible region, the 540 nm peak arises from the bacteriopheophytins, and the 590 nm peak is from all four bacteriochlorophylls. The A and B branch pigments of the same type overlap, except at low temperature where the broad peak in the 540 nm region of the spectrum is resolved into two peaks at 533 nm and 546 nm associated with HB and HA, respectively. The contributions of the tetrapyrrole pigments in the Soret region have been delineated, with H contributing primarily on the
340
Fig. 2. Absorption spectrum of reaction centers from wild-type Rba. sphaeroides. The primary contributions of the bacteriochlorophyll dimer (P), bacteriochlorophyll monomers (B), and bacteriopheophytin monomers (H) to each of the absorption bands are identified. The Soret band arises from all the tetrapyrrole cofactors.
blue side and P absorbing on the red side (Wang et al., 2006). The carotenoid has a major but poorly resolved contribution to the absorption near 500 nm, where absorption of the other cofactors is weak, and can transfer excitation energy to the bacteriochlorophylls, performing a light-harvesting function in addition to photoprotection (Lin et al., 2003). The absorption peaks in the near-infrared region are shifted from the solution spectra, notably in the shift of the bacteriochlorophyll bands to longer wavelengths, presumably because of their protein environment. However, specific alteration of the spectrum generally has not been amenable to mutagenesis. Modeling of the excited states has also proven difficult (Dahlbom and Reimers, 2005). However in certain instances, shifts in the spectrum are attributed to changing particular residues. A notable example is the change in the visible region peak of the bacteriopheophytins due to changing a hydrogen bond to the keto group of HA, which established the assignment of these optical bands to the individual bacteriopheophytins (Bylina et al., 1988). However, changing a hydrogen bond is not typically correlated with a shift in the peak of the tetrapyrrole pigments. Another relatively malleable absorption peak is that of the dimer. It can shift up to approximately 15 nm to shorter wavelengths as a result of mutations, mostly mutations in which a hydrogen bond to the acetyl group is changed, precipitating a rotation of this side group. Major changes in the absorption spectra also occur when substitutions
JoAnn C. Williams and James P. Allen of the cofactors are introduced. For example, the nearinfrared absorption peak of the dimer is significantly different in the heterodimer mutant. Likewise, shifts in the both the visible and near-infrared regions of the spectra are observed in mutants with alterations of ligands to the monomer bacteriochlorophylls that result in pigment changes (Katilius et al., 1999, 2004). Thus, the optical spectrum is a sensitive indicator of the effects of certain types of modifications to the reaction center. The absorption bands in the near-infrared region arise from transitions from the ground state to the first excited state and so are markers of the excited state energy, indicating the maximum amount of energy that can be captured. For the primary donor of Rba. sphaeroides, the absorption peak at 865 nm corresponds to an energy difference of 1.4 eV. The properties of the reaction center can also be characterized from measurement of the spontaneous and stimulated emission of the excited state of the dimer, centered near 915 nm. Because the absorption and emission bands change as the cofactors undergo excitation, oxidation, and reduction, transient optical spectroscopy is one of the major techniques utilized to follow the light-induced transfer of electrons in the reaction center. C. Oxidation/Reduction Midpoint Potentials The oxidation/reduction midpoint potentials of the reaction center cofactors are critical properties for their function as electron transfer components. The midpoint potential of the dimer, at approximately 500 mV, is the only one easily measured directly (Fig. 3). The midpoint potentials of the other cofactors can only be inferred. A change in the chemical nature of a cofactor has a direct effect on its midpoint potential. For example, incorporation of a bacteriopheophytin in the heterodimer mutant increases the potential by approximately 130 mV due to the intrinsically higher potential of bacteriopheophytin (Allen et al., 1996). In addition, the energies of the electronic states are sensitive to the environment so protein interactions with the dimer, including hydrogen bonds and electrostatic forces from charged residues, can affect the dimer midpoint potential and be modulated by mutagenesis. The midpoint potentials of other tetrapyrroles in the reaction center can presumably also be changed by similar modifications, although the evidence is based upon alterations of the electron transfer rates rather than direct measurements.
Chapter 18
Reaction Center Alteration
341
Fig. 3. (Left) The near-infrared region of the absorption spectrum showing a systematic decrease in the band associated with the bacteriochlorophyll dimer as the ambient potential is increased. The absorbance of this band is used to determine the fraction of P+ at any given potential compared to the total amount of P. (Right) Fits of the dependence of the fraction of P+ on the ambient potential are used to determine the oxidation-reduction midpoint potential, Em. A shift in the midpoint potential is observed for a number of different mutants (Allen and Williams, 1995).
Whereas some modifications lead to large changes, analysis of various mutations indicates that most result in small (<50 mV) increases in the P/P+ midpoint potential (Spiedel et al., 2002). The predominance of increases in the midpoint potential of P, rather than a mix of increases and decreases, is probably because the midpoint potential is poised by the protein interactions at a minimum value, and perturbations primarily disrupt this poise. The effect of hydrogen bonds to the conjugated carbonyl molecules on the oxidation/reduction midpoint potential of the dimer is well documented (Allen and Williams, 1995, 2006). An additional hydrogen bond raises the potential by 60 to 120 mV while loss of the hydrogen bond present in the wild type decreases the potential by approximately 80 mV (Stocker et al., 1992; Williams et al., 1992; Murchison et al., 1993). Multiple systematic changes in the hydrogen bonding pattern produce a wide range of midpoint potentials (Lin et al., 1994a) and provide the opportunity to investigate the effect of altered energetics on the properties of the reaction center as discussed below. Hydrogen bonds have been introduced to the monomer bacteriochlorophylls, but with little effect on the primary photochemistry (Chen et al., 2004). A water molecule is found in the wild type between P and BA in hydrogen bonding position to the keto carbonyl of BA. This water molecule can be displaced by mutations (at M203 in Rba. sphaeroides), resulting in alteration in the electron transfer characteristics, including a slowing of the primary electron transfer
rates, suggesting that BA is more difficult to reduce, although other factors, such as an increase in the reorganization energy (see below), would also result in a slower rate (Potter et al., 2005; Yakovlev et al., 2005). Hydrogen bonds to bacteriopheophytins are also probably influencing their midpoint potentials. On the A side, a hydrogen bond between the 131 keto of HA and a glutamic acid residue (L104 in Rba. sphaeroides and Rba. capsulatus) is naturally occurring, and the equivalent situation on the B side can be achieved by mutagenesis (Val M131 to Asp in Rba. capsulatus, M133 in Rba. sphaeroides) (Bylina et al., 1988; Müh et al., 1998; Kirmaier et al., 2002a). The presence of the hydrogen bonds most likely makes the bacteriopheophytins easier to reduce, consistent with the observed changes in electron transfer rates in the mutants. The potential of each cofactor is also shaped by electrostatic interactions with charged and polar amino acid residues. Insertion or removal of ionizable residues at several different locations approximately 10 to 15 Å from P (L135, L155, L164, L170, L247 and M199 in Rba. sphaeroides) leads to a midpoint potential decrease up to 60 mV due to a negative charge and an increase up to 50 mV due to a positive charge (Williams et al., 2001; Johnson and Parson, 2002; Johnson et al., 2002). The effect of the charges on the potential is relatively modest because of screening of the charge-charge interactions by the surrounding protein. The energies of BA and BB can be significantly changed by altering the polarity of residues near
342 these tetrapyrroles, including replacing a conserved tyrosine residue with phenylalanine (M210 in Rba. sphaeroides, M208 in Rba. capsulatus), and changing a conserved phenylalanine to tyrosine (L181 in Rba. sphaeroides and Rba. capsulatus), as measured by changes in the electron transfer rates (Finkele et al., 1990; Nagarajan et al., 1990; Jia et al., 1993) and Stark spectra (Treynor et al., 2004). Theoretical calculations suggest that electrostatic fields near BA and BB are influenced by the polarity of these residues, and consequently mutations result in energetic shifts for the oxidized states (Alden et al., 1996; Gunner et al., 1996). In general, the sensitivity of the cofactors to specific protein interactions provides a means to test the role of the energies of the charge-separated states in determining the electronic structure of the cofactors and the rates of electron transfer as presented in the subsequent sections. D. Modeling the Electronic Structure of the Bacteriochlorophyll Dimer The close overlap of the two tetrapyrroles in P results in a sharing of the electron orbitals and hence changes in the properties of P compared to bacteriochlorophyll monomers. For example, one effect of the dimerization is a shift in the absorption band to a longer wavelength than is observed for the bacteriochlorophyll monomers. In a simple Hückel molecular orbital model, the two conjugated molecules can be considered to be coupled together, resulting in the electrons being distributed over the two bacteriochlorophylls in molecular orbitals that
JoAnn C. Williams and James P. Allen are split by an energy 2β (Fig. 4). An additional term represents the difference in the energies of the two sides arising from the inhomogeneous nature of the protein surrounding P. When P is oxidized, the higher molecular orbital loses an electron, leaving one unpaired electron. In the Hückel model, the predominant contribution to the higher molecular orbital is from the side with the higher energy (Plato et al., 1992). In the wild-type reaction center, measurement of the unpaired electron spin densities using the magnetic resonance technique called electron nuclear double resonance yields a 2:1 ratio for the spin density on the L side bacteriochlorophyll compared to the M side. The energy difference between the two sides can be manipulated by introduction of hydrogen bonds to P, resulting in systematic changes in the molecular energies and hence the ratio of the spin densities (Fig. 4) (Artz et al., 1997; Müh et al., 2002). In general, a hydrogen bond between the side chain of an amino acid residue and the M side of P stabilizes the energy of that side and increases the energy difference of the molecular orbitals, making the spin ratio more asymmetric. A hydrogen bond to the L side of P stabilizes that side and hence decreases the energy difference for the molecular orbitals and produces a more symmetrical spin distribution. In both cases, the additional hydrogen bond lowers the energy of the highest molecular orbital, which is coupled to the midpoint potential, making P more difficult to oxidize. More extensive modeling involving additional factors, such as the contribution of vibrational states, provides estimates for the positions of bands seen in the infrared region of the optical spectrum (Müh et
Fig. 4. A two-orbital Hückel model of the bacteriochlorophyll dimer. In wild type, the energy splitting of the molecular orbitals of P is determined by the energy difference between the two sides of P and the coupling. Mutations that introduce a hydrogen bond to the M side of P stabilize the energy of the M side bacteriochlorophyll, resulting in a more asymmetrical dimer with a larger energy difference for the molecular orbitals and a larger midpoint potential (Em). Introducing a hydrogen bond to the L side results in a more symmetrical dimer while still increasing Em.
Chapter 18
Reaction Center Alteration
343
al., 2002; Reimers and Hush, 2004; Kanchanawong et al., 2006). In comparison to P in the bacterial reaction center, the electronic structure of the primary electron donor of Photosystem I, P700, is very asymmetrical with a ratio of at least 3:1 in spin density distribution over the two tetrapyrroles (Webber and Lubitz, 2001). Two factors contribute to this asymmetry. The primary donor has a heterodimeric nature as the Bside tetrapyrrole is a chlorophyll a´, which contains an epimeric configuration at the 132 position, rather than a chlorophyll a as found on the A side. Whereas both sides are axially coordinated by histidines, only the chlorophyll a´ has a hydrogen bond (between Thr 739 of the PsaA subunit and the 131 position). The presence of this hydrogen bond likely stabilizes the chlorophyll a´ side of P700 resulting in the chlorophyll a side having a much higher energy and hence a much larger unpaired spin density. Mutation of Thr 739 to Ala results in a 60 mV decrease in the oxidation/reduction potential while a 30 mV decrease is observed for a Val mutation (Witt et al., 2002; Li et al., 2004). Associated with the loss of the hydrogen bond is a small redistribution of the electron spin density from the B side to the A side of P700 as expected based upon the theoretical model. Thus, the understanding developed from the bacterial system appears to be applicable to other photosystems. III. Electron Transfer Concepts According to Marcus theory, the rate, k, of electron transfer between a donor and acceptor can be written in terms of the temperature, T, the effective electronic matrix element, or coupling, V, and energetics according to: k=
⎛ ( ∆G o + λ )2 ⎞ ⎟ 4 λRT ⎠
−⎜ 4π 3 V2 e ⎝ 2 h λRT
(1)
with two contributions to the energetics: the free energy difference between the final and initial states, ∆G°, and the reorganization energy, λ, which represents the rearrangement energy associated with the charge transfer (Marcus and Sutin, 1985). The exponential dependence of the rate on the energy terms is due to the activation energy associated with the process. Electrons can flow over long distances in biological systems because of a careful balance,
mediated by the protein, involving the free energy difference, the reorganization energy and the coupling between distant redox cofactors. The factors that influence the energetics and coupling in protein complexes are considered below. A. Energetics The electron transfer rate has an exponential dependence upon the free energy difference and reorganization energy (Eq. 1). The free energy differences between various states formed during electron transfer steps in wild type range from relatively small values of approximately 200 meV for the initial electron transfer to a value of approximately 500 meV for charge recombination from the primary and secondary quinones. Mutants with altered P/P+ midpoint potentials have corresponding changes in the free energy differences for the electron transfer steps. For example, a higher P/P+ midpoint potential increases the free energy difference between the PQA ground state and the P+QA– state. By measuring the rates for different mutants, the dependence of the electron transfer rates on the free energy difference can be experimentally determined. The dependence of the electron transfer rate typically is shown as the logarithm of the rate versus the free energy difference, yielding a parabolic curve (Fig. 5). The value of the free energy difference at the peak of the curve is equal to the reorganization energy, because for that value the exponential term has a maximal value of one. The rate at the peak of the parabola is proportional to the coupling for the reaction. In principle, the energies of the cofactors, including their charged states, can be established by theoretical calculations based upon the three-dimensional structure of the protein. However, for large pigmentprotein complexes, these efforts remain a challenge due to the large number of interactions and the uncertainties in the atomic positions often found in structural models that are limited by the resolution of the X-ray diffraction data. Electrostatic potentials calculated for the reaction center from Blastochloris viridis show a strong asymmetry in the electrostatic potential for the two branches, with the A branch being substantially more positive than the B branch, significantly favoring electron transfer along the A branch (Gunner et al., 1996). The oxidation/reduction midpoint potential for P680 of Photosystem II is calculated to have a significantly higher potential than P in bacterial reaction centers, in agreement with
344
JoAnn C. Williams and James P. Allen
Fig. 5. (Left) The dependence of the electron transfer rate on the free energy difference for four reactions: the initial P* decay, electron transfer from bound cytochrome c2 to P+, and charge recombinations from the primary and secondary quinones. The lines are fits according to the Marcus relationship (Eq. 1). The data points and fits are from previous publications (Lin et al., 1994b; Allen et al., 1998; Haffa et al., 2002). (Right) The maximum rate from each fit is proportional to the coupling for the reaction and yields a rate that is predicted to be exponentially related to the separation distance between the electron donor and acceptor. For the P* decay, the maximum rate is plotted for transfer to two possible acceptors, BA and HA. The fit shown is from Dutton and coworkers (Noy et al., 2006).
the experimental data (Ishikita et al., 2006). These calculations also suggest that the specific orientation of the 132 ester of bacteriochlorophylls has a strong influence on the potential (Ishikita et al., 2005). Calculations making use of dynamics and alternate values for the dielectric constant result in different values for the potentials but are still in agreement with favorable transfer along the A branch (Parson et al., 1990; Gehlen et al., 1994). Experimentally, the strength of the electrostatic fields can be established using a technique known as electroabsorption or Stark spectroscopy, in which changes in the optical spectrum due to the application of an electric field across a sample are measured. These optical changes can be interpreted in terms of the interactions of the applied electric field with the dipole moments of the cofactors. For the bacterial reaction center, such measurements lead to estimates of a strong field of 106 V/cm near the bacteriochlorophyll dimer (Middendorf et al., 1993). The electrostatic field around P changes by approximately 10% when comparing reaction centers with and without the carotenoid (Yanagi et al., 2005). Because the electron transfer rates are insensitive to the presence or absence of the carotenoid, the electrostatic field is not a major determinant for the directionality along the A branch.
Replacement of one of the coordinating histidine residues of P with leucine produces the heterodimer, while mutants with the histidine replaced with glycine are very similar to wild type in their properties, presumably due to the incorporation of water as a ligand in place of the histidine (M202 in Rba. sphaeroides) (Goldsmith et al., 1996). Likewise, replacement of the histidine coordinating BA with a small amino acid residue (His L153 to Ser or Gly in Rba. sphaeroides, Thr or Ser in Rba. capsulatus, and Cys in Blastochloris viridis) results in mutants with properties that are very similar to those of wild type, while mutation to leucine alters the pigment composition (Bylina et al., 1990; Arlt et al., 1996; Katilius et al., 2004). Thus, these bacteriochlorophylls are tolerant of coordination changes, with the energetics of P and BA being only weakly dependent upon the nature of the coordination. Protein interactions such as hydrogen bonds to the intermediate electron acceptors also influence the energetics of electron transfer. Removal of the hydrogen bond between a glutamic acid residue (L104 in Rba. sphaeroides and Rba. capsulatus) and the keto group at the 131 position of HA results in a small decrease in the forward electron transfer rate from P* (Bylina et al., 1988), consistent with a decrease in the free energy difference due to a change in the
Chapter 18
Reaction Center Alteration
energy of HA–. In Photosystem I, loss of the hydrogen bond between a tyrosine residue and either of the chlorophyll monomers located in positions analogous to HA and HB (Tyr 696 of the PsaA subunit near the 131 position of ec3A or Tyr 676 of the PsaB subunit near ec3B) alters the electron transfer ratio of the two branches, suggesting that in Photosystem I, the rates are very sensitive to the energetics established by the pigment-protein interactions (Li et al., 2006). B. Coupling In addition to controlling the energies of the cofactors, the protein must provide an environment that establishes a suitable coupling, or formally the effective electronic matrix element, between the cofactors poised in their excited or charged states for the reactions. In theoretical models, coupling can be considered as arising from a pathway from the donor to the acceptor. In these models, the protein is considered to consist of a complex array of bonds as well as close but non-bonding contacts. A theoretical framework has been developed to express the protein as a network of covalent bonds, hydrogen bonds, and through-space jumps along which electron transfer proceeds in many steps (Lin et al., 2005; Gray and Winkler, 2005). Rather than making use of explicit calculations, coupling is more commonly estimated by assuming that the protein environment is relatively homogeneous, with the coupling being primarily dependent upon only the distance between the donor and acceptor (Gray and Winkler, 2005; Noy et al., 2006). In this simplified case, the coupling exponentially decreases with increasing separation (Fig. 5) showing that the maximal electron transfer rates can span a range of nearly 1012 s–1 for cofactors in close contact to a rate of 25 s–1 when the separation is 23 Å. Because of the exponential dependence on distance, electron transfer in biological systems often makes use of a series of steps involving closely positioned cofactors rather than utilizing a single step between more distantly located cofactors. In order to meet the critical demands for efficient electron transfer, careful positioning of the cofactors and fine control of the energies of the cofactors is essential. For this reason, electron transfer in the reaction center from P* to QB involves a series of intermediate acceptors, BA, HA, and QA. In general, the initial steps have very fast electron transfer rates between closely aligned cofactors. The approximate two-fold symmetry for the two branches suggests comparable couplings
345 as has been found by consideration of the relative rates in various mutants (Katilius et al., 2002b). Comparison of the P* to P+HA– and P* to P+HB– rates and free energy differences indicates that the slower rate for the B-branch transfer is not accounted for entirely by the energetics, and so the coupling does have a contribution, although modest (Kirmaier et al., 2001, 2005). The charge recombination reactions illustrate how the maximal rate is primarily determined by the distance. In the wild type, the two quinones have observed recombination rates that differ by a factor of 10, although the two quinones are approximately equidistant from P. However, charge recombination from P+QB– has a significantly larger reorganization energy than that from P+QA–, as evident by the peak of the parabola being at a higher free energy difference, resulting in very comparable maximum rates as predicted by their similar distances to P (Allen et al., 1998). These theoretical models to describe electron transfer within proteins can be expanded to also understand electron transfer between cofactors found in two different proteins (Lin and Beratan, 2005; Lin et al., 2005; Miyashuta et al., 2005). In interprotein transfer, additional factors must be considered, such as the docking of the proteins and the involvement of water molecules at the protein interface. For example, cytochrome c2 is a water-soluble protein that transfers an electron only after binding to the reaction center (Chapter 17, Axelrod et al.). In this case, coupling is established by the ability of interfacial water molecules to form multiple hydrogen bonding pathways that connect the tunneling pathways for each of the proteins (Miyashita et al., 2005). C. Dynamics The conventional Marcus theory (Eq. 1) rests on the assumption that the electronic transition from the initial to the final state is much faster than nuclear motion. Theories have been developed that explicitly account for dynamics (Sumi and Marcus, 1986; Warshel et al., 1989; Gehlen et al., 1994). However, dynamical contributions are rarely used because of difficulties in experimentally measuring the contributions. Although the initial electron transfer has been well characterized, including the free energy dependence for the rate (Haffa et al., 2002), certain aspects of the transfer have been puzzling. The kinetics of the changes in the spectra of the pigments are complex,
346
JoAnn C. Williams and James P. Allen
and the three-fold increase in the electron transfer rate as the temperature decreases from 298 to 4 K is unusual. To determine the contribution of dynamics to the initial rate, this reaction has been analyzed by probing nearby amino acid residues, measured by absorbance changes in tryptophan residues (Wang et al., 2007). These measurements show that the protein dynamics is the limiting factor in the initial electron transfer rate rather than the energetics. In the classical picture of electron transfer, the protein undergoes structural changes only after the electron transfer to the acceptor is complete (Fig. 6). This observed dependence on the dynamics shows that the protein motions occur during the process and control the observed rate by overcoming the activation barriers described in the classical Marcus theory. IV. Pathways of Electron Transfer A. B-side Electron Transfer In wild-type reaction centers, electron transfer proceeds along the A branch with a nearly 100% efficiency. Manipulation of the various factors influencing electron transfer comes in most dramatically in experiments aimed at getting the electron to go the ‘wrong’ way. Mutations that increase B-branch electron transfer have been reviewed by Wakeham and Jones (2005). Achieving a significant amount of transfer along the B branch requires a combination of mutations, such as blocking the A side by mutations that prevent incorporation of QA or HA, altering the energies of the A-side states to make electron transfer less favorable by mutations near BA and HA, and altering the energies of the B side states to make electron transfer more favorable by mutations near BB (Fig. 7). Although reaction centers from Rba. sphaeroides and Rba. capsulatus have very similar electron transfer rates, the effect of many subtle differences involved in the branching is illustrated by a lower propensity of reaction centers from Rba. sphaeroides for electron transfer along the B branch (Kirmaier et al., 2002b). In early experiments on altering the branching ratios, only electron transfer to HB was observed, but now transfer to QB has been achieved with reasonable efficiency. The difference in the energy levels of the states P+BA– and P+BB– with respect to P* is an important factor in determining the branching ratio. Mutations that only increase the lifetime of P* are insufficient
Fig. 6. Light results in formation of the charge-separated state P+HA– in reaction centers. In the traditional static view of electron transfer, the movement of dipoles from protein side chains and water molecules is considered to be very slow compared to the initial electron transfer rate. In contrast, in the dynamic model of the reaction center, dipoles rearrange in response to the lightinduced formation of P*CT , the excited state of P, which has a charge-transfer character, followed by transfer of the electron. Recent measurements of the dynamics of the reaction center show that these rearrangements play a critical role in determining the rate of the electron transfer process (Wang et al., 2007).
for switching to B-side electron transfer (de Boer et al., 2002). Changing the energetics along the two branches can be accomplished by the introduction of an aspartic acid residue near BA (for Gly M201 in Rba. capsulatus, M203 in Rba. sphaeroides), showing that individual residues can influence the balance between A-side electron transfer, B-side electron transfer, and charge recombination (Heller et al., 1995). Similarly, a lysine residue introduced near BB results in an alteration of the branching ratio (at Ser L178 in Rba. capsulatus) (Kirmaier et al., 1999), as does a combination of mutations that destabilizes P+BA and stabilizes P+BB (at L181 in Rba. sphaeroides and Rba. capsulatus, and at M210 in Rba. sphaeroides, M208 in Rba. capsulatus) (Kirmaier et al., 2004). The φB mutant, in which BB is replaced with a bacteriopheophytin (because of the His M182 to Leu mutation in Rba. sphaeroides), is observed to form the state P+BB– with a yield of 35%, although no electron transfer to HB is observed, probably because
Chapter 18
Reaction Center Alteration
347
Fig. 7. In wild-type reaction centers, electron transfer occurs almost exclusively along the A branch of cofactors. In reaction centers with a set of four mutations, a significant percentage of electron transfer is observed to occur along the B side. For the YFHV mutant in Rba. capsulatus (Kirmaier et al., 2003), the Phe L181 to Tyr mutation lowers the energy of P+BB– while three changes make electron transfer along the A branch unfavorable: the Tyr M208 to Phe mutation raises the energy of BA–, the Leu M212 to His mutation results in HA being replaced with a bacteriochlorophyll (β), and the Trp M250 to Val mutation prevents the incorporation of QA.
the state P+BB– is lower in energy than P+HB– (Katilius et al., 1999, 2002a,b, 2003). Thus raising the free energy of P+BA– and lowering the free energy of P+BB– has been key in promoting electron transfer down the B side to give P+HB–. Yields of P+HB– of greater than approximately 30% are accomplished by combining mutations near the monomer bacteriochlorophylls with mutants that impede electron transfer along the A branch. A number of these combination mutants incorporate the β mutation, in which HA is replaced by a bacteriochlorophyll, resulting in a substantially higher energy for the P+β – state compared to P+HA– (Kirmaier et al., 2002a, 2003; Kee et al., 2006). An exceptionally high yield of 70% P+HB– was achieved by joining together mutations that alter the energies of BA and BB with changes derived from the DLL mutant, for which electron transfer along the A branch is blocked because of the lack of HA in that mutant (Chuang et al., 2006). On the other hand, relatively small changes in the energies of HA and HB, such as by addition of a hydrogen bond to HB to lower the energy level of
P+HB–, are insufficient to increase the yield of B-side transfer (Kirmaier et al., 2002a). A number of mutants with changes near the bacteriochlorophylls and bacteriopheophytins produce P+HB– but do not have significant electron transfer to QB, attributed to the forward reaction being unable to compete with charge recombination (de Boer et al., 2002). The rate of electron transfer from P+HB– to P+QB– is much slower than the rate from P+HA– to P+QA–, and so competes less well with charge recombination (Kirmaier et al., 2003; Kee et al., 2006). Several factors could influence the difference in the two rates (Kirmaier et al., 2003; Paddock et al., 2005). For example, the much larger reorganization energy associated with formation of QB– compared to QA–, as discussed above, would favor a much faster rate for QA– formation. Also contributing could be the free energy differences for the bacteriopheophytin to quinone electron transfer on the two sides being slightly different. Although the bacteriopheophytin to quinone separation distances for the two sides are comparable, differences in the coupling are suggested
348 by distinct interactions of the quinones with nearby aromatic amino acid residues (Trp M252 in Rba. sphaeroides and M250 in Rba. capsulatus near QA and Phe L216 near QB). Electron transfer can get all the way to QB via Bbranch electron transfer if the A branch is blocked (Laible et al., 2003). Removal of QA also allows one to observe spectra of QB generated only by B-side transfer (de Boer et al., 2002; Breton et al., 2004; Paddock et al., 2005, 2006). The yield of QB– when the A branch is blocked depends on what other mutations are present (Wakeham et al., 2003, 2004; Frolov et al., 2005). Mutations near QB that decrease the polarity (Glu L212 to Ala and Asp L213 to Ala in Rba. sphaeroides) increase the yield of QB– by B-branch electron transfer, postulated to be either by changing the rate through altering the oxidation-reduction potential and therefore the driving force or altering the reorganization energy (Wakeham et al., 2003). Although coupling has been modeled extensively, its effects on establishing the branching of electron transfer are somewhat difficult to measure experimentally, as it is not clear how to rationally change the spatial orbital overlaps. However, the heterodimer mutations afford a way to change the electron density distribution, which is also involved in the coupling. Mutants constructed by Kirmaier and coworkers (2005) to assess the coupling involve combining either the A or B side heterodimers and the β mutation that replaces HA with a bacteriochlorophyll. This combination of mutations has the effect of making the states P+BA– and P+HB– approximately equal in energy (thus equalizing the free energy contribution) while raising the monomer states such that the mechanism (superexchange) of the initial electron transfer should be the same on both branches. Although the yield of B-side transfer is low in both cases, it is approximately four-fold less for the mutant with the A-side heterodimer. Because the energies for the heterodimer mutants with the bacteriochlorophyll on either side are assumed to be comparable, the difference in yields is attributed to a small but real difference in coupling of the dimer to HA and HB. This coupling difference should contribute to the A-side preference for electron transfer found in other mutants and wild type. The pathway of electron transfer also appears to be dependent on the wavelength of the excitation energy. Blue light excitation into the Soret band gives rise to a transient BB+HB– state, which decays in picoseconds at room temperature but is longer lived at low temperature (Lin et al., 2001; Haffa et al.,
JoAnn C. Williams and James P. Allen 2003). Although energy transfer is fast, alternate photochemistry upon differing excitation wavelengths shows that energy transfer and electronic relaxation do not always precede electron transfer (Wang et al., 2006). These surrogate conduits for the dissipation of excitation energy suggest that the function of the B branch remains to rapidly quench higher excited states (Lin et al., 2001). Thus the B branch is poised to form charge-separated states but not to result in a terminal electron transfer as the A side is. This is consistent with the location of the carotenoid, also thought to be involved in photoprotection (Lin et al., 2003), on the B branch. The pathway of electron transfer, in addition to being dependent on wavelength, can also be dependent upon electrostatic interactions involving a cofactor and a nearby charged amino acid residue. Introduction of protonatable residues (His L168 to Glu and Asn L170 to Asp in Rba. sphaeroides) results in a pH dependence for the states generated by light excitation (Haffa et al., 2004). Because the charges on the ionizable residues depend on their protonation state, and the protonation state is pH dependent, the pathway then becomes pH dependent, demonstrating an ability to switch the pathway by adjustment of the pH. In summary, the highly efficient transfer of electrons along the A branch in wild type can be manipulated by mutagenesis to generate charge-separated states involving the B-branch cofactors. Overall, electron transfer to HB is largely limited by the unfavorable energy of BB and its limited coupling to P*. Electron transfer to QB can be observed but only when the normal electron transfer pathway along the A branch is blocked. The yield of electron transfer to QB is influenced by several factors, but most likely is limited by the rate being slow due to a high reorganization energy associated with formation of QB–. B. New Electron Transfer Reactions Whereas most of the emphasis concerning electron transfer in reaction centers has been on elucidating the factors that control the rates and the asymmetry of electron transfer, other manipulations have introduced new pathways into the reaction center (see also Chapter 16, Jones). The focus in these experiments has been on exploiting the environment near P so as to introduce novel secondary donors. In Rba. sphaeroides, P is normally reduced by a water-soluble cytochrome c2 (see Chapter 17, Axelrod et al.). In the absence of the cytochrome, P remains oxidized for
Chapter 18
Reaction Center Alteration
approximately 1 s until charge recombination from QB– occurs. By comparison, after light excitation and charge separation, the oxidized primary electron donor of Photosystem II is rapidly reduced by a redox active tyrosine (Tyr 161 of the D1 subunit), which is usually identified as YZ. The electron transfer continues with YZ+ being reduced by the Mn cluster where the four equivalents are stored until oxygen is evolved. Because of the relatively low midpoint potential of 500 mV for P, the bacterial reaction centers are normally not capable of oxidizing tyrosine or manganese. However, after hydrogen bonds are added to the dimer, P becomes highly oxidizing and is capable of oxidizing tyrosine residues, including a tyrosine introduced at a location analogous to that of YZ (Kálmán et al., 1999, 2003a,b; Narváez et al., 2002, 2004). Likewise, the modified reaction centers are capable of oxidizing manganese, which can be bound after amino acid residues acting as ligands are introduced (Fig. 8) (Thielges et al., 2005; Kálmán et al., 2005). These results demonstrate that new electron transfer pathways can be introduced, and such approaches can be used to delineate the complex evolutionary process by which anoxygenic photosynthetic complexes evolved into oxygenic complexes.
349
Fig. 8. Comparison of the three-dimensional structures of the Mn binding mutant (Thielges et al., 2005) and Photosystem II shows many similar features. In both cases, three critical amino acid residues are present, a carboxylate (M173 and D1-170), a tyrosine (M164 and YZ), and a histidine (M193 and D1-190). In addition, both complexes have a bound redox active Mn (a mononuclear and a Mn4 cluster for the mutant and Photosystem II respectively; Mn atoms shown as small spheres) that can serve as a rapid secondary electron donor to the oxidized primary donor.
V. Conclusions When a reaction center is excited by light, the interplay between the cofactors and the surrounding protein determines the route of electron transfer. The remarkable robustness of the reaction center to both biochemical manipulation and mutagenesis provides the opportunity to probe the features that influence electron transfer. The standard route down the A branch is observed to have the characteristics of energetics, coupling and dynamics that make it a durable electron transfer pathway. However multiple perturbation of the system makes possible other opportunistic reactions. Acknowledgments Our work is supported by the National Science Foundation, grant MCB0640002. We thank Aileen Taguchi for Fig. 6.
References Alden RG, Parson WW, Chu ZT and Warshel A (1996) Orientation of the OH dipole of tyrosine (M)210 and its effect on electrostatic energies in photosynthetic bacterial reaction centers. J Phys Chem 100: 16761–16770 Allen JP and Williams JC (1995) Relationship between the oxidation potential of the bacteriochlorophyll dimer and electron transfer in photosynthetic reaction centers. J Bioenerg Biomembr 27: 275–283 Allen JP and Williams JC (2006) The influence of protein interactions on the properties of the bacteriochlorophyll dimer in reaction centers. In: Grimm B, Porra RJ, Rüdiger W and Scheer H (eds) Chlorophylls and Bacteriochlorophylls: Biochemistry, Biophysics, Functions and Applications (Advances in Photosynthesis and Respiration, Vol 25), pp 283–295. Springer, Dordrecht Allen JP, Artz K, Lin X, Williams JC, Ivancich A, Albouy D, Mattioli TA, Fetsch A, Kuhn M and Lubitz W (1996) Effects of hydrogen bonding to a bacteriochlorophyll-bacteriopheophytin dimer in reaction centers from Rhodobacter sphaeroides. Biochemistry 35: 6612–6619 Allen JP, Williams JC, Graige MS, Paddock ML, Labahn A, Feher G and Okamura MY (1998) Free energy dependence of the direct charge recombination from the primary and secondary
350 quinones in reaction centers from Rhodobacter sphaeroides. Photosynth Res 55: 227–233 Arlt T, Dohse B, Schmidt S, Wachtveitl J, Laussermair E, Zinth W and Oesterhelt D (1996) Electron transfer dynamics of Rhodopseudomonas viridis reaction centers with a modified binding site for the accessory bacteriochlorophyll. Biochemistry 35: 9235–9244 Artz K, Williams JC, Allen JP, Lendzian F, Rautter J and Lubitz W (1997) Relationship between the oxidation potential and electron spin density of the primary electron donor in reaction centers from Rhodobacter sphaeroides. Proc Natl Acad Sci USA 94: 13582–13587 Breton J, Wakeham MC, Fyfe PK, Jones MR and Nabedryk E (2004) Characterization of the bonding interactions of QB upon photoreduction via A-branch or B-branch electron transfer in mutant reaction centers from Rhodobacter sphaeroides. Biochim Biophys Acta 1656: 127–138 Bylina EJ and Youvan DC (1988) Directed mutations affecting spectroscopic and electron transfer properties of the primary donor in the photosynthetic reaction center. Proc Natl Acad Sci USA 85: 7226–7230 Bylina EJ, Kirmaier C, McDowell L, Holten D and Youvan DC (1988) Influence of an amino-acid residue on the optical properties and electron transfer dynamics of a photosynthetic reaction centre complex. Nature 336: 182–184 Bylina EJ, Kolaczkowski SV, Norris JR and Youvan DC (1990) EPR characterization of genetically modified reaction centers of Rhodobacter capsulatus. Biochemistry 29: 6203–6210 Chen L, Holten D, Bocian DF and Kirmaier C (2004) Effects of hydrogen bonding and structure of the accessory bacteriochlorophylls on charge separation in Rb. capsulatus reaction centers. J Phys Chem B 108: 10457–10464 Chuang JI, Boxer SG, Holten D and Kirmaier C (2006) High yield of M-side electron transfer in mutants of Rhodobacter capsulatus reaction centers lacking the L-side bacteriopheophytin. Biochemistry 45: 3845–3851 Dahlbom MG and Reimers JR (2005) Successes and failures of time-dependent density functional theory for the low-lying excited states of chlorophylls. Mol Phys 103: 1057–1065 de Boer AL, Neerken S, de Wijn R, Permentier HP, Gast P, Vijgenboom E and Hoff AJ (2002) B-branch electron transfer in reaction centers of Rhodobacter sphaeroides assessed with site-directed mutagenesis. Photosynth Res 71: 221–239 Finkele U, Lauterwasser C, Zinth W, Gray KA and Oesterhelt D (1990) Role of tyrosine M210 in the initial charge separation of reaction centers of Rhodobacter sphaeroides. Biochemistry 29: 8517–8521 Frolov D, Wakeham MC, Andrizhiyevskaya EG, Jones MR and van Grondelle R (2005) Investigation of B-branch electron transfer by femtosecond time resolved spectroscopy in a Rhodobacter sphaeroides reaction centre that lacks the QA ubiquinone. Biochim Biophys Acta 1707: 189–198 Gehlen JN, Marchi M and Chandler D (1994) Dynamics affecting the primary charge transfer in photosynthesis. Science 263: 499–502 Goldsmith JO, King B and Boxer SG (1996) Mg coordination by amino acid side chains is not required for assembly and function of the special pair in bacterial photosynthetic reaction centers. Biochemistry 35: 2421–2428 Gray HB and Winkler JR (2005) Long-range electron transfer. Proc Natl Acad Sci USA 102: 3534–3539
JoAnn C. Williams and James P. Allen Gunner MR, Nicholls A and Honig B (1996) Electrostatic potentials in Rhodopseudomonas viridis reaction centers: Implications for the driving force and directionality of electron transfer. J Phys Chem 100: 4277–4291 Haffa ALM, Lin S, Katilius E, Williams JC, Taguchi AKW, Allen JP and Woodbury NW (2002) The dependence of the initial electron-transfer rate on driving force in Rhodobacter sphaeroides reaction centers. J Phys Chem 106: 7376–7384 Haffa ALM, Lin S, Williams JC, Taguchi AKW, Allen JP and Woodbury NW (2003) High yield of long-lived B-side charge separation at room temperature in mutant bacterial reaction centers. J Phys Chem B 107: 12503–12510 Haffa ALM, Lin S, Williams JC, Bowen BP, Taguchi AKW, Allen JP and Woodbury NW (2004) Controlling the pathway of photosynthetic charge separation in bacterial reaction centers. J Phys Chem B 108: 4–7 Heller BA, Holten D and Kirmaier C (1995) Control of electron transfer between the L- and M-sides of photosynthetic reaction centers. Science 269: 940–945 Ishikita H, Loll B, Biesiadka J, Galstyan A, Saenger W and Knapp EW (2005) Tuning electron transfer by ester-group of chlorophylls in bacterial photosynthetic reaction center. FEBS Lett 579: 712–716 Ishikita H, Saenger W, Loll B, Biesiadka J and Knapp EW (2006) Energetics of a possible proton exit pathway for water oxidation in Photosystem II. Biochemistry 45: 2063–2071 Jackson JA, Lin S, Taguchi AKW, Williams JC, Allen JP and Woodbury NW (1997) Energy transfer in Rhodobacter sphaeroides reaction centers with the initial electron donor oxidized or missing. J Phys Chem B 101: 5747–5754 Jia Y, DiMagno TJ, Chan CK, Wang Z, Du M, Hanson DK, Schiffer M, Norris JR, Fleming GR and Popov MS (1993) Primary charge separation in mutant reaction centers of Rhodobacter capsulatus. J Phys Chem 97: 13180–13191 Johnson ET and Parson WW (2002) Electrostatic interactions in an integral membrane protein. Biochemistry 41: 6483–6494 Johnson ET, Müh F, Nabedryk E, Williams JC, Allen JP, Lubitz W, Breton J and Parson WW (2002) Electronic and vibronic coupling of the special pair of bacteriochlorophylls in photosynthetic reaction centers from wild-type and mutant strains of Rhodobacter sphaeroides. J Phys Chem B 106: 11859–11869 Kálmán L, LoBrutto R, Allen JP and Williams JC (1999) Modified reaction centres oxidize tyrosine in reactions that mirror Photosystem II. Nature 402: 696–699 Kálmán L, Williams JC and Allen JP (2003a) Proton release upon oxidation of tyrosine in reaction centers from Rhodobacter sphaeroides. FEBS Lett 545: 193–198 Kálmán L, LoBrutto R, Narváez AJ, Williams JC and Allen JP (2003b) Correlation of proton release and electrochromic shifts of the optical spectrum due to oxidation of tyrosine in reaction centers from Rhodobacter sphaeroides. Biochemistry 42: 13280–13286 Kálmán L, Thielges MC, Williams JC and Allen JP (2005) Proton release due to manganese binding and oxidation in modified bacterial reaction centers. Biochemistry 44: 13266–13273 Kanchanawong P, Dahlbom MG, Treynor TP, Reimers JR, Hush NS and Boxer SG (2006) Charge delocalization in the specialpair radical cation of mutant reaction centers of Rhodobacter sphaeroides from Stark spectra and nonadiabatic spectral simulations. J Phys Chem B 110: 18688–18702 Katilius E, Turanchik T, Lin S, Taguchi AKW and Woodbury NW
Chapter 18
Reaction Center Alteration
(1999) B-side electron transfer in a Rhodobacter sphaeroides reaction center mutant in which the B-side monomer bacteriochlorophyll is replaced with bacteriopheophytin. J Phys Chem B 103: 7386–7389 Katilius E, Katiliene Z, Lin S, Taguchi AKW and Woodbury NW (2002a) B side electron transfer in a Rhodobacter sphaeroides reaction center mutant in which the B side monomer bacteriochlorophyll is replaced with bacteriopheophytin: Low-temperature study and energetics of charge-separated states. J Phys Chem B 106: 1471–1475 Katilius E, Katiliene Z, Lin S, Taguchi AKW and Woodbury NW (2002b) B-side electron transfer in the HE(M182) reaction center mutant from Rhodobacter sphaeroides. J Phys Chem B 106: 12344–12350 Katilius E, Babendure JL, Katiliene Z, Lin S, Taguchi AKW and Woodbury NW (2003) Manipulations of the B-side chargeseparated states’ energetics in the Rhodobacter sphaeroides reaction center. J Phys Chem B 107: 12029–12034 Katilius E, Babendure JL, Lin S and Woodbury NW (2004) Electron transfer dynamics in Rhodobacter sphaeroides reaction center mutants with a modified ligand for the monomer bacteriochlorophyll on the active side. Photosynth Res 81: 165–180 Kee HL, Laible PD, Bautista JA, Hanson DK, Holten D and Kirmaier C (2006) Determination of the rate and yield of B-side quinone reduction in Rhodobacter capsulatus reaction centers. Biochemistry 45: 7314–7322 Khatypov RA, Vasilieva LG, Fufina TY, Bolgarina TI and Shuvalov VA (2005) Substitution of isoleucine L177 by histidine affects the pigment composition and properties of the reaction center of the purple bacterium Rhodobacter sphaeroides. Biochemistry (Moscow) 70: 1527–1533 King BA, de Winter A, McAnaney TB and Boxer SG (2001) Excited state energy transfer pathways in photosynthetic reaction centers. 4. Asymmetric energy transfer in the heterodimer mutant. J Phys Chem B 105: 1856–1862 Kirmaier C, Holten, D, Bylina EJ and Youvan DC (1988) Electron transfer in a genetically modified bacterial reaction center containing a heterodimer. Proc Natl Acad Sci USA 85: 7562–7566 Kirmaier C, Gaul D, DeBey R, Holten D and Schenck CC (1991) Charge separation in a reaction center incorporating bacteriochlorophyll for photoactive bacteriopheophytin. Science 251: 922–927 Kirmaier C, Weems D and Holten D (1999) M-side electron transfer in reaction center mutants with a lysine near the nonphotoactive bacteriochlorophyll. Biochemistry 38: 11516–11530 Kirmaier C, He C and Holten D (2001) Manipulating the direction of electron transfer in the bacterial reaction center by swapping Phe for Tyr near BChlM (L181) and Tyr for Phe near BChlL (M208). Biochemistry 40: 12132–12139 Kirmaier C, Cua A, He C, Holten D and Bocian DF (2002a) Probing the M-branch electron transfer and cofactor environment in the bacterial photosynthetic reaction center by addition of a hydrogen bond to the M-side bacteriopheophytin. J Phys Chem B 106: 495–503 Kirmaier C, Laible PD, Czarnecki K, Hata AN, Hanson DK, Bocian DF and Holten D (2002b) Comparison of M-side electron transfer in Rb. sphaeroides and Rb. capsulatus reaction centers. J Phys Chem B 106: 1799–1808 Kirmaier C, Laible PD, Hanson DK and Holten D (2003) B-side
351 charge separation in bacterial photosynthetic reaction centers: nanosecond time scale electron transfer from HB– to QB. Biochemistry 42: 2016–2024 Kirmaier C, Laible PD, Hanson DK and Holten D (2004) B-side electron transfer to form P+HB– in reaction centers from the F(L181)Y/Y(M208)F mutant of Rhodobacter capsulatus. J Phys Chem B 108: 11827–11832 Kirmaier C, Bautista JA, Laible PD, Hanson DK and Holten D (2005) Probing the contribution of electronic coupling to the directionality of electron transfer in photosynthetic reaction centers. J Phys Chem B 109: 24160–24172 Laible PD, Kirmaier C, Udawatte CSM, Hofman SJ, Holten D and Hanson DK (2003) Quinone reduction via secondary Bbranch electron transfer in mutant bacterial reaction centers. Biochemistry 42: 1718–1730 Li Y, Lucas MG, Konovalova T, Abbott B, MacMillan F, Petrenko A, Sivakumar V, Wang R, Hastings G, Gu F, van Tol J, Brunel LC, Timkovich R, Rappaport F and Redding K (2004) Mutation of the putative hydrogen-bond donor to P700 of Photosystem I. Biochemistry 43: 12634–12647 Li Y, van der Est A, Lucas MG, Ramesh VM, Gu F, Petrenko A, Lin S, Webber AN, Rappaport F and Redding K (2006) Directing electron transfer within Photosystem I by breaking H-bonds in the cofactor branches. Proc Natl Acad Sci USA 103: 2144–2149 Lin J and Beratan DN (2005) Simulation of electron transfer between cytochrome c2 and the bacterial photosynthetic reaction center: Brownian dynamics analysis of the native proteins and double mutants. J Phys Chem B 109: 7529–7534 Lin J, Balabin IA and Beratan DN (2005) The nature of aqueous tunneling pathways between electron-transfer proteins. Science 310: 1311–1313 Lin S, Katilius E, Haffa ALM, Taguchi AKW and Woodbury NW (2001) Blue light drives B-side electron transfer in bacterial photosynthetic reaction centers. Biochemistry 40: 13767–13773 Lin S, Katilius E, Taguchi AKW and Woodbury NW (2003) Excitation energy transfer from carotenoid to bacteriochlorophyll in the photosynthetic purple bacterial reaction center of Rhodobacter sphaeroides. J Phys Chem B 107: 14103–14108 Lin X, Murchison HA, Nagarajan V, Parson WW, Allen JP and Williams JC (1994a) Specific alteration of the oxidation potential of the electron donor in reaction centers from Rhodobacter sphaeroides. Proc Natl Acad Sci USA 91: 10265–10269 Lin X, Williams JC, Allen JP and Mathis P (1994b) Relationship between rate and free energy difference for electron transfer from cytochrome c2 to the reaction center in Rhodobacter sphaeroides. Biochemistry 33: 13517–13523 Marcus RA and Sutin N (1985) Electron transfers in chemistry and biology. Biochim Biophys Acta 811: 265–322 McDowell LM, Gaul D, Kirmaier C, Holten D and Schenck CC (1991) Investigation into the source of electron transfer asymmetry in bacterial reaction centers. Biochemistry 30: 8315–8322 Middendorf TR, Mazzola LT, Lao K, Steffen MA and Boxer SG (1993) Stark effect (electroabsorption) spectroscopy of photosynthetic reaction centers at 1.5 K: Evidence that the special pair has a large excited-state polarizability. Biochim Biophys Acta 1143: 223–234 Miyashita O, Okamura MY and Onuchic JN (2005) Interprotein electron transfer from cytochrome c2 to photosynthetic reaction
352 center: Tunneling across an aqueous interface. Proc Natl Acad Sci USA 102: 3558–3563 Moore LJ and Boxer SG (1998) Inter-chromophore interactions in pigment-modified and dimer-less bacterial photosynthetic reaction centers. Photosynth Res 55: 173–180 Müh F, Williams, JC, Allen JP and Lubitz W (1998) A conformational change of the photoactive bacteriopheophytin in reaction centers from Rhodobacter sphaeroides. Biochemistry 37: 13066–13074 Müh F, Lendzian F, Roy M, Williams JC, Allen JP and Lubitz W (2002) Pigment-protein interactions in bacterial reaction centers and their influence on oxidation potential and spin density distribution of the primary donor. J Phys Chem B 106: 3226–3236 Murchison HA, Alden RG, Allen JP, Peloquin JM, Taguchi AKW, Woodbury NW and Williams JC (1993) Mutations designed to modify the environment of the primary electron donor of the reaction center from Rhodobacter sphaeroides: Phenylalanine to leucine at L167 and histidine to phenylalanine at L168. Biochemistry 32: 3498–3505 Nagarajan V, Parson WW, Gaul D and Schenck C (1990) Effect of specific mutations of tyrosine-(M)210 on the primary photosynthetic electron-transfer process in Rhodobacter sphaeroides. Proc Natl Acad Sci USA 87: 7888–7892 Narváez AJ, Kálmán L, LoBrutto R, Allen JP and Williams JC (2002) Influence of the protein environment on the properties of a tyrosyl radical in reaction centers from Rhodobacter sphaeroides. Biochemistry 41: 15253–15258 Narváez AJ, LoBrutto R, Allen JP and Williams JC (2004) Trapped tyrosyl radical populations in modified reaction centers from Rhodobacter sphaeroides. Biochemistry 43: 14379–14384 Noy D, Moser CC and Dutton PL (2006) Design and engineering of photosynthetic light-harvesting and electron transfer using length, time, and energy scales. Biochim Biophys Acta 1757: 90–105 Paddock ML, Chang C, Xu Q, Abresch EC, Axelrod HL, Feher G and Okamura MY (2005) Quinone (QB) reduction by B-branch electron transfer in mutant bacterial reaction centers from Rhodobacter sphaeroides: quantum efficiency and X-ray structure. Biochemistry 44: 6920–6928 Paddock ML, Flores M, Isaacson R, Chang C, Abresch EC, Selvaduray P and Okamura MY (2006) Trapped conformational states of semiquinone (D+•QB–•) formed by B-branch electron transfer at low temperature in Rhodobacter sphaeroides reaction centers. Biochemistry 45: 14032–14042 Parson WW, Chu ZT and Warshel A (1990) Electrostatic control of charge separation in bacterial photosynthesis. Biochim Biophys Acta 1017: 251–272 Plato M, Lendzian F, Lubitz W and Möbius K (1992) Molecular orbital study of electronic asymmetry in primary donors of bacterial reaction centers. In: Breton J and Verméglio A (eds) The Photosynthetic Bacterial Reaction Center II: Structure, Spectroscopy, and Dynamics, pp 109–118. Plenum, New York Potter JA, Fyfe PK, Frolov D, Wakeham MC, van Grondelle R, Robert B and Jones MR (2005) Strong effects of an individual water molecule on the rate of light-driven charge separation in the Rhodobacter sphaeroides reaction center. J Biol Chem 280: 27155–27164 Reimers JR and Hush NS (2004) A unified description of the electrochemical, charge distribution, and spectroscopic properties of the special-pair radical cation in bacterial photosynthesis.
JoAnn C. Williams and James P. Allen J Am Chem Soc 126: 4132–4144 Robles SJ, Breton J and Youvan DC (1990) Partial symmetrization of the photosynthetic reaction center. Science 248: 1402–1405 Spiedel D, Jones MR and Robert B (2002) Tuning of the redox potential of the primary electron donor in reaction centres of purple bacteria: Effects of amino acid polarity and position. FEBS Lett 527: 171–175 Stocker JW, Taguchi AKW, Murchison HA, Woodbury NW and Boxer SG (1992) Spectroscopic and redox properties of sym1 and (M)F195H: Rhodobacter capsulatus reaction center symmetry mutants which affect the initial electron donor. Biochemistry 31: 10356–10362 Sumi H and Marcus RA (1986) Dielectric relaxation and intramolecular electron transfers. J Chem Phys 84: 4272–4276 Thielges M, Uyeda G, Cámara-Artigas A, Kálmán L, Williams JC and Allen JP (2005) Design of a redox-linked active metal site: Manganese bound to bacterial reaction centers at a site resembling that of Photosystem II. Biochemistry 44: 7389–7394 Treynor TP, Yoshina-Ishii C and Boxer SG (2004) Probing excited-state electron transfer by resonance Stark spectroscopy: 4. Mutations near BL in photosynthetic reaction centers perturb multiple factors that affect BL* → BL+ HL–. J Phys Chem B 108: 13523–13535 van Brederode ME, van Stokkum IHM, Katilius E, van Mourik F, Jones MR and van Grondelle R (1999) Primary charge separation routes in the BChl:Bphe heterodimer reaction centers of Rhodobacter sphaeroides. Biochemistry 38: 7545–7555 Wakeham MC and Jones MR (2005) Rewiring photosynthesis: Engineering wrong-way electron transfer in the purple bacterial reaction centre. Biochem Soc Trans 33: 851–857 Wakeham MC, Goodwin MG, McKibbin C and Jones MR (2003) Photo-accumulation of the P+QB– radical pair state in purple bacterial reaction centres that lack the QA ubiquinone. FEBS Lett 540: 234–240 Wakeham MC, Breton J, Nabedryk E and Jones MR (2004) Formation of a semiquinone at the QB site by A- or B-branch electron transfer in the reaction center from Rhodobacter sphaeroides. Biochemistry 43: 4755–4763 Wang H, Lin S and Woodbury NW (2006) Electronic transitions of the Soret band of reaction centers from Rhodobacter sphaeroides studied by femtosecond transient absorbance spectroscopy. J Phys Chem B 110: 6956–6961 Wang H, Lin S, Allen JP, Williams JC, Blankert S, Laser C and Woodbury NW (2007) Protein dynamics control the kinetics of initial electron transfer in photosynthesis. Science 316: 747–750 Warshel A, Chu ZT and Parson WW (1989) Dispersed polaron simulations of electron transfer in photosynthetic reaction centers. Science 246: 112–116 Watson AJ, Fyfe PK, Frolov D, Wakeham MC, Nabedryk E, van Grondelle R, Breton J and Jones MR (2005) Replacement or exclusion of the B-branch bacteriopheophytin in the purple bacterial reaction centre: The HB cofactor is not required for assembly or core function of the Rhodobacter sphaeroides complex. Biochim Biophys Acta 1710: 34–46 Webber AN and Lubitz W (2001) P700: The primary electron donor of Photosystem I. Biochim Biophys Acta 1507: 61–79 Williams JC, Alden RG, Murchison HA, Peloquin JM, Woodbury NW and Allen JP (1992) Effects of mutations near the bacteriochlorophylls in reaction centers from Rhodobacter
Chapter 18
Reaction Center Alteration
sphaeroides. Biochemistry 31: 11029–11037 Williams JC, Haffa ALM, McCulley JL, Woodbury NW and Allen JP (2001) Electrostatic interactions between charged amino acid residues and the bacteriochlorophyll dimer in reaction centers from Rhodobacter sphaeroides. Biochemistry 40: 15403–15407 Williams JC, Paddock ML, Way YP and Allen JP (2007) Changes in metal specificity due to iron ligand substitutions in reaction centers from Rhodobacter sphaeroides. Appl Magn Reson 31: 45–58 Witt H, Schlodder E, Teutloff C, Niklas J, Bordignon E, Carbonera D, Kohler S, Labahn A and Lubitz W (2002) Hydrogen bonding to P700: site-directed mutagenesis of threonine A739 of
353 Photosystem I in Chlamydomonas reinhardtii. Biochemistry 41: 8557–8569 Yakovlev AG, Jones MR, Potter JA, Fyfe PK, Vasilieva LG, Shkuropatov AY and Shuvalov VA (2005) Primary charge separation between P* and BA: Electron-transfer pathways in native and mutant GM203L bacterial reaction centers. Chem Phys 319: 297–307 Yanagi K, Shimizu M, Hashimoto H, Gardiner AT, Roszak AW and Cogdell RJ (2005) Local electrostatic field induced by the carotenoid bound to the reaction center of the purple photosynthetic bacterium Rhodobacter sphaeroides. J Phys Chem B 109: 992–998
Chapter 19 Mechanism of Charge Separation in Purple Bacterial Reaction Centers William W. Parson* Department of Biochemistry, Box 357350; University of Washington, Seattle, WA 98195-7350, U.S.A.
Arieh Warshel Department of Chemistry, University of Southern California, Los Angeles, CA 90007, U.S.A.
Summary ............................................................................................................................................................... 355 I. The Reaction Sequence and Kinetics ............................................................................................................ 356 II. Energies of the Radical-Pair Intermediates .................................................................................................... 357 A. P+HA– ................................................................................................................................................. 357 B. P+BA– and P+BB– ................................................................................................................................. 357 III. Unusual Features of the Charge-Separation Reactions................................................................................. 360 A. Speed and Temperature Dependence ............................................................................................ 360 B. Specificity ........................................................................................................................................ 360 C. Multiphasic Kinetics ......................................................................................................................... 361 D. Coupling to Vibrational Wavepackets ............................................................................................. 362 E. Stability to Mutations ........................................................................................................................ 363 IV. Theories of Electron-Transfer Reactions........................................................................................................ 363 A. Marcus Theory and Semiclassical Surface-hopping ....................................................................... 363 B. Coupling to Quantized Vibrational Modes ....................................................................................... 367 C. Density-Matrix Treatments .............................................................................................................. 368 Acknowledgments ................................................................................................................................................ 370 References ............................................................................................................................................................ 370
Summary This chapter discusses the pathway, kinetics and energetics of the initial electron-transfer steps in reaction centers (RCs) from purple photosynthetic bacteria. We consider the unusual dependence of the kinetics on temperature, the strong specificity of the reactions for electron acceptors on the ‘A’ side of the RC, effects of vibrational wavepackets, the multiphasic nature of the kinetics, effects of mutations and pigment substitutions, and links between the kinetics of electron transfer and vibrational equilibration. We then discuss some of the theories that have been used to rationalize the dynamics and temperature dependence of the electron-transfer reactions, including Marcus theory, coupling to quantized vibrational modes and density-matrix treatments. Our discussion illustrates the power of microscopic simulations for connecting structural information with mechanistic interpretations. *Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 355–377. © 2009 Springer Science + Business Media B.V.
356 I. The Reaction Sequence and Kinetics All the known reaction centers (RCs) of purple photosynthetic bacteria (Proteobacteria) contain a bacteriochlorophyll (BChl) dimer that serves as the ‘primary’ electron donor. This ‘special pair’ of BChls (P) contributes the RC’s main absorption band in the region of 865 to 885 nm in species such as Rhodobacter (Rba.) sphaeroides, Rba. capsulatus, Chromatium vinosum and Thermochromatium tepidum, which contain BChl a, and near 960 nm in Blastochloris (Blc.) viridis and other species that contain BChl b. When P is raised to its first excited singlet state (P*), it transfers an electron to a molecule of bacteriopheophytin (BPhe), generating a transient radical-pair state, P+HA– or P+HL–. The reduced BPhe (HA– or HL–) then passes an electron to a quinone (QA). The oxidized special pair (P+) subsequently extracts an electron from a c-type cytochrome, while QA– sends an electron on to a second quinone (QB). At this point, the RC has moved an electron across a membrane that is topologically contiguous with the cell’s plasma membrane, and all the primary reactants are back in their original states, ready to transfer another electron. At room temperature, transfer of an electron from P* to HA takes 3 to 5 ps in RCs of Rba. sphaeroides and Rba. capsulatus and about 2 ps in Blc. viridis (Holten et al., 1980; Paschenko et al., 1985; Woodbury et al., 1985, 1994; Breton et al., 1986; Martin et al., 1986; Wasielewski and Tiede, 1986; Kirmaier and Holten, 1988; Holzapfel et al., 1989, 1990; Lauterwasser et al., 1991; Peloquin et al., 1994; Holzwarth and Muller, 1996; Huppman et al., 2002). Reduction of QA by HA– takes about 200 ps (Kaufmann et al., 1975; Rockley et al., 1975; Holten et al., 1978), whereas the time constant for oxidation of the cytochrome by P+ ranges from about 0.5 to 50 µs depending on the bacterial species (Chapter 17, Axelrod et al.). The kinetics of formation of P+HA– can be measured by following either the decay of stimulated emission from P* or the development of a broad absorption band of HA– in the region of 650 to 670 nm. Other absorbance changes (between 780 and 805 nm in the BChl a species and 800–830 nm in the BChl b Abbreviations: BA, BB – monomeric BChls that serve as electron acceptors in the RC; BChl – bacteriochlorophyll; BPhe – bacteriopheophytin; HA, HB – BPhes that serve as electron acceptors; MD – molecular dynamics; P – the special pair of BChls that acts as the primary electron donor; Pheo – pheophytin; RC – reaction center; UV – ultra-violet
William W. Parson and Arieh Warshel species) signal loss of the Qy absorption band of HA and disruption of interactions of the BPhe with the neighboring BChls. Electron transfer from P* to HA almost certainly occurs in two discrete steps, with the ‘accessory’ BChl BA (also called BL) serving as an intermediary electron carrier. Although the first crystal structures of RCs from Blc. viridis and Rba. sphaeroides showed BA situated between P and HA in what appeared to be a good position to play this role, compelling experimental evidence for the transient reduction of the BChl was slow to emerge. The difficulty evidently is that electron transfer from P* to BA is rate-limiting, so that the product of this first step (P+BA–) decays to P+HA– almost as rapidly as it is formed. In addition, transient bleaching of an absorption band of BA is not necessarily diagnostic for formation of BA– because similar absorbance changes could result from formation of an excited singlet state or a change in exciton interactions of the pigments. The stimulated emission from P* was found to contain a fast decay component that could reflect a multi-step kinetic process (Holzapfel et al., 1989, 1990; Chan et al., 1991a,b; Lauterwasser et al., 1991), but time-dependent shifts in the emission due to vibrational relaxations and the oscillatory motions of nuclear wavepackets made the interpretation of this component ambiguous. The negative results of early attempts to detect P+BA– in pump-probe experiments suggested that this state was not formed as a true intermediate, but rather served to couple P* with P+HA– by mixing with both states quantum mechanically, a process known as superexchange. However, measurements in the region of 1020 nm eventually revealed a transient absorption band that could be assigned persuasively to BA– (Arlt et al., 1993; van Stokkum et al., 1997). The rise and decay of this band in Rba. sphaeroides RCs at room temperature are consistent with a kinetic scheme in which an electron moves from P* to BA with a time constant of about 3.5 ps and then from BA– to HA with a time constant of 0.9 ps (Arlt et al., 1993). Time constants of 1 to 3 ps are obtained for both steps if the reactions are assumed to be reversible (Holzwarth and Muller, 1996; van Stokkum et al., 1997). Perhaps the most convincing experimental support for the two-step mechanism with P+BA– as an intermediate is the observation that the amplitude and lifetime of the absorption signals assigned to P+BA– increase if bacteriopheophytin HA is replaced by pheophytin a (Pheo) (Shkuropatov and Shuvalov, 1993; Schmidt et al., 1994; Schmidt et al., 1995; Kennis et al., 1997)
Chapter 19
Mechanism of Charge Separation
357
or bacteriochlorophyll (Kirmaier et al., 1995a,b). Because Pheo and BChl have more negative reduction potentials (Ems) than BPhe, either of these substitutions would be expected to slow the decay of P+BA–. Replacing a Phe residue near HA by Asp, which should raise the free energy of P+HA– if the Asp is ionized, appears to have a similar inhibitory effect on electron transfer from BA– to HA (Heller et al., 1996; Roberts et al., 2001). Mutations that should raise the free energy of P+BA– also slow charge separation, in this case probably by impairing electron transfer from P* to BA (Nagarajan et al., 1990, 1993; Chan, et al., 1991; Jia et al., 1993; Shochat et al., 1994; Beekman et al., 1996; Roberts et al., 2001; Haffa et al., 2002; Katilius et al., 2004; Chapter 16, Jones). Conversely, a long-lived P+BPhe– radical pair is seen in mutants that incorporate BPhe instead of BChl in the BA site, which should lower the free energy of transferring an electron here (Arlt et al., 1996b). Alternative pathways leading to P+HA– may open up if BA or HA is excited instead of P (van Brederode et al., 1997, 1999), or if RCs are excited with multiple photons (Lin et al., 1999) or with blue light (Lin et al., 2001). BA* evidently can extract an electron directly from P, although it also transfers its excitation energy to P with a time constant of about 0.1 ps. The reaction BA* → P+BA– is seen most clearly at low temperatures in Y(M)210F mutant RCs, where electron transfer from P* to BA is slowed. The higher energy of BA* relative to P* could enable BA* to drive the formation of P+BA– when the reaction from P* fails. Under some conditions, BA* also may transfer an electron directly to HA. However, the quantitative importance of these pathways is unclear, because some of the spectroscopic properties of RCs appear to be inconsistent with fast electron transfer to or from BA* (Zhou and Boxer, 1998a,b; King et al., 2001).
10 to 20 ns that matches the decay of the absorption signals associated with P+HA– (Shuvalov and Klimov, 1976; Godik and Borisov, 1979; Rademaker and Hoff, 1981; Godik et al., 1982; Schenck et al., 1982; Woodbury and Parson, 1984; Hörber et al., 1986; Woodbury et al., 1986; Goldstein and Boxer, 1988; Ogrodnik, 1990; Ogrodnik et al., 1994; Peloquin et al., 1994). Some of the faster decay components probably reflect relaxations of P+HA–. The relative amplitude of the slow component in Rba. sphaeroides RCs at 295 K indicates that the standard free energy of the relaxed P+HA– is between 0.21 and 0.26 eV (4.8 - 6.0 kcal/mol) below that of P*. The free energy of P+HA– also was obtained from the equilibrium constant between this state and an excited triplet state of P (3P). If electron transfer from HA– to QA is blocked, 3P forms from P+HA– with a quantum yield of about 30% in Rba. sphaeroides RCs at 295 K (Parson et al., 1975). The phosphorescence emission spectrum (Takiff and Boxer, 1988) and the activation energy for thermal repopulation of the excited singlet state (Shuvalov and Parson, 1980) put 3P 0.42 eV below P*. 3P decays largely via thermal excitation back to P+HA–, followed by charge recombination to the ground state. The formation and decay of 3P are sensitive to external magnetic fields because they require interconversion between singlet and triplet forms of the radical pair (1[P+HA–] and 3[P+HA–]), which have very similar energies in the absence of an external field but split apart in the presence of a field (Blankenship et al., 1977; Haberkorn and Michel-Beyerle, 1979; Hoff, 1981; Boxer et al., 1983). Analyses of the effects of magnetic fields and temperature on the decay kinetics indicate that P+HA– sits between 0.25 and 0.26 eV below P*, in good accord with the fluorescence measurements (Chidsey et al., 1985; Goldstein and Boxer, 1988; Ogrodnik et al., 1988; Takiff and Boxer, 1988; Boxer et al., 1989).
II. Energies of the Radical-Pair Intermediates
B. P+BA– and P+BB–
+
– A
A. P H
The free energy difference between P* and P+HA– was obtained by measuring the equilibrium between the two states when electron transfer from HA– to QA is prevented by reducing QA beforehand or removing the quinone from the RCs. Under these conditions, fluorescence from P* decays in a multiphasic manner, but most of the decay occurs with a time constant of
Estimates of the free energy of P+BA– rest on weaker ground because of the difficulty of accessing this state experimentally. One approach is to calculate the energy by working backward from the free energy of P+HA–, using the difference between the measured Em values of BChl and BPhe in solution and incorporating calculated solvation energies of BChl– and BPhe– in solution and of P+BA– and P+HA– in the RC (Creighton et al., 1988; Parson et al., 1990; Alden et al., 1995; Parson and Warshel, 2006; Warshel et
358 al., 2006). The calculated solvation energies include electrostatic interactions of P and the electron acceptor with each other, as well as with the surrounding protein, water and (in some treatments) electrolytes. Quantum calculations of the gas-phase energy of P+BA– also have been used instead of the Em values and solvation energies (Scherer and Fischer, 1989, 1998; Thompson et al., 1991; Marchi et al., 1993; Gunner et al., 1996; Blomberg et al., 1998; Hasegawa and Nakatsuji, 1998; Ivashin et al., 1998; Hughes et al., 2001), although this approach probably entails greater uncertainties because of the large size of the BChl cluster and the complex effects of local electric fields on the quantum energies. The effects of dielectric screening in the protein can be evaluated by assigning macroscopic dielectric constants to the protein and solvent and using the Poisson-Boltzmann equation (Gilson et al., 1985; Gunner et al., 1996), or by including microscopic induced dipoles explicitly and representing the solvent by Langevin dipoles (Warshel and Russell, 1984; Parson et al., 1990; Alden et al., 1995). Dielectric effects in periodic systems can be treated by Ewald sums, and this also has been done with RCs (Procacci et al., 1997; Ceccarelli and Marchi, 2003). This approach, however, appears to be less reliable than using a spherically symmetrical model with proper polarization at the boundaries of the region that is treated explicitly (Warshel and Russell, 1984; Hunenberger et al., 1995; Figueirido et al., 1997; Sham and Warshel, 1998; Parson and Warshel, 2006; Warshel et al., 2006). Classical molecular-dynamics (MD) simulations and free-energy perturbation techniques can be used to examine how rearrangement of the protein around the charged species affects the energy of P+BA– (Creighton et al., 1988; Alden et al., 1995; Warshel and Parson, 2001; Parson and Warshel, 2006; Warshel et al., 2006). Because the protein and the surrounding water include a vast number of atoms, it is necessary to define a generalized reaction coordinate for electron transfer rather than focusing on an individual nuclear position or bond length. The most useful choice probably is the difference between the electrostatic energies of the product and reactant states (X = ∆Velec). After recording X as a function of time during an MD trajectory on the potential-energy surface of P* or P+BA–, one can generate a histogram (p(X)) of the probability of finding various values of X. These probabilities can be converted to free energies with the relationship g(X) = go -RTln(p(X)),
William W. Parson and Arieh Warshel where go = 0 for the reactant state and go = ∆ Go (the standard free energy change) for the product state. Better sampling of the relevant conformational space can be obtained by combining results from trajectories on the potential surfaces of both P* and P+BA–, or on a series of intermediate ‘mapping’ potentials comprised of linear combinations of these two potentials (see Warshel and Parson (2001) and Parson and Warshel (2006) for reviews). Comparisons of the calculated free energy of P+HA– with the measured values described above provide a useful check on the methods, although fortuitous agreements here have sometimes been misleading. Figure 1A shows calculated free energy curves for generating P+BA– from P* in the Ermler et al. (1994) crystal structure of Rba. sphaeroides RCs (Warshel and Parson, 2001; Parson and Warshel, 2006). The free energies of P* and P+BA– are plotted as functions of the reaction coordinate, and ∆Go is indicated. The calculations give ∆Go ≈ –2 kcal/mol, putting P+BA– in a good position to act as an intermediate between P* and P+HA–. Calculations on Blc. viridis RCs gave similar results (Parson et al., 1990, 1998; Alden et al., 1995, 1996a). The reorganization energy (λ), which is the free energy change associated with moving along the reaction coordinate from the equilibrium position of P* to that of P+BA–, is about 2 kcal/mol. Because the calculations have an estimated uncertainty of about ±2 kcal/mol, they do not exclude the possibility that P+BA– is slightly above P*. However, calculations that found P+BA– to be far above P* (Marchi et al., 1993) were shown to be in error as a result of an incorrect treatment of dielectric screening (Alden et al., 1995, 1996a; Gunner et al., 1996; Ceccarelli and Marchi, 2003). Other calculations that gave high free energies for P+BA– (Thompson et al., 1991; Gunner et al., 1996; Hasegawa and Nakatsuji, 1998) neglected the selfenergies of the charged species or assigned what now appears to be an unrealistically low dielectric constant to the protein (see Parson and Warshel, 2006). Calculations of the energy for P+BB–, the radical-pair created by electron transfer from P* to the monomeric BChl on the ‘B’ or ‘M’ side of the RC, put this state considerably above P* (Fig. 1B). The consequences of this higher energy, and the electrostatic interactions that underlie the difference between the energies on the two sides, are discussed in Section III. The free energy of P+BA– also has been estimated from the effects of replacing BPhe HA by Pheo, which has a more negative reduction potential than BPhe
Chapter 19
Mechanism of Charge Separation
359
Fig. 1. Calculated free energies of P*, P+BA– and P+BB– in Rba. sphaeroides RCs at 300 K, expressed relative to the minimum free energy of P*. Panel A shows the free energy of P+BA– as a function of the reaction coordinate for P* → P+BA–; panel B, the free energy of P+BB–, as a function of the reaction coordinate for P* → P+BB–. The reaction coordinate is the electrostatic energy difference between the diabatic product and reactant states in each case. The free energies were obtained by free-energy perturbation calculations as described (Warshel and Parson, 2001; Parson and Warshel, 2006), using separate 1-ns MD trajectories in the reactant and product states. The energy of transferring an electron from P* to a BChl at infinite distance in a vacuum was adjusted arbitrarily within the estimated uncertainties of the calculations so that the energies of P+BA– and P* coincide at the origin. This adjustment does not affect the calculated difference between the minimum energies of P+BA– and P+BB– (∆∆Go ≈ 6 kcal/mol). The asymmetry of the individual curves is probably an artifact of inadequate sampling of the conformational space in regions away from the minima (Warshel and Parson, 2001). The standard free energy change (∆Go) and the reorganization energy (λ) are indicated for each reaction.
(Schmidt et al., 1994, 1995). As mentioned above, transient absorbance changes following excitation of the modified RCs show an increased population of P+BA– in equilibrium with P+Pheo–. From the apparent equilibrium constant and the assumed free energy difference between P+Pheo– and P+HA–, P+BA– in native RCs was calculated to be 0.06 eV (1.4 kcal/mol) below P*, in good agreement with the calculations described above. A potential problem with this approach is that the difference between the energies of P+Pheo– and P+HA– depends partly on interactions of Pheo– and HA– with the protein, and so can not be determined simply from the Em values of the pigments in solution. This concern could be addressed by calculations of the solvation energies. Another approach is to use site-directed mutations or pigment exchanges to shift the free energy of P+BA– by predictable or measurable amounts relative to its value in wild-type RCs, and to fit the resulting changes in the rate of electron transfer with a theoretical expression that has the absolute free energy in wild-type RCs as an adjustable parameter (Williams et al., 1992; Jia et al., 1993; Murchison
et al., 1993; Nagarajan et al., 1993; Egger and Mak, 1994; Bixon et al., 1995, 1996; Arlt et al., 1996a; Holzwarth and Muller, 1996; Sim and Makri, 1997; Zusman and Beratan, 1998; Spörlein et al., 2000; Roberts et al., 2001; Haffa et al., 2002; Huppman et al., 2002; Ivashin and Larsson, 2002). A variety of theoretical formalisms have been used for this, including the classical Marcus equation (see Section IV), quantum mechanical expressions with a single high-frequency vibrational mode and a continuum of low-frequency modes, and path-integral treatments with many vibrational modes. These studies appear to have converged on an energy close to, and somewhat below, the energy of P* in native RCs, in agreement with the results described above. Using a phenomenological theoretical expression in this way is not without potential pitfalls, however. The analyses that have been described generally assume that the mutations change only a single parameter and that vibrational equilibration occurs rapidly relative to electron transfer.
360 III. Unusual Features of the ChargeSeparation Reactions A. Speed and Temperature Dependence The reactions that generate P+HA– have several remarkable features. To begin with, they are extremely fast. This in itself is perhaps not surprising, considering that some of the atoms in the π system of BA are only 3 to 4 Å from atoms of P or HA. After all, photosynthetic bacteria have had several billion years to evolve reactions that achieve charge separation with a quantum yield close to 100% (Wraight and Clayton, 1974; Trissl et al., 1990; Volk et al., 1998). But the centers of P and HA are some 18 Å apart. Moving an electron this far creates a large electric dipole that one would expect to interact strongly with the surrounding protein. In solution, such reactions require a major reorganization of the solvent, which usually makes them slow and strongly dependent on temperature. Electron transfer from P* to HA in the RC behaves in just the opposite manner, becoming even faster with decreasing temperature (Woodbury et al., 1985, 1994; Breton et al., 1988; Fleming et al., 1988; Nagarajan et al., 1990, 1993; Lauterwasser et al., 1991; Jia et al., 1993; Haffa et al., 2002; Huppman et al., 2002, 2003). Formation of P+HA– at temperatures below 20 K occurs with an overall time constant of about 1.2 ps in Rba. sphaeroides and 0.9 ps in Blc. viridis. Electron transfer from HA– to QA also speeds up with decreasing temperature (Schenck et al., 1981; Kirmaier et al., 1985b; Gunner and Dutton, 1989). Although the interactions of the radical-pair states with their surroundings are attenuated by delocalization of the charges over the macrocyclic ring systems and the two BChls of P+, the acceleration of the charge-separation reactions with decreasing temperature demands an explanation. We will return to this question in Section IV. B. Specificity Perhaps the most remarkable aspect of the chargeseparation reactions is their high specificity for the electron acceptors on the ‘A’ or ‘L’ side of the RC’s axis of approximate C2 symmetry (BA and HA) (Kirmaier et al., 1985a, 2001, 2004; Kellog et al., 1989; Mar and Gingras, 1990; Robles et al., 1990a; Heller et al., 1995; de Boer et al., 2002a; Katilius et al., 2002; Haffa et al., 2004). HA can be distinguished from HB, the BPhe on the inactive (‘B’ or ‘M’) side, because a
William W. Parson and Arieh Warshel hydrogen bond to a protonated Glu residue shifts the Qx and Qy absorption bands of HA to slightly longer wavelengths relative to those of HB (Bylina et al., 1988). Electron transfer from P* to HB is undetectable in wild-type RCs. The specificity of the reactions is surprising because the arrangement of the electron carriers is highly symmetric, and most of the amino acid residues in homologous positions on the two sides are identical or structurally similar. But as shown in Fig. 1B, the P+BB– radical-pair is calculated to be about 6 kcal/mol above P+BA– in energy, which (if correct) puts P+BB– significantly above P* (Parson et al., 1990; Parson and Warshel, 2006). The difference between the calculated energies of P+BB– and P+BA– results partly from the phenolic –OH group of Tyr M210 in Rba. sphaeroides (M208in Blc. viridis and Rba. capsulatus), which probably is oriented so as to stabilize BA– (Parson et al., 1990; Alden et al., 1996b). The homologous residue on the other side of the RC, Phe L181, cannot stabilize BB– in this way. In agreement with this picture, replacing Tyr M210 by Trp or other amino acids slows charge separation and increases the activation energy (Finkele et al., 1990; Nagarajan et al., 1990, 1993; Chan et al., 1991b; Shochat et al., 1994; Beekman et al., 1996). In Rba. capsulatus, the double mutation Y(M)208F/ F(L)181Y partly reverses the specificity of charge separation, resulting in formation of P+HB– with a quantum yield of 15 to 30% (Kirmaier et al., 2001, 2004). The Y(M)208F and F(L)181Y mutations affect the rate of charge separation in Rba. sphaeroides in essentially the same way, although they do not lead to formation of P+HB– in detectable amounts (Chan et al., 1991b). Other mutations or pigment substitutions designed to stabilize P+BB– relative to P+BA– also have been found to promote electron transfer to the B side (Taguchi et al., 1992, 1996; Heller et al., 1995, 1996; Kirmaier et al., 1995a,b, 1999, 2001, 2002, 2004; Lin et al., 1996b; Czarnecki et al., 1999; Katilius et al., 1999, 2002, 2004; de Boer et al., 2002a,b; Wakeham and Jones, 2005; Chuang et al., 2006; Chapter 16, Jones). Formation of P+HB– in a yield of 70% was achieved by first interchanging multiple residues from the D helices of the L and M subunits, which disrupts the binding of BPhe HA and completely prevents charge separation, and then introducing the F(L)181Y, H(M)195F and H(L)168F mutations to lower the energy of P+BB– (Chuang et al., 2006). Together, these observations provide compelling evidence that
Chapter 19
Mechanism of Charge Separation
the relative energies of P+BA–, P* and P+BB– shown in Fig. 1 are at least qualitatively correct, and that the favorable energy of P+BA– plays a key role in the specificity of the charge-separation reactions in wildtype RCs. A conflicting set of calculations that put P+BB– below both P* and P+BA– in energy (Ceccarelli and Marchi, 2003) could possibly reflect the use of periodic boundary conditions and the problematic Ewald-sum method to treat electrostatic interactions (see Section II.B). Several authors have found that P* has greater orbital overlap with BA than with BB, and have suggested that the resulting difference in the electronic coupling factor or ‘tunneling matrix element’ (J) accounts for the specificity of the initial electrontransfer reaction (Ivashin et al., 1998; Kolbasov and Scherz, 1998, 2000; Zhang and Friesner, 1998). If J is small relative to the energy difference between the reactant and product states, the rate of electron transfer is expected to depend on J 2 and to fall off rapidly with the distance between the electron donor and acceptor (Parson et al., 1987; Warshel et al., 1988; Moser et al., 1992; Gray and Winkler, 1996; Newton, 2003; Skourtis et al., 2005). This explanation of the specificity is difficult to test experimentally because a mutation or pigment substitution that changes J is likely to change the energetics of the reactions as well. However, the density-matrix theory described in Section IV suggests that J is large enough so that the calculated differences in its magnitude would have relatively minor effects on the rate of electron transfer (Parson and Warshel, 2004a,b). Fluctuations in the distances between the electron carriers are likely to reduce the differences in J (Warshel et al., 1988). A difference between the dielectric properties of the protein on the A and B sides also has been suggested to be an important factor in the specificity of electron transfer (Steffen et al., 1994). However, the calculated reorganization energies, which express the energetic costs of rearranging the protein around the charged products, are essentially the same on the two sides (see Fig. 1). The contributions of electronic induced dipoles to the calculated free energies of P+BA– and P+BB– also do not differ significantly (Parson et al., 1990). The favorable free energy of P+BA– thus appears to depend primarily on interactions with pre-arranged groups in the protein, rather than on dielectric effects.
361 C. Multiphasic Kinetics The electron-transfer kinetics are more complex than one might anticipate, even allowing for the formation of P+BA– as an intermediate. In addition to the major component with a time constant of 3 to 4 ps, the decay of the stimulated emission and absorbance changes associated with P* has a small component lasting 10 to 20 ps (Vos et al., 1991, 1992; Du et al., 1992; Hamm et al., 1993; Stanley and Boxer, 1995; Beekman et al., 1996; Holzwarth and Muller, 1996; Ogrodnik et al., 1999; Wang et al., 2007). It is unclear whether the two components represent distinctly different processes, or are early and later parts of a single process with a nonexponential time course. Explanations that have been suggested for the slower component include static or dynamic heterogeneity in the energy of the reaction (Kirmaier and Holten, 1990; Becker et al., 1991; Jia et al., 1993; Wang et al., 1993; Gehlen et al., 1994; Hartwich et al., 1998), relaxations of the protein around P+HA– (Peloquin et al., 1994; Woodbury et al., 1994; Ogrodnik et al., 1999), relaxations of P* (Nagarajan et al., 1993; Wang et al., 2007), and reversible transfer of an electron to BB or HB (Hamm et al., 1993). Spectral hole-burning experiments argue against gross heterogeneity as an explanation for the multiphasic kinetics (Small et al., 1992; Lyle et al., 1993; Small, 1995). The main observation here is that the width and shape of the zero-phonon hole burned by exciting RCs near the 0-0 transition energy at low temperatures agree well with those expected if P* decays with a time constant of about 1 ps. A significant population of RCs that reacted at a much lower rate should give rise to a correspondingly narrower hole that probably would have been resolved experimentally. Further, the electron-transfer rate calculated from the hole width does not vary significantly when the excitation wavelength is scanned across the inhomogeneous distribution of 0-0 transition energies (Small et al., 1992; Lyle et al., 1993). This indicates that the electron-transfer kinetics are not correlated with the 0-0 transition energy, as one might expect them to be if the multiphasic decay kinetics of stimulated emission from P* reflected heterogeneity in the RCs. Infra-red absorbance changes seen at short times after an excitation pulse suggest that P* undergoes an internal relaxation with a time constant of 0.2 ps at 285 K, possibly to a state with increased chargetransfer character (Hamm and Zinth, 1995). However,
William W. Parson and Arieh Warshel
362 the finding that the 1 ps time constant derived from the zero-phonon hole width matches the time constant for electron transfer argues against the possibility that P* evolves into a different electronic state on a time scale much shorter than this, at least at low temperatures (Johnson et al., 1990; Middendorf et al., 1991; Reddy et al., 1992; Lyle et al., 1993). Whether the infra-red absorbance changes occur at low temperatures is not known. Sub-picosecond reorientations of water molecules around P* have been invoked to account for the high efficiency with which RCs trap energy from the antenna (Borisov and Fok, 1999), but have not been studied experimentally. D. Coupling to Vibrational Wavepackets The absorption band at 1020 nm that is assigned to BA– apparently does not simply rise and decay once, as it would in a classical two-step kinetic scheme. It develops and disappears again several times on a picosecond time scale, suggesting that P+BA– is generated in concert with nuclear motions that are launched when RCs are excited with a short pulse of light. (Yakovlev et al., 2000, 2002, 2003; Yakovlev and Shuvalov, 2000; Shuvalov and Yakovlev, 2003). Because a subpicosecond excitation pulse includes a broad band of frequencies, it can project molecules simultaneously into many different vibrational substates of P*. The wavefunctions for a harmonic vibrational mode (m) have the form Ψ nm ( rm , t ) = χ nm ( rm ) exp ⎡⎣ −iEmn (t + δ ) / ប ⎤⎦ = χ nm ( rm ) exp ⎡⎣ −iω m ( n + 1 / 2)(t + δ ) / ប ⎤⎦
(1) where n = 0, 1, 2, …, χnm (rm) is a spatial (harmonic oscillator) wavefunction of nuclear coordinate rm, Enm is the energy of level n, ωm is the wavenumber (reciprocal wavelength) of the mode, ប is Planck’s constant divided by 2π, t is time, and δ is a phase shift. Population of several such levels at the same time (δ ≈ 0) creates a linear combination of these wavefunctions (wavepacket) that moves back and forth on rm much like a classical particle. As the wavepacket moves, its spatial overlap with various vibrational levels of the ground electronic state changes, causing the emission spectrum of the excited state to oscillate. Such oscillations have been seen in both the stimulated and spontaneous emission of P* (Vos et al., 1991,
1992, 1993, 1994a,b, 1996, 2000; Stanley and Boxer, 1995; Martin and Vos, 1992; Rischel et al., 1998; Spörlein et al., 1998; Vos and Martin, 1999). Fourier analysis of the oscillations reveals components with energies ranging from below 30 cm–1 to about 250 cm–1, with the upper limit probably being set by the spectral width of the excitation pulse. Shuvalov and coworkers (Yakovlev et al., 2000; Yakovlev and Shuvalov, 2000; Yakovlev et al., 2002, 2003; Shuvalov and Yakovlev, 2003) have suggested that P+BA– forms reversibly each time the P* wavepacket crosses the potential energy surface of P+BA–, but returns to P* and is regenerated several times before it is trapped by relaxations of the surrounding protein. Transient peaks in the amount of P+BA– also could be explained by stepwise formation of P+BA– followed by rapid conversion to P+HA– rather than return to P* (Parson and Warshel, 2004a, 2004b). The oscillations of P* and P+BA– have not yet been assigned definitively to particular motions of the BChls or the surrounding protein. Normal-mode calculations and resonance Raman spectra of isotopically labeled BChl suggest that out-of-plane deformation of the C2 acetyl groups contributes a strong mode at 35-cm–1, whereas several types of deformation of the macrocyclic rings contribute to a band around 130 cm–1 (Czarnecki et al., 1997). Motions of the protein or bound water molecules also appear to be important, however, because mutations can modify the oscillation frequencies substantially (Rischel et al., 1998; Yakovlev et al., 2003). At low temperatures where charge separation occurs with a time constant of about 1 ps, the wavepacket motions of P* disappear with essentially the same kinetics as the excited state itself. In RCs carrying a point mutation that slows charge-separation, damping of the wavepacket precedes electron transfer but still occurs on the time scale of 1 to 2 ps, with high-frequency components of the motions decaying more rapidly than low-frequency components (Vos et al., 1996). The time constant for damping of the wavepacket under these conditions probably reflects stochastic relaxations that take the vibrational sublevels of P* to thermal equilibrium and are analogous to T1 processes in NMR. Vibrational coherence also can decay by ‘pure dephasing’ caused by fluctuations of the vibrational energies (T2-type processes), but this process likely is slower under most conditions. The observed time constants of 1 to 2 ps therefore indicate that vibrational relaxations of P* take place on the same time scale as electron transfer.
Chapter 19
Mechanism of Charge Separation
By extension, vibrational relaxations of P+BA– and P+HA– probably occur on the same time scale, which raises the question of whether they limit the overall rate of charge separation. Several models in which relaxations of the product state are rate-limiting have been suggested (Kitzing and Kühn, 1990; Peloquin et al., 1994; Woodbury et al., 1994; Lin, Taguchi et al., 1996; Yakovlev et al., 2000, 2002, 2003; Yakovlev and Shuvalov, 2000; Shuvalov and Yakovlev, 2003). However, the density-matrix analysis described in Section IV suggests that these relaxations are fast enough so that further increases in their speed would not increase the rate of charge separation greatly, and could even decrease it (Parson and Warshel, 2004a, 2004b). In an attempt to probe relaxations of the protein more directly, Wang et al. (2007) have measured absorbance changes at 280 nm that could reflect the effects of changing electric fields on tryptophan residues near the electron carriers. Absorbance changes at this wavelength develop with little or no lag after RCs are excited, and then decay with a multi-exponential time course with time constants on the order of 3, 10 and 190 ps. The rise and decay kinetics are not affected by mutations that alter the rate of charge separation, suggesting that the signals are initiated by the initial excitation of P rather than the formation of P+BA– or P+HA–. Wang et al. (2007) therefore interpret the 3- and 10-ps components of the decay kinetics as reflecting dielectric relaxation of the protein around P*. The 190-ps component appears to accompany electron transfer from HA– to QA because it has a much smaller amplitude in RCs that are depleted of the quinone. Studies of mutants in which individual Trp residues near P, BA or HA are replaced by Tyr or Phe should help to identify the residues that are responsible for the absorbance changes, and to elucidate why these residues appear to respond more strongly to P* than to the much larger dipoles of P+BA– and P+HA–. E. Stability to Mutations Charge separation in RCs is surprisingly robust to mutations. Although mutations that raise the energy of P+BA– relative to P* slow charge separation, photoreduction of HA still occurs with a relatively high quantum yield as long as the RC assembles completely and contains P, BA and HA (Robles et al., 1990a,b; Lin et al., 1996b; Taguchi et al., 1996). No single amino acid residue appears to be essential for charge separation to the level of P+HA–. For example,
363 mutations that reverse the charge of arginine L135 or M164, the ionizable residues closest to P, have very little effect on the kinetics (Johnson et al., 2003). Mutations that add or remove hydrogen bonds between the electron carriers and the protein generally have larger, but still relatively modest effects (Williams et al., 1992; Haffa et al., 2002; Huppman et al., 2002). And although mutations that change the relative energies of P+BA– and P+BB– can increase the probability of electron transfer to the acceptors on the B side of the RC, inducing a significant fraction of the charge-separation to proceed in this direction requires multiple mutations (Taguchi et al., 1992, 1996; Lin et al., 1996b; Kirmaier et al., 2001, 2004; de Boer et al., 2002b; Chuang et al., 2006). Even mutations that raise the energy of P+BA– enough to slow the initial electron-transfer by a factor of 10 or more generally do not prevent charge separation from occurring at cryogenic temperatures (Nagarajan et al., 1993; Haffa et al., 2002). IV. Theories of Electron-Transfer Reactions A. Marcus Theory and Semiclassical Surfacehopping Much of the discussion of the charge-separation reactions has been based on the general theory of electron-transfer processes developed by R. A. Marcus (Marcus, 1956, 1964, 1993). Consider the free energy surfaces of the diabatic electronic states P* and P+BA– shown in Fig. 1. The term diabatic means that a system in one of these states is not stationary, but rather has some probability of making a transition to the other state. In a quantum mechanical description, P* and P+BA– do not diagonalize the Hamiltonian of the system. Overlap of the molecular orbitals of P* and BA– gives rise to off-diagonal terms in the Hamiltonian, and it is these off-diagonal terms (the electronic coupling factor J and its complex conjugate) that bring about transitions between the two states. The stationary states that diagonalize the interaction Hamiltonian are described as adiabatic. Classically, transitions between two states occur only at the intersection of the potential energy surfaces, where a transition conserves energy and momentum. For parabolic surfaces with the same curvature, the free energy of the intersection relative to the minimum free energy of the reactant state, i.e., the activation free energy of the reaction (∆G‡),
William W. Parson and Arieh Warshel
364 is given by ∆G‡ = (∆Go + λ)2/4λ, where ∆Go is the free energy difference between the two minima and λ again is the reorganization energy (see Fig. 1). In transition-state theory, the probability of finding the system at the transition point is taken to be exp(–∆G‡/kBT), where kB is the Boltzmann constant and T the temperature. (Marcus, 1956, 1964) wrote the rate constant for electron transfer as the product of this probability and a factor that depends mainly on the electronic coupling:
(
)
(
k12 = π1/ 2 J 2 ប λk BT exp ⎡ − ∆G o + λ ⎣⎢
)
2
4λk BT ⎤ ⎦⎥ (2)
This expression assumes that the system has only a small probability of making a transition from one diabatic state to the other each time it passes through the intersection point, which will be the case if J is much smaller than the fluctuating interactions of the system with its surroundings. Electron transfer in this regime is said to be nonadiabatic. In the opposite limit of very strong coupling, the system tends to remain on the lower adiabatic surface as it approaches the intersection, so that transitions between the diabatic states have a high probability; electron transfer then is described as adiabatic. According to Eq. (2), k12 has a maximal value when ∆Go = –λ. The diabatic energy surfaces then intersect at the minimum of state 1; ∆G‡ is zero; and the pre-exponential factor causes the rate to increase weakly with decreasing temperature. The calculated free energy curves for P* and P+BA– are qualitatively consistent with this situation (Fig. 1A). The Marcus equation thus can rationalize the unusual temperature dependence of the charge-separation reactions. It also can account for the sensitivity of the rate and specificity of the initial reaction to changes in the energy of P+BA– or P+BB–. However, the Marcus equation generally underestimates the rate at low temperatures where nuclear tunneling and zero-point vibrational energies become important (Warshel et al., 1989). In particular, it fails to explain how charge separation can continue to occur at low temperatures in RCs with mutations that raise the energy of P+BA– (Haffa et al., 2002). Equation (2) also rests on the assumption that the vibrational sublevels of each electronic state reach thermal equilibrium rapidly relative to the rate of electron transfer. As discussed in Section III, this assumption appears to break down for the initial steps of charge separation. The Marcus equa-
tion nevertheless provides a useful, semiquantitative basis for discussing the rate and specificity of charge separation. It is important to distinguish thermodynamic quantities such as ∆Go and ∆G‡ from the microscopic properties of an individual RC. Thermodynamic quantities represent mean values of fluctuating microscopic quantities over a large ensemble of particles, which are equivalent to averages over an extended period of time for an individual particle. The fluctuating gap (X(t)) between the electrostatic energies of P+BA– and P* can be recorded as a function of time during a classical MD simulation of an individual RC, and can be used to simulate a nonadiabatic electron-transfer reaction with a semiclassical ‘surface-hopping’ model that converges with the Marcus equation at high temperatures. Transitions between the two diabatic states again occur mainly when the energy gap is close to zero (Tully and Preston, 1971; Miller and George, 1972; Warshel, 1982; Warshel and Hwang, 1986; Warshel and Parson, 2001). Figs 2A and B show typical records of time-dependent energy gaps for forming P+BA– and P+BB– during MD simulations of Rba. sphaeroides RCs at 300 K, and Fig. 3A shows the predicted kinetics of electron transfer from P* to BA in a simplified surface-hopping model. In these simulations, the energy gap for forming P+BA– (Fig. 2A) crossed zero about 20 times per ps (2 × 1013 s–1), while that for P+BB– (Fig. 2B) had no crossings (<109 s–1). If the energy gap for P+BB– fluctuates rapidly over a range of about ±2 kcal/mol, as these and other MD simulations indicate, mutations that change the mean value of the energy gap by less than this should have only minor effects on the electron-transfer kinetics. This point is illustrated in Fig. 3A, where increasing the mean energy gap by 1 kcal/mol has virtually no effect whereas shifting the mean value by more than 2 kcal/mol slows the reaction considerably. Surfacehopping simulations of charge separation by a reversible, two-step kinetic model with a small negative ∆Go in each step give dynamics similar to those seen experimentally (Creighton et al., 1988; Warshel and Parson, 1991). Including a more rigorous integration of the transition probabilities near the crossings has only minor effects on the results. The Marcus equation has been extended to include a time-dependent solvent coordinate with diffusional kinetics and its own reorganization energy, in addition to the rapid fluctuations illustrated in Figs. 2A and B (Sumi and Marcus, 1986; Okada, 2000; Ando and
Chapter 19
Mechanism of Charge Separation
365
Fig. 2. (A) Calculated time-dependent energy gap (∆Velec) between P+BA– and P* in Rba. sphaeroides RCs during a 1-ns MD trajectory at 300 K (Warshel and Parson, 2001). The potential surface for the ground state was used as a surrogate for the potential surface of P*. (B) Same as A but for the energy gap between P+BB– and P* during a similar trajectory. (C) A portion of the autocorrelation function of the energy gap shown in A, normalized to 1.0 at t = 0. (D) A portion of the normalized autocorrelation function of the energy gap shown in B. The autocorrelation functions also have a component with an ultrafast (~50 fs) decay that is off scale at the top of the graphs. This component reflects dephasing of oscillations at different frequencies and is independent of vibrational relaxations (Parson and Warshel, 2004a). (E) Fourier transform of the autocorrelation function shown in C. Insert: Fourier transform of the autocorrelation function of fluctuations of the CE2-CZ-OH-HOH dihedral angle of Tyr M210 during a similar trajectory. (F) Fourier transform of the autocorrelation function shown in D.
Sumi, 2003). Motions of an ensemble of particles on such a coordinate can make ∆Go and the total reorganization energy, and thus ∆G‡, evolve with time. Wang et al. (2007) have used the decay kinetics of the ultra-violet (UV) absorbance changes initiated by excitation of P (see above) as a measure of the mean position along a similar ‘slow’ coordinate. They were able to reproduce the nonexponential kinetics of charge separation for both wild-type RCs and mutants with altered values of the initial ∆Go. It was necessary
to adjust the initial ∆Go phenomenologically for each mutant, but the adjustments were generally consistent with other observations. Since Wang et al. (2007) obtain the relaxation kinetics experimentally, the essential feature of their analysis is not that some of the RCs’ motions resemble diffusion, but simply that the mean energy difference between P* and P+BA– changes with time as the protein relaxes around P*. The dynamics of this evolution can be explored by MD simulations in several different
366
William W. Parson and Arieh Warshel
Fig. 3. (A) Simulated kinetics of electron transfer from P* to BA in a surface-hopping model with a single, irreversible step. The population of P* was assumed to split each time the fluctuating energy gap between P+BA– and P* passed through zero in either direction, with 3% of the population at that point going to P+BA– and 97% remaining as P*. The resulting decay curve was averaged over all possible starting times in a cyclically permuted record of the energy gap from a 1-ns MD trajectory in the reactant state (Fig. 2A). For curve 1, the mean energy gap was taken to be zero; for curves 2, 3 and 5, the mean was set at 1.0, 2.0 or 2.4 kcal/mol, respectively. For curve 4 (dashed line), the mean energy gap was increased from 2.0 to 2.4 kcal/mol with a time constant of 1.5 ps to simulate an exponential decrease in the energy of P*. (B) Simulated kinetics of electron transfer from P* to BA in a density-matrix model in which electron transfer was coupled to a single vibrational mode with wavenumber 30 cm–1. The minimum energies of P+BA– and P+HA– were set, respectively, 175 and 1000 cm–1 (0.5 and 2.86 kcal/mol) below that of P*. The dimensionless displacement of P* from the ground state on the vibrational coordinate was 2.5, the displacement of P+BA– relative to P* was 3.7, the electronic coupling factor (Jel) was 6 cm–1, and T1o was 2.5 ps. The RC was assumed to start with the vibrational sublevels of the ground state in thermal equilibrium at 80 K, and was excited either coherently (solid lines) or incoherently (dashed lines). The excitation pulse had a Gaussian spectral distribution with full width at halfmaximum amplitude of 150 cm–1 centered on the 0-0 transition energy. P+BA– was assumed to decay stochastically to P+HA– with a time constant of 0.5 ps. The parameter values used in this one-mode model were chosen to illustrate aspects of the density-matrix treatment, and are not estimates of the values for any particular vibrational mode of the RC. (C) The time-dependent expectation value of the energy gap (X(t)) for RCs in the P* state in the density-matrix model considered in B.
ways. According to the fluctuation-dissipation theorem (Kubo, 1966), the linear response of an ensemble of systems to a small, external perturbation such as impulsive creation of an electric dipole is given by the autocorrelation function of the fluctuations of the ensemble at equilibrium. The normalized autocorrelation function of a fluctuating function x(t) is defined as C(t) = 〈x(t´)x(t´+t)〉/〈x(t´)x(t´)〉, where 〈...〉 denotes the average value of the quantity in the brackets. Panels C and D of Fig. 2 show autocorrelation functions for the energy gaps for forming P+BA– and P+BB– in Rba. sphaeroides RCs at 300 K (Parson and Warshel, 2004a, 2004b). Following an ultrafast step that is off scale in the figures, the autocorrelation functions exhibit complex oscillations that are damped on the
time scale of 1 to 10 ps. The center of these oscillations decays to zero on the same time scale. Essentially identical dynamics are obtained by averaging the time-dependent energy gaps in many independent MD trajectories of P+BA– beginning when this state is created from P* (Parson and Warshel, 2004a). Judging from the decay of the autocorrelation functions, an ensemble of RCs thus appears to complete most of its response to an impulsive electrostatic perturbation within about 10 ps. This is consistent with the relaxations of the UV absorbance changes described by Wang et al. (2007), which probably provide a more sensitive probe of small components of the relaxation that continue on longer time scales. The effects of a slow change in the mean energy
Chapter 19
Mechanism of Charge Separation
gap between P* and P+BA– can be illustrated by the simple surface-hopping model used in Fig. 3A. For the simulation shown as curve 4, the mean energy gap (〈X 〉) was written as 〈X(t)〉 = 2.4 – exp(–t/τ) kcal/mol, with a relaxation time constant (τ) of 1.5 ps. The mean gap thus began with the value used for curve 3 but increased progressively to that used for curve 5. The resulting electron-transfer kinetics are decidedly non-exponential, in qualitative accord with the kinetics seen experimentally. Note that the rate constants for the fast and slow component of the electron-transfer kinetics depend on the random fluctuations about the initial and asymptotic values of 〈X 〉, not the time constant for approaching the asymptote. If the initial value of 〈X 〉 were negative rather than positive as assumed here, the rate of electron transfer would increase with time and then could decrease again when 〈X 〉 became positive. Note also that non-exponential kinetics will be seen only if the relaxation occurs on the same time scale as the faster component of electron transfer, as appears to be the case in RCs. B. Coupling to Quantized Vibrational Modes To deal with quantum mechanical effects that are missing in the treatments described above, BornOppenheimer vibronic wavefunctions commonly are written in the form ⎧ ⎫ ψ k (r, t ) = ⎨φk ∏ χ mn( m,k ) ( rm ) ⎬ exp( −iEk t / ប ) ⎩ m ⎭
(3)
Here φk is an electronic wavefunction, χmn(m,k) (rm) is a harmonic-oscillator wavefunction of vibrational coordinate rm, r is the complete vector of coordinates, and n(m,k) is the excitation level (number of phonons) of mode m in vibronic state k. Ek, the energy of vibronic state k, is Ek = Ek ,0 + ∑ [ n( m, k ) + 1 / 2]បω m
(4)
m
where Ek,0 is an electronic energy. We have neglected phase shifts for simplicity. In a system with a single vibrational mode (r1), the interaction matrix element for transitions from a vibronic state comprised of electronic state j and u phonons to a state comprised of electronic state k and n phonons is the product of a purely electronic factor (Jel) and a vibrational overlap integral,
367 F1n,u =
∞
∫ χ1 χ1 dr1 n
u
−∞
The Franck-Condon factor for the transition is |F1n,u|2, which can differ from zero only if wavefunction χ1n is displaced from χ1u along r1. The requirement for a non-zero Franck-Condon factor replaces the classical principle that transitions occur only at intersections of the diabatic energies. Overlapping tails of the reactant and product wavefunctions thus allow electron transfer to occur at low temperatures even if the only vibrational levels of the reactant state with significant populations lie below the intersection of the diabatic surfaces. More generally, the Franck-Condon factor involves a product of many vibrational overlap integrals, one for each vibrational mode that is coupled to the transition, i.e., for each coordinate along which the energy minima of the reactant and product states are displaced. The overall rate of electron transfer depends on a sum of Franck-Condon factors for transitions between various possible initial and final vibrational states, with each Franck-Condon factor weighted by the population of the initial state. Several expressions for the rate constant have been developed using this general formalism. Kubo and Toyozawa (1955) gave the exact expression for a multidimensional harmonic system. Jortner, Bixon and coworkers have derived expressions for a single high-frequency mode and a continuum of low-frequency modes, and have used them to fit the dependence of the charge-separation kinetics on temperature and on mutations that change ∆Go (Bixon and Jortner, 1968, 1986, 1999; Kestner et al., 1974; Rips and Jortner, 1988). Instead of the single maximum predicted by Eq. (2), the quantum treatment with even a single vibrational mode can give a series of maxima where the energy of the initial state matches the energies of successive steps in the ladder of possible final states (Sarai, 1980). Information on the vibrations that are coupled to charge separation can be obtained from molecular dynamics (MD) simulations of the reaction. In the dispersed polaron theory developed by Warshel and coworkers, the fluctuations of the time-dependent energy gap (X(t)) recorded during an MD trajectory are related to the frequencies of the vibrational motions that are coupled to the reaction (Churg and Warshel, 1985; Hwang and Warshel, 1987; Hwang et al., 1988; Warshel et al., 1989; Warshel and Parson, 1991). A Fourier transform of the autocorrelation function of X(t) has a peak at the frequency of each
368 of these vibrational modes, and the amplitude of the peak is proportional to the square of the displacement of the energy minima along the corresponding vibrational coordinate. The Franck-Condon factors for transitions between different vibronic states of the system can be obtained straightforwardly from the vibrational frequencies and displacements. The frequencies and displacements can be used directly in Kubo and Toyozawa’s (1955) expression for the rate constant (Warshel et al., 1989), or can be used to build a density-matrix model as discussed in the following section. The dispersed polaron treatment is formally identical to the spin boson treatment (Garg et al., 1985; Leggett et al., 1987; Bader et al., 1990; Schulten and Tesch, 1991), which was developed about the same time but initially did not capitalize on the potential of using classical MD simulations to address quantum effects in condensed phases. See Hwang and Warshel (1997) for a discussion of the equivalence of the two treatments and Warshel and Parson (2001) for a review. In retrospect, the important advance was not the development of a formal expression for the rate of electron transfer, since Kubo had already provided that, but the realization that the door was open to microscopic simulations of quantum effects on electron-transfer reactions in proteins. Figure 2 illustrates a dispersed-polaron analysis of the vibrational modes that are coupled to formation of P+BA– and P+BB– in Rba. sphaeroides. The time-dependent energy gaps (panels A and B) and their autocorrelation functions (C and D ) have been discussed above. Fourier transforms of the autocorrelation functions are shown in panels E and F. The bands between 300 and 400 cm–1 that appear prominently in Fig. 2E but not in Fig. 2F can be assigned to torsional motions of the phenolic –OH group of Tyr M208, because peaks at the same frequencies appear in Fourier transforms of the correlation function of the torsional angle itself (Fig. 2E, insert), and both sets of peaks are lost if the Tyr is replaced by Phe in the computer model (unpublished observations). As discussed above, the calculated energy of P+BA– depends strongly on the orientation of the –OH group, while the energy of P+BB– does not. Motions of the –OH group are not coupled strongly to excitation of P, and so do not figure significantly in the wavepackets discussed in Section III. (They also have too high a frequency for many levels to be excited coherently by the excitation flashes that have been used so far.) The bands in the regions of 35, 70
William W. Parson and Arieh Warshel and 130 cm–1 have frequencies similar to some of the oscillations seen experimentally in P* and P+BB–, but have not yet been assigned to particular motions of the protein. Although the fine structures of the spectra for P+BA– (Fig. 2E) and P+BB– (Fig. 2F) differ in these regions, the overall shapes and amplitudes of the two spectra are similar. In principle, the dispersed-polaron/spin-boson treatment provides the frequencies and displacements of all the vibrational modes that are coupled to an electron-transfer reaction, just as the autocorrelation function of the energy gap contains the complete dielectric response of the system to a perturbation. Capturing very low-frequency modes and very slow relaxations of course requires sufficiently long MD trajectories. High-frequency intramolecular modes of the electron carriers also are unlikely to be captured properly by classical MD simulations, but can be treated by simulations that use mixed quantumclassical fields. Mercer et al. (1999) have described such a treatment of vibronic coupling in excitation of BChl. In addition, the vibrational frequencies and displacements and other information obtained from a classical MD simulation at room temperature can be transformed to a quantum mechanical model that should remain valid at low temperatures (Parson and Warshel, 2004a). C. Density-Matrix Treatments Density-matrix formalisms (Slichter, 1963; Redfield, 1965; Davidson, 1976; Blum, 1996) provide a way to introduce stochastic vibrational relaxations and dephasing in the context of the dispersed-polaron/ spin-boson theory, and so to obviate the assumption that vibrational thermalization occurs rapidly relative to electron transfer. In this approach, the wavefunction for an ensemble of RCs is written as a linear combination of vibronic wavefunctions (see Eqs. 3 and 4): ⎧ ⎫ Ψ(r, t ) = ∑ ck (t ) ⎨φk ∏ χ nm( m,k ) ( rm ) ⎬ = k ⎩ m ⎭ ⎧ ⎫ ∑ Ck (t ) exp(−iEk t / ប ) ⎨φk ∏ χnm( m,k ) (rm ) ⎬ k ⎩ m ⎭
(5) The reduced density matrix (ρ) of the system then can be defined by the elements
Chapter 19
Mechanism of Charge Separation
ρ j ,k (t ) = c j (t ) ck *(t ) = C j (t ) Ck *(t ) exp ⎡⎣ −i( E j − Ek )t / ប ⎤⎦
(6)
where the bar denotes an average over an ensemble of molecules. In this representation, (t) contains the complete dependence of the ensemble on time. The diagonal matrix element ρj,j(t) gives the population of vibronic state j at time t, while the off-diagonal element ρj,k(t) conveys the coherence between states j and k (i.e., the ensemble average of exp[–i(Ej – Ek)t/ ប]). The expectation value of any dynamic observable can be computed as the trace of the product (t)⋅õ, where õ is a matrix representation of the operator corresponding to the observable. The time dependence of the density matrix is given by the stochastic Liouville equation, ∂ρ j ,k ∂t
=
i ∑ ⎡ρ ⋅ J - J j ,u ⋅ ρu,k ⎤⎦ + ប u ⎣ j ,u u ,k ∑ ∑ R j ,k ,u,v ⋅ ρu,v u
(7)
v
where Ju,k is the coupling factor for transitions between states k and u (the product of an electronic factor and the vibrational overlap integrals) and the Rj,k,u,v are elements of the Redfield relaxation tensor (Slichter, 1963; Redfield, 1965; Jean et al., 1992; Walsh and Coalson, 1992; Jean and Fleming, 1995; Pollard et al., 1996; Parson and Warshel, 2004a, 2004b). Rj,j,k,k is the rate constant for stochastic transitions from vibronic state k to state j; Rk,k,k,k is –1 times the sum of the rate constants for transitions from state k to all the other states; Rj,k,j,k is –1 times the rate constant for decay of coherence between states j and k; and Ri,j,k,l is for transfer of coherence from one pair of states (k and l) to another (i and j). The density matrix at time t can be obtained by starting at t = 0 with an initial density matrix that reflects the absorption spectrum of P and the energy and shape of a given excitation pulse, and then integrating Eq. 7 numerically up to time t (Parson and Warshel, 2004b). To evaluate the elements of the relaxation tensor, one needs the rate constants for stochastic transitions between different sublevels of each vibrational mode that is coupled to electron transfer. The necessary rate constants can be obtained by calculating the autocorrelation function of the energy gap (X(t)) both from the MD simulations and from a density-matrix model
369 with equivalent vibrational modes, and adjusting the microscopic rate constants in the density-matrix model to make the two functions similar. Although this might seem to require a vast number of free parameters, it can be done surprisingly well by adjusting a single parameter, the time constant for equilibration of the two lowest levels of a vibrational mode (T1o) (Parson and Warshel, 2004a,b). The time constant for equilibration of any other pair of levels can be related straightforwardly to T1o and the numbers of phonons in these levels. The necessary value of T1o for Rba. sphaeroides RCs is on the order of 1 to 2 ps at 80 K and is not very dependent on temperature. Values of T1o in this range are consistent with the damping of the oscillations of stimulated emission and excitedstate absorption seen after RCs are excited with a short flash (see Section II.D). The overall decay of the oscillations of the energy gap (Fig. 2C, D) can take longer than T1o, because it represents the approach to thermal equilibrium among a large number of microscopic states. Figures 3B and C illustrate the behavior of a simple density-matrix model in which electron transfer from P* to BA is coupled to a single vibrational mode with an energy of 30 cm–1, the fundamental vibrational transition time (T1o) is 2.5 ps, and P+BA– decays to P+HA– stochastically with a time constant of 0.5 ps. When the ensemble of RCs is excited coherently (solid lines), the population of P+BA– exhibits damped oscillations during the first few picoseconds after the excitation pulse, with the rate of electron transfer peaking whenever the energy gap goes to, or close to, zero. Only small hints of the oscillatory features can be seen if the ensemble is excited incoherently with a pulse of the same energy (dashed lines), although the overall rate of electron transfer is essentially the same except at the shortest times. With the parameter values used in Figs. 3B and C, the oscillating energy gap following coherent excitation makes only one excursion across zero and back before vibrational relaxations destroy most of the ensemble’s coherence. However, nuclear tunneling allows electron transfer to continue with a lower rate constant even though the energy gap is not zero. The density-matrix model thus can reproduce the biphasic kinetics induced by shifting the average value of the energy gap in the surface-hopping model (Fig. 3A), as well as the more complex, oscillatory behavior at short times. Energy is conserved by interactions of the system with the surroundings, which do not need to be enumerated explicitly.
370 In a model with five vibrational modes and T1o ≈ 2 ps, the density-matrix treatment reproduces the measured rate of electron transfer from P* to BA well if the purely electronic part of the coupling factor (Jel) is taken to be on the order of 20 cm–1 (Parson and Warshel, 2004b). This agrees with the values that have been obtained by molecular orbital theory (Ivashin et al., 1998; Kolbasov and Scherz, 1998, 2000; Zhang and Friesner, 1998). The theory also accounts easily for the increase in the rate with decreasing temperature (Parson and Warshel, 2004a,b), but the explanation for the unusual temperature dependence differs from that offered by Marcus theory. In the density-matrix theory, when vibrational relaxation occurs more slowly than electron transfer, temperature affects the rate mainly through the Boltzmann distribution of P among the vibrational sublevels of the ground state before the RC is excited, rather than by affecting P* directly. For example, a vibrational mode of P that is at its zero-point level at low temperatures can be projected into various higher levels in P*, depending on the corresponding Franck-Condon factors. The electron-transfer rate therefore is not necessarily maximal when ∆Go = –λ as in the Marcus theory. Similarly, the maximal electron-transfer rate in the density-matrix theory generally does not occur when P* is at its potential minimum, but rather when the vibrational energy is somewhat higher than this. The rate therefore should fall off if the initial energy of P* is lowered by moving the excitation flash to the red of the 0-0 excitation energy. However, this effect probably would be detectable only at temperatures below about 10 K, and would be lost if Jel is large enough to drive electron transfer between vibronic states that are well off resonance. It has not been seen experimentally (Peloquin et al., 1995, 1996; Lin et al., 1996a; H. Wang, S. Lin and N. Woodbury, personal communication). Modest changes of T1o in the region of 2 ps have little effect on the calculated rate of electron transfer in the density-matrix model. However, decreasing T1o below about 0.5 ps can slow electron-transfer (Parson and Warshel, 2004a,b). This is partly because relaxations of P* compete with electron transfer from ‘hot’ vibrational states that are populated by the excitation pulse. In addition, because transitions from P* to P+BA– occur only by way of off-diagonal elements of ρ, they are thwarted if the off-diagonal elements decay too rapidly. The overall rate of electron transfer also decreases if the off-diagonal elements of ρ are destroyed rapidly by fluctuations of the
William W. Parson and Arieh Warshel vibronic energies (pure dephasing) or by electron transfer from BA– to HA. The last effect, an interesting illustration of the ‘quantum Zeno’ or ‘watched pot’ paradox, is unknown in classical chemical kinetics, where increasing the rate constant for the second of a two-step process never decreases the overall rate. The density-matrix model thus appears to capture the unusual features of the initial steps of charge separation reasonably well. It is particularly satisfying that values for almost all the parameters of the model can be obtained from MD simulations and quantum calculations based on the actual structure of the RC. The main obstacle to using MD simulations to find the rate constants for vibrational relaxations probably is the difficulty of extending MD trajectories much beyond several ns. Without sufficiently long trajectories, small, long-lived components of the autocorrelation function are difficult to distinguish from noise. It should be possible to use UV absorbance changes (Wang et al., 2007) or other experimental measures of dielectric relaxation to refine the relaxation tensor, but this probably will require additional adjustable parameters that are not determined uniquely. Acknowledgments Preparation of this chapter was supported by National Science Foundation grant MCB-9904618 to W. W. P. and National Institutes of Health grants (GM24492 and GM40283) to A. W. We thank Neal Woodbury for helpful discussions and communicating recent results from his laboratory. References Alden RG, Parson WW, Chu ZT and Warshel A (1995) Calculations of electrostatic energies in photosynthetic reaction centers. J Am Chem Soc 117: 12284–12298 Alden RG, Parson WW, Chu ZT and Warshel A (1996a) Macroscopic and microscopic estimates of the energetics of charge separation in bacterial reaction centers. In: Michel-Beyerle ME (ed) Reaction Centers of Photosynthetic Bacteria: Structure and Dynamics, pp 105–116. Springer-Verlag, Berlin Alden RG, Parson WW, Chu ZT and Warshel A (1996b) Orientation of the OH dipole of tyrosine (M)210 and its effect on electrostatic energies in photosynthetic bacterial reaction centers. J Phys Chem 100: 16761–16770 Ando K and Sumi H (2003) Path-integral Monte Carlo calculation of reaction-diffusion equation. J Chem Phys 118: 8315–8320 Arlt T, Schmidt S, Kaiser W, Lauterwasser C, Meyer M, Scheer H and Zinth W (1993) The accessory bacteriochlorophyll: A real electron carrier in primary photosynthesis. Proc Natl Acad
Chapter 19
Mechanism of Charge Separation
Sci USA 90: 11757–11762 Arlt T, Bibikova M, Penzkofer H, Oesterhelt D and Zinth W (1996a) Strong acceleration of primary photosynthetic electron transfer in a mutated reaction center of Rhodopseudomonas viridis. J Phys Chem 100: 12060–12065 Arlt T, Dohse B, Schmidt S, Wachtveitl J, Laussermaier E, Zinth W and Oesterhelt D (1996b) Electron transfer dynamics of Rhodopseudomonas viridis reaction centers with a modified binding site for the accessory bacteriochlorophyll. Biochemistry 35: 9235–9244 Bader JS, Kuharski RA and Chandler D (1990) Role of nuclear tunneling in aqueous ferrous-ferric electron transfer. J Chem Phys 93: 230–236 Becker M, Nagarajan V, Middendorf D, Parson WW, Martin JE and Blankenship RE (1991) Temperature dependence of the initial electron-transfer kinetics in photosynthetic reaction centers of Chloroflexus aurantiacus. Biochim Biophys Acta 1057: 299–312 Beekman LMP, van Stokkum IHM, Monshouwer R, Rijnders AJ, McGlynn P, Visschers RW, Jones MR and van Grondelle R (1996) Primary electron transfer in membrane-bound reaction centers with mutations at the M210 position. J Phys Chem 100: 7256–7268 Bixon M and Jortner J (1968) Intramolecular radiationless transitions. J Chem Phys 48: 715–726 Bixon M and Jortner J (1986) Coupling of protein modes to electron transfer in bacterial photosynthesis. J Phys Chem 90: 3795–3800 Bixon M and Jortner J (1999) Electron transfer. From isolated molecules to biomolecules. Adv Chem Phys 106: 35–202 Bixon M, Jortner J and Michel-Beyerle ME (1995) A kinetic analysis of the primary charge separation in bacterial photosynthesis. Energy gaps and static heterogeneity. Chem Phys 197: 389–404 Bixon M, Jortner J and Michel-Beyerle ME (1996) Energetics of the primary charge separation in bacterial photosynthesis. In: Michel-Beyerle ME (ed) The Reaction Center of Photosynthetic Bacteria: Structure and Dynamics, pp 287–296. Springer-Verlag, Berlin Blankenship RE, Schaafsma TJ and Parson WW (1977) Magnetic field effects on radical-pair intermediates in bacterial photosynthesis. Biochim Biophys Acta 461: 297–305 Blomberg MRA, Siegbahn PHM and Babcock GT (1998) Modeling electron transfer in biochemistry: A quantum chemical study of charge separation in Rhodobacter sphaeroides and photosystem II. J Am Chem Soc 120: 8812–8824 Blum K (1996) Density Matrix Theory and Applications. Plenum Press, New York Boxer SG, Chidsey ED and Roelofs MG (1983) Magnetic field effects on reaction yields in the solid state: An example from photosynthetic reaction centers. Ann Rev Phys Chem 34: 389–417 Boxer SG, Goldstein RA, Lockhart DJ, Middendorf TR and Takiff L (1989) Excited states, electron-transfer reactions, and intermediates in bacterial photosynthetic reaction centers. J Phys Chem 93: 8280–8294 Breton J, Martin J-L, Migus A, Antonetti A and Orszag A (1986) Femtosecond spectroscopy of excitation electron transfer and initial charge separation in the reaction center of the photosynthetic bacterium Rhodopseudomonas viridis. Proc Natl Acad Sci USA 83: 5121–5175
371 Breton J, Martin J-L, Fleming GR and Lambry J-C (1988) Low temperature femtosecond spectroscopy of the initial step of electron transfer in reaction centers from photosynthetic purple bacteria. Biochemistry 27: 8276–8284 Bylina EJ, Kirmaier C, McDowell L, Holten D and Youvan DC (1988) Influence of an amino-acid residue on the optical properties and electron transfer dynamics of a photosynthetic reaction centre complex. Nature 336: 182–184 Ceccarelli M and Marchi M (2003) Simulation and modeling of the Rhodobacter sphaeroides bacterial reaction center II: Primary charge separation. J Phys Chem B 107: 5630–5641 Chan CK, DiMagno TJ, Chen LXQ, Norris JR and Fleming GR (1991a) Mechanism of the initial charge separation in bacterial photosynthetic reaction centers. Proc Natl Acad Sci USA 88: 11202–11206 Chan CK, Chen LXQ, Dimagno TJ, Hanson DK, Nance SL, Schiffer M, Norris JR and Fleming GR (1991b) Mechanism of the initial charge separation in photosynthetic bacterial reaction centers. Chem Phys Lett 176: 366–372 Chidsey CED, Takiff L, Goldstein RA and Boxer SG (1985) Effect of magnetic fields on the triplet state lifetime in photosynthetic reaction centers: Evidence for thermal repopulation of the initial radical pair. Proc Natl Acad Sci USA 82: 6850–6854 Chuang JI, Boxer SG, Holten D and Kirmaier C (2006) High yield of M-side electron transfer in mutants of Rhodobacter capsulatus reaction centers lacking the L-side bacteriopheophytin. Biochemistry 45: 3845–3851 Churg AK and Warshel A (1985) Modeling the activation energy and dynamics of electron transfer reactions in proteins. In: Clementi E, Corongiu G, Sarma MH and Sarma RH (ed) Structure and Motion: Membranes, Nucleic Acids and Proteins, pp 361. Adenine Press, Guilderland, NY Creighton S, Hwang J-K, Warshel A, Parson WW and Norris J (1988) Simulating the dynamics of the primary charge separation process in bacterial photosynthesis. Biochemistry 27: 774–781 Czarnecki K, Chynwat V, Erickson JP, Frank HA and Bocian DF (1997) Characterization of the strongly coupled, low-frequency vibrational modes of the special pair of photosynthetic reaction centers via isotopic labeling of the cofactors. J Am Chem Soc 119: 415–426 Czarnecki K, Kirmaier C, Holten D and Bocian DF (1999) Vibrational and photochemical consequences of an Asp residue near the photoactive accessory bacteriochlorophyll in the photosynthetic reaction center. J Phys Chem A 103: 2235 –2246 Davidson ER (1976) Reduced Density Matrices in Quantum Chemistry. Academic Press, London de Boer AL, Neerken S, de Wijn R, Permentier HP, Gast P, Vijgenboom E and Hoff AJ (2002a) High yield of B-branch electron transfer in a quadruple reaction center mutant of the photosynthetic bacterium Rhodobacter sphaeroides. Biochemistry 41: 3081–3088 de Boer AL, Neerken S, de Wijn R, Permentier HP, Gast P, Vijgenboom E and Hoff AJ (2002b) B-branch electron transfer in reaction centers of Rhodobacter sphaeroides assessed with site-directed mutagenesis. Photosynth Res 71: 221–239 Du M, Rosenthal SJ, Xie X, DiMagno TJ, Schmidt M, Hanson DK, Schiffer M, Norris JR and Fleming GR (1992) Femtosecond spontaneous emission studies of reaction centers from photosynthetic bacteria. Proc Natl Acad Sci USA: 8517–8521 Egger R and Mak CH (1994) Dissipative three-state system and
372 the primary electron transfer in the bacterial photosynthetic reaction center. J Phys Chem 98: 9903–9918 Ermler U, Fritzsch G, Buchanan SK and Michel H (1994) Structure of the photosynthetic reaction centre from Rhodobacter sphaeroides at 2.65 Å resolution: Cofactors and protein-cofactor interactions. Structure 2: 925–936 Figueirido F, Del Buono GS and Levy RM (1997) On the finite size corrections to the free energy of ionic hydration. J Phys Chem B 101: 5622–5623 Finkele U, Lauterwasser C, Zinth W, Gray KA and Oesterhelt D (1990) Role of tyrosine M210 in the initial charge separation of reaction centers of Rhodobacter sphaeroides. Biochemistry 29: 8517–8521 Fleming GR, Martin J-L and Breton J (1988) Rates of primary electron transfer in photosynthetic reaction centers and their mechanistic implications. Nature 333: 190–192 Garg A, Onuchic JN and Ambegaokar B (1985) Effect of friction on electron transfer in macromolecules. J Chem Phys 83: 4491–4503 Gehlen JN, Marchi M and Chandler D (1994) Dynamics affecting the primary charge-transfer in photosynthesis. Science 263: 499–502 Gilson MK, Rashin A, Fine R and Honig B (1985) On the calculation of electrostatic interactions in proteins. J Mol Biol 184: 503–516 Godik VI and Borisov AY (1979) Short-lived delayed luminescence of photosynthetic organisms. I. Nanosecond afterglows in purple bacteria at low temperatures. Biochim Biophys Acta 548: 296–308 Godik VI, Kotova EA and Borisov AY (1982) Nanosecond recombination luminescence of purple bacteria. The lifetime temperature dependence in Rhodospirillum rubrum chromatophores. Photobiochem Photobiophys 4: 219–226 Goldstein RA and Boxer SG (1988) The effect of very high magnetic fields on the reaction dynamics in bacterial reaction centers: Implications for the reaction mechanism. Biochim Biophys Acta 977: 70–77 Gray HB and Winkler JR (1996) Electron transfer in proteins. Ann Rev Biochem 65: 537–561 Gunner MR and Dutton PL (1989) Temperature and -∆Go dependence of the electron transfer from BPh• – to QA in reaction center protein from Rhodobacter sphaeroides with different quinones as QA. J Am Chem Soc 111: 3400–3412 Gunner M, Nichols A and Honig B (1996) Electrostatic potentials in Rhodopseudomonas viridis reaction centers: Implications for the driving force and directionality of electron transfer. J Phys Chem 100: 4277–4291 Haberkorn R and Michel-Beyerle ME (1979) On the mechanism of magnetic field effects in bacterial photosynthesis. Biophys J 26: 489–498 Haffa ALM, Lin S, Katilius E, Williams JC, Taguchi AKW, Allen JP and Woodbury NW (2002) The dependence of the initial electron-transfer rate on driving force in Rhodobacter sphaeroides reaction centers. J Phys Chem B 106: 7376–7384 Haffa ALM, Lin S, Williams JC, Bowen BP, Taguchi AKW, Allen JP and Woodbury NW (2004) Controlling the pathway of photosynthetic charge separation in bacterial reaction centers. J Phys Chem B 108: 4–7 Hamm P and Zinth W (1995) Ultrafast initial reaction in bacterial photosynthesis revealed by femtosecond infrared spectroscopy. J Phys Chem 99: 13537–13544
William W. Parson and Arieh Warshel Hamm P, Gray KA, Oesterhelt D, Feik R, Scheer H and Zinth W (1993) Subpicosecond emission studies of bacterial reaction centers. Biochim Biophys Acta 1142: 90–105 Hartwich G, Lossau H, Michel-Beyerle ME and Ogrodnik A (1998) Nonexponential fluorescence decay in reaction centers of Rhodobacter sphaeroides reflecting dispersive charge separation up to 1 ns. J Phys Chem B 102: 3815–3820 Hasegawa J and Nakatsuji H (1998) Mechanism and unidirectionality of the electron transfer in the photosynthetic reaction center of Rhodopseudomonas viridis: SAC-CI theoretical study. J Phys Chem B 102: 10420–10430 Heller BA, Holten D and Kirmaier C (1995) Control of electron transfer between the L- and M-sides of the photosynthetic reaction center. Science 269: 940–945 Heller BA, Holten D and Kirmaier C (1996) Effects of Asp residues near the L-side pigments in bacterial reaction centers. Biochemistry 35: 15418–15427 Hoff AJ (1981) Magnetic field effects on photosynthetic reaction centers. Quart Rev Biophys 14: 599–665 Holten D, Windsor MW, Parson WW and Thornber JP (1978) Primary photochemical processes in isolated reaction centers of Rhodopseudomonas viridis. Biochim Biophys Acta 501: 112–126 Holten D, Hoganson C, Windsor MW, Schenck GC, Parson WW, Migus A, Fork RL and Shank CV (1980) Subpicosecond and picosecond studies of electron transfer intermediates in Rhodopseudomonas sphaeroides reaction centers. Biochim Biophys Acta 592: 461–477 Holzapfel W, Finkele U, Kaiser W, Oesterhelt D, Scheer H, Stilz HU and Zinth W (1989) Observation of a bacteriochlorophyll anion radical during the primary charge separation in a reaction center. Chem Phys Lett 160: 1–7 Holzapfel W, Finkele U, Kaiser W, Oesterhelt D, Scheer H, Stilz HU and Zinth W (1990) Initial electron transfer in the reaction center from Rhodobacter sphaeroides. Proc Natl Acad Sci USA 87: 5168–5172 Holzwarth AR and Muller MG (1996) Energetics and kinetics of radical pairs in reaction centers from Rhodobacter sphaeroides. A femtosecond transient absorption study. Biochemistry 35: 11820–11831 Hunenberger PH, Mark AE and van Gunsteren WF (1995) Fluctuation and cross-correlation analysis of protein motions observed in nanosecond molecular dynamics simulations. J Mol Biol 252: 492–503 Huppman P, Arlt T, Penzkofer H, Schmidt S, Bibikova M, Dohse B, Oesterhelt D, Wachtveit J and Zinth W (2002) Kinetics, energetics, and electronic coupling of the primary electron transfer reactions in mutated reaction centers of Blastochloris viridis. Biophys J 82: 3186–3197 Huppmann P, Sporlein S, Bibikova M, Oesterhelt D, Wachtveitl J and Zinth W (2003) Electron transfer in reaction centers of Blastochloris viridis: Photosynthetic reactions approximating the adiabatic regime. J Phys Chem A 107: 8302–8309 Hwang J-K and Warshel A (1987) Microscopic examination of free energy relationships for electron transfer in polar solvents. J Am Chem Soc 109: 715–720 Hwang J-K and Warshel A (1997) On the relationship between the dispersed polaron and spin-boson models. Chem Phys Lett 271: 223–225 Hwang J-K, Creighton S, King G, Whitney D and Warshel A (1988) Effects of solute-solvent coupling and solvent saturation
Chapter 19
Mechanism of Charge Separation
on solvation dynamics of charge transfer reactions. J Chem Phys 89: 859–865 Hörber JKH, Göbel W, Ogrodnik A, Michel-Beyerle M-E and Cogdell RJ (1986) Time-resolved measurements of fluorescence from reaction centers of Rhodopseudomonas sphaeroides. FEBS Lett 198: 273–278 Hughes JM, Hutter, MC, Reimers JR and Hush NS (2001) Modeling the bacterial photosynthetic reaction center. 4. The structure, electrochemical, and hydrogen-bonding properties of 22 mutants of Rhodobacter sphaeroides. J Am Chem Soc 123: 8550–8563 Ivashin N and Larsson S (2002) Vibrational mechanism for primary charge separation in the reaction center of Rhodobacter sphaeroides. J Phys Chem B 106: 3996–4009 Ivashin N, Källenbring B, Larsson S and Hansson Ö (1998) Charge separation in photosynthetic reaction centers. J Phys Chem B 102: 5017–5022 Jean JM and Fleming GR (1995) Competition between energy and phase relaxation in electronic curve crossing processes. J Chem Phys 103: 2092–2101 Jean JM, Friesner RA and Fleming GR (1992) Application of a multilevel Redfield theory to electron-transfer in condensed phases. J Chem Phys 96: 5827–5842 Jia Y, DiMagno TJ, Chan C-K, Wang Z, Du M, Hanson DK, Schiffer M, Norris JR, Fleming GR and Popov MS (1993) Primary charge separation in mutant reaction centers of Rhodobacter capsulatus. J Phys Chem 97: 13180–13191 Johnson ET, Nagarajan V, Zazubovich V, Riley K, Small GJ and Parson WW (2003) Effects of ionizable residues on the absorption spectrum and initial electron-transfer kinetics in the photosynthetic reaction center of Rhodobacter sphaeroides. Biochemistry 42: 13673–13683 Johnson SG, Tang D, Jankowiak R, Hayes JM, Small GJ and Tiede DM (1990) Primary donor state mode structure and energy transfer in bacterial reaction centers. J Phys Chem 94: 5849–5855 Katilius E, Turanchik T, Lin S, Taguchi AKW and Woodbury NW (1999) B-side electron transfer in a Rhodobacter sphaeroides reaction center mutant in which the B-side monomer bacteriochlorophyll is replaced with bacteriopheophytin. J Phys Chem B 103: 7386–7389 Katilius E, Katiliene Z, Lin S, Taguchi AKW and Woodbury NW (2002) B-side electron transfer in the HE(M182) reaction center mutant from Rhodobacter sphaeroides. J Phys Chem B 106: 12344–12350 Katilius E, Babendure JL, Lin S and Woodbury NW (2004) Electron transfer dynamics in Rhodobacter sphaeroides reaction center mutants with a modified ligand for the monomer bacteriochlorophyll on the active side. Photosynth Res 81: 165–180 Kaufmann KJ, Petty KM, Dutton PL and Rentzepis PM (1975) Picosecond kinetics of events leading to reaction center bacteriochlorophyll oxidation. Science 188: 1301–1304 Kellog EC, Kolaczkowski S, Wasielewski MR and Tiede DM (1989) Measurement of the extent of electron transfer to the bacteriopheophytin in the M-subunit in reaction centers of Rhodopseudomonas viridis. Photosynth Res 22: 47–59 Kennis JT, Shkuropatov AY, van Stokkum IHM, Gast P, Hoff AJ, Shuvalov VA and Aartsma TJ (1997) Formation of a long-lived P+BA– state in plant pheophytin-exchanged reaction centers of Rhodobacter sphaeroides R26 at low temperature. Biochemistry
373 36: 16231–16238 Kestner NR, Logan J and Jortner J (1974) Thermal electron transfer reactions in polar solvents. J Phys Chem 78: 2148–2166 King BA, de Winter A, McAnaney T and Boxer SG (2001) Excited state energy transfer pathways in photosynthetic reaction centers. 4. Asymmetric energy transfer in the heterodimer mutant. J Phys Chem B 105: 1856–1862 Kirmaier C and Holten D (1988) Subpicosecond spectroscopy of charge separation in Rhodobacter capsulatus reaction centers. Israel J Chem 28: 79–85 Kirmaier C and Holten D (1990) Evidence that a distribution of bacterial reaction centers underlies the temperature and detection-wavelength dependence of the rates of the primary electron transfer reactions. Proc Natl Acad Sci USA 87: 3552–3556 Kirmaier C, Holten D and Parson WW (1985a) Picosecond photodichroism studies of the transient states in Rhodopseudomonas sphaeroides reaction centers at 5 K: effects of electron transfer on the six bacteriochlorin pigments. Biochim Biophys Acta 810: 49–61 Kirmaier C, Holten D and Parson WW (1985b) Temperature and detection-wavelength dependence of the picosecond electron-transfer kinetics measured in Rhodopseudomonas sphaeroides reaction centers. Resolution of new spectral and kinetic components in the primary charge-separation process. Biochim Biophys Acta 810: 33–48 Kirmaier C, Laporte L, Schenck CC and Holten D (1995a) The nature and dynamics of the charge-separated intermediate in reaction centers in which bacteriochlorophyll replaces the photoactive bacteriopheophytin. 1. Spectral characterization of the transient state. J Phys Chem 99: 8903–8909 Kirmaier C, Laporte L, Schenck CC and Holten D (1995b) The nature and dynamics of the charge-separated intermediate in reaction centers in which bacteriochlorophyll replaces the photoactive bacteriopheophytin. 2. The rates and yields of charge separation and recombination. J Phys Chem 99: 8910–8917 Kirmaier C, Weems D and Holten D (1999) M-side electron transfer in reaction center mutants with a lysine near the nonphotoactive B bacteriochlorophyll. Biochemistry 38: 11516–11530 Kirmaier C, He C and Holten D (2001) Manipulating the direction of electron transfer in the bacterial reaction center by swapping Phe for Tyr near BChlM (L181) and Tyr for Phe near BChlL (M208). Biochemistry 40: 12132–12139 Kirmaier C, Laible PD, Czarnecki K, Hata AN, Hanson DK, Bocian DF and Holten D (2002) Comparison of M-side electron transfer in Rba. sphaeroides and Rba. capsulatus reaction centers. J Phys Chem B 106: 1799–1808 Kirmaier C, Laible PD, Hanson DK and Holten D (2004) B-side electron transfer to form P+HB– in reaction centers from the F(L181)Y/Y(M208)F mutant of Rhodobacter capsulatus. J Phys Chem B 108: 11827–11832 Kitzing E and Kühn H (1990) Primary electron transfer in photosynthetic reaction centers. J Phys Chem 94: 1699–1702 Kolbasov D and Scherz A (1998) Matrix elements play a significant role in asymetric electron transfer in bacterial reaction centers. In: Garab G (ed) Photosynthesis: Mechanisms and Effects, Vol II, pp 719–722. Kluwer Academic Publishers, Dordrecht Kolbasov D and Scherz A (2000) Asymmetric electron transfer in reaction centers of purple bacteria strongly depends on different electron matrix elements in the active and inactive branches. J Phys Chem B 104: 1802–1809
374 Kubo R (1966) The fluctuation-dissipation theorem. Rept Progr Theor Phys 29: 255–284 Kubo R and Toyozawa Y (1955) Application of the method of generating function to radiative and non-radiative transitions of a trapped electron in a crystal. Prog Theor Phys 13: 160–182 Lauterwasser C, Finkele U, Scheer. H and Zinth W (1991) Temperature dependence of the primary electron transfer in photosynthetic reaction centers from Rhodobacter sphaeroides. Chem Phys Lett 183: 471–477 Leggett AJ, Chakravarty S, Dorsey AT, Fisher MPA, Garg A and Zwerger W (1987) Dynamics of the dissipative two-state system. Rev Mod Phys 59: 1–85 Lin S, Taguchi AKW and Woodbury NW (1996a) Excitation wavelength dependence of energy transfer and charge separation in reaction centers from Rhodobacter sphaeroides: Evidence for adiabatic electron transfer. J Phys Chem 100: 17067–17078 Lin S, Xiao W, Eastman JE, Taguchi AKW and Woodbury NW (1996b) Low-temperature femtosecond-resolution transient absorption spectroscopy of large-scale symmetry mutants of bacterial reaction centers. Biochemistry 35: 3187–3196 Lin S, Jackson JA, Taguchi AKW and Woodbury NW (1999) B-side electron transfer promoted by absorbance of multiple photons in Rhodobacter sphaeroides R-26 reaction centers. J Phys Chem B 103: 4757 –4763 Lin S, Katilius E, Haffa ALM, Taguchi AKW and Woodbury NW (2001) Blue light drives B-side electron transfer in bacterial photosynthetic reaction centers. Biochemistry 40: 13767 –13773 Lyle PA, Kolaczkowski SV and Small GJ (1993) Photochemical hole-burned spectra of protonated and deuterated reaction centers of Rhodobacter sphaeroides. J Phys Chem 97: 6924–6933 Mar T and Gingras G (1990) Relative phototrapping rates of the two bacteriopheophytins in the photoreaction center of Ectothiorhodospira sp. Biochim Biophys Acta 1017: 112–117 Marchi M, Gehlen JN, Chandler D and Newton M (1993) Diabatic surfaces and the pathway for primary electron transfer in a photosynthetic reaction center. J Am Chem Soc 115: 4178–4190 Marcus RA (1956) On the theory of oxidation-reduction reactions involving electron-transfer. I. J Chem Phys 24: 966–978 Marcus RA (1964) Chemical and electrochemical electron transfer theory. Ann Rev Phys Chem 15: 155–196 Marcus RA (1993) Electron-transfer reactions in chemistry— theory and experiment (Nobel lecture). Angew Chem Int Ed Engl 32: 1111–1121 Martin J-L, Breton J, Hoff AJ, Migus A and Antonetti A (1986) Femtosecond spectroscopy of electron transfer in the reaction center of the photosynthetic bacterium Rhodopseudomonas sphaeroides R-26: Direct electron transfer from the dimeric bacteriochlorophyll primary donor to the bacteriopheophytin acceptor with a time constant of 2.8 ± 0.2 psec. Proc Natl Acad Sci USA 83: 957–961 Martin J-L and Vos MH (1992) Femtosecond biology. Ann Rev Biophys Biomol Struct 21: 199–222 Mercer IP, Gould IR and Klug DR (1999) A quantum mechanical/molecular mechanical approach to relaxation dynamics: Calculation of the optical properties of solvated bacteriochlorophyll-a. J Phys Chem B 103: 7720–7727 Middendorf TR, Mazzola LT, Gaul DF, Schenck CC and Boxer SG (1991) Photochemical hole-burning spectroscopy of a
William W. Parson and Arieh Warshel photosynthetic reaction center mutant with altered charge separation kinetics: Properties and decay of the initially excited state. J Phys Chem 95: 10142–10151 Miller WH and George TF (1972) Semiclassical theory of electronic transitions in low energy atomic and molecular collisions involving several nuclear degrees of freedom. J Chem Phys 56: 5637–5652 Moser CC, Keske JM, Warncke K, Farid RS and Dutton PL (1992) Nature of biological electron transfer. Nature 355: 796–802 Murchison HA, Alden RG, Allen JP, Peloquin JM, Taguchi AK, Woodbury NW and Williams JC (1993) Mutations designed to modify the environment of the primary electron donor of the reaction center from Rhodobacter sphaeroides: Phenylalanine to leucine at L167 and histidine to phenylalanine at L168. Biochemistry 32: 3498–3505 Nagarajan V, Parson WW, Gaul D and Schenck C (1990) Effect of directed mutations of the tyrosine at site (M)210 on the primary photosynthetic electron transfer process in Rhodobacter sphaeroides. Proc Natl Acad Sci USA 87: 7888–7892 Nagarajan V, Parson WW, Davis D and Schenck CC (1993) Kinetics and free energy gaps of electron-transfer reactions in Rhodobacter sphaeroides reaction centers. Biochemistry 32: 12324–12336 Newton MD (2003) Electronic coupling in electron transfer and the influence of nuclear modes: theoretical and computational probes. Theor Chem Acc 110: 307–321 Ogrodnik A (1990) The free energy difference between the excited primary donor 1P* and the radical pair state P+H– in reaction centers of Rhodobacter sphaeroides. Biochim Biophys Acta 1020: 65–71 Ogrodnik A, Volk M, Letterer R, Feik R and Michel-Beyerle M-E (1988) Determination of free energies in reaction centers of Rb. sphaeroides. Biochim Biophys Acta 936: 361–371 Ogrodnik A, Keupp W, Volk M, Auermeier G and Michel-Beyerle ME (1994) Inhomogeneity of radical pair energies in photosynthetic reaction centers revealed by differences in recombination dynamics of P+HA– when detected in delayed emission and in absorption. J Phys Chem 98: 3432–3439 Ogrodnik A, Hartwich G, Lossau H and Michel-Beyerle ME (1999) Dispersive charge separation and conformational cooling of P+HA– in reaction centers of Rb. sphaeroides R26: A spontaneous emission study. Chem Phys 244: 461–478 Okada A (2000) Qualitative features of electron transfer reaction for high viscosity of solvent and low activation barrier. J Phys Chem A 104: 7744–7750 Parson WW and Warshel A (2004a) A density-matrix model of photosynthetic electron transfer with microscopically estimated vibrational relaxation times. Chem Phys 296: 201–206 Parson WW and Warshel A (2004b) Dependence of photosynthetic electron-transfer kinetics on temperature and energy in a density-matrix model. J Phys Chem B 108: 10474–10483 Parson WW and Warshel A (2006) Calculations of electrostatic energies in proteins using microscopic, semimicroscopic and macroscopic models and free-energy perturbation approaches. In: Aartsma TJ and Matysik J (eds) Biophysical Techniques in Photosynthesis, Vol 2. in press. Springer, Dordrecht Parson WW, Clayton RK and Cogdell RJ (1975) Excited states of photosynthetic reaction centers at low redox potentials. Biochim Biophys Acta 387: 265–278 Parson WW, Creighton S and Warshel A (1987) Calculations of spectroscopic properties and electron transfer kinetics of
Chapter 19
Mechanism of Charge Separation
photosynthetic bacterial reaction centers. In: Kobayashi T (ed) Primary Processes in Photobiology, pp 43–51. SpringerVerlag, Berlin Parson WW, Chu ZT and Warshel A (1990) Electrostatic control of charge separation in bacterial photosynthesis. Biochim Biophys Acta 1017: 251–272 Parson WW, Chu ZT and Warshel A (1998) Reorganization energy of the initial electron-transfer step in photosynthetic bacterial reaction centers. Biophys J 74: 182–191 Paschenko VZ, Korvatovskii BN, Kononenko AA, Chamorovsky SK and Rubin AB (1985) Estimation of the rate of photochemical charge separation in Rhodopseudomonas sphaeroides reaction centers by fluorescence and absorption picosecond spectroscopy. FEBS Lett 191: 245–248 Peloquin JM, Williams JC, Lin X, Alden RG, Taguchi AKW, Allen JP and Woodbury NW (1994) Time-dependent thermodynamics during early electron transfer in reaction centers from Rhodobacter sphaeroides. Biochemistry 33: 8089–8100 Peloquin JM, Lin S, Taguchi AKW and Woodbury NW (1995) Excitation wavelength dependence of bacterial reaction center photochemistry. 1. Ground state and excited state evolution. J Phys Chem 99: 1349–1356 Peloquin JM, Lin S, Taguchi AKW and Woodbury NW (1996) Excitation wavelength dependence of bacterial reaction center photochemistry. 2. Low-temperature measurements and spectroscopy of charge separation. J Phys Chem 100: 14228–14235 Pollard WT, Felts AK and Friesner RA (1996) The Redfield equation in condensed-phase quantum dynamics. Adv Chem Phys 93: 77–134 Procacci P, Darden TA, Paci E and Marchi M (1997) ORAC: A molecular dynamics program to simulate complex molecular systems with realistic electrostatic interactions. J Comp Chem 18: 1848–1862 Rademaker J and Hoff AJ (1981) The balance between primary forward and back reaction in bacterial photosynthesis. Biophys J 34: 325–344 Reddy NRS, Lyle PA and Small GJ (1992) Applications of spectral hole burning spectroscopies to antenna and reaction center complexes. Photosynth Res 31: 167–194 Redfield AG (1965) The theory of relaxation processes. Adv Magn Res 1: 1–32 Rips I and Jortner J (1988) Activationless solvent-controlled electron transfer. J Chem Phys 88: 818–822 Rischel C, Spiedel D, Ridge JP, Jones MR, Breton J, Lambry JC, Martin J-L and Vos MH (1998) Low frequency vibrational modes in proteins: Changes induced by point-mutations in the protein-cofactor matrix of bacterial reaction centers. Proc Natl Acad Sci USA 95: 12306–12311 Roberts JA, Holten D and Kirmaier C (2001) Primary events in photosynthetic reaction centers with multiple mutations near the photoactive electron carriers. J Phys Chem B 105: 5575–5584 Robles SJ, Breton J and Youvan DC (1990a) Partial symmetrization of the photosynthetic reaction center. Science 248: 1402–1405 Robles SJ, Breton J and Youvan DC (1990b) Transmembrane helix exchanges between quasi-symmetric subunits of the photosynthetic reaction center. In: Michel-Beyerle M-E (ed) Reaction Centers of Photosynthetic Bacteria, pp 283–291. Springer, Berlin
375 Rockley MG, Windsor MW, Cogdell RJ and Parson WW (1975) Picosecond detection of an intermediate in the photochemical reaction of bacterial photosynthesis. Proc Natl Acad Sci USA 72: 2251–2255 Sarai A (1980) Possible role of protein in photosynthetic electron transfer. Biochim Biophys Acta 589: 71–82 Schenck CC, Parson WW, Holten D, Windsor MW and Sarai A (1981) Temperature dependence of electron transfer between bacteriopheophytin and ubiquinone in protonated and deuterated reaction centers of Rhodopseudomonas sphaeroides. Biophys J 36: 479–489 Schenck CC, Blankenship RE and Parson WW (1982) Radical-pair decay kinetics, triplet yields, and delayed fluorescence from bacterial reaction centers. Biochim Biophys Acta 680: 44–59 Scherer POJ and Fischer SF (1989) Long-range electron transfer within the hexamer of the photosynthetic reaction center Rhodopseudomonas viridis. J Phys Chem 93: 1633–1637 Scherer POJ and Fischer SF (1998) Charge transfer states of the reaction center. Spectrochim Acta A 54A: 1191–1199 Schmidt S, Arlt T, Hamm P, Huber H, Naegele T, Wachtveitl J, Meyer M, Scheer H and Zinth W (1994) Energetics of the primary electron transfer reaction revealed by ultrafast spectroscopy on modified bacterial reaction centers. Chem Phys Lett 223: 116–120 Schmidt S, Arlt T, Hamm P, Huber H, Nägele T, Wachtveitl J, Meyer M, Scheer H and Zinth W (1995) Primary electron-transfer dynamics in modified bacterial reaction centers containing pheophytin-a instead of bacteriopheophytin-a. Spectrochim Acta A 51A: 1565–1578 Schulten K and Tesch M (1991) Coupling of protein motion to electron transfer: Molecular dynamics and stochastic quantum mechanics study of photosynthetic reaction centers. Chem Phys 158: 421–446 Sham YY and Warshel A (1998) The surface constrained all atom model provides size-independent results in calculations of hydration free energies. J Chem Phys 109: 7940–7944 Shkuropatov AY and Shuvalov VA (1993) Electron transfer in pheophytin a-modified reaction centers from Rhodobacter sphaeroides (R-26). FEBS Lett 322: 168–172 Shochat S, Arlt T, Francke C, Gast P, van Noort PI, Otte SC, Schelvis HPM, Schmidt S, Vijgenboom E, Vrieze J, Zinth W and Hoff AJ (1994) Spectroscopic characterization of reaction centers of the (M)Y210W mutant of the photosynthetic bacterium Rhodobacter sphaeroides. Photosynth Res 40: 55–66 Shuvalov VA and Klimov VV (1976) The primary photoreactions in the complex cytochrome-P-890 . P-760 (bacteriopheophytin760) of Chromatium minutissimum at low redox potentials. Biochim Biophys Acta 440: 587–599 Shuvalov VA and Parson WW (1980) Energies and kinetics of radical pairs involving bacteriochlorophyll and bacteriopheophytin in bacterial reaction centers. Proc Natl Acad Sci USA 78: 957–961 Shuvalov VA and Yakovlev AG (2003) Coupling of nuclear wavepacket motion and charge separation in bacterial reaction centers. FEBS Lett 540: 26–34 Sim E and Makri N (1997) Path integral simulation of charge transfer dynamics in photosynthetic reaction centers. J Phys Chem B 101: 5446 –5458 Skourtis SS, Balabin IA, Kawatsu T and Beratan DN (2005) Protein dynamics and electron transfer: Electronic decoherence and non-Condon effects. Proc Natl Acad Sci USA 102:
376 3552–3557 Slichter CP (1963) Principles of Magnetic Resonance with Examples from Solid State Physics. Harper & Row, New York Small GJ (1995) On the validity of the standard model for primary charge separation in the bacterial reaction center. Chem Phys 197: 239–257 Small GJ, Hayes JM and Silbey RJ (1992) The question of dispersive kinetics for the initial phase of charge separation in bacterial reaction centers. J Phys Chem 96: 7499–7501 Spörlein S, Zinth W and Wachtveitl J (1998) Vibrational coherence in photosynthetic reaction centers observed in the bacteriochlorophyll anion band. J Phys Chem B 102: 7492–7496 Spörlein S, Zinth W, Meyer M, Scheer H and Wachtveitl J (2000) Primary electron transfer in modified bacterial reaction centers: Optimization of the first events in photosynthesis. Chem Phys Lett 322: 454–464 Stanley RJ and Boxer SG (1995) Oscillations in spontaneous fluorescence from photosynthetic reaction centers. J Phys Chem 99: 859–863 Steffen MA, Lao KQ and Boxer SG (1994) Dielectric asymmetry in the photosynthetic reaction center. Science 264: 810–816 Sumi H and Marcus RA (1986) Dynamical effects in electron transfer reactions. J Chem Phys 84: 4894–4914 Taguchi AKW, Stocker JW, Alden RG, Causgrove TP, Peloquin JM, Boxer SG and Woodbury NW (1992) Biochemical characterization and electron-transfer reactions of sym1, a Rhodobacter capsulatus reaction center symmetry mutant which affects the initial electron donor. Biochemistry 31: 10345–10355 Taguchi AKW, Eastman JE, Gallo DM, Sheagley E, Xiao W and Woodbury NW (1996) Asymmetry requirements in the photosynthetic reaction center of Rhodobacter capsulatus. Biochemistry 35: 3175–3186 Takiff L and Boxer SG (1988) Phosphorescence from the primary electron donor in Rhodobacter sphaeroides and Rhodopseudomonas viridis reaction centers. Biochim Biophys Acta 932: 325–334 Thompson MA, Zerner MC and Fajer J (1991) A theoretical examination of the electronic structure and spectroscopy of the photosynthetic reaction center from Rhodopseudomonas viridis. J Am Chem Soc 113: 8210–8215 Trissl HW, Breton J, Deprez J, Dobek A and Leibl W (1990) Trapping kinetics, annihilation and quantum yield in the photosynthetic bacterium Rhodopseudomonas viridis as revealed by electric measurement of the primary charge separation. Biochim Biophys Acta 1015: 322–333 Tully JC and Preston RM (1971) Trajectory surface hopping approach to nonadiabatic molecular collisions: The reaction of H+ with D2. J Chem Phys 55: 562–572 van Brederode ME, Jones MR, van Mourik F, van Stokkum IHM and van Grondelle R (1997) A new pathway for transmembrane electron transfer in photosynthetic reaction centers of Rhodobacter sphaeroides not involving the excited special pair. Biochemistry 36: 6855–6861 van Brederode ME, van Mourik F, van Stokkum IHM, Jones MR and van Grondelle R (1999) Multiple pathways for ultrafast transduction of light energy in the photosynthetic reaction center of Rhodobacter sphaeroides. Proc Natl Acad Sci USA 96: 2054–2059 van Stokkum IHM, Beekman LMP, Jones MR, van Brederode ME and van Grondelle R (1997) Primary electron transfer kinetics in membrane-bound Rhodobacter sphaeroides reaction centers:
William W. Parson and Arieh Warshel A global and target analysis. Biochemistry 36: 11360–11368 Volk M, Aumeier G, Langenbacher T, Feick R, Ogrodnik A and Michel-Beyerle M-E (1998) Energetics and mechanism of primary charge separation in bacterial photosynthesis. A comparative study on reaction centers of Rhodobacter sphaeroides and Chloroflexus aurantiacus. J Phys Chem B 102: 735 –751 Vos M and Martin J-L (1999) Femtosecond processes in proteins. Biochim Biophys Acta 1411: 1–20 Vos MH, Lambry JC, Robles SJ, Youvan DC, Breton J and Martin J-L (1991) Direct observation of vibrational coherence in bacterial reaction centers using femtosecond absorption spectroscopy. Proc Natl Acad Sci USA 88: 8885–8889 Vos MH, Lambry JC, Robles SJ, Youvan DC, Breton J and Martin J-L (1992) Femtosecond spectral evolution of the excited state of bacterial reaction centers at 10 K. Proc Natl Acad Sci USA 89: 613–617 Vos MH, Rappaport F, Lambry J-H, Breton J and Martin J-L (1993) Visualization of coherent nuclear motion in a membrane protein by femtosecond spectroscopy. Nature 363: 320–325 Vos MH, Jones MR, Hunter CN, Breton J and Martin J-L (1994a) Coherent nuclear dynamics at room temperature in bacterial reaction centers. Proc Natl Acad Sci USA 91: 12701–12705 Vos MH, Jones MR, Hunter CN, Breton J, Lambry J-C and Martin J-L (1994b) Coherent dynamics during the primary electron-transfer reaction in membrane-bound reaction centers of Rhodobacter sphaeroides. Biochemistry 33: 6750–6757 Vos MH, Jones MR, Breton J, Lambry JC and Martin J-L (1996) Vibrational dephasing of long- and short-lived primary donor excited states in mutant reaction centers of Rhodobacter sphaeroides. Biochemistry 35: 2687–2692 Vos MH, Rischel C, Jones MR and Martin J-L (2000) Electrochromic detection of a coherent component in the formation of the charge pair P+HL– in bacterial reaction centers. Biochemistry 39: 8353 –8361 Wakeham MC and Jones MR (2005) Rewiring photosynthesis: Engineering wrong-way electron transfer in the purple bacterial reaction center. Biochem Soc Trans 133: 851–857 Walsh AM and Coalson RD (1992) Redfield theory is quantitative for coupled harmonic oscillators. Chem Phys Lett 198: 293–299 Wang HW, Lin S, Allen JP, Williams JC, Blankert S, Laser C and Woodbury NW (2007) Protein dynamics control the kinetics of initial electron transfer in photosynthesis. Science 316: 747–750 Wang Z, Pearlstein RM, Jia Y, Fleming GR and Norris JR (1993) Inhomogeneous electron-transfer kinetics in reaction centers of bacterial photosynthesis. Chem Phys 176: 421–425 Warshel A (1982) Dynamics of reactions in polar solvents. Semiclassical trajectory studies of electron-transfer and protontransfer reactions. J Phys Chem 86: 2218–2224 Warshel A and Hwang J-K (1986) Simulation of the dynamics of electron transfer reactions in polar solvents: semiclassical trajectories and dispersed polaron approaches. J Chem Phys 84: 4938–4957 Warshel A and Parson WW (1991) Computer simulations of electron transfer reactions in solution and photosynthetic reaction centers. Ann Rev Phys Chem 42: 279–309 Warshel A and Parson WW (2001) Dynamics of biochemical and biophysical reactions: insight from computer simulations. Quart Rev Biophys 34: 563–679 Warshel A and Russell ST (1984) Calculations of electrostatic
Chapter 19
Mechanism of Charge Separation
interactions in biological systems and in solutions. Quart Rev Biophys 17: 283–290 Warshel A, Creighton S and Parson WW (1988) Electron-transfer pathways in the primary event of bacterial photosynthesis. J Phys Chem 92: 2698–2701 Warshel A, Chu ZT and Parson WW (1989) Dispersed polaron simulations of electron transfer in photosynthetic reaction centers. Science 246: 112–116 Warshel A, Sharma PK, Kato M and Parson WW (2006) Modeling electrostatic effects in proteins. Biochim Biophys Acta 1764: 1647–1676 Wasielewski MR and Tiede DM (1986) Sub-picosecond measurements of primary electron transfer in Rhodopseudomonas viridis reaction centers using near-infrared excitation. FEBS Lett 204: 368–372 Williams JC, Alden RG, Murchison HA, Peloquin JM, Woodbury NW and Allen JP (1992) Effects of mutations near the bacteriochlorophylls in reaction centers from Rhodobacter sphaeroides. Biochemistry 31: 11029–11037 Woodbury NWT and Parson WW (1984) Nanosecond fluorescence from isolated reaction centers of Rhodopseudomonas sphaeroides. Biochim Biophys Acta 767: 345–361 Woodbury NW, Becker M, Middendorf D and Parson WW (1985) Picosecond kinetics of the initial photochemical electrontransfer reaction in bacterial photosynthetic reaction centers. Biochemistry 24: 7516–7521 Woodbury NW, Parson WW, Gunner MR, Prince RC and Dutton PL (1986) Radical-pair energetics and decay mechanisms in reaction centers containing anthraquinones, naphthoquinones or benzoquinones in place of ubiquinone. Biochim Biophys Acta 851: 6–22 Woodbury NW, Peloquin JM, Alden RG, Lin X, Taguchi AKW, Williams JC and Allen JP (1994) Relationship between thermodynamics and mechanism during photoinduced charge separation in reaction centers from Rhodobacter sphaeroides. Biochemistry 33: 8101–8112
377 Wraight CA and Clayton RK (1974) The absolute quantum efficiency of bacteriochlorophyll photooxidation in reaction centres of Rhodopseudomonas spheroides. Biochim Biophys Acta 333: 246–260 Yakovlev AG and Shuvalov VA (2000) Formation of bacteriochlorophyll anion band at 1020 nm produced by nuclear wavepacket motion in bacterial reaction centers. J Chin Chem Soc 47: 1–6 Yakovlev AG, Shkuropatov AC and Shuvalov AV (2000) Nuclear wavepacket motion producing a reversible charge separation in bacterial reaction centers. FEBS Lett 466: 209–212 Yakovlev AG, Shkuropatov AY and Shuvalov VA (2002) Nuclear wavepacket motion between P* and P+BA– potential surfaces with subsequent electron transfer to HA in bacterial reaction centers. 1. Room temperature. Biochemistry 41: 2667–2674 Yakovlev AG, Vasilieva LG, Shkuropatov AY, Bolgarina TI, Shkuropatova VA and Shuvalov VA (2003) Mechanism of charge separation and stabilization of separated charges in reaction centers of Chloroflexus aurantiacus and of YM210W(L) mutants of Rhodobacter sphaeroides excited by 20 fs pulses at 90 K. J Phys Chem A 107: 8330–8338 Zhang LY and Friesner RA (1998) Ab initio calculation of electronic coupling in the photosynthetic reaction center. Proc Natl Acad Sci USA 95: 13603–13605 Zhou H and Boxer SG (1998a) Probing excited-state electron transfer by resonance Stark spectroscopy. 1. Experimental results for photosynthetic reaction centers. J Phys Chem B 102: 9139–9147 Zhou H and Boxer SG (1998b) Probing excited-state electron transfer by resonance Stark spectroscopy. 2. Theory and application. J Phys Chem B 102: 9148–9160 Zusman LD and Beratan DN (1998) Electron transfer in the photosynthetic reaction center: Mechanistic implications of mutagenesis studies. Spectrochim Acta A 54A: 1211–1218
Chapter 20 The Acceptor Quinones of Purple Photosynthetic Bacteria— Structure and Spectroscopy Colin A. Wraight* Department of Biochemistry, and Center for Biophysics & Computational Biology, University of Illinois, Urbana, IL 61801 U.S.A.
Marilyn R. Gunner Department of Physics, City College of New York, New York, NY 10031 U.S.A.
Summary ............................................................................................................................................................... 379 I. Introduction..................................................................................................................................................... 380 II. The Acceptor Quinone Reactions .................................................................................................................. 382 III. Structural Features of the Acceptor Quinone Binding Sites ........................................................................... 383 A. The Primary Quinone, QA ................................................................................................................ 384 1. The Methoxy Groups of QA .................................................................................................... 384 B. The Secondary Quinone, QB ........................................................................................................... 385 1. The Two QB Binding Sites Found in the X-ray Structures ....................................................... 387 2. The Methoxy Groups of QB .................................................................................................... 388 C. Hydrogen Bonding in the Quinone Binding Sites ............................................................................. 389 IV. Spectroscopy of the Acceptor Quinones ........................................................................................................ 389 A. The Primary Quinone, QA .................................................................................................... 390 B. The Reduced Primary Quinone, QA– ................................................................................................ 392 1. Magnetic Resonance Studies of QA– ....................................................................................... 393 2. Infra-red Spectroscopic Studies of QA– .................................................................................... 394 C. The Secondary Quinone, QB ........................................................................................................... 395 D. The Semireduced Secondary Quinone, QB– .................................................................................... 395 1. Magnetic Resonance Studies of QB– ....................................................................................... 395 2. Vibrational Spectroscopy of QB– .............................................................................................. 396 V. Functionality of the Two Quinone Positions in the QB Site ............................................................................. 396 VI. Conclusions .................................................................................................................................................... 398 Note Added in Proof .............................................................................................................................................. 398 Acknowledgments ................................................................................................................................................. 398 References ............................................................................................................................................................ 399
Summary Type II reaction centers (RCs) have two acceptor quinones that act in series. The primary quinone, QA, cycles between the oxidized quinone and singly reduced semiquinone. QA is tightly bound to the protein as a prosthetic group. The secondary quinone, QB, is reduced by QA–, first to the semiquinone and then to the doubly reduced, fully protonated quinol, QH2. QB freely associates with the protein in the quinone and quinol states. The prop*Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 379–405. © 2009 Springer Science + Business Media B.V.
Colin A. Wraight and Marilyn R. Gunner
380
erties of the two quinones that facilitate this process are largely determined by the nature of the two quinone binding sites. Many reaction center crystal structures show these interactions, although there are significant uncertainties in the conformations of the two quinones, and in the significance of the variable location of QB in the protein. Consequently the influence of structure on quinone function is only very crudely understood. These issues are discussed with emphasis on the quinone reactions in the reaction center from the photosynthetic bacterium, Rhodobacter (Rba.) sphaeroides, which is the best characterized. The structural features of the quinones and their local protein environments are examined in the light of extensive spectroscopic studies, especially by Fourier transform infra-red spectroscopy (FTIR), electron paramagnetic resonance (EPR) and electron nuclear double resonance (ENDOR), on the quinones in their functional redox states. I. Introduction The initial events of photosynthesis, in plants and bacteria, consist of electron transfer from an excited state of (bacterio)chlorophyll, followed by stabilization of the separated charges. This is achieved with retention of a significant portion of the photon free energy. The reaction sequence occurs within a reaction center (RC), a membrane pigment-protein of varying complexity in different species but in all cases based on a homologous pattern of cofactors and protein arrangements with strong two-fold symmetry. There appear to be only two basic types of RC, defined by the nature of their terminal electron acceptor species — Type I and Type II (Heathcote et al., 2002). Type I RCs use iron sulfur complexes to transfer electrons from an intermediate acceptor quinone to a soluble ferredoxin, for export of reducing equivalents one at a time. They operate at low redox potentials and are found in green and brown sulfur bacteria (Chlorobiaceae, and Heliobacteriaceae), and in Photosystem I of oxygenic bacteria (Cyanobacteria) and plants. Type II RCs use the asymmetric function of two acceptor quinones to export reducing equivalents in pairs, as quinol. They operate at higher redox potentials and are found in all purple bacteria (sulfur-oxidizing Chromatiaceae and nonsulfur Rhodobacteraceae, etc), in green, filamentous bacteria (Chloroflexaceae), and in the water-oxidizing Photosystem II of oxygenic photosynthesis. In purple photosynthetic bacteria, the functional core of the RC complex is a heterodimer of similar, Abbreviations: BChl – bacteriochlorophyll; Bphe – bacteriopheophytin; Blc. – Blastochloris; DFT – density functional theory; ENDOR – Electron Nuclear Double Resonance; EPR – Electron Paramagnetic Resonance; FTIR – Fourier Transform Infrared Spectroscopy; IR – infra-red; NMR – Nuclear Magnetic Resonance; QA,B – primary and secondary quinone acceptors; Rba. – Rhodobacter; RC – reaction center; UQ – ubiquinone (2,3-dimethoxy -5-methyl-6-isoprenyl-1,4-benzoquinone)
but non-identical, L and M subunits1 that bind all the active cofactors: four bacteriochlorophylls (BChl), two bacteriopheophytins (BPhe), two quinones and an iron atom. A third polypeptide, the H-subunit, contributes stability to the whole complex and is involved in proton (H+-ion) binding and transfer processes coupled to the electron transfer between the two quinones. The arrangement of cofactors and protein subunits is now well known from the X-ray structures of RC complexes from Blastochloris (Blc.) (formerly Rhodopseudomonas) viridis (Deisenhofer et al., 1985; Deisenhofer and Michel, 1989b, 1991; Lancaster and Michel, 1996a,b), Rba. sphaeroides (Allen et al., 1988; Chang et al., 1991; Ermler et al., 1994; Stowell et al., 1997), and Thermochromatium tepidum (Nogi et al., 2000) (Fig. 1). The A and B cofactor branches show a high degree of rotational (C2) symmetry, as do the surrounding L and M subunits, which also have substantial sequence homology. However, electron transfer from the primary donor, P, is not symmetrical and occurs exclusively via BChlA and BPheA on the L-subunit, to the primary quinone, QA, and then to the secondary quinone, QB. The function of Type II RCs (hereafter, just RC) is to produce fully reduced quinol (or hydroquinone, QH2), as a reductant to drive electron transfers in other membrane-bound enzymes. This results in the formation of an electrochemical gradient of protons across the membrane that drives the chemiosmotic processes of ATP synthesis, ion and substrate transport, flagellar rotation, and reversed electron transport. To produce quinol, the acceptor quinones of RCs, QA and QB, act in series to accumulate two reducing equivalents from the one-electron photoreactions, with two protons taken up from the aqueous solution on the cytoplasmic side of the membrane. After two RC turnovers, QH2 is released from the QB site, which is then refilled by an oxidized quinone to complete the 1
The L and M polypeptides that bind QA and QB are 33.9% identical in Rba. sphaeroides RCs (Williams et al., 1986)
Chapter 20
Acceptor Quinones of Purple Photosynthetic Bacteria
Fig. 1. The reaction center (RC) complex from Rhodobacter sphaeroides comprises three subunits, a heterodimer of similar, but non-identical L and M subunits, and subunit H, which caps LM on the cytosolic side of the membrane. The LM dimer binds all the cofactors, while subunit H stabilizes the structure and is involved in H+-ion uptake and transfer associated with electron transfer to the quinones. The L and M subunits and all associated cofactors are arranged around a quasi-2-fold rotational symmetry axis, normal to the plane of the membrane and passing through the primary donor (P), the special pair dimer of bacteriochlorophylls (BChl), and a ferrous (Fe2+) iron midway between the two quinones. Electron transfer proceeds from the excited singlet state of the primary donor (P*), via the A-branch of cofactors — monomer BChl (BA) and BPhe (HA), bound to the L subunit — to the primary quinone, QA, which is bound in a fold of the M subunit. From QA– the electron crosses the symmetry axis to the secondary quinone, QB, bound in a similar fold in the L subunit. (Figure prepared in VMD.) See also Color Plate 8, Fig. 12.
acceptor quinone cycle (reviewed in Okamura and Feher, 1992, 1995; Shinkarev and Wraight, 1993; Okamura et al., 2000; Paddock et al., 2003; Wraight, 2004, 2005) (Fig. 2). The reduction of quinone to quinol in the QB site relies on significant functional differences between the two quinones (Wraight, 1982, 2004; Shinkarev and Wraight, 1993). QA interfaces between the subnanosecond one-electron transfers needed for initial charge stabilization and the micro- to millisecond reduction of QB, where two electrons are accumulated to produce the stable product, QH2. QA is a tightly bound, one-electron redox couple, operating between the quinone and semiquinone forms, while QB is a reversibly bound, two-electron couple, with
381
Fig. 2. The acceptor quinone cycle of Type II reaction centers. Following each flash, an electron is transferred to the acceptor quinones. (The photooxidized primary donor, P+, is re-reduced by secondary donor (Cyt c2 in vivo in bacteria), and is omitted in the scheme). After the first flash, the anionic semiquinone charge induces pKa changes in some ionizable amino acid residues and H+ are taken up to the protein. This subsequently provides the 2nd proton, H+(2), transferred to QB, after the second flash. After a second flash, the transfer of the second electron from QA– only takes place after QB– has been protonated.
quinone, semiquinone and quinol states all involved in turnover. QA– reduces both the quinone and semiquinone in the QB site. Furthermore, in many species, including Rba. sphaeroides, QA and QB are chemically identical. For both reduction steps to be energetically favorable, significant alterations are required in the quinone physico-chemical properties compared to their behavior in solution. These distinct properties must be imparted by interactions with the protein in their respective binding sites. Thus, the structural and compositional symmetry of the RC, which extends to all the cofactors, contrasts with its functional asymmetry, making the RC a powerful model for studying the molecular basis of protein function. In this chapter, we summarize current knowledge of the structural features of the RC acceptor quinone complex, derived from X-ray crystallography and spectroscopy. We focus on the RC from Rba. sphaeroides, but some reference will be made to other species where useful comparisons can be made.
382
Colin A. Wraight and Marilyn R. Gunner
II. The Acceptor Quinone Reactions When the RC is activated by light, an electron is transferred in about 200 ps from the excited singlet state of the primary donor (P*), via BChlA and BPheA in the active path, to QA, and thence to QB, if available, in ≈100 µs (Fig. 3). Light activation generates the charge separated state, P+QA–, even at temperatures as low as 1 K (Arnold and Clayton, 1960) — and at a rate that is actually three-fold faster at cryogenic temperatures (Schenck et al., 1981; Woodbury et al., 1985; Kirmaier and Holten, 1990; Woodbury and Allen, 1995). Thus, the changes in protein and quinone conformation needed to form P+QA– can be accomplished in a frozen sample, without the full relaxation of the product that would occur at room temperature (McMahon et al., 1998; Xu and Gunner, 2000). The QA site has been shown to bind a wide range of quinones and quinonoid compounds, many of which reconstitute function, and such studies provided some of the first clues to the nature of the QA binding site (Okamura et al., 1975; Gunner et al., 1985, 1986; Woodbury et al., 1986; Warncke et al., 1987; Warncke and Dutton, 1993b; Labahn et al., 1995; Graige et al., 1998; Schmid et al., 1999; Hucke et al., 2002). For example, compounds as large as substituted anthraquinones are able to bind with good affinity, but double methyl-substitution at the 5 and 8 positions results in failure to bind (Gunner et al., 1985). Many non-quinone compounds also bind, even if functionally inactive, and reveal additional aspects of hydrogen bonding requirements of the QA site (Gunner et al., 1985; Warncke and Dutton, 1992, 1993a,b). Quinone reconstitution has also been used to vary the electrochemistry of QA, modifying the free energy of the QA dependent electron transfer reactions (see Fig. 3), including P+H–QA → P+HQA– (Gunner and Dutton, 1989), P+HQA– → PHQA (Gunner and Dutton, 1989; Franzen et al., 1990; Lin et al., 1994), P+HQA– QB → PHQAQB– (Li et al., 2000) and P+HQAQB– → PHQAQB (Allen et al., 1998; Schmid and Labahn, 2000), where H is the BPhe electron acceptor. Together with studies on mutant RCs, all of these reaction rates have been shown to vary as described by Marcus electron transfer theory (Gunner and Dutton, 1989; Moser et al., 1992; Lin et al., 1994; Schmid and Labahn, 2000). Thus, in these reactions, where the electron transfer process itself is rate determining, the rate is maximal at a free energy (–∆G°) equal to the reorganization
Fig. 3. Electron transfer pathways in Rba. sphaeroides reaction centers with no external electron donor, i.e., the fate of the first electron. Dashed lines show routes for the uphill thermal back reactions (charge recombination). Following light absorption, the singlet excited state of P (P*) reduces H, the A-branch bacteriopheophytin, in 3 ps. The electron is transferred to QA in 200 ps and then to QB in 5–200 µs. At each stage in the reaction cycle there are competing charge recombination reactions that transfer the electron back to P+. If there is no QA present, this will occur from H– in ≈10 ns (kHP is 7 × 107 s–1). Once the electron reaches the quinones it can return to P+ either by direct tunneling (kAP in P+QA– or kBP in P+QB– RCs) or via a thermal back reaction, e.g., kAHP is the rate of P+QA– charge recombination via P+H–QA, while kBAP represents the back reaction from P+QAQB– via P+QA–QB; the obs obs or kBP are the sum of the rates by all routes observed rates kAP (see Shinkarev, 1993 for a detailed description). In native RCs, the back reaction from UQA– (i.e., when there is no QB) occurs obs obs ≈ kAP is 10 s–1), and from UQB– in ≈1 s (kBP ≈ in ≈100 ms (kAP kBAP is 1 s–1).
energy, λ, and it slows as the driving force either increases or decreases (DeVault, 1980; Marcus, 1964; Marcus and Sutin, 1985; Moser et al., 1992). In contrast to QA, QB cannot easily be reconstituted by other quinones. Comparative studies indicate that the affinity of substituted benzoquinones and naphthoquinones bind to the QB site 7–30-fold weaker than to the QA site (McComb et al., 1990; Warncke et al., 1994; X. Zhang and M. R. Gunner, unpublished). In
Chapter 20
Acceptor Quinones of Purple Photosynthetic Bacteria
some part, the limited range of effective substitution is due to this weaker binding and the low solubility of many analogs of interest, but an important energetic constraint is also evident. No quinones other than ubiquinone (UQ – 2,3-dimethoxy-5-methyl6-isoprenyl-1,4-benzoquinone) and rhodoquinone (2-methoxy-3-amino-5-methyl-6-isoprenyl-1,4-benzoquinone) have been found to effectively replace both QA and QB function simultaneously in the same protein (McComb et al., 1990; Graige et al., 1999; X. Zhang and M. R. Gunner, unpublished). However, several studies have shown that replacing QA by a quinone with a midpoint potential (Em) lower than UQA reveals functional electron transfer to some non-native quinones as QB (Giangiacomo and Dutton, 1989; Graige et al., 1996; Li et al., 1998; Li et al., 2000; X. Zhang and M. R. Gunner, unpublished). Thus, the modulation of redox potential to allow favorable forward electron transfer is peculiar to UQ. Only the head group is required as the tailless 2,3-dimethoxy, 5-methyl benzoquinone (UQ-0) is active in homologous electron transfer (McComb et al., 1990; Warncke et al., 1994). The electron transfer reaction from QA– to QB is the first in the forward sequence to exhibit significant activation enthalpy, slowing as the temperature is lowered (Mancino et al., 1984), with no reaction seen below 200 K (Kleinfeld et al., 1984a; Xu and Gunner, 2001). Thus, there are rearrangements of the quinone or protein that cannot be accomplished at low temperature before QA– returns its electron to P+, in ≈100 ms. However, when the RCs are illuminated while freezing, they are trapped in the charge separated state P+QB–, which slowly undergoes charge recombination to the ground state. These RCs are now capable of full QB reduction at temperatures as low as 50 K (Kleinfeld et al., 1984a; Xu and Gunner, 2001, 2002). Thus, sufficient rearrangements are frozen into the protein-quinone conformation, leaving a new ground state conformation where any necessary conformational changes can be achieved at low temperature. Room temperature experiments have been carried out with different quinones in the QA site reducing UQB. The electron transfer reaction is largely independent of the driving force ∆GAB° (Graige et al., 1998), providing support that the rate determining step for this reaction is a conformational change rather the electron transfer reaction itself. However, a fast component has been reported to exhibit ∆G°AB dependence (Li et al., 2000). On the basis of the
383
slower, ∆G°AB-independent kinetics, the reaction mechanism for QA– QB → QA QB– has been described as being ‘gated’, i.e., other changes such as in quinone position, protein conformation or protonation, rather then the electron transfer itself, are the rate determining step (Tiede et al., 1996; Graige et al., 1998; Li et al., 1998, 2000; Wraight, 2004). Ultimately, the equilibrium properties of QA and QB, and the rates, temperature and pH dependence, of their oxidation/reduction reactions should be understandable in terms of the structure and dynamics of the protein. III. Structural Features of the Acceptor Quinone Binding Sites X-ray crystallography has shown the acceptor quinones to be symmetrically positioned about a nonheme iron atom (FeII, S = 2), which is liganded by four histidines (residues L190, L230, M219, M266) and a glutamate (M234), giving a net charge of +1. In this ‘acceptor quinone complex’, each quinone is bound in the fold of an interhelical loop — QA in the loop between the D and E helices of the M subunit and QB in the symmetry related position between helices D and E of the L subunit. Both quinones are involved in rather similar hydrogen bond interactions between the two carbonyl groups and a His side chain and an amide backbone. Much of this has been reviewed, based on the original X-ray structures of Blc. viridis (PDB entry 1prc; Deisenhofer et al., 1985; Deisenhofer and Michel, 1989b) at a resolution of 2.8 Å, and Rba. sphaeroides (PDB entries 4rcr (Allen et al., 1987a,b) and 2rcr (Chang et al., 1991) for the carotenoidless strain R26, and 1yst (Arnoux et al., 1995) and 1pss (Chirino et al., 1994) for carotenoid-containing strains), at resolutions between 2.8–3.3 Å. There are currently 47 structures of Rba. sphaeroides RCs, 20 of them with resolution between 1.8 and 2.5 Å (as of March 2007). The quinone binding sites are deeply buried in bacterial RCs, but the protein environment is very rich in ionizable amino acids, with highly networked interactions. In the RC 33% of the titratable residues are buried, a proportion similar to that found in the average protein (Gunner et al., 2000; Kim et al., 2005). The QB site is separated from the aqueous phase by a region of unusually high charge density (Zhu and Karlin, 1996), comprising a web of more than 20 residues whose ionization is strongly coupled
384 together (Beroza et al., 1995; Sebban et al., 1995; Lancaster et al., 1996; Alexov and Gunner, 1999; Ishikita et al., 2003). This feature is not apparent in Photosystem II (pdbs: 1s5l and 2axt), which lacks an H subunit homolog, and provides access to the surface through a single residue, His 252 of the D1 subunit (homologous with L) (Ferreira et al., 2004; Loll et al., 2005; Krivanek et al., 2007). However, tight membrane appression in plants and bulky phycobilisomes in cyanobacteria may serve to restrict solvent accessibility in Photosystem II. One result of the internal buffering in the bacterial RC is that mutations made to modify the charge environment of the quinone often have far smaller impact than might be expected for simple charge removal (Alexov et al., 2000). A. The Primary Quinone, QA In bacterial Type II RCs, QA may be a menaquinone (e.g., Blc. viridis, Allochromatium vinosum, Thermochromatium tepidum) or a ubiquinone (e.g., Rba. sphaeroides, Rba. capsulatus and Rhodospirillum rubrum; Parson, 1978). In Rba. sphaeroides, QA is ubiquinone-10 (UQ-10) and the X-ray structures show it to be hydrogen bonded through both carbonyl oxygens. The average lengths of the two hydrogen bonds are almost identical: 2 O1 to the peptide NH of Ala M260 is 2.79±0.09 Å, and O4 to NδH of His M219 is 2.80±0.15 Å. In Blc. viridis, where QA is menaquinone-9 (MQ-9) (Shopes and Wraight, 1985; Gast et al., 1985), the hydrogen bond arrangement is essentially the same; the hydrogen bonds are longer, at 2.98 ± 0.14 Å and 2.84 ± 0.07 Å, respectively, although the difference is not statistically significant. The main features of the quinone binding sites in Rba. sphaeroides are shown in Fig. 4. For the QA site, the major contributions to the immediate binding site (atoms within 4 Å of the QA headgroup) are from residues (sidechain or backbone) Met M218, His M219, Thr M222, Ala M248, Ala M249, Trp M252, Asn M259, Ala M260, Thr M261, Met M262 and 2 Two atom numbering conventions are in common use for ubiquinone in the biochemical literature. One, based on benzoquinone as the parent compound, identifies C1 as the carbonyl carbon proximal to the isoprene chain (which is considered attached to C6) and the other carbonyl as C4. The other convention, which is used in several X-ray structures, is based on toluene as the parent compound and identifies the methylsubstituted carbon as C1. The carbonyl groups are then at C2 (equivalent to C4) and C5 (equivalent to C1). In this article the former convention is used.
Colin A. Wraight and Marilyn R. Gunner Ile M265, with single atom contacts from Leu M215 and Ile M223 in some structures. Trp M252, which has been implicated in the rapid electron transfer from BPheA– to QA (Coleman et al., 1990b; Stilz et al., 1994), is in close contact and almost parallel to the quinone plane, and is held in this position by a hydrogen bond between the indole nitrogen and the OH of Thr M222 (Stilz et al., 1994). Also, Phe M251 is perpendicular to the Trp ring in a classic aromatic bond configuration (Burley and Petsko, 1985). On the other side of the quinone ring, Ile M265 points directly at the methoxy C3M, the C4 carbonyl, C5 and the C5´-methyl group. The first two isoprene units of the tail of QA are contacted by Leu M214, Leu M215, Met M218, Met M256, Phe M258, Trp M268 and Met M272. Only the first two isoprene units are well packed in the protein. Thus, the first isoprene methyl (C10) has 6.6 ± 1.7 protein atom neighbors within 4.0 Å while the second methyl (C15) has only 1.1 ± 0.7 neighbors. The extended tail can be seen lying outside the protein structure. The tail kinks to bend around Met M256 as it passes through the binding channel (Warncke et al., 1994). The first 2 isoprene units are also bounded by Trp M268 and Met M218. The lack of interaction of the later isoprene groups is seen by the affinity becoming independent of tail length after the third isoprene unit (McComb et al., 1990; Warncke et al., 1994). 1. The Methoxy Groups of QA The environments of the two methoxys are quite different, including a significant difference in packing around them. There are 7.4 ± 1.2 atoms within 4 Å of C2M and 3.9 ± 1.4 within 4 Å of C3M. However, the environment is somewhat variable in the different structures. Thus, Met M262 is within 4 Å of C3M in 79% of the structures, His M219 in 74%, Ile M265 in 65%, and Ile M223 in 47% of the structures. Thr M221, Thr M222, and Ala M223 are in contact in less than 10% of the structures. Likewise, there is variability in the contacts for C2M with Ala M260 (94%), Thr M261 (74%), Ala M249 (65%), Ala M245 (32%), Ala M248 (21%) and Asn M259 (9%) within 4 Å in the different structures. Of the 34 structures with resolution ≤2.8 Å (see Table 1) there is a probable long hydrogen bond, with an average length of 3.5 Å, between the 2-methoxy oxygen (O2M) and the amide N of Ala M249, with 40% also having the amide N of Ala M260 within 4 Å. In some structures, the angles
Chapter 20
Acceptor Quinones of Purple Photosynthetic Bacteria
385
Fig. 4. The quinone binding site — QA (top) and QB (bottom) — are related by the two-fold, rotational symmetry axis of the reaction center, which passes through the iron atom of the acceptor quinone complex. Both quinones are hydrogen bonded through the carbonyl oxygens — hydrogen bonds shown in purple for QA (top) and in green for QB (bottom). The binding sites are predominantly hydrophobic, with the notable exception of GluL212, AspL213 and SerL223 in the QB site, which constitute terminal components of the H+ delivery pathway to QB. Note that the orientations of the methoxy groups of QA and QB do not follow the two-fold rotational symmetry. (Figure prepared in VMD.) See also Color Plate 8, Fig. 13.
are poor for both donors. In contrast, the 3-methoxy oxygen (O3M) has no available hydrogen bond donors within 4 Å. Although the methoxy dihedral angles are not definitive for either quinone, it is noteworthy that the orientation of the 3-methoxy of QA is the best established by the X-ray data, with a dihedral angle of –77 ± 8°, in spite of the lack of any hydrogen bond donor to the oxygen. The 2-methoxy, on the other hand, is less well defined, although it seems to have a weak hydrogen bonding partner within reach (peptide N of Ala M249 at about 3.5 Å). With a dihedral angle of 139 ± 25° it is somewhat closer to being in-plane, and a few structures have the 2-methoxy almost fully in-plane and potentially hydrogen bonded to both Ala M249 and Ala M260. B. The Secondary Quinone, QB In all purple bacteria known to date, QB is a ubi-
quinone (UQ-10 in Rba. sphaeroides), although it is menaquinone in the Type II RC of Chloroflexus aurantiacus (Bruce et al., 1982). The functionally reversible binding of QB leads to partial extraction during isolation, and the X-ray diffraction data show variable and incomplete occupancy of this site. In the initial X-ray structures of Blc. viridis and Rba. sphaeroides, QB occupancy was largely restored by soaking the crystals with short chain analogs (UQ-1 or UQ-2). The headgroup was then modeled into the QB pocket with H-bonds between O4 and Nδ of His L190 and between O1 and Oγ of Ser L223 (Allen et al., 1987a,b; Deisenhofer et al., 1985; Deisenhofer and Michel, 1989a; El-Kabbani et al., 1991). This has become known as the proximal position. Later, an independent, higher resolution Rba. sphaeroides structure (1pcr) revealed QB to be positioned quite differently, in a ‘distal’ position with the headgroup rotated almost 180° and 5 Å away from the original
Colin A. Wraight and Marilyn R. Gunner
386 Table 1. Structure files with resolution ≤ 2.8 Å.1 PDB Rhodobacter sphaeroides 1RZH 1AIJ 1YF6 1E6D 1RZZ 1DS8 1DV6 2GMR 1M3X 1FNP 1FNQ 1PCR 1RVJ 1SOO 1OGV 1L9B 1KBY 2BNS3 1MPS 1JGY 1JGZ 2BNP 1JGW 1UMX 1QOV 1RY5 2GNU 2BOZ 1DV3 1RG5 1AIG 1E14 1RQK 1F6N 1RGN Blastochloris viridis4 1DXR
R
QA
QB
Modifications & Additions2
1.8 2.2 2.3 2.3 2.4 2.5 2.5 2.5 2.5 2.6 2.6 2.7 2.8 2.6 2.3 2.4 2.5 2.5 2.5 2.7 2.7 2.7 2.8 2.8 2.1 2.1 2.2 2.4 2.5 2.5 2.6 2.7 2.7 2.8 2.8
UQ10 UQ10 None UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 None UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 UQ10 UQ10
Distal Distal Distal Distal Distal Distal Distal Distal Distal Distal Distal Distal Distal Distal&Proximal None None None None None None None None None None Proximal Proximal Proximal Proximal Proximal Proximal Proximal Proximal Proximal Proximal Proximal
L213DN, M233RC R26 L181YF, M203DG, M210FY, M214HL, M260WA M197RF, M115FW L213DN, M233RC R26 + Cd2+ R26 + Zn2+ L210DN M197FR, M177YF L209PF L209PE R26 L213DN, H177RH, H8GQ L213DN, M233RC R26 R26; with Cyt c2 M202HL (heterodimer) R26. P+QA– M197FR, M177YF M76YF M76YK R26 M21TL M267RL M260AW L213DN R26 M203GL R26 + Cd2+, P+QB– R-26.1 R26, P+QB– R26; with Cyt c2 R-26.1, + 3,4-dihydrospheroidene L209PY R-26.1, + spheroidene
2.0
MQ9
MST
L168HF, + 2-t-butylamino-4-ethylamino-6-methylthio1,3,5-triazine
1VRN 1PRC 2PRC 3PRC 4PRC 5PRC
2.2 2.3 2.5 2.4 2.4 2.4
MQ9 MQ7 MQ7 MQ7 MQ7 MQ7
UQ7(Prox) UQ1(Prox) UQ2(Prox) None SMA ATZ
6PRC
2.3
MQ7
CEB
+ stigmatellin A + 2-chloro-4-isopropylamino-6-ethylamino-1,3,5-triazine + 2-chloro-4-ethylamino-6-(s(-)-2´-cyano-4-butylamino)-1,3,5-triazine
Chapter 20
Acceptor Quinones of Purple Photosynthetic Bacteria
387
Table 1. Continued PDB 7PRC
R 2.7
QA MQ7
QB CET
Modifications & Additions2 + 2-chloro-4-ethylamino-6-(r(+)-2´-cyano-4-butylamino)-1,3,5-triazine
2I5N 2.0 MQ9 UQ1(Prox) Thermochromatium tepidum 1EYS 2.3 MQ8 None 1 Of the 47 Rba. sphaeroides structures as of March 2007, 36 have a resolution ≤2.8 Å. 2 Non-mutant structures are from R26 or R26.1 (carotenoidless) strains; mutations are in non-R26 (carotenoid-containing) backgrounds e.g., strains Ga or 2.4.1 3 Crystal structures of protein maintained in the P+QA– state under illumination at 100 K (2bns) do show larger structural changes in the H subunit and increased susceptibility to protease digestion (Katona et al., 2005). 4 In the earlier Bcl. viridis structures (1PRC-7PRC) QA was identified as MQ7, but this should probably read MQ9, which is the major prenylog in this species (Parson, 1978; Gast et al., 1985; Shopes and Wraight, 1985).
determination (Ermler et al., 1994). In this position the quinone has moved along the channel holding the tail in the proximal structure, with the headgroup now visible from outside the protein. The novel position was tentatively ascribed to QB being in the quinol state, reduced by ascorbic acid in the isolation medium (Ermler et al., 1994). Subsequently, however, this arrangement was also found for the oxidized quinone (Stowell et al., 1997), although the occupancy is variable (Fritzsch et al., 2002; Pokkuluri et al., 2004). Reexamination of the Blc. viridis structure revealed a similar binding configuration for the native UQ-9, with partial occupancy (Lancaster and Michel, 1997), whereas UQ-2, added to saturate the site, appeared to bind proximally (Lancaster, 1998). A matched pair of X-ray structures was prepared from Rba. sphaeroides RCs frozen in the dark in the PQAQB state and frozen in the light, presumably in the P+QAQB– state (Stowell et al., 1997). Freezing RCs in the charge separated state traps the protein/quinone complex in a conformation that is capable of QB reduction (Kleinfeld et al., 1984a; Xu and Gunner, 2001). The frozen-in-the-dark structure (1aij) has the quinone in the outer, distal position. The frozen-inthe-light structure (1aig) has it in the inner, proximal position. No other significant structural changes in the QB binding site region accompany the distal-toproximal movement of the quinone, and the RMSD between the two L subunits is only 0.3 Å. However, more significant changes have since been reported in the H-subunit (Fritzsch et al., 2002; Katona et al., 2005). 1. The Two QB Binding Sites Found in the X-ray Structures The inner, proximal position of QB is shown in
Fig. 4. The hydrogen bond from O4 to His L190 is 2.69±0.23 Å, perhaps not significantly different from the distance to His M219 in the QA site, but notably not longer; it is also similar in Blc. viridis RCs. In different structures, O1 can make a hydrogen bond to the side chain of Ser L223 (2.81±0.25 Å) or the backbone of Ile L224 (2.91±0.24 Å) or Gly L225 (3.09 ± 0.16 Å), and usually more than one. The Ser Oγ is closest to the carbonyl in 75% of the structures. Residues within 4 Å of the QB headgroup in the proximal site in most structures are Leu L189, His L190, Leu L193, Glu L212, Asp L213, Ser L223, Ile L224, Gly L225, Thr L226 and Ile L229, a single atom in Val L194, and one or two water molecules. Phe L216 is symmetry-related to Trp M252 in the QA site, but it does not appear to be close enough (≥ 4.2 Å) or appropriately oriented to make either a parallel or perpendicular ‘aromatic’ interaction with QB in the proximal position. However, Phe L216does contact the first isoprene unit of the ubiquinone tail. Additional contacts with the first two isoprene units of the tail are made by Leu L185, Ala L186, Tyr L222, Ile L224, Ile L229 and Leu L232. The conversion between proximal and distal QB positions is not a simple translation, with most reports describing at as a ‘propeller twist,’ requiring also a 180° flip of the headgroup around the isoprene chain to switch the O4 carbonyl H-bond from His L190 to Ile L224 (Stowell et al., 1997). In the distal position, O4 is 2.89 ± 0.22 Å from the Ile L224 backbone NH, although in two structures (1m3x and 1fnp) the Ser L223 hydroxyl oxygen is closer. There is now no good hydrogen bond donor for O1. Ala L186 is the closest, but the average distance is >4 Å. The C1=O is the closest quinone carbonyl to Nδ of His L190 but is at a distance of 6.7 Å. The quinone appears to be stabilized by π- π interactions with Phe L216,
388 which is roughly parallel to and above the plane of the quinone ring, 3.8–4.2 Å away. Additional contacts are made with residues Ala L186, Leu L189, Tyr L222, Ile L224, and Ile L229. In the distal position, the tail is almost completely out of the RC protein structure — beyond the first isoprene unit it is only contacted on one side. The orientations of the methoxy groups in the distal quinone are very ill-defined, with large standard deviations. The distal position for QB is presumed to be nonfunctional. The distance from QA is 1.6 Å further than for the proximal position, measured between nearest carbonyl oxygens, or 2.4 Å larger measured from ring center to center. As electron transfer slows exponentially with distance, the rate kAB is expected to be 1–2 orders of magnitude slower than to the proximal site (Marcus and Sutin, 1985; Moser et al., 1995; Moser et al., 2003). Also, the calculated Em in the distal site is 250 mV lower than in the proximal site, largely because it is further from the non-heme iron and lacks hydrogen bonds to one carbonyl (Rabenstein et al., 2000; Zhu and Gunner, 2005). Recent work with mutant RCs that exhibit electron transfer to QB from HB, along the B branch, indicates that even though the distal site is closer to HB it is not active, suggesting that the driving force for quinone reduction by HB– is inadequate to compete with recombination from P+HB– (Laible et al., 2003; Breton et al., 2004; Wakeham et al., 2004; Paddock et al., 2005; Paddock et al., 2006). The difficulties in defining the QB binding configurations are well illustrated by a very recent collection of high resolution structures, determined by Fritzsch and coworkers, of RCs frozen in the dark or light, and at several pH values (Koepke et al., 2007). They report roughly equal populations of proximal and distal occupancies under all conditions, i.e., including RCs frozen in the light, which are supposedly trapped in the P+QB– state. The presence of both QB positions means that even at the very good resolutions obtained (down to 1.87 Å), the orientation of the ring and many of the substituent interactions are ambiguous. Nevertheless, they conclude that the distal position is not flipped relative to the proximal position. The earlier work on dark-frozen RCs, at lower resolution, also either had a mixture of QB positions, or additional water molecules modeled in to account for electron density (Stowell et al., 1997; Pokkuluri et al., 2002, 2004). Thus, this issue must be considered open.
Colin A. Wraight and Marilyn R. Gunner Occupancy of both QB sites in the light-frozen samples, as suggested by Koepke et al. (2007), would have major consequences for functional models of the acceptor quinones, and for understanding the origin of the quinone properties. However, in spite of the high overall resolution of these data, the QB binding site structures have some physically highly improbable features, including hydrogen bonds as short as 2.26 Å (heavy atom distance), and substantial distortions of some quinone ring substituents. 2. The Methoxy Groups of QB In the proximal QB site, the orientations of the methoxy groups of QB are uncertain, as is the case for QA. For QB, this problem is exacerbated by the smaller set of structures and less well resolved quinone position. Nevertheless, ignoring the earliest structures, which have gross anomalies in the quinone placements, the large majority of structures with QB in the proximal position (13 structures) yields a consensus with both methoxy groups strongly out of plane, with dihedral angles of –90 ± 9° and 88 ± 20° for 2- and 3-methoxy, respectively, which is superficially similar to the conformation for QA. (Several new structures from Koepke et al. (2007) show a very wide range of conformations for the 3-methoxy group, and are not included in this average). The similarity of the QA and QB infra-red (IR) bands at 1263 and 1287 cm–1 supports similar methoxy orientations in the two sites (Remy et al., 2003) (but see below). However, the amino acid composition of the QB site and a departure from the approximate C2 symmetry of the overall RC structure provide very distinct hydrogen-bonding patterns for the methoxy groups of QA and QB. The 2-methoxy oxygen of proximal QB is well endowed with potential donors, notably the peptide nitrogens of Gly L225 and Thr L226. For the 3-methoxy group, the oxygen lone pairs generally point away from the carboxyl group of Glu L212 and the most prevalent hydrogen bonding interaction is with Nδ of His L190,. In the somewhat better defined QB site structures of the Blc. viridis RC, this histidine forms bifurcated H-bonds to the carbonyl O4 and methoxy O3. It is noteworthy that this interaction cannot occur for QA in Rba. sphaeroides because the methoxy groups of QA and QB are almost 180° out of phase and the 3-methoxy lone pairs of QA are too distant (>4.4 Å) and point away from His M219. This represents a striking departure from the overall
Chapter 20
Acceptor Quinones of Purple Photosynthetic Bacteria
389
Fig. 5. View down the two-fold rotational symmetry axis, showing that the methoxy groups (C2M and C3M) of QA (left) and QB (right) violate the symmetry that is followed by the isoprene tail and the major features of the reaction center organization (cofactor positions and architecture of the L and M subunits).
C2 symmetry relating the two sites (Fig. 5). C. Hydrogen Bonding in the Quinone Binding Sites There is a variety of evidence for the importance of hydrogen bonding of the two quinones of the RC. The carbonyls and the methoxy oxygens are both nominally capable of functioning as hydrogen bond acceptors, and hydrogen bonds to the carbonyls, in particular, are expected to affect the affinity and specificity of QA and QB binding and to modify their physico-chemical properties to be sufficient for the reduction of quinone to quinol. In general, however, the resolution of the X-ray structures is too low to allow unequivocal assignment of the interactions between the quinones and their binding sites, and the energetics and specific properties of, e.g., hydrogen bonds, must be determined from spectroscopy and computation. The crystal structures, measurements of affinity of modified quinones, electron paramagnetic resonance (EPR), electron nuclear double resonance (ENDOR), solid state nuclear magnetic resonance (NMR) and Fourier transfer infra-red (FTIR) spectroscopy all offer complementary pictures of the quinone interactions. Fruitful comparisons might also be expected with an increasing selection of approximately 30 structurally defined quinone and 10 inhibitor binding sites in other proteins (Fisher and Rich, 2000), including cytochromes bc1 and b6 f, succinate dehydrogenase and fumarate reductase, as well as the binding of inhibitors as quinone or quinol analogs such as stigmatellin in RCs and cytochromes bc1 and b6 f complexes. However, the crystal structures residing in the protein databank in March 2007 reveal a very wide variation of hydrogen bond partners to quinone carbonyls, with no evident motifs defining a canonical quinone binding site. Thus, the quinone site architecture found in RCs is not replicated in
any other known quinone binding sites and useful correlations will require considerable sophistication, if they exist at all. IV. Spectroscopy of the Acceptor Quinones Vibrational spectroscopy provides an independent source of data on the quinone binding interactions, and FTIR spectroscopy in particular has provided significant insights, while magnetic resonance methods have been applied with great effect to the semiquinones. An advantage of IR absorbance and Raman over magnetic resonance spectroscopies is that the former can see all redox states of the quinones, not just the radical species. The use of light-induced (or redox potential-induced) difference spectra has allowed excellent resolution of the difference between quinone and semiquinone states in RCs, against the much greater background of mostly unchanging protein vibrations (for reviews, see Breton and Nabedryk, 1996; Nabedryk and Breton, 2008). The theoretical and experimental bases for the IR spectra of ubiquinones, in solution, have been systematically explored by Breton and Nonella and coworkers (Nonella and Brändli, 1996; Burie et al., 1997; Nonella, 1997, 1998; Nonella et al., 2003), and others (Boesch and Wheeler, 1997; O’Malley, 2001; Wheeler, 2001). The IR absorbance spectrum of isoprenyl ubiquinone (UQ-n) in solution, at room temperature, is characterized by a C=C stretch at 1610 cm–1 and two C=O stretches at 1665 and 1650 cm–1. The C=C and C=O modes are substantially coupled to each other but those of the two carbonyls are not. The splitting indicates asymmetry of substitution, which, for ubiquinone, arises from the conformation of the two methoxy groups, since an in-plane methoxy donates electrons into the ring while those out of plane are electron withdrawing (Burie et al.,
390 1997; Meyerson, 1985). The lone pair electrons of an in-plane methoxy oxygen are conjugated to the quinone C=C—C=O π electron system, decreasing the bond order, downshifting the distal C=O stretch frequency and upshifting the proximal C=O (Nonella and Brändli, 1996; Burie et al., 1997; Boullais et al., 1998; Nonella et al., 2003). The in-plane conformation is favored by the stabilizing electron resonance but is countered by unfavorable van der Waals interactions with any adjacent substituents.3 The steric clash is diminished by the larger bond angle arising from increased sp2 character with hydridization of the inplane methoxy oxygen. There is only room for one of the adjacent methoxy groups in UQ to be in-plane at a time, so the two carbonyls are not equivalent. However, both carbonyls contribute equally to both C=O stretch frequencies since neither methoxy has precedence in solution (Burie et al., 1997; Boullais et al., 1998; Remy et al., 2003). The quinone C=O and C=C bands are generally not very sensitive to changing from a protic (hydrogen bonding) to an aprotic solvent (Bauscher and Mäntele, 1992; Burie et al., 1995). This reflects the relatively weak hydrogen bonding propensity of β-unsaturated carbonyls (Rasmussen et al., 1949; Bellamy, 1968; 1975). Extensive IR studies have been made of the quinones in bacterial RCs (reviewed in Breton and Nabedryk, 1996; Boullais et al., 1998). The two C=O vibrations are very weakly coupled, if at all, indicating at least as much conformational and environmental asymmetry as is found in solution. Also as in solution, both carbonyls are substantially mixed with the C=C stretch. Assignment of the dominant C1=O, C4=O and C2=C3 and C5=C6 stretch bands has been made through the use of site-specific isotopic labels (Breton et al., 1994a, 1995; Brudler et al., 1994, 1995). A. The Primary Quinone, QA For QA, the quinone mode dominated by C1=O, associated with a hydrogen bond to the backbone of Ala M260, is found at 1660 cm–1, close to the average of the two solution values. A band at 1628 cm–1 is identified as having major C=C character, signifi3 Both the methyl and the non-conjugated lone pair electrons of the methoxy oxygen interact with adjacent groups — the proximal carbonyl oxygen and the other methoxy group. In the absence of a neighboring substituent, the preferred in-plane conformation is with the methoxy methyl group pointing away from the carbonyl. However, even a methyl in the adjacent position hinders this conformer significantly more than the other in-plane alignment, which is therefore preferred.
Colin A. Wraight and Marilyn R. Gunner cantly up-shifted from solution (1610 cm–1) but still recognizable. However, the C4=O stretch, hydrogen bonded to His M219, has been convincingly assigned to a band at 1601 cm–1, which is downshiftedfrom the solution value by 50–65 cm–1 (Breton et al., 1994a; Brudler et al., 1994). The relative positions of the C1=O and C4=O stretch frequencies are consistent with the C2- and C3-methoxy groups being roughly in-plane and outof-plane, respectively, but the magnitude of the C4=O downshift is far larger than can be accounted for by this effect. Furthermore, a substantially in-plane orientation for either group is not well supported by a consensus of the more recent X-ray structures, nor by an FTIR analysis of the methoxy marker bands at 1260-1290 cm–1 (Remy et al., 2003). However, FTIR spectra of RCs reconstituted with monomethoxy-ubiquinone-4 analogs (with tetra-isoprenyl sidechains) show that absence of the 2-methoxy group does result in an upshift of the C4=O stretch, although only by about 10 cm–1 (A. Vakkasoglu, B. Lipshutz and C. A. Wraight, unpublished). Consistent with such a role, tetramethylbenzoquinone (duroquinone) does not exhibit a substantially downshifted C=O stretch, although small, tailless quinones also appear to be distinct from isoprenylated analogs in other ways (Breton et al., 1994b). Strong hydrogen bonding has been suggested as the origin of the C4=O stretch frequency, but it is a very large shift, especially since hydrogen bonding by protic solvents has little influence on the C=O stretch frequency (Bauscher and Mäntele, 1992). Linear relationships between hydrogen bond strength and relative frequency shift (∆νC=O /νC=O), have been shown; a specific dependence of 4 x 10–3 (kcal/mol)–1, reported by Zadorozhnyi and Ishchenko (1965) for both intra- and intermolecular hydrogen bonds, has been widely cited. Using the relationship of Zadorozhnyi and Ishchenko, the 50–65 cm–1 frequency shift for C4=O of QA (depending on whether 1665 or 1650 cm–1 is used as a reference, see above) would correspond to a hydrogen bond strength (enthalpy) of about 7.5–10 kcal/mol. This is a large value, but IR and Raman band shifts have suggested hydrogen bond strengths of this magnitude in other enzyme systems (Deng and Callender, 1999). However, other solution studies have shown that, unlike the classic Badger-Bauer relationship for νOH, different carbonyl species do not fall on a unique line but exhibit distinct linear dependencies (Bellamy and Pace, 1970; Thijs and Zeegers-Huyskens, 1984; Meyerson, 1985).
Chapter 20
Acceptor Quinones of Purple Photosynthetic Bacteria
Thus, caution is needed in applying the Zadorozhnyi and Ishchenko linear relationship, as it has not been validated for quinones Some support for one strong hydrogen bond to QA comes from measurements of the binding affinity of the QA site of Rba. sphaeroides RCs for a series of compounds with different numbers of carbonyls (Gunner et al., 1985; Warncke and Dutton, 1992; 1993a). Thus, anthrone (one C=O) is found to bind 1.2 kcal/mol weaker than 9,10-anthraquinone, while 9,10-dimethyl-anthracene (no C=O) binds more than 6.5 kcal/mol weaker (Gunner et al., 1985). This is consistent with quinone binding utilizing one strong and one weak hydrogen bond but does not prove it, since the removal of one carbonyl could strengthen the remaining interaction. Despite the evidence of the affinity measurements, assigning the C=O band shift in the protein to hydrogen bonding is not well founded for quinones, which show only weak interactions with hydrogen bonding solvents (Bauscher and Mäntele, 1992). Large downshifts of the carbonyl bands are seen in quinones capable of intramolecular hydrogen bonding, such as anthraquinones with OH-substituted in the α position adjacent to the carbonyl (Flett, 1948). However, this is thought not to be primarily due to the hydrogen bonding, per se, but to the electronic redistribution made possible through resonance in a dimer that is stabilized by H-bonding (Rasmussen et al., 1949; Bloom et al., 1959). In contrast, the O-H band for these compounds is greatly shifted by hydrogen bonding, in a typical Badger-Bauer relationship (Badger and Bauer, 1937; Gordy, 1940; Rao et al., 1975). Similar conclusions have been drawn from studies on quinhydrones (dimers of quinone and hydroquinone in a charge transfer complex), which suggest that π electronic interactions make the major contribution to the C=O downshift rather than a strong H-bonding between the quinone carbonyl and adjacent hydroquinone hydroxyl (Bloom et al., 1959; Slifkin and Walmsley, 1969; Kruk et al., 1993). Other evidence that the downshift of the C4=O stretch is not due to strong hydrogen bonding is that incubation with D2O has no effect on the 1601 cm–1 QA band (Breton and Nabedryk, 1995; Wells et al., 2003). It was suggested that no (or very slow) H/D exchange occurred at His M219. Slow exchange is supported by ENDOR data, which showed very slow D2O exchange of two proton hyperfine interactions with identical half times of ≈3 hrs (Okamura and Feher, 1986). More recent ENDOR studies found
391
one of the two hyperfine interactions (assigned to the peptide NH of Ala M260) to exchange in about 1 hour, significantly more quickly than the other (assigned to NδH of His M219), which takes several hours at room temperature (Flores et al., 2006).4 However, H/D shifts are seen in the semiquinone IR spectrum (Breton and Nabedryk, 1995; Wells et al., 2003), suggesting that the light-dark cycling used in IR difference spectroscopy may accelerate the exchange. Thus, the lack of a significant D/H-shift in the quinone C=O stretch at 1600 cm–1 may not be due to lack of exchange, but to insensitivity to the isotopic substitution. This would not be supportive of an unusually strong H-bond. Considering X-ray structures with a resolution ≤2.8 Å, the heavy atom distances from O4 to Nδ of His M219 and O1 to the carbonyl of Ala M260 are both about 2.8 Å — neither bond is particularly short and both are of similar length. In addition neither Ala nor His donor N is in plane with the carbonyl, which is not optimal. The hydrogen bonding strength has been estimated using a standard molecular mechanics force field, which considers the Lennard-Jones attractive and repulsive terms and the electrostatic interactions that the quinone makes with the protein. In 1aig the interactions with the His sidechain are estimated at only –1.6 kcal/mol, and –0.4 kcal/mol for the amide backbone, neither of which are especially strong (Zhu and Gunner, 2005). Solid state NMR studies on 13C-labeled quinones in RCs also provide little support for the strong hydrogen bond to C4=O (van Liemt et al., 1995), although de Groot (1995) has made a brave effort to show that the expected chemical shift could be masked by other effects. Thus, although the FTIR and NMR results may not be fundamentally inconsistent, it seems fair to say that the origin of the C4=O frequency shift for 4 A mutant RC with His M266 changed to cysteine (mutant M266HC) has been widely used for ENDOR studies of the semiquinones, because it readily incorporates Zn in place of Fe (Williams et al., 1991). Although no substantial perturbations are seen in the X-ray structure (E. C. Abresch, unpublished, reported in Lubitz and Feher, 1999), Paddock et al. (1999) have reported much faster H/D exchange (<10 s) for one carbonyl, suggesting some perturbation of the acceptor quinone complex. This is also supported by electron spin echo envelope modulation (ESEEM studies, which show M266HC RCs to be almost completely lacking the 14N quadrupole couplings assigned to His M219 (Spoyalov et al., 1996), even though no kinetic disparities have been reported. The g-tensor values are also very similar to wild type, with only gx somewhat decreased (Lubitz and Feher, 1999).
392 QA is far from clear and other explanations should be considered. An abnormally strong interaction between the quinone and His M219 could arise from the His being liganded to the high spin FeII that lies between the two quinones. The pKa of the NδH group (imidazolate NH/N–) is normally in excess of 14 in solution, but can be lowered by several pH units when ligated to some metals. However, there is little precedent for this from model compounds with FeII, in contrast to FeIII (Sundberg and Martin, 1974). Enhanced acidity of the NδH proton would make it a more effective hydrogen bond donor, but one might still expect this to show up as the significantly shorter of the two carbonyl linkages, and to manifest effects of H/D exchange on the hydrogen-bonded C=O stretch. Furthermore, QB in the proximal position is very similarly arranged, making a hydrogen bond to His L190, which is one of the other ligands to the non-heme iron, but the QB C4=O stretch is not shifted by an unusual amount. Another source of the carbonyl frequency shift might be the π-π interaction between QA and Trp M252, which is roughly parallel to the quinone headgroup and approaches to within 3.5 Å. Tryptophan M252 is between QA and the active BPhe and has been shown to contribute to the fast electron transfer to QA (Coleman et al., 1990a; Stilz et al., 1994), as well as to the binding of the quinone (Stilz et al., 1994). This residue is electronically coupled to both QA and the BPhe (Plato et al., 1989), although not so strongly as to significantly perturb the visible spectra. Furthermore, mutating this residue to tyrosine was found to have negligible effect on the hyperfine parameters (nitrogen quadrupole couplings) of QA– (Spoyalov et al., 1996). Calculations of the electronic structure parameters of QA– in bacterial RCs showed only minor effects of both Trp M252 and Ile M265 (Fritscher et al., 2006). Nevertheless, a role for tryptophan-(semi)quinone interactions has been suggested for quinone function in bacterial RCs and Photosystem I on the basis of density functional theory (DFT) calculations, despite the analysis showing that the in vivo π-stacked configuration has only a small effect on the semiquinone g-tensor (Kaupp, 2002; Kacprzak and Kaupp, 2004). The remaining and likely influence is proteinimposed constraints on the dihedral angle of the methoxy groups. This affects not only the bond order (and hence vibrational frequencies) of the C=C and C=O bonds, as discussed above, but also the electron affinity (redox potential) of the quinone and the ba-
Colin A. Wraight and Marilyn R. Gunner sicity of the carbonyl oxygens (Breen, 1975; Prince et al., 1983; Prince et al., 1988; Robinson and Kahn, 1990). The lower frequency of the C4=O stretch is consistent with a more nearly in-plane position for the 2-methoxy, allowing the lone pair to enter the conjugated C2=C3–C4=O π-bond system. This would also enhance the charge density on the C4=O carbonyl oxygen, making it more basic and a stronger hydrogen bond acceptor. However, an in-plane conformation for the 2-methoxy group is only approached in very few RC crystal structures. Furthermore, the IR bands of the methoxy C–O stretch modes at 1263 and 1287 cm–1 are very similar (±1 cm–1) for both QA and QB, for pool quinone in the native membrane (Mezzetti et al., 2003), and for ubiquinone in solution where the methoxy groups are out-of-plane, suggesting they have the same conformation in the RCs. On this basis, Gerwert and coworkers have concluded that the methoxy group dihedral angles cannot contribute significantly to the differences between QA and QB properties, such as the IR spectra or redox energies (Remy et al., 2003). This is at odds with the observation that only ubiquinones can simultaneously function as QA and QB. Furthermore, recent results with mono-methoxy ubiquinone-4 analogs show that simultaneous function as QA and QB specifically requires only the 2-methoxy group (A. Vakkasoglu, B. Lipshutz and C. A. Wraight, unpublished). This suggests that the 2-methoxy group is involved in generating the necessary difference in redox potential between QA and QB, which is also consistent with the ability of rhodoquinone to function simultaneously as QA and QB (Graige et al., 1999). The necessary redox potential difference is quite small (approx. 75 mV for the 1-electron couples, Q/Q–, involved in the first electron transfer (Wraight, 1979; Blankenship and Parson, 1979; Kleinfeld et al., 1984b; Mancino et al., 1984; Shinkarev and Wraight, 1997)) and might not result in a significant spectroscopic difference. In fact, whatever the origin of the unusual C4=O IR band, which has garnered a lot of attention, it may not be indicative of something essential for normal function. In a mutant with Ala M260 replaced by cysteine, the C4=O frequency is upshifted by 30–40 cm–1, with only minor perturbations of the QA to QB electron transfer (Breton et al., 2007). It is now only 15–20 cm–1 downshifted from the solution value. B. The Reduced Primary Quinone, QA– The semiquinone QA– can be generated in RCs by light
Chapter 20
Acceptor Quinones of Purple Photosynthetic Bacteria
excitation and charge separation (forming P+QA–) at temperatures down to 1K (Arnold and Clayton, 1960), or by chemical reduction. PQA– can also be trapped for minutes in samples where an electron donor is added to reduce P+. Optical and IR spectroscopy, which can distinguish between Q– and QH, have shown that the equilibrium species for the QA semiquinone is the unprotonated anion, QA–. The hydrogen bond and matrix environment of QA– has been explored by magnetic resonance measurements and by FTIR. 1. Magnetic Resonance Studies of QA– The EPR spectrum of QA– in native RCs is extremely broad and unrecognizable as a free radical signal, due to strong distortion by the integral spin system of the FeII atom (S = 2). It is centered at g = 1.82, with a gx component at g = 1.68 in X-band EPR (Feher et al., 1972; Leigh and Dutton, 1972). The origin of this signal has been explored in many studies, most extensively by Butler et al. (1980, 1984), who showed that the iron was equidistant from QA and QB and not directly coordinated to either. However, the complexity of the system does not allow extraction of any electronic information pertaining to the quinones themselves. Fortunately, the iron can be exchanged for other metals, including diamagnetic zinc (see Section IV.A and Footnote 4, above), which unveils the semiquinone free radical signal and allows the whole panoply of magnetic resonance methods to be applied. The disposition of hydrogen bonds to the semiquinones was first investigated in ENDOR and EPR studies by Feher and coworkers (Feher et al., 1985, 1988; Lubitz et al., 1985; Isaacson et al., 1996), who quantified the hyperfine interactions for exchangeable and non-exchangeable protons, as well as for 17O in the carbonyl groups. A marked asymmetry is evident in the spin density distribution of the QA– radical, distinct from solution behavior and consistent with unequal hydrogen bond lengths to the two ‘carbonyl’ oxygens of QA– — one suggestive of a strong bond, and one indicating a more typical H-bond. However, the calculation assumed the O—H bond axis to His M219 or to the Ala M260 backbone is in the plane of the quinone ring, which, from the X-ray structures, is likely not true. The hyperfine coupling of the methyl (C5´) group is weaker than that seen in solution studies, implying a lower spin density on C5 (and, hence, higher on
393
C4 and C6). In good agreement with this, Q-band EPR studies of 13C-labeled Q-10 in RCs, show the spin density to be enhanced on C2, C4, C6 and O1 (van den Brink et al., 1994; Lubitz and Feher, 1999). The anionic charge is distributed in a complementary fashion5, with more on O4 than O1, promoting stronger hydrogen bonding to C4– –O. Specific identification of the hydrogen bonding partner was made by the pulsed EPR method, electron spin echo envelope modulation (ESEEM) (Bosch et al., 1995), which revealed 14N nuclear quadrupole coupling parameters that are consistent only with a strong histidine/QA– interaction. Subsequent work revealed the weaker coupling of C1– –O to a peptide nitrogen (Spoyalov et al., 1996). Thus, hydrogen bonding in the semiquinone state appears to be asymmetrical, with C4– –O more strongly bonded to Nδ-His M219 than is C1– –O to N-Ala M260. Recent ENDOR experiments (Flores et al., 2003; Flores et al., 2006, 2007) and DFT studies (Fritscher et al., 2006; Sinnecker et al., 2006) have substantially refined the earlier results, and a qualitative distinction between the hydrogen bonds to the two quinone oxygens is confirmed. Consensus values, calculated without using the point dipole approximation, are: for the C4– –O bond to His M219, the H-bond length, rO–H is 1.60-1.66 Å and the heavy atom distance (O–N) is 2.66-2.72 Å; for C1– –O and Ala M260, the H-bond length, rO–H is 1.71-1.77 Å and the heavy atom distance (O–N) is 2.78-2.84 Å. All studies confirmed that the H-bonds were significantly out of the plane of the quinone ring, rendering a pointdipole calculation inaccurate due to the enhanced interaction between the proton and the p-π electron density on the oxygens. The out-of-plane configuration is consistent with an increased sp3 hybridization on the oxygens expected for the radical state, and as previously calculated for benzoquinone in solution (O’Malley, 1997b, 2001). DFT calculations on a model QA– binding site showed that a positive charge on the histidine ligand (modeled as imidazoleH+) is necessary to approximate the correct spin distribution in QA– (O’Malley, 1997a, 2003). This suggests a role for charge delocalization from the divalent Fe2+ — or Zn2+, which 5 If this seems counterintuitive, imagine the neutral semiquinone: the H-O covalent bond will require a valence pair of electrons, so the unpaired spin cannot predominantly reside there. However, when the proton is dissociated to form the semiquinone anion, the negative charge will be largely on this oxygen.
394 is often used as a diamagnetic substitute in the RC. Consistent with this, Fritscher et al. (2006) found a metal ion, in this case Zn2+, to be necessary to reproduce the asymmetry parameter for His M219 in DFT calculations, but the geometry optimization did not show the expected shortening of the His-quinone hydrogen bond length. Interestingly, geometry optimization in DFT indicates there should be a small rotation of QA upon reduction, leading to more linear — and therefore stronger — hydrogen bonds (Fritscher et al., 2006). (See Note Added in Proof.) Nevertheless, the calculated bond from QA– to His M219 does not appear to be unusually strong. Furthermore, the optimized geometry for oxidized QA shows the expected lengthening of the heavy atom distances for the neutral quinone, by 0.17 Å for C4=O to His M219, and 0.13 Å for C1=O to Ala M260 (Sinnecker et al., 2006). This also does not seem to provide much support for the strong H-bond interpretation of the unusual IR frequency of the C4=O stretch of QA. 2. Infra-red Spectroscopic Studies of QA– The IR spectra of semiquinones are sensitive to the protonation state of the radical. In alcoholic solution, protonated ubisemiquinone, UQH, absorbs at ≈1520 cm–1 (C– –O), 1475 cm–1 (C– –O) and 1415 cm–1 (C– –C) (Burie et al., 1995), while the anion, UQ–, exhibits a single main peak at ≈1465 cm–1 of complex origin (Bauscher and Mäntele, 1992). The hydrogen bond strengths of the anionic semiquinone carbonyls are expected to be significantly stronger than for neutral quinone and the semiquinone IR spectra are found to be quite sensitive to hydrogen bonding, with the anion band at ≈1485 cm–1 in aprotic solvent (Bauscher and Mäntele, 1992). In RCs, the IR spectrum of QA– is more complex than in vitro and the features between 1420-1480 cm–1 are referred to as the ‘anion band’. Unlike the independent C=O bands of the neutral quinone, the semiquinone C– –O vibrations are substantially coupled as well as exhibiting greater mixing of the C– –C and C– –O modes, as clearly demonstrated by the frequency shifts in response to specific isotopic substitution (Breton et al., 1994a; Brudler et al., 1994). For unlabeled ubiquinone-10, the main QA– anion band is at 1466 cm–1, with smaller peaks or shoulders at 1484, 1457, 1445 and 1422 cm–1. Isotopic (13C and 18 O) labeling identifies the main band (1466 cm–1)
Colin A. Wraight and Marilyn R. Gunner as predominantly C– –O, appropriate for a hydrogen bonded group (Breton et al., 1994a; Brudler et al., 1994), and the small band at 1422 cm–1 may be predominantly C– –C in character (Bauscher and Mäntele, 1992; Breton et al., 1994a; Brudler et al., 1994). The 1484 cm–1 band seems best assigned to a C– –C ring mode (Breton and Nabedryk, 1996). Breton and Nabedryk suggested that the 1466 and 1445 cm–1 bands correspond to the C1– –O and C4– –O groups, respectively, based on the effect of specific 13 C labeling at the C1 and C4 positions and on the fact that the 1445 cm–1 band shifts to ≈ 1432 cm–1 upon H2O/D2O exchange (Breton and Nabedryk, 1996). This is consistent with the proposed strongly hydrogen bonded character of the semiquinone C4=O group. Clouding the issue is the fact that the 1466 cm–1 band does not shift in D2O, although a shoulder at 1456 cm–1 shifts to higher frequency, by about 8 cm–1 (Wells et al., 2003). Thus, the anion band assignments are not completely satisfying and it is possible that some of the complexity, which is almost unique to the isoprenyl-ubiquinones (i.e., not including methyl-Q-0) (Breton et al., 1994b), arises from subtle matching of modes, i.e., Fermi resonance. Raman spectroscopy, which could resolve some of these issues, has been applied to the RC quinones only to a very limited extent. Raman-active bands at 1600–1620 cm–1 (C– –C), expected for the semiquinone species (Schuler et al., 1983; Burie et al., 1995; Zhao et al., 1997a; Zhao and Kitagawa, 1998), are seen in a strong resonance Raman peak at 1607 cm–1 for QA– (Zhao et al., 1997b). This band is not seen in the IR spectrum, in spite of the low symmetry of the molecule and its environment. If the assignments by Breton and coworkers (Breton and Nabedryk, 1996) of the C1– –O and C4– –O IR bands of the QA semiquinone are correct, the smaller splitting (∆ν ≈20 cm–1), compared to the C1=O and C4=O quinone bands (∆ν ≈60 cm–1), would be consistent with more symmetrical binding of the semiquinone than the quinone. In Blc. viridis, where QA is menaquinone, or in Rba. sphaeroides with naphthoquinone derivatives substituted as QA, the very distinctive C– –O contribution to the anion band (at 1478 cm–1) is not split at all in the QA– spectrum, although the quinone C=O bands are split for QA (Breton et al., 1994b; Breton et al., 1994c). No significant differences in the positions of QA– and QA are detected in the crystal structures (Abresch et al., 1998; Stowell et al., 1997), or by comparison of EPR
Chapter 20
Acceptor Quinones of Purple Photosynthetic Bacteria
(Van der Est et al., 1997), ENDOR (Isaacson et al., 1995) or FTIR (Breton et al., 2002) signatures of samples prepared at low and ambient temperatures. However, the symmetrization of QA– vs. QA could easily be achieved with very little movement, beneath the detection limits of these comparisons. (See Note Added in Proof.) C. The Secondary Quinone, QB For the oxidized QB spectrum, Breton and coworkers found only one C=O IR stretch, at 1641 cm–1 (Breton et al., 1995). Upon labeling at both carbonyls with 18 O this was shifted to 1621 cm–1, a relatively small shift indicative of substantial mixing with the C=C mode. Specific 13C-labeling at either C1 or C4 also demonstrated the equivalence of the two carbonyls, with downshifts of about 20 cm–1. Using the scale of Zadorozhnyi and Ishchenko (1965), with the caveat that the relationship was not derived from quinone species, as discussed above, the QB carbonylfrequency shift from the solution values of 10-25 cm–1 (depending on which peakis used as a reference) corresponds to hydrogen bond strengths of about 1.5–3.75 kcal/ mol. This is consistent with the order of magnitude weaker binding affinity of quinones measured for the QB site, compared to the QA site (McComb et al., 1990; Warncke et al., 1994; X. Zhang and M. R. Gunner, unpublished). Unexpectedly, Breton and coworkers found that the C=C stretch was not equivalently affected by labeling at C1and C4, but was downshifted (from ≈1616 cm–1 to 1600 cm–1) only when C1 was labeled with 13C. They suggested that binding of QB holds the 5-methyl group (Breton et al., 1995) or 3-methoxy (Breton and Nabedryk, 1996) very tightly, thereby restricting its motion coupled to C4=O. The same effect was seen in Blc. viridis (Breton et al., 1995). In contrast, Gerwert and coworkers reported two distinct IR signatures for QB in Rba. sphaeroides (Brudler et al., 1995). The dominant one (≈75%) exhibited a single C=O stretch frequency of 1641 cm–1, as observed by Breton et al. (1995), but a minor population (≈25%) showed split peaks at 1664/1651 cm–1, very close to free quinone. Because of this distinction and, especially, because they assigned different isotope shifts to the main 1641 cm–1 band, Brudler et al. did not report any unusual C1 vs. C4 labeling effect on the C=C stretch as proposed by Breton et al., but interpreted the differences in the
395
spectra as being due to cancellation of overlapping spectral contributions from two populations of QB.6 Furthermore, they found that specific 13C-labeling at C5 or C6 has the same effect on the C=C stretch. Thus, any asymmetry in the coupling between C=O and C=C modes, as described by Breton et al. (1995), would have to involve the C2=C3 bond, uniquely, therefore implicating the 3-methoxy rather than the 5-methyl group. In view of the X-ray studies showing two positions for QB and the possibility of a large scale motion of QB after light activation, the mixed population of QB reported by Brudler et al. (1995) takes on particular significance. No sub-populations were evident in the QB– semiquinone IR spectrum. This is consistent with a dark-adapted state where the quinone equilibrates between distal and proximal positions, and a light state in which QB– is locked into the single, functional proximal site. The free carbonyl character of the minor quinone population, at 1664/1651 cm–1, could be consistent with the nature of the distal binding site, in which only one carbonyl (C4=O) is hydrogen bonded, to the peptide NH of Ile L224. This is discussed further in Section V. D. The Semireduced Secondary Quinone, QB– Although the secondary quinone functions in all three redox states, quinone, semiquinone and quinol, no spectroscopic signatures of the quinol, QBH2, have been reported. However, Mezzetti et al. (2003) have reported the IR spectrum of pool quinones in native membranes of Rba. sphaeroides. Notably, the methoxy marker bands at 1264 and 1288 cm–1 of oxidized pool UQ are indistinguishable from those of QA and QB. Under prolonged illumination, reduction to quinol of up to 20 pool quinones per RC was observed, with main bands at 1491, 1470, 1433, and 1388–1375 cm–1. 1. Magnetic Resonance Studies of QB– Magnetic resonance studies of QB– are not as extensive as for QA–, but they serve to make some important distinctions. In native RCs, the magnetic interaction between QB– and FeII yields an EPR signal like that for QA–, centered at g = 1.82, but even broader with 6 It may be helpful to note that Brudler et al. (1994, 1995), Remy et al. (2003) and van Liemt et al. (1995) use yet another quinone atom numbering system (see Footnote 2).
396 a high field component at g = 1.63 in X-band EPR (Wraight, 1978; Butler et al, 1984). In zinc substituted RCs, the QA– and QB– free radical signals are also very similar, but the small difference in gx-tensor component is consistent with a more polar environment for QB– (Isaacson et al., 1995). Hyperfine coupling constants (hfcs) for QA– and QB–, determined using 17O-labeled (Feher et al., 1985; Lubitz and Feher, 1999) and 13C-labeled (Isaacson et al., 1996) ubiquinones, showed the spin densities on the carbonyls of QB– to be much less asymmetric than in QA–. Similarly, ENDOR studies of QB– show hyperfine splittings similar to those of UQ– in alcoholic solution (Lubitz and Feher, 1999). Potential hydrogen bonding interactions for QB– were indicated by two significant hyperfine couplings seen in 1H-ENDOR using M266HC mutant RCs, which provide a high yield of Zn-replacement of the Fe (Lubitz and Feher, 1999). Both lines disappeared when subjected to D2O/H2O, one much faster than the other, as also seen for QA–. (Note, however, that M266HC mutant RCs exhibit much greater proton lability than WT RCs — see Footnote 4). Only one set of 14N-nuclear quadrupole parameters have been detected for QB–, assigned to the NδH of a histidine (Lendzian et al., 1996), but the absence of any additional resonances is not diagnostic. Indeed, the almost equal spin densities on the carbonyls strongly imply hydrogen bonds of similar strength at each carbonyl oxygen, but the proton hyperfine coupling to QB– has not yet been evaluated with the sophistication applied to QA–. The X-ray structures, which show at least two and possibly three H-bond donors to the C1=O carbonyl, suggest that a refined analysis of the geometry may be difficult. However, Paddock et al. (2007) have identified a new hyperfine coupling in Q-band ENDOR, which they assign to H-bonding from Ser L223 on the basis of comparisons with mutant RCs. They used a quintuple mutant lacking QA but able to reduce QB through the normally inactive B-pathway, both at room temperature and at low temperature. The novel hfc was seen when the state P+QB– is permanently trapped in the quintuple mutant by illuminating while freezing, but not when P+QB– is generated de novo at low temperature, nor in the light trapped state in RCs with Ser L223 mutated to Ala. It was suggested that the coupling reflects a hydrogen bond between Ser L223 and QB– contributes to stabilizing the semiquinone state when formed at permissive temperatures. However, it was previously shown for
Colin A. Wraight and Marilyn R. Gunner wild-type RCs, using high-field ENDOR, that there is no detectable difference between the stable, freezetrapped P+QB– state and one generated by a different illumination regime that decays non-exponentially in 10 –100 s at low temperature (Utschig et al., 2005). This suggests that stabilization is contributed by multiple factors. 2. Vibrational Spectroscopy of QB– In contrast to the QB spectrum, the major features of the QB– semiquinone IR spectrum are largely agreed upon (Breton et al., 1995; Brudler et al., 1995). As with QA–, QB– is unprotonated. The main peak at 1479 cm–1 arises from both C– –O groups, which remain equivalent in 18O and uniform 13C-labeling. The position of this band is at higher frequency than those of QA– (1466 and 1445 cm–1), consistent with weaker hydrogen bonding for QB–. Upon uniform 13C-labeling, the 1479 cm–1 band downshifts by ≈ 52 cm–1, showing this mode to be strongly coupled to C– –C modes (Breton et al., 1995). Selective 13C-labeling at either C1 or C4 shifts and splits the C– –O band equivalently, with new peaks at 1438 and ≈1417 cm–1 (possibly shifted from a minor band at 1461 cm–1) confirming the near equivalence of the two C– –O (and C=O) environments. However, specific labeling at C5 and C6 gave no significant isotopic shifts in the semiquinone anion band (Brudler et al., 1995). This may indicate that the C– –O/C– –C coupling is so extensive over the ring system as to dilute any single atom labeling below detection. Alternatively, the C– –O/C– –C coupling might involve the C2– –C3 bond only. The C– –C and ring modes of the QB– semiquinone are weak but are tentatively assigned, with varying admixtures of C– –O, at 1490, 1461, 1441 and 1422 cm–1, similar to QA–. Also similarly to QA–, QB– exhibits a strong resonance Raman band at 1613 cm–1 (C– –C) (Zhao et al., 1997b) that is not seen in the IR spectrum. V. Functionality of the Two Quinone Positions in the QB Site The motion of QB between the distal and proximal positions was suggested to represent a rate limiting ‘gating’ event that precedes the first electron transfer from QA– (Stowell et al., 1997; Graige et al., 1998), but its functional significance has been challenged by subsequent crystal structures, and by FTIR and kinetic studies. Currently (Spring 2007) there are 12
Chapter 20
Acceptor Quinones of Purple Photosynthetic Bacteria
structures with resolution ≤2.8 Å with the quinone in the inner, proximal site, 12 with it in the outer distal site and 10 with no quinone. One structure (1s00) is modeled with a 50:50 mixture of distal and proximal position. While the structures frozen in the light (2bns, 1dv3, and 1aig) have the quinone in the proximal position, a number of dark structures do as well. At slightly lower resolution, Schiffer and colleagues found that RCs in the trigonal crystal form frozen in the dark exhibit a 4:1 proximal:distal configuration (Pokkuluri et al., 2004). It was noted that the quinone position may be sensitive to the crystal packing (the Stowell structure is from tetragonal crystals), to whether the measurements are made at room temperature or in frozen crystals, and to the presence of cryoprotectant (Pokkuluri et al., 2004). Fritzsch et al. (2002) have found both quinone positions occupied in the dark adapted structure and an increased occupancy of proximal quinone in the ‘charge separated’ protein. This may be partly dependent on the length of the tail as UQ-2 appears to be more localized in the proximal site of Blc. viridis RCs (2prc) than UQ-9 (Lancaster, 1998). However, time resolved X-ray diffraction studies of Blc. viridis RCs have shown that long tail quinone (native UQ-9 and reconstituted UQ-10) can be in the proximal site in the light or dark, with no change in occupancy of the proximal position at least during a 6 s cycle time (Baxter et al., 2004). Also, Rba. sphaeroides RCs with a mutation of Pro L209 to Tyr have been found to have the quinone in the proximal position, while for mutation to Phe or Glu it is mostly distal (Kuglstatter et al., 2001); nevertheless the first electron transfer reaction is very similar in mutants that exhibit quite different occupancies of the two QB sites (Tandori et al., 1999, 2002) even at cryogenic temperatures (Xu et al., 2002). Simulations have suggested that the relative occupancy of the distal and proximal positions is influenced by the ionization state of acidic residues such as Glu L212 and Asp L213 in the QB site (Grafton and Wheeler, 1999; Zachariae and Lancaster, 2001; Taly et al., 2003). This is very weakly supported, if at all, by crystallographic studies (Koepke et al., 2007). Computational analyses of the Em of the ubiquinone in the distal and proximal sites support the assumption that only the proximal site can support quinone reduction, i.e., the free energy for quinone reduction is calculated to be significantly more unfavorable than in the proximal site (Rabenstein et al., 2000; Zhu and Gunner, 2005). The ability to move
397
from the distal to proximal sites has been subjected to modeling and molecular dynamics simulations. With the quinone headgroup in the proximal-like orientation, i.e., with O1 directed towards His L190, translation encounters no significant barriers and QB can very rapidly equilibrate between the two sites (Grafton and Wheeler, 1999), although translation could be controlled by other features of the protein conformation including protonation (Walden and Wheeler, 2002). Molecular mechanics calculations on QB translation in Blc. viridis RCs showed that there is a significant barrier if the quinone must flip over while confined in or near the distal binding site (Zachariae and Lancaster, 2001). However, it may be that the quinone simply leaves the protein if it enters the binding channel incorrectly. Spectroscopic studies also provide strong argument against a kinetic role or relevance of the distal to proximal motion in QB reduction. First, the Pro L209→Tyr mutant discussed above, for which the X-ray crystal structure shows a proximal QB position, yields a single C=O IR frequency for QB (1641 cm–1), essentially unchanged from the wild type (Breton et al., 2002). An extensive FTIR study of wild type RCs also showed no difference in the QB/QB– spectrum when activated at room temperature (when such motion would be allowed) or at 90K after freezing in the light (when QB would be trapped in the light-activated, proximal position) (Breton, 2004). These results do not support the occurrence of one hydrogen bonded and one non-bonded C=O in the wild type RC. Kinetic behavior also mitigates against a functionally significant influence of the proximal-distal transition. Short chain ubiquinones have been cited as favoring the proximal position (Lancaster and Michel, 1996a, 1997), possibly consistent with a proposed anticooperative influence of the headgroup and hydrocarbon tail on binding (McComb et al., 1990). However, the length of the tail has no significant effect on the first electron transfer rate, kAB(1), at room temperature (McComb et al., 1990), and the rate and temperature dependence are the same in RCs with UQ-10 and UQ-1 down to 160 K (Xu and Gunner, 2002). On the other hand, UQ-0, lacking any substituent at position 6, exhibits significant differences in the FTIR spectrum, indicating some difference in binding configuration (Breton et al., 1994b). UQ-0 also has a two-fold larger first electron transfer equilibrium constant, K(1) AB, than the tailed ubiquinones (McComb et al., 1990). Finally, in chromatophore membranes of Rba.
398 capsulatus, Lavergne et al. (1999) found that the first and second electron transfers exhibit essentially identical kinetic properties. Since QB is clearly bound proximally (as QB–) for the second electron transfer, they argue that neither electron transfer is gated by any physical motion of the quinone, in vivo. Thus, in spite of the clear evidence for proximal and distal binding in many X-ray crystal structures, the preponderance of computational, spectroscopic and time resolved studies indicate that the movement, per se, does not have an essential role in single turnover studies, and is not generally seen in multiple turnovers required for signal averaging. Thus, the equilibrium position for quinone binding must generally favor the proximal position. Indeed, it may be that the distribution of distal:proximal occupancy can be modulated by factors unique to the crystallization procedure. Nevertheless, some non-electron transfer event does control the first electron transfer in RCs, as the major phases of this reaction are independent of the driving force (Graige et al., 1998). The nature of these changes, be they changes in protonation or hydrogen bonding networks in surrounding residues, or other conformational changes, remains an open question. VI. Conclusions Ultimately, the functional behavior and properties of the acceptor quinones arise from the environment provided by the protein. This, however, is not a static entity. As impressive and beguiling as the structural pictures are, a great deal of uncertainty resides in the interpretation of the spectroscopic markers, and the relationship between structure and the function of the acceptor quinones is still obscure. This is both exciting and frustrating, but reflects the complexity of events that are intimately coupled to protein conformational dynamics and to the solution through proton uptake. It is certainly clear that a great deal of work remains for us to understand ‘structure-function’ in this system, and the structural data must be supported and extended by computation and spectroscopy, in order to develop the energetics and dynamics of the interactions.
Colin A. Wraight and Marilyn R. Gunner that QA may move upon reduction, have gained some experimental support from two separate high-field, pulsed EPR studies, reported in 2007, although they are not in full agreement. Using orientation-resolved pulsed electron dipolar high-field EPR spectroscopy, Möbius and coworkers (Savitsky et al., 2007) have determined the relative orientations of the partners in the spin-correlated state of P+QA– of dark frozen RC samples at 150 K. Comparison of the positions and relative orientations of the precursor cofactors, P and QA, known from X-ray crystallography, shows a small (14°) rotation of the QA– headgroup around the x-axis (joining the carbonyl groups) compared to QA. In contrast, Kothe and coworkers (Heinen, et al., 2007) determined the structure of P+QA–, within 200 nanoseconds of flash photoexcitation from dark frozen RC samples at 70 K, using time-resolved (high time resolution) high-field EPR. Their results show a 60° rotation about the QA– z-axis (normal to the ring plane), relative to the quinone position seen in crystal structures. They suggest that the structure they observe represents the ‘initial’ neutral QA position prior to photoexcitation of the RC, and the 60° rotation constitutes the event that gates electron transfer from QA– to QB. Such a significant motion is likely to alter the H-bonding interactions at one or both carbonyl groups. It may be worth noting that the first crystal structure of Rba. sphaeroides RCs showed QA in position to H-bond between C4=O and Thr M222 rather than His M219 (Allen et al., 1988). Also, molecular dynamics simulations can result in switching the H-bond from His M219 to Thr M222 (L.Rinyuand P. Maróti, personal communication). To account for the position of QA determined from the crystal structure, Heinen et al. (2007) suggest that radiation-induced chemistry may lead to a reduced QA, in many X-ray structures, and this could explain the observation that the crystal structures are similar for room temperature and low temperature; as otherwise, the proposed gating mechanism of Heinen et al. (2007) would seem to require that the rotation of QA– occurs at low temperature. Regardless, these new studies reinforce the conclusion, above, that much remains to be learned in this seemingly welldefined system. Acknowledgments
Note Added in Proof The calculations of Fritscher et al. (2006), suggesting
The preparation of this chapter was supported by NSF grant MCB 03-44449 to CAW, and by USDA
Chapter 20
Acceptor Quinones of Purple Photosynthetic Bacteria
grant CREES 2003-02123 to MRG. Infrastructure support from NIH 5G12 RR03060 (National Center for Research Resources) to MRG is also gratefully acknowledged. References Abresch EC, Paddock ML, Stowell MHB, McPhillips TM, Axelrod HL, Soltis SM, Rees DC, Okamura MY and Feher G (1998) Identification of proton transfer pathways in the X-ray crystal structure of the bacterial reaction center from Rhodobacter sphaeroides. Photosynth Res 55: 119–125 Alexov E and Gunner MR (1999) Calculated protein and proton motions coupled to electron transfer: Electron transfer from QA– to QB in bacterial photosynthetic reaction centers. Biochemistry 38: 8254–8270 Alexov E, Miksovska J, Baciou L, Schiffer M, Hanson DK, Sebban P and Gunner MR (2000) Modeling effects of mutations on the free energy of the first electron transfer from QA– to QB in photosynthetic reaction centers. Biochemistry 39: 5940–5952 Allen JP, Feher G, Yeates TO, Komiya H and Rees DC (1987a) Structure of the reaction center from Rhodobacter sphaeroides R-26: The cofactors. Proc Natl Acad Sci USA 84: 5730–5734 Allen JP, Feher G, Yeates TO, Komiya H and Rees DC (1987b) Structure of the reaction center from Rhodobacter sphaeroides R-26: The protein subunits. Proc Natl Acad Sci USA 84: 6162–6166 Allen JP, Feher G, Yeates TO, Komiya H and Rees DC (1988) Structure of the Reaction Center from Rhodobacter sphaeroides R-26: Protein-cofactor (quinones and Fe2+) Interactions. Proc Natl Acad Sci USA 85: 8487–8491 Allen JP, Williams JC, Graige M, Paddock ML, Labahn A, Feher G and Okamura MY (1998) Free energy dependence of the direct charge recombination from the primary and secondary quinones in reaction centers from Rhodobacter sphaeroides. Photosynth Res 55: 227–233 Arnold W and Clayton RK (1960) The first step in photosynthesis: Evidence for its electronic nature. Proc Natl Acad Sci USA 46: 769–776 Arnoux B, Gaucher JF, Ducruix A and Reiss-Husson F (1995) Structure of the photochemical-reaction center of a spheroidenecontaining purple bacterium, Rhodobacter sphaeroides Y, at 3 Å resolution. Acta Crystallographica D 51: 368–379 Badger RM and Bauer SH (1937) Spectroscopic studies of the hydrogen bond. II. The shift of the O-H vibrational frequency in the formation of the hydrogen bond. J Chem Phys 5: 839–851 Bauscher M and Mäntele W (1992) Electrochemical and infraredspectroscopic characterization of redox reactions of p-quinones. J Phys Chem 96: 11101–11108 Baxter RHG, Ponomarenko N, Srajer V, Pahl R, Moffat K and Norris JR (2004) Time-resolved crystallographic studies of light-induced structural changes in the photosynthetic reaction center. Proc Nat Acad Sci USA 101: 5982–5987 Bellamy LJ (1968) Advances in Infrared Group Frequencies. Methuen & Co. Ltd, London Bellamy LJ (1975) The Infrared Spectra of Complex Molecules. Chapman and Hall, London, New York
399
Bellamy LJ and Pace RJ (1970) Hydrogen bonding in alcohols and phenols. III. Hydrogen bonds between alcohols and carbonyl groups. Spectrochimica Acta A 27: 705–713 Beroza P, Fredkin DR, Okamura MY and Feher G (1995) Electrostatic calculations of amino acid titration electron transfer, QA–QB → QAQB–, in the reaction center. Biophys J 68: 2233–2250 Blankenship RE and Parson WW (1979) The involvement of iron and ubiquinone in electron transfer reactions mediated by reaction centers from photosynthetic bacteria. Biochim Biophys Acta 545: 429–444 Bloom H, Briggs LH and Cleverley B (1959) Physical properties of anthraquinone and its derivatives. Part I. Infrared spectra. J Chem Soc 1959: 178–185 Boesch SE and Wheeler RA (1997) Structures and properties of ubiquinone-1 and its radical anion predicted from a hybrid Hartree-Fock/density functional method. J Phys Chem A 101: 5799–5804 Bosch MK, Gast P, Hoff AJ, Spoyalov AP and Tsvetkov YD (1995) The primary acceptor quinone QA in reaction centers of Rhodobacter sphaeroides R26 is hydrogen bonded to the Nδ(1)-H of His M219. An electron spin echo study of QA–•. Chem Phys Lett 239: 306–312 Boullais C, Nabedryk E, Burie J-R, Nonella M, Mioskowski C and Breton J (1998) Site-specific isotope labeling demonstrates a large mesomeric resonance effect of the methoxy groups on the carbonyl frequency in ubiquinones. Photosynth Res 55: 247–252 Breen, DL (1975) Coenzyme Q: A molecular orbital study. J Theor Biol 53: 101–113 Breton J (2004) Absence of large-scale displacement of quinone QB in bacterial photosynthetic reaction centers. Biochemistry 43: 3318–3326 Breton J and Nabedryk E (1995) Protein and bacteriopheophytin response to QA reduction in photosynthetic bacterial reaction centers from Rb. sphaeroides and Rp. viridis investigated by 1 H/2H exchange and light-induced FTIR difference spectroscopy. In: Mathis P (ed) Photosynthesis: From Light to Biosphere, Vol I, pp 395–400. Kluwer Academic Publishers, Dordrecht Breton J and Nabedryk E (1996) Protein-quinone interactions in the bacterial photosynthetic reaction center: Light-induced FTIR difference spectroscopy of the quinone vibrations. Biochim Biophys Acta 1275: 84–90 Breton J, Boullais C, Burie J, Nabedryk E and Mioskowski C (1994a) Binding sites of quinones in photosynthetic bacterial reaction centers investigated by light-induced FTIR difference spectroscopy: Assignment of the interactions of each carbonyl of QA in Rhodobacter sphaeroides using site-specific 13C-labeled ubiquinone. Biochemistry 33: 14378–14386 Breton J, Burie J, Boullais C, Berger G and Nabedryk E (1994b) Binding sites of quinones in photosynthetic bacterial reaction centers investigated by light-induced FTIR difference spectroscopy: Binding of chainless symmetrical quinones to the QA site of Rhodobacter sphaeroides. Biochemistry 33: 12405–12415 Breton J, Burie J-R, Berthomieu C, Berger G and Nabedryk E (1994c) The binding sites of quinones in photosynthetic bacterial reaction centers investigated by light-induced FTIR difference spectroscopy: Assignment of the QA vibrations in Rhodobacter sphaeroides using 18O- or 13C-labeled ubiquinone and vitamin K1. Biochemistry 33: 4953–4965 Breton J, Boullais C, Berger G, Mioskowski C and Nabedryk E
400 (1995) Binding sites of quinones in photosynthetic bacterial reaction centers investigated by light induced FTIR difference spectroscopy: symmetry of the carbonyl interactions and close equivalence of the QB vibrations in Rhodobacter sphaeroides and Rhodopseudomonas viridis probed by isotopic labeling. Biochemistry 34: 11606–11616 Breton J, Boullais C, Mioskowski C, Sebban P, Baciou L and Nabedryk E (2002) Vibrational spectroscopy favors a unique QB binding site at the proximal position in wild-type reaction centers and in the Pro-L209→Tyr mutant from Rhodobacter sphaeroides. Biochemistry 41: 12921–12927 Breton J, Wakeham MC, Fyfe PK, Jones MR and Nabedryk E (2004) Characterization of the bonding interactions of QB upon photoreduction via A-branch or B-branch electron transfer in mutant reaction centers from Rhodobacter sphaeroides. Biochim Biophys Acta 1656: 127–138 Breton J, Lavergne J, Wakeham MC, Nabedryk E and Jones MR (2007) The unusually strong hydrogen bond between the carbonyl of QA and His M219 in the Rhodobacter sphaeroides reaction center is not essential for efficient electron transfer from QA– to QB. Biochemistry 46: 6468–6476 Bruce BD, Fuller RC and Blankenship RE (1982) Primary photosynthesis in the facultative aerobic green photosynthetic bacterium Chloroflexus aurantiacus. Proc Natl Acad Sci USA 79: 6532–6536 Brudler R, de Groot HJM, van Liemt WBS, Steggerda WF, Esmeijer R, Gast P, Hoff AJ, Lugtenburg J and Gerwert K (1994) Asymmetric binding of the 1- and 4-C=O groups of QA in Rhodobacter sphaeroides R26 reaction centers monitored by Fourier transform infra-red spectroscopy using site-specific isotopically labelled ubiquinone-10. EMBO J 13: 5523–5530 Brudler R, de Groot HJM, van Liemt WBS, Hoff AJ, Lugtenburg JL and Gerwert K (1995) FTIR spectroscopy shows weak symmetric hydrogen bonding of the QB carbonyl groups in Rhodobacter sphaeroides R26 reaction centres. FEBS Lett 370: 2–14 Burie J-R, Boussac A, Boullas C, Berger G, Mattioli T, Mioskowski C, Nabedryk E and Breton J (1995) FTIR spectroscopy of UVgenerated quinone radicals: Evidence for an intramolecular hydrogen atom transfer in ubiquinone, naphthoquinone, and plastoquinone. J Phys Chem B 99: 4059–4070 Burie J-R, Boullais C, Nonella M, Mioskowski C, Nabedryk E and Breton J (1997) Importance of the conformation of methoxy groups on the vibrational and electrochemical properties of ubiquinones. J Phys Chem B 101: 6607–6617 Burley SK and Petsko GA (1985) Aromatic-aromatic interaction: A mechanism of protein structure stabilization. Science 229: 23–28 Butler WF, Johnston DC, Shore HB, Fredkin DR, Okamura MY and Feher G (1980) The electronic structure of Fe2+ in reaction centers from Rhodopseudomonas sphaeroides I. Static magnetization measurements. Biophys J 32: 967–992 Butler WF, Calvo R, Fredkin DR, Isaacson RA, Okamura MY and Feher G (1984) The electronic structure of Fe2+ in reaction centers from Rhodopseudomonas sphaeroides III. EPR measurements of the reduced acceptor complex. Biophys J 45: 947–973 Chang C-H, El-Kabbani O, Tiede D, Norris J and Schiffer M (1991) Structure of the membrane-bound protein photosynthetic reaction center from Rhodobacter sphaeroides. Biochemistry 30: 5352–5360
Colin A. Wraight and Marilyn R. Gunner Chirino AJ, Lous EJ, Huber M, Allen JP, Schenck CC, Paddock ML, Feher G and Rees DC (1994) Crystallographic analyses of site-directed mutants of the photosynthetic reaction center from Rhodobacter sphaeroides. Biochemistry 33: 4584–4593 Coleman WJ, Bylina EJ and Youvan DC (1990a) Reconstitution of photochemical activity in Rhodobacter capsulatus reaction centers containing mutations at trytophan M-250 in the primary quinone binding site. In: Baltscheffsky M (ed) Current Research in Photosynthesis, Vol 1, pp 149–152. Kluwer Academic Publishers, Dordrecht Coleman WJ, Youvan DC, Aumeier W, Eberl U, Volk M, Lang E, Siegl J, Heckmann R, Lersch W, Ogrodnik A and Michel-Beyerle ME (1990b) How conclusive is mutagenic replacement of Trp M250 in photosynthetic reaction centers? In: Baltscheffsky M (ed) Current Research in Photosynthesis, Vol 1, pp 153–156. Kluwer Academic Publishers, Dordrecht De Groot HJM (1995) Asymmetric primary quinone hydrogen bonding in Rhodobacter sphaeroides reaction centers. In: Mathis P (ed) Photosynthesis: From Light to Biosphere, I, pp 401–406. Kluwer Academic Publishers, Dordrecht Deisenhofer J and Michel H (1989a) The photosynthetic reaction center from the purple bacterium Rhodopseudomonas viridis. Science 245: 1463–1473 Deisenhofer J and Michel H (1989b) The photosynthetic reaction centre from the purple bacterium Rhodopseudomonas viridis. EMBO J 8: 2149–2170 Deisenhofer J and Michel H (1991) High-resolution structures of photosynthetic reaction centers. Annu Rev Biophys Chem 20: 247–266 Deisenhofer J, Epp O, Miki R, Huber R and Michel H (1985) Structure of the protein subunits in the photosynthetic reaction center of Rhodopseudomonas viridis at 3 Å resolution. Nature 318: 618–624 Deng H and Callender R (1999) Raman spectroscopic studies of the structure, energetics, and bond distortions of substrates bound to enzymes. Methods Enzymol 308: 176–201 DeVault D (1980) Quantum-mechanical tunneling in biological systems. Q Rev Biophys 13: 387–564 El-Kabbani O, Chang C-H, Tiede D, Norris J and Schiffer M (1991) Comparison of reaction centers from Rhodobacter sphaeroides and Rhodopseudomonas viridis: Overall architecture and protein-pigment interactions. Biochemistry 30: 5361–5369 Ermler U, Fritzsch G, Buchanan SK and Michel H (1994) Structure of the photosynthetic reaction center from Rhodobacter sphaeroides at 2.65 Å resolution: Cofactors and protein-cofactor interactions. Structure 2: 925–936 Feher G, Okamura MY and McElroy JD (1972) Identification of an electron acceptor in reaction centers of Rhodopseudomonas spheroides by EPR spectroscopy. Biochim Biophys Acta 267: 222–226 Feher G, Isaacson RA, Okamura MY and Lubitz W (1985) ENDOR of semiquinones in reaction centers from Rhodopseudomonas sphaeroides. In: Michel-Beyerle ME (ed) Antennas and Reaction Centers of Photosynthetic Bacteria—Structure, Interactions and Dynamics, pp 174–189. Springer-Verlag, Berlin Feher G, Isaacson RA, Okamura MY and Lubitz W (1988). ENDOR of exchangeable protons of the reduced intermediate acceptor in reaction centers from Rhodobacter sphaeroides R-26. In: Breton J and Verméglio A (eds) The Photosynthetic Bacterial Reaction Center, Cadarache, France, Plenum Press
Chapter 20
Acceptor Quinones of Purple Photosynthetic Bacteria
Ferreira KN, Iverson TM, Maghlaoui K, Barber J and Iwata S (2004) Architecture of the photosynthetic oxygen-evolving center. Science 303: 1831–1838 Fisher N and Rich PR (2000) A motif for quinone binding sites in respiratory and photosynthetic systems. J Mol Biol 296: 1153–1162 Flett MSC (1948) The application of infra-red spectroscopy to structural problems in the anthraquinone field. J Chem Soc 1948: 1441–1448 Flores M, Isaacson RA, Calvo R, Feher G and Lubitz W (2003) Probing hydrogen bonding to quinone anion radicals by 1H and 2H ENDOR spectroscopy at 35 GHz. Chem Phys 294: 401–413 Flores M, Isaacson R, Abresch E, Calvo R, Lubitz W and Feher G (2006) Protein-cofactor interactions in bacterial reaction centers from Rhodobacter sphaeroides R-26: I. Identification of the ENDOR lines associated with the hydrogen bonds to the primary quinone QA–. Biophys J 90: 3356–3362 Flores M, Isaacson R, Abresch E, Calvo R, Lubitz W and Feher G (2007) Protein-cofactor interactions in bacterial reaction centers from Rhodobacter sphaeroides R-26: II. Geometry of the hydrogen bonds to the primary quinone QA•– by 1H and 2H ENDOR spectroscopy. Biophys J 92: 671–682 Franzen S, Goldstein RF and Boxer SG (1990) Electric field modulation of electron transfer reaction rates in isotropic systems: Long distance charge recombination in photosynthetic reaction centers. J Phys Chem 94: 5135–5149 Fritscher J, Prisner TF and MacMillan F (2006) Theoretical investigation of QA–ligand interactions in bacterial reaction centers of Rhodobacter sphaeroides. Appl Mag Res 30: 251–268 Fritzsch G, Koepke J, Diem R, Kuglstatter A and Baciou L (2002) Charge separation induces conformational changes in the photosynthetic reaction centre of purple bacteria. Acta Cryst D 58: 1660–1663 Gast P, Michalski, TJ, Hunt JE and Norris JR (1985) Determination of the amount and the type of quinones present in single crystals from reaction center protein from the photosynthetic bacterium Rhodopseudomonas viridis. FEBS Lett 179: 325–328 Giangiacomo KM and Dutton PL (1989) In photosynthetic reaction centers, the free energy difference for electron transfer between quinones bound at the primary and secondary quinonebinding sites governs the observed secondary site specificity. Proc Natl Acad Sci USA 86: 2658–2662 Gordy W (1940) Spectroscopic evidence for hydrogen bonds: Effects of chelation on the carbonyl frequency. J Chem Phys 8: 516–519 Grafton AK and Wheeler RA (1999) Amino acid protonation states determine binding sites of the secondary ubiquinone and its anion in the Rhodobacter sphaeroides photosynthetic reaction center. J Phys Chem B 103: 5380–5387 Graige MS, Paddock ML, Bruce JM, Feher G and Okamura MY (1996) Mechanism of proton-coupled electron transfer for quinone (QB) reduction in reaction centers of Rb. sphaeroides. J Am Chem Soc 118: 9005–9016 Graige MS, Feher G and Okamura MY (1998) Conformational gating of the electron transfer reaction QA– •QB → QAQB– • in bacterial reaction centers of Rhodobacter sphaeroides determined by a driving force assay. Proc Natl Acad Sci USA 95: 11679–11684 Graige MS, Paddock ML, Feher G and Okamura MY (1999) Observation of the protonated semiquinone intermediate in
401
isolated reaction centers from Rhodobacter sphaeroides: Implications for the mechanism of electron and proton transfer in proteins. Biochemistry 38: 11465–11473 Gunner MR and Dutton PL (1989) Temperature and –∆G dependence of the electron transfer from BPh•– to QA in reaction center protein from Rhodobacter sphaeroides with different quinones as QA. J Am Chem Soc 111: 3400–3412. Gunner MR, Braun BS, Bruce JM and Dutton PL (1985) The characterization of the QA binding site of the reaction center of Rhodopseudomonas sphaeroides. In: Michel-Beyerle ME (ed) Antennas and Reaction Centers of Photosynthetic Bacteria: Structure, Interactions and Dynamics, pp 298–304. SpringerVerlag, Berlin Gunner MR, Robertson DE and Dutton PL (1986) Kinetic studies on the reaction center protein from Rhodopseudomonas sphaeroides: The temperature and free energy dependence of electron transfer between various quinones in the QA site and the oxidized bacteriochlorophyll dimer. J Phys Chem 90: 3783–3795 Gunner MR, Saleh MA, Cross E, ud-Doula A and Wise M (2000) Backbone dipoles generate positive potential in all proteins: Origins and implications of the effect. Biophys J 78: 1126-1144 Heathcote P, Fyfe PK and Jones MR (2002) Reaction centres: The structure and evolution of biological solar power. Trends Biochem Sci 27: 79–87 Heinen U, Utschig LM, Poluektov OG, Link G, Ohmes E and Kothe G (2007) Structure of the charge separated state P865+QA– in the phootosynthetic reaction centers of Rhodobacter sphaeroides by quantum beat oscillations and high-field electron paramagnetic resonance: Evidence for light-induced QA– reorientation. J Am Chem Soc 129: 15935–15946 Hucke O, Schmid R and Labahn A (2002) Exploring the primary electron acceptor (QA)-site of the bacterial reaction center from Rhodobacter sphaeroides. Binding mode of vitamin K derivatives. Eur J Biochem 269: 1096–1108 Isaacson RA, Lendzian F, Abresch C, Lubitz W and Feher G (1995) Electronic structure of Q–A in reaction centers from Rhodobacter sphaeroides. I. Electron paramagnetic resonance in single crystals. Biophys J 69: 311–322 Isaacson RA, Abresch EC, Lendzian F, Boullais C, Paddock ML, Mioskowski C, Lubitz W and Feher G (1996). Asymmetry of the binding sites of QA– and QB– in reaction centers of Rb. sphaeroides probed by Q-Band EPR with 13C-labeled quinones. In: Michel-Beyerle ME (ed) The Reaction Center of Photosynthetic Bacteria: Structure and Dynamics, pp 353–368. Springer-Verlag, Berlin Ishikita H, Morra G and Knapp E-W (2003) Redox potential of quinones in photosynthetic reaction centers from Rhodobacter sphaeroides: Dependence on protonation of Glu-L212 and Asp-L213. Biochemistry 42: 3882–3892 Kacprzak S and Kaupp M (2004) Electronic g-tensors of semiquinones in photosynthetic reaction centers. A density functional study. J Phys Chem B 108: 2464–2469 Katona G, Snijder A, Gourdon P, Andréasson U, Hansson Ö, Andréasson L-E and Neutze R (2005) Conformational regulation of charge recombination in a photosynthetic bacterial reaction center. Nature Struct Mol Biol 12: 630-631 Kaupp M (2002) The function of Photosystem I. Quantum chemical insight into the role of tryptophan-quinone interactions. Biochemistry 41: 2895–2900
402 Kim J, Mao J and Gunner MR (2005) Are acidic and basic groups in buried proteins predicted to be ionized? J Mol Biol 348: 1283–1298 Kirmaier C and Holten D (1990) Evidence that a distribution of bacterial reaction centers underlies the temperature and detection-wavelength dependence of the rates of the primary electrontransfer reactions. Proc Natl Acad Sci USA 87: 3552–3556 Kleinfeld D, Okamura MY and Feher G (1984a) Electron-transfer kinetics in photosynthetic reaction centers cooled to cryogenic temperatures in the charge separated state: Evidence for lightinduced structural changes. Biochemistry 23: 5780–5786 Kleinfeld D, Okamura MY and Feher G (1984b) Electron transfer in reaction centers of Rhodopseudomonas sphaeroides: I. Determination of the charge recombination pathway of D+QAQB– and free energy and kinetic relations between QA– QB and QAQB–. Biochim Biophys Acta 766: 126–140 Koepke J, Krammer EM, Klingen AR, Sebban P, Ullmann GM and Fritzsch G (2007) pH modulates the quinone position in the photosynthetic reaction center from Rhodobacter sphaeroides in the neutral and charge separated states. J Mol Biol 371: 396–409 Krivanek R, Kern J, Zouni A, Dau H and Haumann M (2007) Spare quinones in the QB cavity of crystallized Photosystem II from Thermosynechococcus elongatus. Biochim Biophys Acta 1767: 520–527 Kruk J, Strzalka K and Leblanc RM (1993) Fourier transform infrared studies on charge-transfer interactions of plastoquinones and α-tocopherol quinone with their hydroquinone forms and monogalactosyldiacylglycerol. Biophysical Chemistry 45: 235–244 Kuglstatter A, Ermler U, Michel H, Baciou L and Fritzsch G (2001) X-ray structure analysis of photosynthetic reaction center variants from Rhodobacter sphaeroides: Structural changes induced by point mutations at position L209 modulate electron and proton transfer. Biochemistry 40: 4253–4260 Labahn A, Bruce JM, Okamura MY and Feher G (1995) Direct charge recombination from D+QAQB– to DQAQB in bacterial reaction centers from Rhodobacter sphaeroides containing low potential quinone in the QA site. Chem Phys 197: 355–366 Laible PD, Kirmaier C, Udawatte CS, Hofman SJ, Holten D and Hanson DK (2003) Quinone reduction via secondary Bbranch electron transfer in mutant bacterial reaction centers. Biochemistry 42: 1718–1730 Lancaster CRD (1998) Ubiquinone reduction and protonation in photosynthetic reaction centres from Rhodopseudomonas viridis — X-ray structures and their functional implications. Biochim Biophys Acta 1365: 143–150 Lancaster CRD and Michel H (1996a) New Insights into the X-ray structure of the reaction center of Rhodopseudomonas viridis. In: Michel-Beyerle ME (ed) The Reaction Center of Photosynthetic Bacteria: Structure and Dynamics, pp 23–35. Springer-Verlag, Berlin Lancaster CRD and Michel H (1996b) Three-dimensional structures of photosynthetic reaction centers. Photosynth Res 48: 65–74 Lancaster CRD and Michel H (1997) The coupling of lightinduced electron transfer and proton uptake as derived from crystal structures of reaction centres from Rhodopseudomonas viridis modified at the binding site of the secondary quinone, QB. Structure 5: 1339–1359 Lancaster CRD, Michel H, Honig B and Gunner MR (1996)
Colin A. Wraight and Marilyn R. Gunner Calculated coupling of electron and proton transfer in the photosynthetic reaction center of Rhodopseudomonas viridis. Biophys J 70: 2469–2492 Lavergne J, Matthews C and Ginet N (1999) Electron and proton transfer on the acceptor side of the reaction center in chromatophores of Rhodobacter capsulatus: Evidence for direct protonation of the semiquinone state of QB. Biochemistry 38: 4542–4552 Leigh JS and Dutton PL (1972) The primary electron acceptor in photosynthesis. Biochem Biophys Res Comm 46: 414–421 Lendzian F, Rautter J, Käß H, Gardiner A and Lubitz W (1996) ENDOR and pulsed EPR studies of photosynthetic reaction centers: Protein-cofactor interactions. Ber Bunsenges Phys Chem 100: 2036–2040 Li J, Gilroy D, Tiede DM and Gunner MR (1998) Kinetic phases in the electron transfer from P+QA– QB to P+QAQB– and the associated processes in Rhodobacter sphaeroides R-26 reaction centers. Biochemistry 37: 2818–2829 Li J, Takahashi E and Gunner MR (2000) –∆GABo and pH dependence of the electron transfer from P+QA– QB to P+QAQB– in Rhodobacter sphaeroides reaction centers. Biochemistry 39: 7445–7454 Lin X, Murchison HA, Nagarajan V, Parson WW, Allen JP and Williams JC (1994) Specific alterations of the oxidation potential of the electron donor in reaction centers from Rhodobacter sphaeroides. Proc Natl Acad Sci USA 91: 10265–10269 Loll B, Kern J, Saenger W, Zouni A and Biesiadka J (2005) Towards complete cofactor arrangement in the 3.0 Å resolution structure of photosystem II. Nature 438: 1040–1044 Lubitz W and Feher G (1999) The primary and secondary acceptors in bacterial photosynthesis III. Characterization of the quinone radicals QA–• and QB–• by EPR and ENDOR. Appl Magn Res 17: 1–48 Lubitz W, Abresch EC, Debus RJ, Isaacson RA, Okamura MY and Feher G (1985) Electron nuclear double resonance of semiquinones in reaction centers of Rhodopseudomonas sphaeroides. Biochim Biophys Acta 808: 464–469 Mancino LJ, Dean DP and Blankenship RE (1984) Kinetics and thermodynamics of the P870+QA– → P870+QB– reaction in isolated reaction centers from the photosynthetic bacterium Rhodopseudomonas sphaeroides. Biochim Biophys Acta 764: 46–54 Marcus RA (1964) Chemical and electrochemical electron-transfer theory. Ann Rev Phys Chem 15: 155–196 Marcus RA and Sutin N (1985) Electron transfers in chemistry and biology. Biochim Biophys Acta 811: 265–322 McComb JC, Stein RR and Wraight CA (1990) Investigations on the influence of headgroup substitution and isoprene side-chain length in the function of primary and secondary quinones of bacterial reaction centers. Biochim Biophys Acta 1015: 156–171 McMahon BH, Müller JD, Wraight CA and Nienhaus GU (1998) Electron transfer and protein dynamics in the photosynthetic reaction center. Biophys J 74: 2567–2587 Meyerson ML (1985) A quantitative model for carbonyl infrared frequencies of substituted quinones. Spectrochim Acta A 41: 1263–1267 Mezzetti A, Leibl W, Breton J and Nabedryk E (2003) Photoreduction of the quinone pool in the bacterial photosynthetic membrane: Identification of infrared marker bands for quinol formation. FEBS Lett 537: 161–165 Moser CC, Keske JM, Warncke K, Farid RS and Dutton PL (1992)
Chapter 20
Acceptor Quinones of Purple Photosynthetic Bacteria
Nature of biological electron transfer. Nature 355: 796–802 Moser CC, Page CC, Farid R and Dutton PL (1995) Biological electron transfer. J Bioenerg Biomemb 27: 263–274 Moser CC, Page CC, Cogdell RJ, Barber J, Wraight CA and Dutton PL (2003) Length, time and energy scales of photosystems. Advances Protein Chem 63: 71–109 Nabedryk E and Breton J (2008) Coupling of electron transfer to proton uptake at the QB site of the bacterial reaction center: A perspective from FTIR difference spectroscopy. Biochim Biophys Acta, in press Nogi T, Fathir I, Kobayashi M, Nozawa T and Miki K (2000) Crystal structures of photosynthetic reaction center and highpotential iron-sulfur protein from Thermochromatium tepidum: Thermostability and electron transfer. Proc Nat Acad Sci USA 97: 13561–13566 Nonella M (1997) Structure and vibrational spectra of p-benzoquinone in different oxidation and protonation states: A density functional study. J Phys Chem B 101: 1235–1246 Nonella M (1998) A quantum chemical investigation of structures, vibrational spectra and electron affinities of the radicals of quinone model compounds. Photosynth Res 55: 253–259 Nonella M and Brändli C (1996) Density functional investigation of methoxy-substituted p-benzoquinones: Conformational analysis and harmonic force fields of 2-methoxy- and 2,3-dimethoxy-1,4-benzoquinone. J Phys Chem 100: 14549–14559 Nonella M, Mathias G, Eichinger M and Tavan P (2003) Structures and vibrational frequencies of the quinones of Rb. sphaeroides derived by a combined density functional/molecular mechanics approach. J Phys Chem B 107: 316–322 Okamura MY and Feher G (1986) Isotope effect on electron transfer in reaction centers from Rhodopseudomonas sphaeroides. Proc Natl Acad Sci USA 83: 8152–8156 Okamura MY and Feher G (1992) Proton transfer in reaction centers from photosynthetic bacteria. Annu Rev Biochem 61: 861–896 Okamura MY and Feher G (1995) Proton-coupled electron transfer reactions of QB in reaction centers from photosynthetic bacteria. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 577–593. Kluwer Academic Publishers Okamura MY, Isaacson RA and Feher G (1975) The primary acceptor in bacterial photosynthesis: Obligatory role of ubiquinone in photoactive reaction centers of Rhodopseudomonas sphaeroides. Proc Natl Acad Sci USA 72: 3492–3496 Okamura MY, Paddock ML, Graige MS and Feher G (2000) Proton and electron transfer in bacterial reaction centers. Biochim Biophys Acta 1458: 148–163 O’Malley PJ (1997a) A density functional study of the effect of reduction on the geometry and electron affinity of hydrogen bonded 1,4-benzoquinone. Implications for quinone reduction and protonation in photosynthetic reaction centers. Chem Phys Lett 274: 251–254 O’Malley PJ (1997b) Hybrid density functional study of the pbenzosemiquinone anion radical: The influence of hydrogen bonding on geometry and hyperfine couplings. J Phys Chem A 101: 6334–6338 O’Malley PJ (2001) Electronic structure studies of quinones and semiquinones: Accurate calculation of spin densities and electron paramagnetic resonance parameters. Antioxid Redox Signal 3: 825–838
403
O’Malley PJ (2003) The origin of the spin density asymmetry at the QA binding site of type II photosynthetic reaction centres. Chem Phys Lett 379: 277–281 Paddock ML, Abresch EC, Isaacson RA, Lubitz W, Okamura MY and Feher G (1999) Identification of hydrogen bonds to QA– in RCs of Rb. sphaeroides by ENDOR spectroscopy. Biophys J 76: A141 Paddock ML, Feher G and Okamura MY (2003) Proton transfer pathways and mechanism in bacterial reaction centers. FEBS Lett 555: 45–50 Paddock ML, Flores M, Isaacson R, Chang C, Selvaduray P, Feher G and Okamura MY (2005) Hydrogen bond reorientation upon QB reduction revealed by ENDOR spectroscopy in reaction centers from Rhodobacter sphaeroides. Biophys J 88: 204a Paddock ML, Flores M, Isaacson R, Chang C, Abresch EC, Selvaduray P and Okamura MY (2006) Trapped conformational states of semiquinone (D+•QB–•) formed by B-branch electron transfer at low temperature in Rhodobacter sphaeroides reaction centers. Biochemistry 45: 14032–14042 Paddock ML, Flores M, Isaacson R, Chang C, Abresch EC and Okamura MY (2007) ENDOR spectroscopy reveals light induced movement of the H-bond from Ser-L223 upon forming the semiquinone (QB–•) in reaction centers from Rhodobacter sphaeroides. Biochemistry 46: 8234–8243 Parson WW (1978) Quinones as secondary electron acceptors. In: Clayton RK and Sistrom WR (eds) The Photosynthetic Bacteria, pp 455–469. Plenum, New York Plato M, Michel-Beyerle ME, Bixon M and Jortner J (1989) On the role of tryptophan as a superexchange mediator for quinone reduction in photosynthetic reaction centers. FEBS Lett 249: 70–74 Pokkuluri PR, Laible PD, Deng Y-L, Wong TN, Hanson DK and Schiffer M (2002) The structure of a mutant photosynthetic reaction center shows unexpected changes in main chain orientations and quinone position. Biochemistry 41: 5998–6007 Pokkuluri PR, Laible PD, Crawford AE, Mayfield JF, Yousef MA, Ginell SL, Hanson DK and Schiffer M (2004) Temperature and cryoprotectant influence secondary quinone binding position in bacterial reaction centers. FEBS Lett 570: 171–174 Prince RC, Dutton PL and Bruce JM (1983) Electrochemistry of ubiquinones, menaquinones and plastoquinones in aprotic solvents. FEBS Lett 160: 273–276 Prince RC, Halbert TR and Upton TH (1988) Structural influences on the electrochemistry of ubiquinone. In: Kim CH, Tedeschi H, Diwan JJ and Salerno JC (eds) Advances in Membrane Biochemistry and Bioenergetics, pp 469–478. Plenum Press, New York Rabenstein B, Ullmann GM and Knapp E-W (2000) Electron transfer between the quinones in the photosynthetic reaction center and its coupling to conformational changes. Biochemistry 39: 10487–10496 Rao CNR, Dwivedi PC, Ratajczak H and Orville-Thomas WJ (1975) Relation between O-H stretching frequency and hydrogen bond energy: Re-examination of the Badger-Bauer rule. J Chem Soc Faraday Trans II 71: 955–966 Rasmussen RS, Tunnicliff DD and Brattain RR (1949) Infrared and ultraviolet spectroscopic studies on ketones. J Am Chem Soc 71: 1068–1072 Remy A, Boers RB, Egorova-Zachernyuk T, Gast P, Lugtenberg J and Gerwert K (2003) Does different orientation of the methoxy groups of ubiquinone-10 in the reaction center of
404 Rhodobacter sphaeroides cause different binding at QA and QB? Eur J Biochem 270: 3603–3609 Robinson HH and Kahn SD (1990) Interplay of substitutent conformation and electron affinity in quinone models of quinone reductases. J Am Chem Soc 112: 4728–4731 Savitsky A, Dubinskii AA, Flores M, Lubitz W and Möbius K (2007) Orientation-resolving pulsed electron dipolar high-field EPR spectroscopy on disordered solids: I. Structure of spincorrelated radical pairs in bacterial photosynthetic reaction centers. J Phys Chem B 111: 6245–6262 Schenck CC, Parson WW, Holten D, Windsor MW and Sarai A (1981) Temperature dependence of electron transfer between bacteriopheophytin and ubiquinone in protonated and deuterated reaction centers of Rhodopseudomonas sphaeroides. Biophys J 36: 479–489 Schmid R and Labahn A (2000) Temperature and free energy dependence of the direct charge recombination rate from the secondary quinone in bacterial reaction centers from Rhodobacter sphaeroides. J Phys Chem B 104: 2928–2936 Schmid R, Goebel F, Warnecke A and Labahn A (1999) Synthesis and redox potentials of methylated vitamin K derivatives. J Chem Soc, Perkin Trans 2 1199–1202 Schuler RH, Tripathi GNR, Prebenda MF and Chipman DM (1983) Resonance Raman and molecular orbital studies of the effects of deuteration on the vibrational structure of the p-benzoquinone radical anion. J Phys Chem 87: 5357–5361 Sebban P, Maróti P, Schiffer M and Hanson DK (1995) Electrostatic dominoes: Long distance propagation of mutational effects in photosynthetic reaction centers of Rhodobacter capsulatus. Biochemistry 34: 8390–8397 Shinkarev VP and Wraight CA (1993) Electron and proton transfer in the acceptor quinone complex of reaction centers of phototrophic bacteria. In: Deisenhofer J and Norris JR (eds) The Photosynthetic Reaction Center, 1, pp 193–255. Academic Press, San Diego Shinkarev VP and Wraight CA (1997) The interaction of quinone and detergent with reaction centers of purple bacteria. I. Slow quinone exchange between reaction center micelles and pure detergent micelles. Biophys J 72: 2304–2319 Shopes RJ and Wraight CA (1985) The acceptor quinone complex of Rhodopseudomonas viridis reaction centers. Biochim Biophys Acta 806: 348–356 Sinnecker S, Flores M and Lubitz W (2006) Protein-cofactor interactions in bacterial reaction centers from Rhodobacter sphaeroides R-26: Effect of hydrogen bonding on the electronic and geometric structure of the primary quinone. A density functional theory study. Phys Chem Chem Phys 8: 5659–5670 Slifkin MA and Walmsley RH (1969) Infra-red studies of quinhydrone type complexes. Spectrochim Acta A 26: 1237–1242 Spoyalov AP, Hulsebosch RJ, Shochat S, Gast P and Hoff AJ (1996) Evidence that Ala M260 is hydrogen-bonded to the reduced primary acceptor quinone QA–• in reaction centers of Rb. sphaeroides. Chem Phys Lett 263: 715–720 Stilz HU, Finkele U, Holzapfel W, Lauterwasser C, Zinth W and Oesterhelt D (1994) Influence of M subunit Thr222 and Trp252 on quinone binding and electron transfer in Rhodobacter sphaeroides reaction centres. Eur J Biochem 223: 233–242 Stowell MHB, McPhillips TM, Rees DC, Soltis SM, Abresch E and Feher G (1997) Light-induced structural changes in photosynthetic reaction center: Implications for mechanism of electron-proton transfer. Science 276: 812–816
Colin A. Wraight and Marilyn R. Gunner Sundberg RJ and Martin RB (1974) Interactions of histidine and other imidazole derivatives with transition metal ions in chemical and biological systems. Chem Rev 74: 471–517 Taly A, Sebban P, Smith JC and Ullmann GM (2003) The position of QB in the photosynthetic reaction center depends on pH: A theoretical analysis of the proton uptake upon QB reduction. Biophys J 84: 2090–2098. Tandori J, Sebban P, Michel H and Baciou L (1999) In Rhodobacter sphaeroides reaction centers, mutation of proline L209 to aromatic residues in the vicinity of a water channel alters the dynamic coupling between electron and proton transfer processes. Biochemistry 38: 13179–13187 Tandori J, Maróti P, Alexov E, Sebban P and Baciou L (2002) Key role of proline L209 in connecting the distant quinone pockets in the reaction center of Rhodobacter sphaeroides. Proc Natl Acad Sci USA 99: 6702–6706 Thijs R and Zeegers-Huyskens T (1984) Infrared and Raman studies of hydrogen bonded complexes involving acetone, acetophenone and benzophenone. I. Thermodynamic constants and frequency shifts of the vOH and vC=O stretching vibrations. Spectrochim Acta 40A: 307–313 Tiede DM, Vazquez J, Cordova J and Marone PA (1996) Timeresolved electrochromism associated with the formation of quinone anions in the Rhodobacter sphaeroides R26 reaction center. Biochemistry 35: 10763–10775 Utschig LM, Thurnauer MC, Tiede DM and Poluektov OG (2005) Low-temperature interquinone electron transfer in photosynthetic reaction centers from Rhodobacter sphaeroides and Blastochloris viridis: Characterization of QB– states by high-frequency electron paramagnetic resonance (EPR) and electron-nuclear double resonance (ENDOR). Biochemistry 44: 14131–14142 Van den Brink JS, Hulsebosch RJ, Gast P, Hore PJ and Hoff AJ (1994) QA binding in reaction in reaction centers of the photosynthetic purple bacterium Rhodobacter sphaeroides R26: Investigation with electron spin polarization spectroscopy. Biochemistry 33: 13668–13677 Van der Est A, Prisner T, Bittl R, Fromme P, Lubitz W, Möbius K and Stehlik D (1997) Time resolved X-, K- and W- band EPR •+ •– A1 of photosystem I in comparison of the radical pair state P700 •+ •– with P865 QA of bacterial reaction center. J Phys Chem B 101: 1437–1443 Van Liemt WBS, Boender GJ, Gast P, Hoff AJ, Lugtenburg J and De Groot HJM (1995) 13C Magic angle spinning NMR characterization of the functionally asymmetric QA binding in Rhodobacter sphaeroides R26 photosynthetic reaction centers using site-specific 13C-labeled ubiquinone-10. Biochemistry 34: 10229–10236 Wakeham MC, Breton J, Nabedryk E and Jones MR (2004) Formation of a semiquinone at the QB site by A- or B-branch electron transfer in the reaction center from Rhodobacter sphaeroides. Biochemistry 43: 4755–4763 Walden SE and Wheeler RA (2002) Protein conformational gate controlling binding site preference and migration for ubiquinone-B in the photosynthetic reaction center of Rhodobacter sphaeroides. J Phys Chem B 106: 3001-3006 Warncke K and Dutton PL (1992) Both carbonyl groups are necessary for strong interaction of the semiquinone at the QA site of photosynthetic reaction center protein. Biophys J 57: 571a Warncke K and Dutton PL (1993a) Experimental resolution of the free energies of aqueous solvation contributions to
Chapter 20
Acceptor Quinones of Purple Photosynthetic Bacteria
ligand-protein binding: quinone-QA site interactions in the photosynthetic reaction center protein. Proc Natl Acad Sci USA 90: 2920–2924 Warncke K and Dutton PL (1993b) Influence of QA site cofactor structure on equilibrium binding, in situ electrochemistry, and electron-transfer performance in the photosynthetic reaction center protein. Biochemistry 32: 4769–4779 Warncke K, Gunner MR and Dutton PL (1987) Effect of hydrocarbon tail structure on the affinity of substituted quinones for the QA site in reaction centers of Rhodopseudomonas sphaeroides R-26. In: Biggins J (ed) Progress in Photosynthesis Research, pp 217–220. Martinus Nijhoff, Dordrecht Warncke K, Gunner MR, Braun BS, Gu L, Yu C-A, Bruce JM and Dutton PL (1994) Influence of hydrocarbon tail structure on quinone binding and electron-transfer performance at the QA and QB sites of the photosynthetic reaction center protein. Biochemistry 33: 7830–7841 Wells TA, Takahashi E and Wraight CA (2003) Protein control of the redox potential of the primary quinone acceptor in reaction centers from Rhodobacter sphaeroides. Biochemistry 42: 4064–4074 Wheeler RA (2001) Quinones and quinoidal radicals in photosynthesis. In: Eriksson LA (ed) Theoretical Biochemistry — Processes and Properties, pp 655–690. Elsevier, Amsterdam Williams JC, Steiner LA and Feher G (1986) Primary structure of the reaction center from Rhodopseudomonas sphaeroides. Proteins 1: 312–325 Williams JC, Paddock ML, Feher G and Allen JP (1991) Effects of iron ligand substitutions in reaction centers from Rhodobacter sphaeroides. Biophys J 59: 142a Woodbury NW and Allen JP (1995) The pathway, kinetics and thermodynamics of electron transfer in wild type and mutant reaction centers of purple nonsulfur bacteria. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria, pp 527–557. Kluwer Academic Publishers, Dordrecht Woodbury NW, Becker M, Middendorf D and Parson WW (1985) Picosecond kinetics of the initial photochemical electrontransfer reaction in bacterial photosynthetic reaction centers. Biochemistry 24: 7516–7521 Woodbury NW, Parson WW, Gunner MR, Prince RC and Dutton PL (1986) Radical-pair energetics and decay mechanisms in reaction center containing anthraquinones or benzoquinones in place of ubiquinone. Biochim Biophys Acta 851: 6–22 Wraight CA (1978) Iron-quinone interactions in the electron acceptor region of bacterial photosynthetic reaction centers. FEBS Lett 93: 283—288 Wraight CA (1979) Electron acceptors of bacterial photosynthetic reaction centers II. H+ binding coupled to secondary electron transfer in the quinone acceptor complex. Biochim Biophys Acta 548: 309–327 Wraight CA (1982) The involvement of stable semiquinones in the two-electron gates of plant and bacterial photosystems. In:
405
Trumpower BL (ed) Function of Quinones in Energy Conserving Systems, pp 181–197. Academic Press, New York Wraight CA (2004) Proton and electron transfer in the acceptor quinone complex of bacterial photosynthetic reaction centers. Frontiers Biosci 9: 309–327 Wraight CA (2005) Intraprotein proton transfer — Concepts and realities from the bacterial photosynthetic reaction center. In: Wikström M (ed) Biophysical and Structural Aspects of Bioenergetics, pp 273–313. Royal Society of Chemistry, Cambridge, U.K. Xu Q and Gunner MR (2000) Temperature dependence of the free energy, enthalpy, and entropy of P+QA– charge recombination in Rhodobacter sphaeroides R-26 reaction centers. J Phys Chem B 104: 8035–8043 Xu Q and Gunner MR (2001) Trapping conformational intermediate states in the reaction center protein from photosynthetic bacteria. Biochemistry 40: 3232–3241 Xu Q and Gunner MR (2002) Exploring the energy profile of the QA– to QB electron transfer reaction in bacterial photosynthetic reaction centers: pH dependence of the conformational gating step. Biochemistry 41: 2694–2701 Xu Q, Baciou L, Sebban P and Gunner MR (2002) Exporing the energy landscape for QA– to QB electron transfer in bacterial photosynthetic reaction centers: Effect of substrate position and tail length on the conformational gating step. Biochemistry 41: 10021–10025 Zachariae U and Lancaster CRD (2001) Proton uptake associated with the reduction of the primary quinone QA influences the binding site of the secondary quinone QB in Rhodopseudomonas viridis photosynthetic reaction centres. Biochim Biophys Acta 1505: 280–290 Zadorozhnyi BA and Ishchenko IK (1965) Hydrogen bond energies and shifts of the stretching vibration bands of C=O groups. Optics & spectroscopy (Eng Translation) 19: 306–308 Zhao X and Kitagawa T (1998) Solvent effects of 1,4-benzoquinone and its anion radicals probed by resonance Raman and absorption spectra and their correlation with redox potentials. J Raman Spectrosc 29: 773–780 Zhao X, Imahori H, Zhan C-G, Sakata Y, Iwata S and Kitagawa T (1997a) Resonance Raman and FTIR spectra of isotope-labeled reduced 1,4-benzoquinone and its protonated forms in solution. J Phys Chem A 101: 622–631 Zhao X, Ogura T, Okamura M and Kitagawa T (1997b) Observation of the resonance Raman spectra of the semiquinones QA•– and QB•– in photosynthetic reaction centers from Rhodobacter sphaeroides R26. J Am Chem Soc 119: 5263–5264 Zhu Z and Gunner MR (2005) The energetics of quinone dependent electron and proton transfers in Rhodobacter sphaeroides photosynthetic reaction centers. Biochemistry 44: 82–96 Zhu Z-Y and Karlin S (1996) Clusters of charged residues in protein three-dimensional structures. Proc Natl Acad Sci USA 93: 8350–8355
Chapter 21 Biogenesis of c-type Cytochromes and Cytochrome Complexes Carsten Sandersa, Serdar Turkarslana, Ozlem Ondera, Elaine R. Frawleyb, Robert G. Kranzb, Hans Georg Kochc and Fevzi Daldala* a
University of Pennsylvania, Department of Biology, Plant Science Institute, Philadelphia, PA 19104, U.S.A.; bWashington University, Department of Biology, Saint Louis, 63130 MI, U.S.A.; c University of Freiburg, Institut für Biochemie und Molekularbiologie, Zentrum für Biochemie und Molekulare Zellforschung, 79104 Freiburg, Germany
Summary ............................................................................................................................................................... 407 I. Introduction..................................................................................................................................................... 408 II. Maturation of c-type Cytochromes: Ccm-system I and Ccs-system II .......................................................... 409 A. General Features of Cytochrome c Maturation ................................................................................ 409 B. Heme b Translocation and Delivery ................................................................................................. 410 C. Thio-oxidoreduction of Apocytochrome c ........................................................................................ 412 D. Heme Ligation Complex ................................................................................................................... 414 E. Not All Phototrophs Use Ccm-system I: The Simpler Ccs-System II ............................................... 414 III. Biogenesis of Cytochrome Complexes .......................................................................................................... 415 A. Salient Features of aa3- and cbb3-type Cytochrome c Oxidase in Rhodobacter Species ............... 415 B. Assembly of the aa3-type Cytochrome c Oxidase in Rhodobacter sphaeroides ............................ 417 C. Assembly of the cbb3-type Cytochrome c Oxidase in Rhodobacter Species.................................. 418 Acknowledgments ................................................................................................................................................. 421 References ............................................................................................................................................................ 421
Summary In most anoxygenic phototrophic bacteria, apocytochromes c are synthesized in the cytoplasm, translocated across the membrane and matured to holocytochromes c on the periplasmic side. The extracytoplasmic maturation process requires a complex biogenesis pathway (Ccm-system I) consisting of up to ten components to carry out specific steps. These include translocation and delivery of the heme, chaperoning and thio-oxidoreduction of the apocytochrome c as well as heme ligation steps. Because cytochromes are often part of multi-subunit protein complexes, matured holocytochromes may be assembled into active enzyme complexes. Examples include the ubihydroquinone:cytochrome c oxidoreductase (cytochrome bc1 complex) or the aa3- or cbb3-type cytochrome c oxidases. Assemblies of these complexes are tightly coordinated and often require additional specific biogenesis components. In this chapter, we discuss the maturation process of c-type cytochromes and the assembly pathways of the aa3- and cbb3-type cytochrome c oxidases in anoxygenic phototrophic bacteria, with a specific focus on Rhodobacter species.
*Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 407–423. © 2009 Springer Science + Business Media B.V.
408
Carsten Sanders, Serdar Turkarslan, Ozlem Onder, Elaine R. Frawley, Robert G. Kranz, Hans Georg Koch and Fevzi Daldal
I. Introduction Anoxygenic phototrophic bacteria contain a variety of cytochromes (Cyts) that are hemo-proteins primarily involved in electron transfer for photosynthesis or respiration. Depending on the nature of the heme prosthetic groups that Cyts contain, they are referred to as the a-, b-, c-, o- or d-type Cyts. The prosthetic group of the b- and c-type Cyts is iron-protoporphyrin IX, or heme b, whereas in the remaining Cyts they are heme b derivatives modified at different positions of the tetrapyrrole ring system. Hemes a and o have a farnesyl hydroxyethyl side chain at ring A and heme a has an additional formyl group at ring D. Heme d has two hydroxyl groups at ring C. In a-, b-, d- and o-type Cyts the respective heme groups are axially coordinated (i.e., the heme iron has out of plane coordination to amino acid side chains) and their topological localizations within the proteins are variable (Fig. 1). In the c-type Cyts, which are often subunits of multimeric enzymes, the heme cofactor is covalently and stereo-specifically attached via thioether bonds between its vinyl-2 and -4 groups and the cysteine thiols of a conserved Cys1Xxx1-Xxx2-Cys2-His (CXXCH) motif within the apoprotein. This chapter describes the current state of our understanding of the Cyt c maturation process and of the assembly of mature Cyts (holocyts) c into active multi-subunit membrane protein complexes in Rhodobacter species. Over the past two decades, Rhodobacter (Rba.) species (mainly Rba. capsulatus) have emerged as model organisms for studying Cyt c biogenesis in anoxygenic phototrophic and other Gram-negative bacteria. Rba. capsulatus produces a variety of membrane-bound and soluble c-type Cyts, including the Cyts c1, c2, c´, cy, co and cp, to sustain its versatile growth modes (Davidson and Daldal, 1987; Zannoni and Daldal, 1993; Gray et al., 1994; Koch et al., 1998a). Of these proteins, Cyt c1 (Davidson and Daldal, 1987) and either Cyts c2 or cy are required Abbreviations: aa3-Cox – aa3-type Cyt c oxidase; apoCyt – immature Cyt c without its heme cofactor attached; B. – bacillus; BN-PAGE – blue-native polyacrylamide gel electrophoresis; cbb3-Cox – cbb3-type Cyt c oxidase; Ccm – cytochrome c maturation system I; Ccs – cytochrome maturation system II; Cyt – cytochrome; holo- and apo-CcmE – CcmE with and without its heme cofactor, respectively; holoCyt – mature Cyt c with its heme cofactor attached; pre-apoCyt – apoCyt c precursor with its signal sequence; Ps – photosynthetic; Rba. – Rhodobacter; Res – respiratory; TPR – Tetratricopeptide repeats; WWD – Tryptophan rich motif
for photosynthetic (Ps) growth (Zannoni and Daldal, 1993), while Cyts co and cp are involved in respiratory (Res) growth (Koch et al., 1998a) as subunits of the cbb3-type Cyt c oxidase (cbb3-Cox) (Gray et al., 1994). The cbb3-Cox activity can be directly detected by staining colonies with the Nadi reagents (α-naphthol + dimethylphenylenediamine → indophenol blue + H2O) (Keilin, 1966). Rba. capsulatus mutants deficient in holocyt c production exhibit Ps– and Nadi– dual phenotypes (Koch et al., 1998b), but despite the absence of the cbb3-Cox, they can still grow under Res conditions via a Cyt c-independent hydroquinone oxidase (La Monica and Marrs, 1976; Zannoni et al., 1976). These phenotypes have been crucial for the identification of components that are essential for Cyt c maturation in this species (Kranz, 1989; Beckman et al., 1992; Lang et al., 1996; Monika et al., 1997; Goldman et al., 1998; Deshmukh et al., 2000). In addition, Rba. capsulatus is an organism of choice for studying the biogenesis pathways specific for cbb3-Cox, because it contains only a cbb3-Cox (Gray et al., 1994) and naturally lacks an aa3-type Cyt c oxidase (aa3-Cox), unlike other related anoxygenic phototrophic bacteria such as Rba. sphaeroides. Indeed, exploration of Rba. capsulatus mutants that are Nadi– but still Ps+, led to the genes for structural (Koch et al., 1998b) and assembly components required for the production of an active cbb3-Cox (Koch et al., 2000), as described in this chapter. Rba. sphaeroides also provides an excellent bacterial model to investigate the enzymatic mechanism of and the biogenesis pathway specific to, the aa3-Cox due to its structural similarity to the catalytic core of the mitochondrial Cox (Complex IV) (SvenssonEk et al., 2002; Bratton et al., 2003). Remarkably, homologs of several of the eukaryotic components involved in the assembly of the mitochondrial Cox are also found in Rba. sphaeroides and some of the biogenesis intermediates have been defined (Cao et al., 1992; Hiser and Hosler, 2001; Richter and Ludwig, 2003; Smith et al., 2005). Finally, a great deal of knowledge is available on the biogenesis of the Cyt bc1 complex in Rhodobacter species. However, this topic is beyond the consideration of this chapter and for an overview, the reader is referred to Gray and Dadal (1995), and to the chapters dealing with the Cyt bc1 complex in this book (Chapter 22, Berry et al.; Chapter 23, Kramer et al.).
Chapter 21
Cytochrome c Maturation and Assembly
409
Fig. 1. Chemical structures of cytochrome heme groups. All heme derivatives have the same basic structure formed of a cyclic tetrapyrrole found in a b-type heme (or protoporphyrin IX). The four rings of the porphyrin macrocycle and the different substitutions at various side chains are indicated as A, B, C and D and R1, R2 and R3, respectively. The different chemical moieties occupying the R1, R2 and R3 positions of different hemes are listed on the right. Note that heme d has two OH groups at ring C that are not shown.
II. Maturation of c-type Cytochromes: Ccmsystem I and Ccs-system II A. General Features of Cytochrome c Maturation Cyt c maturation refers to the process of covalent addition of heme b group(s) to immature Cyts c without their heme cofactors attached (apoCyts), yielding holocyts and shares several general features in all organisms. First, cellular compartments where preapoCyts c are synthesized are not identical to those where the holoCyts c act as electron transfer proteins. Thus, pre-apoCyts c are translocated across at least one cellular membrane to the maturation sites, making Cyt c maturation a process that occurs after translation and translocation (see Figs. 2 and 3 for diagrammatic overviews of Ccm and Ccs systems). In anoxygenic phototrophic bacteria, Cyt c maturation occurs in the periplasmic space, which is the functional compartment for holoCyts c (Kranz et al., 1998). Second, apoCyt synthesis and translocation are performed independently from the biosynthesis and transport of heme b. In anoxygenic phototrophic bacteria, heme biosynthesis occurs in the cytoplasm and multiple components have been proposed to transport heme across the cytoplasmic membrane and deliver it to the heme ligation sites (Goldman et al., 1998; Goldman and Kranz, 2001). Third, the CXXCH heme binding motif within the apoCyts, as well as the heme-iron of heme b (Fig. 1), have to be in their reduced states prior to covalent thio-ether bond formation (Kranz et al., 1998). In anoxygenic phototrophic bacteria, several of the components required for the reduction
steps and for delivery of the reduced apoCyts to the heme attachment sites are known (Beckman et al., 1992; Beckman and Kranz, 1993; Monika et al., 1997; Deshmukh et al., 2000). On the other hand, how the heme-iron reduction is accomplished in vivo and which component(s) are involved is unknown. Finally, stereo-specific covalent heme ligation to the CXXCH motif within the apoCyts c is necessary for proper folding into the native holoCyts c conformations (Kranz et al., 1998). The activity that catalyzes heme attachment to apoCyt c is called Cyt c heme lyase or Cyt c synthetase and is proposed to consist of at least three membrane-bound components in Rba. capsulatus (Sanders et al., 2005a). In mitochondria (e.g., yeast or human) the similar activity is associated with apparently a single protein (Dumont et al., 1987; Schaefer et al., 1996). Studies conducted in recent years on Cyt c maturation yielded three evolutionarily distinct enzymatic systems in various organisms (Kranz et al., 1998). The most sophisticated biogenesis machinery (Ccm-system I) is used by α- and γ-proteobacteria, deinococci and mitochondria of plants and protozoa. Ccm-system I consists of up to ten genes designated as ccmABCDEFGHI and ccdA (or dsbD) (Fig. 2 and Table 1). These genes encode membrane-bound components that act on either apoCyts c or heme b subsequent to their translocation across the cytoplasmic membrane to yield holoCyts c (Kranz et al., 1998; Thony-Meyer, 2002; Allen et al., 2003b). The second biogenesis apparatus (Ccs-system II) is present in Gram-positive bacteria, cyanobacteria, chloroplasts of plants and algae and some β-, δ- and ε-proteobacteria and the green sulfur phototrophic
410
Carsten Sanders, Serdar Turkarslan, Ozlem Onder, Elaine R. Frawley, Robert G. Kranz, Hans Georg Koch and Fevzi Daldal
Fig. 2. The Ccm-system I components for c-type Cyt maturation. All components of the c-type Cyt maturation, except the thiol-disulfide oxidoreductase DsbA, are located in the cytoplasmic membrane. Both apoCyt c and heme follow different routes to the heme ligation core complex, composed of CcmI, CcmH and CcmF. Apocyt c is translocated via the Sec pathway; its cysteine thiols in the conserved CXXCH motif are first oxidized by the DsbA-DsbB pathway and then reduced by the Cyt c maturation specific CcdA-CcmG and/or CcmH thio-reductive pathway. CcmI is involved in delivering apoCyt c to the core heme ligation complex via its different domains. Heme is translocated across the membrane, possibly via ABC–type transporter CcmABCD and is covalently attached to the conserved His residue of the heme chaperone CcmE. CcmC is involved in attaching heme to CcmE and CcmD enhances holo-CcmE production. CcmA and CcmB promote release of holo-CcmE from CcmC and CcmD. Upon formation of the thioether bonds between the apoCyt c and the heme vinyls, catalyzed by the CcmH-CcmI-CcmF complex, mature holoCyt c is released. See also Color Plate 9, Fig. 14.
bacterium Chlorobium tepidum. A minimum of four components (CcsA, CcsB, CcsX and DsbD or CcdA) (Fig. 3) are essential for various extra-cytoplasmic steps during Cyt c maturation (Beckett et al., 2000; Le Brun et al., 2000; Kranz et al., 2002); see below for a description of the Ccs-system II). The third biogenesis system (system III) is confined to mitochondria of some lower eukaryotes, invertebrates and vertebrates. It includes either one component with defined Cyt c heme lyase activity, or sometimes two that differ in their apoCyt c substrate specificities (HCCS/CCHL and CC1HL, respectively) (Dumont et al., 1987; Drygas et al., 1989; Zollner et al., 1992). An additional accessory factor (Cyc2) with a proposed heme reductase activity is also involved under some conditions (Dumont et al., 1987; Drygas et al., 1989; Zollner et al., 1992; Bernard et al., 2005). This chapter deals mainly with the Ccm-system I of anoxygenic phototrophic bacteria, which is most intensively studied using Rhodobacter species. In
these species, the Ccm-system I consists so far of ten membrane-bound components (CcmABCDEFGHI and CcdA) (Table 1) that ensure the transport and delivery of heme b to the periplasmic heme ligation sites; in addition these components chaperone the apoCyts following their translocation across the membrane, rendering them competent for ligation via thio-oxidoreduction reactions and they also mediate the recognition and covalent ligation of both apoCyts c and heme b substrates to produce holoCyts c (Table 1) B. Heme b Translocation and Delivery Of the Cyt c biogenesis components, the CcmABCD products have been suggested to form an ATP-binding cassette (ABC) containing transporter complex (Beckman et al., 1992; Goldman et al., 1997, 1998; Schulz et al., 1998) which translocates heme b across the cytoplasmic membrane. CcmA contains
Chapter 21
Cytochrome c Maturation and Assembly
411
Table 1. Ccm-system I cytochrome c biogenesis components Gene name ccmA
Conserved motifs in gene product
Proposed role in Cyt c biogenesis
Salient topological or other properties
Walker-A and -B motifs (LS[A/G]GQ)
ATP-dependent release of holoCcmE from CcmC-CcmD
ABC-type transporter
ccmB
FXXDXXDGSL motif
Involved in release of holoCcmE from CcmC-CcmD
ABC-type transporter, six transmembrane helices
ccmC
Tryptophan rich WWD motif (WGX[F,Y,W]WXWDXRLT)
Heme transfer to CcmE
Six transmembrane helices, two conserved histidines in the first and third periplasmic loop
ccmD
No recognizable motif
Part of CcmABCD complex required for holoCcmE release
Small 52 residue (in Rba. capsulatus) protein with a single hydrophobic helix
ccmE
(LPDLFREG) and(VLAKH*DE) motifs (*covalent heme binding site)
Heme chaperone
One transmembrane helix, a single conserved histidine that binds heme via its vinyl-2 group
ccmF
Tryptophan rich WWD motif, (WGGXWFWDPVEN)
Heme ligation
Eleven transmembrane helices, four conserved histidines
ccmG
CXXC motif
Reduction of apoCyt c CXXCH and/or CcmH LRCXXC motifs
One transmembrane helix, with periplasmic thioredoxin motif
ccmH
LRCXXC
Heme ligation and thio-reduction
One transmembrane helix with periplasmic CXXC motif
ccmI
A cytoplasmic leucine zipperlike and three to four periplasmic TPR-motifs
ApoCyt c chaperone
Two transmembrane helices with a large C-terminal periplasmic loop
ccdA
Two conserved cysteines
Thio-reduction of CcmG
Six transmembrane helices, homologous to the central domain of E. coli DsbD
dsbA
Thioredoxin-fold CXXC
Thio-oxidation of apoCyt c CXXCH motif
Soluble, periplasmic
dsbB
CXXC
Re-oxidation of DsbA
Four transmembrane helices and two periplasmic loops
the conserved ATP binding (Walker A and B) motifs and an additional signature sequence that are required for function (Goldman et al., 1997). CcmE is a monotopic membrane protein with a periplasmic heme binding domain (Reid et al., 1998; Schulz et al., 1998; Thony-Meyer, 2003). Experimental data indicate that CcmC and CcmD are necessary and sufficient for attachment of heme b to the periplasmic heme chaperone CcmE (Goldman et al., 1997, 1998; Goldman and Kranz, 2001; Thony-Meyer, 2003), which is considered to be the final heme donor to the ligation complex. In Rba. capsulatus, CcmB and CcmC, both of which have six transmembrane helices (Goldman et al., 1997; Thony-Meyer, 1997; Goldman et al., 1998), are required for membrane localization of CcmA (Goldman et al., 1997). In addition, both CcmC and CcmD co-immunoprecipitate with CcmA (Goldman et al., 1997), suggesting that they form a
multi-subunit CcmABCD complex. CcmC possesses a periplasmic tryptophan-rich (WWD) motif and two conserved histidine residues, which are required for function (Goldman et al., 1998). These residues may be important for the stable interaction of CcmC with heme b and transfer of this cofactor to CcmE (Ren and Thony-Meyer, 2001). The chemical nature of heme-CcmE ligation is unusual, via a critical His residue (His130 in E. coli CcmE, equivalent to His123 in Rba. capsulatus CcmE) (Deshmukh et al., 2000) which covalently links to the heme vinyl-2 group (Schulz et al., 1998; Thony-Meyer, 2003; Lee et al., 2005). At least in E. coli, attachment of heme b to CcmE is independent of CcmA and CcmB (Schulz et al., 1999), but requires CcmC and is enhanced by CcmD (Ahuja and Thony-Meyer, 2005), suggesting that CcmC is a CcmE-specific heme lyase (Ren and Thony-Meyer, 2001). Thus, based on the ability to
412
Carsten Sanders, Serdar Turkarslan, Ozlem Onder, Elaine R. Frawley, Robert G. Kranz, Hans Georg Koch and Fevzi Daldal
Fig. 3. The Ccs-system II components for c-type Cyt maturation. Similarly to system I, all components of the system II c-type Cyt maturation pathway are located in the cytoplasmic membrane, with the exception of DsbA. The apoCyt c is translocated across the membrane by the Sec machinery and is oxidized by DsbA, then reduced by CcsX. The transmembrane proteins CcsB and CcsA form a complex that represents the system II synthetase (heme ligation core complex). The reduced apoCyt c is likely bound by the periplasmic domain of CcsB. Heme is translocated by the CcsBA complex, probably through CcsA and binds to the CcsA periplasmic WWD domain before ligation to the reduced apoCyt c. Following covalent ligation of the heme vinyl groups to the reduced cysteines of the apoCyt c CXXCH motif, the holoCyt c folds into its mature form with release from the complex. See also Color Plate 9, Fig. 15.
bind heme b, CcmC alone has been proposed to be sufficient for heme translocation from the cytoplasm to CcmE, implying that CcmA and CcmB might have a function which differs from acting as a part of a heme transporter (Ren and Thony-Meyer, 2001). Recent work, reconciling these apparent discrepancies, revealed that the E. coli CcmABCD forms a complex and that, in addition to their implication in heme b transport and ligation to CcmE, an ATP-dependent release of holoCcmE from CcmC occurs (Feissner et al., 2006a). A model depicting a CcmABCD-dependent heme delivery pathway of the Ccm-system I is shown in Fig. 2. Accordingly, CcmC translocates heme b across the cytoplasmic membrane and ligates it covalently to CcmE with the help of CcmD, while CcmA-CcmB catalyzes ATP-dependent release of holoCcmE from CcmCD to become a heme b donor to the apoCyt c heme ligation complex.
C. Thio-oxidoreduction of Apocytochrome c Most c-type Cyts, including those from Rba. capsulatus, have a typical Sec-type signal sequence, hence most pre-apoCyts c are translocated across the cytoplasmic membrane via the Sec-dependent pathway and their signal peptides are either processed or used as N-terminal membrane anchors. As disulfide bonds are typically formed between the cysteines of exported proteins via the DsbA-DsbB pathway for oxidative protein folding (Collet and Bardwell, 2002; Kadokura et al., 2003), the presence of such a bond within the CXXCH motifs of apoCyts c was also assumed (Metheringham et al., 1996; Sambongi and Ferguson, 1996). Because ligation to heme requires reduced cysteines, a periplasmic apoCyt c thioreduction pathway, including CcdA, CcmG and CcmH, has been proposed to reduce cysteines in Rba. capsulatus (Beckman and Kranz, 1993; Monika et al., 1997). CcdA contains six transmembrane helices
Chapter 21
Cytochrome c Maturation and Assembly
(Deshmukh et al., 2000), with two invariant cysteines in the first and fourth of its helices that are essential for mediating transfer of reducing equivalents from the cytoplasm to the periplasm. CcmG and CcmH are monotopic membrane proteins, each of which has a single thioredoxin-like (CXXC) domain facing the periplasmic space (Monika et al., 1997). The reactive cysteines within the thioredoxin domains of these proteins, and also those at the apoCyt c heme binding motifs (Allen et al., 2002), are required for holocyt c production in an otherwise wild type background (Beckman and Kranz, 1993; Monika et al., 1997). Mixed disulfide bond formation in vivo has been detected between Rba. capsulatus CcmG and CcdA, indicating that the former is a specific substrate for the oxidoreductase activity of the latter (Katzen et al., 2002). Furthermore, in vitro studies with purified soluble variants of CcmG and CcmH suggested that CcmG can reduce CcmH, while CcmH can reduce an apoCyt c fragment carrying a CXXCH motif (Monika et al., 1997; Setterdahl et al., 2000). Thus, in the Ccm-system I, electrons are thought to be shuttled from the cytoplasmic thioredoxin TrxA across the membrane via CcdA (homologous to the central domain of DsbD used in some other Ccm system organisms like E. coli) to CcmG and then to CcmH to reduce the disulfide bond at the apoCyt c heme binding sites (Monika et al., 1997; Deshmukh et al., 2000; Fabianek et al., 2000; Setterdahl et al., 2000; Reid et al., 2001; Katzen et al., 2002). More recent work with Rba. capsulatus demonstrated that null mutants of DsbA and DsbB produce various ctype holocyts (Deshmukh et al., 2003), pointing out that the reduction and not the formation of disulfide bonds at the heme ligation motifs is crucial for Cyt c maturation. Studies in other species using a Ccm system, including E. coli, have confirmed this finding (Reid et al., 2001; Allen et al., 2003a). Moreover, Rba. capsulatus DsbA- or DsbB-null mutants can suppress the Cyt c deficiency phenotype of CcdA-null mutants (Deshmukh et al., 2003), indicating that in the absence of the thio-oxidative DsbA-DsbB pathway, the thio-reductive pathway becomes dispensable for Cyt c maturation. Recent work further revealed that in mutants lacking DsbA, both CcdA and CcmG become unnecessary for holocyt c production, although important growth phenotype differences were seen between CcdA-DsbA versus CcmG-DsbA double knockout mutants (S. Turkarslan, C. Sanders and F. Daldal, unpublished). Similar compensatory effects have also been described for Bacillus (B.) subtilis,
413
which uses the Ccs-system II (see below), but contains homologs of thio-oxidation (DsbA/BdbD and DsbB/ BdbC) and thio-reduction (CcdA and CcmG/CcsX, also called ResA) components (Erlendsson and Hederstedt, 2002; Erlendsson et al., 2003). It therefore appears that the activities of CcdA and CcmG/CcsX are critical for Cyt c maturation only in the presence of an active thio-oxidative pathway (DsbA/BdbD and DsbB/BdbC) (Erlendsson et al., 2003). Recently, a mechanistic view has been proposed which implies that the thio-oxidative and thio-reductive components might affect the stereo-selectivity of heme attachment to apoCyts c, thereby rationalizing the decreased efficiency of Cyt c maturation in the absence of thioredox components (Sanders et al., 2005b). It was noted that in Rba. capsulatus a mutant form of CcmG lacking its active site cysteines, although defective for Cyt c maturation, improves both growth and Cyt c production of DsbA-CcmG double knockout mutants (S. Turkarslan, C. Sanders and F. Daldal, unpublished). This observation suggests that during Cyt c maturation CcmG might be important beyond its thioredoxin function. In contrast to CcdA and CcmG, under any growth condition (even using thiolreactive reagents) neither reversion of CcmH knock out mutants (Monika et al., 1997), nor suppression between the DsbA- or DsbB-CcmH double mutants has been detected (S. Turkarslan, C. Sanders and F. Daldal, unpublished). Similarly, no mixed disulfide bond formation between CcmG and CcmH has been observed so far to support direct electron transfer between these proteins in vivo. Thus, the role of CcmH in the thio-reduction pathway of Cyt c maturation remains to be elucidated. Recently, using Ccs-system II, the reduced form of B. subtilis CcmG homolog CcsX was proposed to recognize apoCyts c that contain disulfide bonds and to release them in their reduced forms to the heme ligation sites (Crow et al., 2005; Colbert et al., 2006). As both Ccm and Ccs systems contain homologs of CcmG, but not of CcmH, it is intriguing to consider that CcmG also plays a similar role in the Ccm-system I (Fig. 2). Although it remains to be determined which component reduces specifically the oxidized apoCyt c in vivo, Rba. capsulatus CcmG clearly binds apoCyt c in vitro (C. Sanders, S. Turkarslan and F. Daldal unpublished). If this binding activity contributes to the suppressor ability of the CcmG protein lacking its thio-reactive cysteines in the DsbA-CcmG double knockout mutant, then an apoCyt c chaperone-like function, similar to that proposed for CcmI (Lang et
414
Carsten Sanders, Serdar Turkarslan, Ozlem Onder, Elaine R. Frawley, Robert G. Kranz, Hans Georg Koch and Fevzi Daldal
al., 1996), might be attributed to CcmG in addition to its thio-reduction function. D. Heme Ligation Complex In Rba. capsulatus, the CcmF, CcmH and CcmI components have been proposed to form a heme ligation core complex (Sanders et al., 2005a). CcmI is an essential Cyt c maturation component, thought to be involved in chaperoning apoCyts c to the ligase complex following their translocation across the membrane (Lang et al., 1996). In many species, CcmI is a bipartite membrane protein with two amino-terminal transmembrane helices encompassing a leucine zipper-like motif in its cytoplasmic loop (CcmI-1 domain) and a large periplasmic carboxylterminal extension (CcmI-2 domain) that contains four tetratricopeptide repeat (TPR)-like motifs (Lang et al., 1996; Sanders et al., 2005a). CcmI mutants devoid of the CcmI-2 domain exhibit a ‘growth medium-dependent’ Cyt c maturation phenotype: they can produce some c-type Cyts when grown on minimial medium, however, they lack all but holocyt c1 during growth on enriched medium (Lang et al., 1996). CcmI mutants devoid of the CcmI-2 domain exhibit growth medium-dependent Cyt c maturation and lack all holocyts, with the exception of holocyt c1, during growth on enriched medium (Lang et al., 1996). Thus, CcmI-1 appears to be needed for the production of all Cyts c, whereas CcmI-2 is dispensable for the C-terminally membrane-anchored Cyt c1 (Lang et al., 1996; Sanders et al., 2005a). No CcmI-1 homolog has been found in some other species such as E. coli, where the homolog of CcmI-2 is naturally fused C-terminally to the CcmH homolog, yielding a bifunctional protein. In E. coli, lack of CcmI-1 is consistent with the lack of C-terminally membrane attached c-type Cyts such as Cyt c1 In Rba. capsulatus, Cyt c maturation defects due to the lack of CcmI can be readily suppressed by an over-abundance of CcmF, CcmH and CcmG via distinct and additive events. Overproduction of CcmF and CcmH overcomes the need for CcmI on minimal, but not enriched, media (Deshmukh et al., 2002) and additional overproduction of CcmG ensures sufficient holocyt c production on all media (Sanders et al., 2005a). Similarly, overproduction of the CcmI-1 domain alone (like overproduction of CcmF and CcmH) partly bypasses Cyt c deficiency in the absence of CcmI on minimal medium and additional overproduction of either CcmI-2 or apoCyt
c2 behaves in the same way as CcmG overproduction by further complementing CcmI-null mutants on all media. These findings suggest that CcmI-1, CcmF and CcmH on the one hand and CcmI-2 and CcmG on the other, work together during Cyt c maturation (Sanders et al., 2007). If this is indeed correct, then the membrane-spanning CcmI-1 and the periplasmic CcmI-2 domains of CcmI could be the junction point between the CcdA- and CcmG-dependent apoCyt c thio-reduction and the CcmF and CcmH heme ligation processes (Fig. 2). CcmF is a large integral membrane protein with 11 transmembrane helices and, like CcmC, has a tryptophan-rich (WWD) signature motif and four conserved histidine residues facing the periplasm. The WWD motif and the conserved histidines are required for the function of CcmF and have been proposed to bind heme b to ligate it to apoCyt c (Goldman et al., 1998; Ren and Thony-Meyer, 2001). In E. coli, CcmF interacts with both the holoform of CcmE and the thioredoxin-like CcmH (Ren et al., 2002) and has been proposed to form the apoCyt c heme lyase of Ccm-system I together with CcmH. Similarly, analyses of CcmI-null mutant suppressors pointed out that in Rba. capsulatus CcmF forms a heme ligation core complex together with CcmH and CcmI (Sanders et al., 2005a,b). Clearly, biochemical work is needed to define the CcmHIF-containing complex and to elucidate the mechanism of heme ligation to apoCyts c during their maturation (Sanders et al., 2005b). E. Not All Phototrophs Use Ccm-system I: The Simpler Ccs-System II While this chapter deals mainly with the Ccm-system I pathway, some phototrophic bacteria use the simpler Ccs-system II (Fig. 3). Besides the phototrophic cyanobacteria and the green sulfur anoxygenic chlorobi, the non-photosynthetic β- and ε-proteobacteria (e.g., Bordetella pertussis and Helicobacter species, respectively) contain Ccs-system II genes. Recently the genomic sequences of two phototrophic β-proteobacteria have been completed. Surprisingly, these two, Rhodoferax ferrireducens (http://genome.jgipsf.org/draft_microbes/rhofe/rhofe.home.html) and Rubrivivax gelatinosus have Ccm-system I but not Ccs-system II genes (http://genome.jgi-psf.org/finished_microbes or http://www.ncbi.nlm.nih.gov/entrez). Several non-phototrophic β-proteobacteria also use Ccm-system I for Cyt c biogenesis, including
Chapter 21
Cytochrome c Maturation and Assembly
Dechloromonas, Nitrosomonas and Nitrosospira species. Interestingly, only a few bacterial genomes out of hundreds currently available encode both the Ccm and Ccs systems. These include the β-proteobacteria Bordetella parapertussis and Bordetella bronchiseptica. Although it is unknown whether both systems are functional in these bacteria, the presence of both Ccm and Ccs systems tends to suggests that the β-group has undergone lateral transfer of one system. Most likely, natural selection in various habitats might have selected for one or the other system in the majority of species. This raises the question of whether the phototrophic, anoxygenic proteobacteria benefit from the use of Ccm-system I over Ccs-system II. The functional reconstitution of recombinant Ccssystem II in E. coli, which contains naturally only a Ccm-system I, suggests that the Gram-negative cytoplasmic membrane and periplasm are appropriate compartments for the function of either system (Feissner et al., 2006b). Results with these recombinant Ccm or Ccs systems in the same E. coli background suggest that organisms with Ccm-system I can use endogenous (Feissner et al., 2006b) and exogenous (Richard-Fogal et al., 2007) heme at lower concentrations than those with Ccs-system II. Therefore, a selective factor for organisms with Ccm-system I may be heme limitation. Iron shortage limits heme biosynthesis and recently, it was found that Cyt c produced by recombinant Ccm-system I is not impacted at low iron concentrations where Ccs-system II is affected (San Francisco and R. G. Kranz, unpublished). An additional advantage of Ccm-system I is that the holoCcmE can act as a heme reservoir, again facilitating Cyt c synthesis when faced with heme and iron limitation (Feissner et al., 2006b). Extension of these findings to the anoxygenic proteobacteria is consistent with the idea that these species might be limited for iron in their habitats and hence retained the Ccm-system I pathway to assure Cyt c synthesis under such starvation conditions. Why is the Ccm-system I more complicated than Ccs-system II and how does Ccs-system II operate with a smaller number of proteins? If Ccm-system I evolved to deliver heme at low levels, biogenesis might require carrier proteins that bind heme with high affinity and then pass it on to the apoCyt c. However, this presents a problem, as the carrier proteins cannot have very high affinity for heme, because the bound heme must be released to the apoCyt c. One solution to this dilemma is the release of heme to a chaperone (CcmE) using the CcmABCD complex,
415
where ATP hydrolysis is used for this purpose. In Ccs-system II, CcsA with its periplasmic WWD domain appears to have a low affinity for heme and its partner CcsB, an integral membrane protein, presumably binds the CXXCH motif of apoCyt c, with the reduced cysteines of this motif positioned at the two vinyl groups of heme for ligation (Fig. 3). A possibility is that subsequent holocyt c folding provides enough energy to release heme from CcsA and that the CcsA-CcsB complex supplies a scaffold for apoCyt c and heme in the reduced state. The mechanism of reduction of the apoCyt c cysteine residues is thought to be similar to that used in the Ccm-system I. A membrane integral DsbD (or CcdA) moves reducing equivalents from a cytoplasmic thioredoxin to a periplasmic thioredoxin CcsX (ResA) and then CcsX reduces the apoCyt c cysteine residues. The recent crystal structure of CcsX suggests how the specificity towards the apoCyt c CXXCH might occur (Lewin et al., 2006). Organisms that are not faced with iron limitation could benefit from the simpler Ccs-system II to maintain heme in a reduced state necessary for ligation. Clearly, as more phototrophic proteobacteria are discovered and as more genomes are sequenced, it will become clearer whether all of them have Ccmsystem I. Consideration of the ecological aspects of these bacteria may yield a full understanding of how and why one system or another might be advantageous for certain lifestyles. III. Biogenesis of Cytochrome Complexes Biogenesis of membrane-bound multi subunit Cyt c complexes is an intrinsically complex process. The insertion of the apoproteins into the lipid phase has to be tightly coordinated with their maturation (e.g., incorporation of heme groups for Cyts and availability and insertion of Cu atoms for Cox). Moreover, matured subunits have to find their partners accurately to yield active oligomeric protein complexes. This chapter focuses mainly on the assembly processes of the aa3-Cox and cbb3-Cox of Rhodobacter species. A. Salient Features of aa3- and cbb3-type Cytochrome c Oxidase in Rhodobacter Species The best studied examples of Cox assembly are the aa3-Cox, which are the terminus of the aerobic elec-
Carsten Sanders, Serdar Turkarslan, Ozlem Onder, Elaine R. Frawley, Robert G. Kranz, Hans Georg Koch and Fevzi Daldal
416
tron transfer chain in mitochondria and many bacteria (Richter and Ludwig, 2003). The aa3-Cox of Rba. sphaeroides consists of four subunits (Table 2A), of which the largest three are homologous to the core subunits of the eukaryotic aa3-Cox (Svensson-Ek et al., 2002). Subunit IV is a small single spanning membrane protein with unknown function (Witt and Ludwig, 1997). Subunit I contains the low-spin heme a and the high-spin heme a3-CuB binuclear center where oxygen reduction to water occurs. These cofactors are deeply buried within the membrane and connected to a network of α-helical transmembrane domains (Iwata et al., 1995). The CuA-center of subunit II on the other hand is embedded in a large β-barrel-like structure that is exposed to the periplasmic side of the membrane, which favors its interaction with the donor c-type Cyts. Rba. sphaeroides also contains a cbb3-type Cyt c oxidase (cbb3-Cox), which is assumed to be involved in microaerobic (low oxygen tension) respiration (Takamiya, 1983;
Garcia-Horsman et al., 1994). Both cbb3-Cox and aa3-Cox belong to the superfamily of heme-copper oxidases, but they exhibit significant differences in terms of subunit composition and cofactor content (Table 2). Subunit I of cbb3-Cox (CcoN) contains the conserved histidine residues that are diagnostic for the heme-copper oxidase superfamily. These residues ligate a low-spin heme b and a high-spin heme b3-CuB binuclear center. Like ubihydroquinone oxidases, cbb3-Cox lack a CuA-containing subunit and are instead composed of two membrane-bound c-type Cyts, the mono-heme subunit II (CcoO) and the di-heme subunit III (CcoP) (Gray et al., 1994). Both CcoO and CcoP are required for electron transfer from a donor Cyt c to the binuclear center of CcoN, but the precise electron transfer pathway internal to cbb3-Cox is unknown. A fourth subunit (CcoQ) is present in most cbb3-Cox, which, like subunit IV of aa3-Cox, is a small single-spanning membrane protein. CcoQ is not essential for activity but might be involved in
Table 2A. Subunits and their cofactors of the aa3- and cbb3-Cox Subunits Subunit I
Subunit II
Subunit III
Subunit IV
Oxidase Type aa3
Gene name
Proposed role
ctaD
Conserved motifs in gene product Six conserved histidines (H102, 284, 333, 334, 419, 421) in CoxI
cbb3
ccoN
Six conserved histidines
O2 to H2O conversion site
Multiple transmembrane helices with a low spin heme b and a high spin heme b3-CuB binuclear center
aa3
ctaC
No recognizable motif
Electron transfer from donor Cyt c to the binuclear center of CoxI
Two transmembrane helices with a periplasmic domain and a CuA center
cbb3
ccoO
One CXXCH motif
Electron transfer from donor Cyt c to the binuclear center of CcoN
Membrane-anchored monoheme Cyt c
aa3
ctaE (coxIII)
No recognizable motif
Not required for assembly of the redox centers and not involved directly in electron transfer
Seven transmembrane helices, no cofactor
cbb3
ccoP
Two CXXCH motifs
Electron transfer from donor Cyt c to the binuclear center of CcoN
Membrane-anchored diheme Cyt c
aa3
coxIV
No recognizable motif
Unknown function, not essential for activity
Small single transmembrane helix, no cofactor
cbb3
ccoQ
No recognizable motif
Regulation and stabilization of cbb3-Cox under aerobic condition, not essential for activity
Small single transmembrane helix, no cofactor
O2 to H2O conversion site
Salient topological or other properties 12 transmembrane helices with a low spin heme a and a high spin heme a3-CuB binuclear center
Chapter 21
Cytochrome c Maturation and Assembly
417
Table 2B. Genes involved in mitochondrial and bacterial Cox biogenesis Gene name
Conserved motifs in gene product
Proposed role in Cyt oxidase
Salient topological or other properties
cox17
Cupredoxin-fold
Delivers Cu1+ to Sco1 and Cox11
Soluble protein of the mitochondrial intermembrane space
sco1 (senC, prrC, ypmQ)
Conserved copper-binding motif (CXXXC)
Formation of the CuA center of Cox2
One transmembrane helix with the Cu binding periplasmic domain at C-terminal
cox11
Conserved Cysteine residues for Cu binding and a conserved metal binding motif (CFCF)
Formation of the CuB center of Cox1
Single transmembrane helix with a large globular domain
surf1
No recognizable motif
heme a3 insertion in aa3-Cox
Two transmembrane helices at the C- and N-termini of the protein
ccoG
Two (4Fe-4S) cluster binding motifs
Not essential for assembly or stability
Five transmembrane helices
ccoH
No recognizable motif
Stabilize the CcoNOQ sub-complex and CcoP to facilitate their assembly into an active complex
A C-terminal transmembrane-helix with a type I orientation (Nout-Cin)
ccoI
Homologous to Cu transporting CPX-type ATPases with Cu binding motif
Involved in Cu acquisition for CuB center, required for steady-state amount of the enzyme
Eight transmembrane helices
ccoS
No recognizable motif
Critical for proper maturation of CcoN and its binuclear center
Small protein with a single transmembrane helix
aa3-type Cox
cbb3-type Cox
the regulation and stabilization of the cbb3-Cox under aerobic conditions (Oh and Kaplan, 2002). The structural genes ccoNOQP of cbb3-Cox in both Rba. capsulatus and Rba. sphaeroides are clustered in the genome and even their 5´ upstream and 3´ downstream regions are highly conserved (Koch et al., 1998a). An ORF encoding for a putative membrane protein of 277 amino acids (ORF277) is located upstream of the ccoNOQP cluster and its protein product has recently been identified by mass spectrometry in a partially purified cbb3-Cox (Kulajta et al., 2006). In Rba. capsulatus, mutation of ORF277 does not impair cbb3-Cox activity, although association with cbb3-Cox suggests that ORF277 might have a function related to cbb3-Cox (Koch et al., 1998a). ORF277 is homologous to members of the universal stress protein A (UspA) superfamily, an ancient and conserved group of proteins found in bacteria, archaea, fungi, flies and plants (Kvint et al., 2003). The role of these proteins is enigmatic, but apparently linked to resistance to oxidative stress and respiratory uncouplers.
B. Assembly of the aa3-type Cytochrome c Oxidase in Rhodobacter sphaeroides Rba. sphaeroides provides an invaluable model for studying aa3-Cox assembly, because in contrast to the mitochondrial enzyme, assembly intermediates lacking all or some of the cofactors can accumulate. Most of the components involved in aa3-Cox assembly in eukaryotes are conserved in the Rba. sphaeroides genome and are amenable to biochemical and genetic studies. An exception is Cox17, the key copper metallochaperone of the mitochondrial inter membrane space, which is apparently absent in prokaryotes (Carr and Winge, 2003). So far, Cox17 orthologs have been found exclusively in eukaryotes, although proteins which contain a cupredoxin-fold, like Cox17, have been identified in bacteria and are proposed to be the functional homologs of eukaryotic Cox17 (Banci et al., 2005). In mitochondria, Cox17 delivers Cu1+ to two additional Cu1+ binding proteins, Sco1 and Cox11, which in contrast to Cox17 are present in both eukaryotic and prokaryotic cells (Horng et al., 2004). Sco1 appears to be specific for formation
418
Carsten Sanders, Serdar Turkarslan, Ozlem Onder, Elaine R. Frawley, Robert G. Kranz, Hans Georg Koch and Fevzi Daldal
of the CuA center of Cox2, while Cox11 is required for that of the CuB center (Carr and Winge, 2003). Cox11 interacts with ribosomal proteins and is even fused to a ribosomal subunit in Schizosaccharomyces pombe (Khalimonchuk et al., 2005), suggesting that CuB addition to Cox1 might be coupled to its co-translational membrane integration. Consistent with the role of Cox11 in CuB center formation, Rba. sphaeroides Cox11 mutants lack the CuB center in their aa3-type Cox (Hiser et al., 2000), but in these mutants subunit assembly and insertion of the CuA, heme a and a3 cofactors are not impaired. Thus, differently from eukaryotes, Cox11 might be involved in a later step of assembly in bacteria, or the absence of an intact CuB center does not arrest subsequent assembly steps. The exact role of Cox11 may be to provide a binding site for Cu together with distant His residues (Banci et al., 2006). In eukaryotes, Sco1 is known to interact with Cox2 and a mutant lacking ScoI has significantly reduced aa3-Cox activity causing fatal infantile hepatoencephalomyopathy in humans (Barrientos et al., 2002a). In human cell lines, the absence of Sco1 can be partially suppressed by addition of exogenous copper (Jaksch et al., 2000), suggesting that Sco1 is involved in CuA center formation. However, as Sco1 homologs also exhibit significant similarities to peroxiredoxins and thiol:disulfide oxidoreductases (Chinenov, 2000), ScoI might also have catalytic functions, such as reduction of the closely spaced Cys residues of the Cu binding motif in Cox2 (Banci et al., 2006). Moreover, Sco1 might be involved in redox sensing (Williams et al., 2005) and the aa3Cox deficiency in its absence might be a downstream regulatory effect. The possibility that Sco1 might be involved in multiple steps during aa3-Cox biogenesis is also reflected in the different phenotypes of bacterial mutants lacking Sco1. In B. subtilis, absence of Sco1 (YpmQ) leads to a loss of Cox activity, but does not impair the activity of the menahydroquinone oxidase, which lacks a CuA center (Mattatall et al., 2000). In Rba. sphaeroides, the Sco1 homolog PrrC has been implicated in both Cu binding (McEwan et al., 2002) and in signal transduction (Eraso and Kaplan, 2000). However, in mutants lacking PrrC, the aa3-Cox activity is not significantly impaired (Smith et al., 2005), leaving the role of Sco1 obscure. In Rba. capsulatus, which does not have an aa3-Cox, absence of the Sco1-homolog SenC affects the activity of the cbb3-Cox (Borsetti et al., 2005; Swem et al., 2005), as mentioned below.
As for the CuB center of aa3-Cox, it is assumed that the heme a and a3 cofactors are inserted cotranslationally into Cox1. In Cox10 mutants of yeast, which are devoid of heme a synthase and in similar Paracoccus denitrificans mutants, Cox1 is translated, but is almost undetectable in membranes (Nobrega et al., 1990), suggesting that heme insertion is required for correct folding and stability of Cox1. In Rba. sphaeroides, assembly of the aa3-Cox proceeds even in the absence of heme a (Hiser and Hosler, 2001), suggesting that an alternative pathway to yield the active enzyme might exist. It is still largely unknown whether specific assembly proteins aid the insertion of the non-covalently linked heme groups into Cox1. Recently, a protein called Surf1 has been implicated in heme a3 insertion in Rba. sphaeroides aa3-Cox (Smith et al., 2005). Surf1 is a conserved integral membrane protein, found in the inner mitochondrial, or bacterial cytoplasmic membrane. Mutations in the human gene encoding Surf1 lead to the Leigh syndrome, characterized by Cox deficiency (Barrientos et al., 2002b). However, in the absence of Surf1 some active aa3-Cox is still found both in mitochondria and bacteria, suggesting that either additional assembly factors exist, or that spontaneous heme a3 insertion can yield some active aa3-Cox. Finally, as Surf1 is implicated in heme a3 insertion only, it remains to be seen how the low-spin heme a is inserted into Cox1. C. Assembly of the cbb3-type Cytochrome c Oxidase in Rhodobacter Species In Rba. capsulatus membranes, cbb3-Cox forms a catalytically active complex composed of four subunits, running at an apparent molecular weight of 230 kDa in Blue-Native-polyacrylamide gel electrophoresis (BN-PAGE) (Kulajta et al., 2006). In addition to the 230 kDa complex, BN-PAGE reveals the presence of an inactive 210 kDa sub-complex, composed of only CcoN, CcoO and CcoQ. A large pool of monomeric, mature CcoP is also present. The presence of the 210 kDa sub-complex in a mutant lacking CcoP strongly suggests that the 210 kDa species corresponds to an assembly intermediate (Kulajta et al., 2006), which is converted into the catalytically active 230 kDa complex by recruiting CcoP. As a CcoNOQ sub-complex has also been proposed to exist in Bradyrhizobium japonicum (Zufferey et al., 1996), it is likely that the cbb3-Cox assembly proceeds via a conserved and sequential pathway. This contrasts to the assembly of the aa3-Cox in Rba. sphaeroides, which does not ap-
Chapter 21
Cytochrome c Maturation and Assembly
pear to follow a sequential pathway (Hiser and Hosler, 2001). Because the aa3-Cox and cbb3-Cox are highly similar to each other, it is important to determine if they share any assembly factors. A Rba. sphaeroides mutant lacking Cox11 was not impaired in cbb3-Cox activity (Hiser et al., 2000), suggesting that although both of the aa3-Cox and cbb3-Cox contain a CuB center, maturation of their active sites follows different molecular modes. It is unknown whether Surf1, suggested to be involved in heme a3 insertion (see above), is also present in the Rba. capsulatus genome, leaving its role in cbb3-Cox assembly open. The role of Rhodobacter Sco1 homologs (PrrC/SenC) in cbb3Cox assembly is also unclear. In Rba. sphaeroides, absence of PrrC does not seem to affect cbb3-Cox activity or assembly (Eraso and Kaplan, 2000), as this enzyme has no CuA center. In contrast, absence of SenC in Rba. capsulatus decreases drastically the activity and steady-state assembly of cbb3-Cox (Borsetti et al., 2005; Swem et al., 2005), which is consistent with the possibility that Sco1 homologs might be involved in Cu traffic and delivery. Specific functions of Rhodobacter Sco1 homologs remain to be determined. It was recently observed that Rba. capsulatus mutants lacking DsbA, which is a periplasmic dithiol: disulfide oxidoreductase involved in the oxidative folding of extracytoplasmic proteins (Kadokura et al., 2003), are temperature-sensitive for respiratory growth, especially under Cu limiting conditions (Deshmukh et al., 2003). Comparative extracytoplasmic sub-proteomes of such mutants and their revertants revealed that a periplasmic protease, DegP is highly overproduced in these mutants, leading to a complex set of phenotypes, including absence of an active cbb3-Cox (O. Onder and F. Daldal, unpublished). In Rba. capsulatus mutants lacking DsbA, the c-type Cyts subunits (CcoO and CcoP) of the cbb3-Cox are correctly matured and antigens corresponding to CcoN are also detectable. However, whether the CcoN subunit contains its hemes b, b3 and the CuB cofactor is unknown. Remarkably, the cbb3-Cox activity of these mutants can be restored upon addition of exogenous Cu, like mutants devoid of SenC (Swem et al., 2005). Although Mn or Zn addition cannot rescue this mutant phenotype (O. Onder and F. Daldal, unpublished), it remains to be seen whether Cu addition promotes oxidation in the absence of DsbA of an unknown component required for Cu traffic, or Cu acts as a missing specific cofactor for cbb3-Cox. Clearly disulfide bond formation
419
and Cu traffic and insertion into the cbb3-Cox appear to be linked. A requirement for different molecular machineries for the maturation and assembly of cbb3- and aa3-Cox has emerged with the identification of the ccoGHIS genes in Rba. capsulatus (Koch et al., 1998a, 2000). This gene cluster is located downstream of ccoNOQP operon in most cbb3-Cox containing organisms (Kahn et al., 1989; de Gier et al., 1996; Preisig et al., 1996; Koch et al., 1998a, 2000) and encodes four predicted integral membrane proteins that specifically affect the assembly of cbb3-Cox, but not that of aa3-Cox. A model illustrating how the individual subunits of cbb3-Cox are assembled into an active enzyme is depicted in Fig. 4 and the roles of the ccoGHIS products in this process are discussed below. CcoG is a highly conserved ferredoxin-like protein, which is not essential for cbb3-Cox assembly or stability (Koch et al., 2000), although it affects the onset of Cox activity in Rba. capsulatus. In contrast, absence of CcoH abolishes completely both the 230 kDa CcoNOQP and the 210 kDa CcoNOQ complexes and also decreases the amounts of monomeric CcoP as seen by BN-PAGE (Koch et al., 2000; Kulajta et al., 2006). CcoH has a single transmembrane domain with a type I orientation (Nout-Cin), with a C-terminal hydrophobic stretch (initially thought to be a second transmembrane domain) exposed to the cytoplasm as a potential dimerization domain. CcoH interacts specifically with the 210 kDa CcoNOQ sub-complex and probably also with monomeric CcoP, but is not a part of the 230 kDa complex. Thus, CcoH might stabilize both the 210 kDa intermediate and the monomeric CcoP and facilitate their assembly into the active 230 kDa CcoNOQP complex, possibly via its dimerization. A Rba. capsulatus mutant lacking CcoI produces neither the 230 kDa CcoNOQP nor the 210 kDa CcoNOQ complexes in BN-PAGE analysis, but contains quasi-native amounts of CcoP (Kulajta et al., 2006). CcoI shows high sequence similarity to Cu-transporting CPx-type ATPases (Kahn et al., 1989). Absence of CcoI, or a mutation in the predicted Cu-binding motif of CcoI, mimics closely the cbb3-Cox phenotype of a wild type Rba. capsulatus strain grown under Cu-depletion conditions (Koch et al., 2000), although biochemical evidence that CcoI indeed transports Cu is lacking. Surprisingly, absence of CcoI does not impair aa3-Cox activity in Rba. sphaeroides, suggesting that Cu delivery to cbb3-Cox is different from that to aa3-Cox, which depends on Cox11 and Sco1. The
420
Carsten Sanders, Serdar Turkarslan, Ozlem Onder, Elaine R. Frawley, Robert G. Kranz, Hans Georg Koch and Fevzi Daldal
Fig. 4. Assembly pathway of the Cyt cbb3 oxidase enzyme. A proposed stepwise pathway for the assembly of an active Cyt cbb3 oxidase is depicted. Step 1: After translation and membrane integration of the individual subunits, the c-type Cyts subunits are matured via the Ccm-system I pathway yielding the di-heme Cyt CcoP and mono-heme Cyt CcoO. This maturation step appears to be crucial for the stability of both proteins and for their subsequent assemblies. Step 2: First, the CcoN, CcoO and CcoQ proteins form a stable subcomplex of about 210 kDa, which has no Cox activity. This sub-complex is produced efficiently only in the presence of the putative CPX-type Cu-uptake ATPase, CcoI, suggesting that Cu might be inserted early during assembly. The 210 kDa CcoNOQ sub-complex is only detectable in the presence of CcoH, which specifically interact with it to probably stabilize it. In the membranes, CcoP is found in a monomeric form, which is independent of CcoI but requires CcoH for its stability. Like CcoH, CcoS can also be found associated with the 210 kDa CcoNOQ sub-complex and is required for the formation of an active heme b3-CuB binuclear center. It is possible that CcoS is involved in heme b acquisition. Remarkably, in the absence of CcoS, an inactive cbb3-Cox can still be assembled. Step 3: The active 230 kDa CcoNOQP complex is formed by recruiting a monomeric CcoP into the 210 kDa CcoNOQ sub-complex, possibly via dimerization of CcoH. Apparently, neither CcoH nor CcoS are part of the 230 kDa active cbb3-Cox complex, suggesting that both probably dissociate after its formation.
predicted Cu-binding sites of the latter proteins face the periplasm, suggesting that Cu is collected from the periplasm, as in mitochondria where the Cu-binding motifs of Cox11 and Sco1 are also exposed to the inter membrane space (Carr and Winge, 2003). Although currently it is thought that Cu is inserted from the periplasmic side into aa3-Cox, it is unclear whether this event occurs co- or post-translationally (Carr and Winge, 2003) and apparently, no specific Cu transporter is required in either mitochondria or bacteria. The predicted Cu-binding motif of CcoI differs from those of Cox11 and ScoI and it faces the cytoplasm. If CcoI is indeed a Cu transporter specialized for cbb3-Cox assembly, then a different mechanism for Cu insertion from the cytoplasmic side of the membrane might be involved. Consistent with this view, the absence of Cox11 in Rba. sphaeroides impairs aa3-Cox, but not cbb3-Cox activity (Hiser et al., 2000). Clearly, despite the fact that cbb3-Cox is widespread among bacteria, little is known about how Cu is inserted into this enzyme. CcoS is a small membrane protein (51 amino acids) with no homology to any protein outside the CcoS/FixS group and has no motif that could suggest a possible function. Yet in its absence, although the presence of the 230 kDa and 210 kDa complexes
in BN-PAGE analysis is not significantly reduced, cbb3-Cox activity is completely lost due to the lack of a functional heme b3-CuB binuclear center (Koch et al., 2000; Kulajta et al., 2006). Thus CcoS, which does not contain any metal binding site or motif, is not involved in either the biogenesis of the c-type Cyt subunits of cbb3-Cox or their assemblies with CcoN, but is critical for proper maturation of CcoN and its binuclear center. The finding that absence of the putative Cu transporter CcoI also blocks cbb3-Cox assembly suggests that CcoS might be involved in heme b insertion, about which very little is known (Koch et al., 1998a; Koch et al., 2000; Kulajta et al., 2006). Interestingly, production of an inactive cbb3-Cox lacking hemes in the absence of CcoS suggests that heme b insertion to CcoN subunit is not a prerequisite for assembly, unlike the heme c insertion into CcoO or CcoP subunits (Koch et al., 1998a; Koch et al., 2000). Similarly, absence of heme a synthase does not significantly impair the assembly of aa3-Cox in Rba. sphaeroides (Hiser and Hosler, 2001). Much remains to be learned about the assembly of oligomeric, cofactor-containing membrane proteins in general. The c-type Cyts and Cyt c-containing multi subunit protein complexes of facultative phototrophic bacteria of Rhodobacter
Chapter 21
Cytochrome c Maturation and Assembly
species provide excellent experimental systems for future studies of these essential biological processes, as illustrated in this chapter. Acknowledgments This work is supported by grants to R. G. Kranz (NIH, GM47909), H-G. Koch (DFG, SFB388) and F. Daldal (DOE, 91ER 20052 and NIH, GM 38237). References Ahuja U and Thony-Meyer L (2005) CcmD is involved in complex formation between CcmC and the heme chaperone CcmE during cytochrome c maturation. J Biol Chem 280: 236–243 Allen JW, Tomlinson EJ, Hong L and Ferguson SJ (2002) The Escherichia coli cytochrome c maturation (Ccm) system does not detectably attach heme to single cysteine variants of an apocytochrome c. J Biol Chem 277: 33559–33563 Allen JW, Barker PD and Ferguson SJ (2003a) A cytochrome b562 variant with a c-type cytochrome CXXCH heme-binding motif as a probe of the Escherichia coli cytochrome c maturation system. J Biol Chem 278: 52075–52083 Allen JW, Daltrop O, Stevens JM and Ferguson SJ (2003b) Ctype cytochromes: Diverse structures and biogenesis systems pose evolutionary problems. Philos Trans R Soc Lond B Biol Sci 358: 255–266 Banci L, Bertini I, Ciofi-Baffoni S, Katsari E, Katsaros N, Kubicek K and Mangani S (2005) A copper(I) protein possibly involved in the assembly of CuA center of bacterial cytochrome c oxidase. Proc Natl Acad Sci USA 102: 3994–3999 Banci L, Bertini I, Cantini F, D’Amelio N and Gaggelli E (2006) Human SOD1 before harboring the catalytic metal: Solution structure of copper-depleted, disulfide-reduced form. J Biol Chem 281: 2333–2337 Barrientos A, Barros MH, Valnot I, Rotig A, Rustin P and Tzagoloff A (2002a) Cytochrome oxidase in health and disease. Gene 286: 53–63 Barrientos A, Korr D and Tzagoloff A (2002b) Shy1p is necessary for full expression of mitochondrial COX1 in the yeast model of Leigh’s syndrome. EMBO J 21: 43–52 Beckett CS, Loughman JA, Karberg KA, Donato GM, Goldman WE and Kranz RG (2000) Four genes are required for the system II cytochrome c biogenesis pathway in Bordetella pertussis, a unique bacterial model. Mol Microbiol 38: 465–481 Beckman DL and Kranz RG (1993) Cytochromes c biogenesis in a photosynthetic bacterium requires a periplasmic thioredoxinlike protein. Proc Natl Acad Sci USA 90: 2179–2183 Beckman DL, Trawick DR and Kranz RG (1992) Bacterial cytochromes c biogenesis. Genes Dev 6: 268–283 Bernard DG, Quevillon-Cheruel S, Merchant S, Guiard B and Hamel PP (2005) Cyc2p, a membrane-bound flavoprotein involved in the maturation of mitochondrial c-type cytochromes. J Biol Chem 280: 39852–39859 Borsetti F, Tremaroli V, Michelacci F, Borghese R, Winterstein C, Daldal F and Zannoni D (2005) Tellurite effects on Rhodobacter capsulatus cell viability and superoxide dismutase activity under
421
oxidative stress conditions. Res Microbiol 156: 807–813 Bratton M, Mills D, Castleden CK, Hosler J and Meunier B (2003) Disease-related mutations in cytochrome c oxidase studied in yeast and bacterial models. Eur J Biochem 270: 1222–1230 Cao J, Hosler J, Shapleigh J, Revzin A and Ferguson-Miller S (1992) Cytochrome aa3 of Rhodobacter sphaeroides as a model for mitochondrial cytochrome c oxidase. The coxII/coxIII operon codes for structural and assembly proteins homologous to those in yeast. J Biol Chem 267: 24273–24278 Carr HS and Winge DR (2003) Assembly of cytochrome c oxidase within the mitochondrion. Acc Chem Res 36: 309–316 Chinenov YV (2000) Cytochrome c oxidase assembly factors with a thioredoxin fold are conserved among prokaryotes and eukaryotes. J Mol Med 78: 239–242 Colbert CL, Wu Q, Erbel PJ, Gardner KH and Deisenhofer J (2006) Mechanism of substrate specificity in Bacillus subtilis ResA, a thioredoxin-like protein involved in cytochrome c maturation. Proc Natl Acad Sci USA 103: 4410–4415 Collet JF and Bardwell JC (2002) Oxidative protein folding in bacteria. Mol Microbiol 44: 1–8 Crow A, Le Brun NE and Oubrie A (2005) The role of ResA in type II cytochrome c maturation. Biochem Soc Trans 33: 149–151 Davidson E and Daldal F (1987) Primary structure of the bc1 complex of Rhodopseudomonas capsulata. Nucleotide sequence of the pet operon encoding the Rieske cytochrome b and cytochrome c1 apoproteins. J Mol Biol 195: 13–24 de Gier JW, Schepper M, Reijnders WN, van Dyck SJ, Slotboom DJ, Warne A, Saraste M, Krab K, Finel M, Stouthamer AH, van Spanning RJ and van der Oost J (1996) Structural and functional analysis of aa3-type and cbb3-type cytochrome c oxidases of Paracoccus denitrificans reveals significant differences in proton-pump design. Mol Microbiol 20: 1247–1260 Deshmukh M, Brasseur G and Daldal F (2000) Novel Rhodobacter capsulatus genes required for the biogenesis of various c-type cytochromes. Mol Microbiol 35: 123–138 Deshmukh M, May M, Zhang Y, Gabbert KK, Karberg KA, Kranz RG and Daldal F (2002) Overexpression of ccl1-2 can bypass the need for the putative apocytochrome chaperone CycH during the biogenesis of c-type cytochromes. Mol Microbiol 46: 1069–1080 Deshmukh M, Turkarslan S, Astor D, Valkova-Valchanova M and Daldal F (2003) The dithiol:disulfide oxidoreductases DsbA and DsbB of Rhodobacter capsulatus are not directly involved in cytochrome c biogenesis, but their inactivation restores the cytochrome c biogenesis defect of CcdA-null mutants. J Bacteriol 185: 3361–3372 Drygas ME, Lambowitz AM and Nargang FE (1989) Cloning and analysis of the Neurospora crassa gene for cytochrome c heme lyase. J Biol Chem 264: 17897–17906 Dumont ME, Ernst JF, Hampsey DM and Sherman F (1987) Identification and sequence of the gene encoding cytochrome c heme lyase in the yeast Saccharomyces cerevisiae. EMBO J 6: 235–241 Eraso JM and Kaplan S (2000) From redox flow to gene regulation: Role of the PrrC protein of Rhodobacter sphaeroides 2.4.1. Biochemistry 39: 2052–2062 Erlendsson LS and Hederstedt L (2002) Mutations in the thioldisulfide oxidoreductases BdbC and BdbD can suppress cytochrome c deficiency of CcdA-defective Bacillus subtilis cells. J Bacteriol 184: 1423–1429
422
Carsten Sanders, Serdar Turkarslan, Ozlem Onder, Elaine R. Frawley, Robert G. Kranz, Hans Georg Koch and Fevzi Daldal
Erlendsson LS, Acheson RM, Hederstedt L and Le Brun NE (2003) Bacillus subtilis ResA is a thiol-disulfide oxidoreductase involved in cytochrome c synthesis. J Biol Chem 278: 17852–17858 Fabianek RA, Hennecke H and Thony-Meyer L (2000) Periplasmic protein thiol:disulfide oxidoreductases of Escherichia coli. FEMS Microbiol Rev 24: 303–316 Feissner RE, Richard-Fogal CL, Frawley ER and Kranz RG (2006a) ABC transporter-mediated release of a haem chaperone allows cytochrome c biogenesis. Mol Microbiol 61: 219–231 Feissner RE, Richard-Fogal CL, Frawley ER, Loughman JA, Earley KW and Kranz RG (2006b) Recombinant cytochromes c biogenesis systems I and II and analysis of haem delivery pathways in Escherichia coli. Mol Microbiol 60: 563–577 Garcia-Horsman JA, Berry E, Shapleigh JP, Alben JO and Gennis RB (1994) A novel cytochrome c oxidase from Rhodobacter sphaeroides that lacks CuA. Biochemistry 33: 3113–3119 Goldman BS and Kranz RG (2001) ABC transporters associated with cytochrome c biogenesis. Res Microbiol 152: 323–329 Goldman BS, Beckman DL, Bali A, Monika EM, Gabbert KK and Kranz RG (1997) Molecular and immunological analysis of an ABC transporter complex required for cytochrome c biogenesis. J Mol Biol 268: 724–738 Goldman BS, Beck DL, Monika EM and Kranz RG (1998) Transmembrane heme delivery systems. Proc Natl Acad Sci USA 95: 5003–5008 Gray KA, Grooms M, Myllykallio H, Moomaw C, Slaughter C and Daldal F (1994) Rhodobacter capsulatus contains a novel cb-type cytochrome c oxidase without a CuA center. Biochemistry 33: 3120–3127 Gray KA and Daldal F (1995) Mutational Studies of the Cytochrome bc1 complex. In: Blankenship RE, Madigan, MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol. 2), pp 747–774. Kluwer Academic Publishers, Dordrecht Hiser L, Di Valentin M, Hamer AG and Hosler JP (2000) Cox11p is required for stable formation of the CuB and magnesium centers of cytochrome c oxidase. J Biol Chem 275: 619–623 Hiser L and Hosler JP (2001) Heme A is not essential for assembly of the subunits of cytochrome c oxidase of Rhodobacter sphaeroides. J Biol Chem 276: 45403–45407 Horng YC, Cobine PA, Maxfield AB, Carr HS and Winge DR (2004) Specific copper transfer from the Cox17 metallochaperone to both Sco1 and Cox11 in the assembly of yeast cytochrome c oxidase. J Biol Chem 279: 35334–35340 Iwata S, Ostermeier C, Ludwig B and Michel H (1995) Structure at 2.8 Å resolution of cytochrome c oxidase from Paracoccus denitrificans. Nature 376: 660–669 Jaksch M, Ogilvie I, Yao J, Kortenhaus G, Bresser HG, Gerbitz KD and Shoubridge EA (2000) Mutations in SCO2 are associated with a distinct form of hypertrophic cardiomyopathy and cytochrome c oxidase deficiency. Hum Mol Genet 9: 795–801 Kadokura H, Katzen F and Beckwith J (2003) Protein disulfide bond formation in prokaryotes. Annu Rev Biochem 72: 111–135 Kahn D, David M, Domergue O, Daveran ML, Ghai J, Hirsch PR and Batut J (1989) Rhizobium meliloti fixGHI sequence predicts involvement of a specific cation pump in symbiotic nitrogen fixation. J Bacteriol 171: 929–939 Katzen F, Deshmukh M, Daldal F and Beckwith J (2002) Evolutionary domain fusion expanded the substrate specificity of
the transmembrane electron transporter DsbD. EMBO J 21: 3960–3969 Khalimonchuk O, Ostermann K and Rodel G (2005) Evidence for the association of yeast mitochondrial ribosomes with Cox11p, a protein required for the CuB site formation of cytochrome c oxidase. Curr Genet 47: 223–233 Koch HG, Hwang O and Daldal F (1998a) Isolation and characterization of Rhodobacter capsulatus mutants affected in cytochrome cbb3 oxidase activity. J Bacteriol 180: 969–978 Koch HG, Myllykallio H and Daldal F (1998b) Using genetics to explore cytochrome function and structure in Rhodobacter. Methods Enzymol 297: 81–94 Koch HG, Winterstein C, Saribas AS, Alben JO and Daldal F (2000) Roles of the ccoGHIS gene products in the biogenesis of the cbb3-type cytochrome c oxidase. J Mol Biol 297: 49–65 Kranz RG (1989) Isolation of mutants and genes involved in cytochromes c biosynthesis in Rhodobacter capsulatus. J Bacteriol 171: 456–464 Kranz R, Lill R, Goldman B, Bonnard G and Merchant S (1998) Molecular mechanisms of cytochrome c biogenesis: Three distinct systems. Mol Microbiol 29: 383–396 Kranz RG, Beckett CS and Goldman BS (2002) Genomic analyses of bacterial respiratory and cytochrome c assembly systems: Bordetella as a model for the system II cytochrome c biogenesis pathway. Res Microbiol 153: 1–6 Kulajta C, Thumfart JO, Haid S, Daldal F and Koch HG (2006) Multi-step assembly pathway of the cbb3-type cytochrome c oxidase complex. J Mol Biol 355: 989–1004 Kvint K, Nachin L, Diez A and Nystrom T (2003) The bacterial universal stress protein: Function and regulation. Curr Opin Microbiol 6: 140–145 La Monica RF and Marrs BL (1976) The branched respiratory system of photosynthetically grown Rhodopseudomonas capsulata. Biochim Biophys Acta 423: 431–439 Lang SE, Jenney FE, Jr. and Daldal F (1996) Rhodobacter capsulatus CycH: A bipartite gene product with pleiotropic effects on the biogenesis of structurally different c-type cytochromes. J Bacteriol 178: 5279–5290 Le Brun NE, Bengtsson J and Hederstedt L (2000) Genes required for cytochrome c synthesis in Bacillus subtilis. Mol Microbiol 36: 638–650 Lee D, Pervushin K, Bischof D, Braun M and Thony-Meyer L (2005) Unusual heme-histidine bond in the active site of a chaperone. J Am Chem Soc 127: 3716–3717 Lewin A, Crow A, Oubrie A and Le Brun NE (2006) Molecular basis for specificity of the extracytoplasmic thioredoxin ResA. J Biol Chem 281: 35467–35477 Mattatall NR, Jazairi J and Hill BC (2000) Characterization of YpmQ, an accessory protein required for the expression of cytochrome c oxidase in Bacillus subtilis. J Biol Chem 275: 28802–28809 McEwan AG, Lewin A, Davy SL, Boetzel R, Leech A, Walker D, Wood T and Moore GR (2002) PrrC from Rhodobacter sphaeroides, a homologue of eukaryotic Sco proteins, is a copper-binding protein and may have a thiol-disulfide oxidoreductase activity. FEBS Lett 518: 10–16 Metheringham R, Tyson KL, Crooke H, Missiakas D, Raina S and Cole JA (1996) Effects of mutations in genes for proteins involved in disulphide bond formation in the periplasm on the activities of anaerobically induced electron transfer chains in Escherichia coli K12. Mol Gen Genet 253: 95–102
Chapter 21
Cytochrome c Maturation and Assembly
Monika EM, Goldman BS, Beckman DL and Kranz RG (1997) A thioreduction pathway tethered to the membrane for periplasmic cytochromes c biogenesis; in vitro and in vivo studies. J Mol Biol 271: 679–692 Nobrega MP, Nobrega FG and Tzagoloff A (1990) COX10 codes for a protein homologous to the ORF1 product of Paracoccus denitrificans and is required for the synthesis of yeast cytochrome oxidase. J Biol Chem 265: 14220–14226 Oh JI and Kaplan S (2002) Oxygen adaptation. The role of the CcoQ subunit of the cbb3 cytochrome c oxidase of Rhodobacter sphaeroides 2.4.1. J Biol Chem 277: 16220–16228 Preisig O, Zufferey R and Hennecke H (1996) The Bradyrhizobium japonicum fixGHIS genes are required for the formation of the high-affinity cbb3-type cytochrome oxidase. Arch Microbiol 165: 297–305 Reid E, Eaves DJ and Cole JA (1998) The CcmE protein from Escherichia coli is a haem-binding protein. FEMS Microbiol Lett 166: 369–375 Reid E, Cole J and Eaves DJ (2001) The Escherichia coli CcmG protein fulfils a specific role in cytochrome c assembly. Biochem J 355: 51–58 Ren Q and Thony-Meyer L (2001) Physical interaction of CcmC with heme and the heme chaperone CcmE during cytochrome c maturation. J Biol Chem 276: 32591–32596 Ren Q, Ahuja U and Thony-Meyer L (2002) A bacterial cytochrome c heme lyase. CcmF forms a complex with the heme chaperone CcmE and CcmH but not with apocytochrome c. J Biol Chem 277: 7657–7663 Richard-Fogal CL, Frawley ER, Feissner RE and Kranz RG (2007) Heme concentration dependence and metalloporphyrin inhibition of the system I and II cytochrome c assembly pathways. J Bacteriol 189: 455–463 Richter OM and Ludwig B (2003) Cytochrome c oxidase–structure, function and physiology of a redox-driven molecular machine. Rev Physiol Biochem Pharmacol 147: 47–74 Sambongi Y and Ferguson SJ (1996) Mutants of Escherichia coli lacking disulphide oxidoreductases DsbA and DsbB cannot synthesise an exogenous monohaem c-type cytochrome except in the presence of disulphide compounds. FEBS Lett 398: 265–268 Sanders C, Deshmukh M, Astor D, Kranz RG and Daldal F (2005a) Overproduction of CcmG and CcmFH(Rc) fully suppresses the c-type cytochrome biogenesis defect of Rhodobacter capsulatus CcmI-null mutants. J Bacteriol 187: 4245–4256 Sanders C, Turkarslan S and Daldal F (2005b) Periplasmic oxidative folding and cytochrome c maturation: A mechanistic view of stereo-selective heme attachment. In: van der Est A and Bruce D (eds) Photosynthesis: Fundamental Aspects and Global Perspectives, pp 421–424. Allen Press, Lawrence Sanders C, Boulay C and Daldal F (2007) Membrane-spanning and periplasmic segments of CcmI have distinct functions during cytochrome c biogenesis in Rhodobacter capsulatus. J Bacteriol 189: 789–800 Schaefer L, Ballabio A and Zoghbi HY (1996) Cloning and characterization of a putative human holocytochrome c-type synthetase gene (HCCS) isolated from the critical region for microphthalmia with linear skin defects (MLS). Genomics 34: 166–172
423
Schulz H, Hennecke H and Thony-Meyer L (1998) Prototype of a heme chaperone essential for cytochrome c maturation. Science 281: 1197–1200 Schulz H, Fabianek RA, Pellicioli EC, Hennecke H and ThonyMeyer L (1999) Heme transfer to the heme chaperone CcmE during cytochrome c maturation requires the CcmC protein, which may function independently of the ABC-transporter CcmAB. Proc Natl Acad Sci USA 96: 6462–6467 Setterdahl AT, Goldman BS, Hirasawa M, Jacquot P, Smith AJ, Kranz RG and Knaff DB (2000) Oxidation-reduction properties of disulfide-containing proteins of the Rhodobacter capsulatus cytochrome c biogenesis system. Biochemistry 39: 10172–10176 Smith D, Gray J, Mitchell L, Antholine WE and Hosler JP (2005) Assembly of cytochrome-c oxidase in the absence of assembly protein Surf1p leads to loss of the active site heme. J Biol Chem 280: 17652–17656 Svensson-Ek M, Abramson J, Larsson G, Tornroth S, Brzezinski P and Iwata S (2002) The X-ray crystal structures of wild-type and EQ(I-286) mutant cytochrome c oxidases from Rhodobacter sphaeroides. J Mol Biol 321: 329–339 Swem DL, Swem LR, Setterdahl A and Bauer CE (2005) Involvement of SenC in assembly of cytochrome c oxidase in Rhodobacter capsulatus. J Bacteriol 187: 8081–8087 Takamiya K. (1983) Properties of the cytochrome c oxidase activity of cytochrome b561 from photoanaerobically grown Rhodopseudomonas sphaeroides. Plant and Cell Physiology 24: 1457–1462 Thony-Meyer L (1997) Biogenesis of respiratory cytochromes in bacteria. Microbiol Mol Biol Rev 61: 337–376 Thony-Meyer L (2002) Cytochrome c maturation: A complex pathway for a simple task? Biochem Soc Trans 30: 633–638 Thony-Meyer L (2003) A heme chaperone for cytochrome c biosynthesis. Biochemistry 42: 13099–13105 Williams JC, Sue C, Banting GS, Yang H, Glerum DM, Hendrickson WA and Schon EA (2005) Crystal structure of human SCO1: Implications for redox signaling by a mitochondrial cytochrome c oxidase ‘assembly’ protein. J Biol Chem 280: 15202–15211 Witt H and Ludwig B (1997) Isolation, analysis and deletion of the gene coding for subunit IV of cytochrome c oxidase in Paracoccus denitrificans. J Biol Chem 272: 5514–5517 Zannoni D, Melandri BA and Baccarini-Melandri A (1976) Energy transduction in photosynthetic bacteria. X. Composition and function of the branched oxidase system in wild type and respiration deficient mutants of Rhodopseudomonas capsulata. Biochim Biophys Acta 423: 413–430 Zannoni D and Daldal F (1993) The role of c-type cytochromes in catalyzing oxidative and photosynthetic electron transport in the dual functional plasmamembrane of facultative phototrophs. Arch Microbiol 160: 413–423 Zollner A, Rodel G and Haid A (1992) Molecular cloning and characterization of the Saccharomyces cerevisiae CYT2 gene encoding cytochrome-c1-heme lyase. Eur J Biochem 207: 1093–1100 Zufferey R, Preisig O, Hennecke H and Thony-Meyer L (1996) Assembly and function of the cytochrome cbb3 oxidase subunits in Bradyrhizobium japonicum. J Biol Chem 271: 9114–9119
Chapter 22 Structural and Mutational Studies of the Cytochrome bc1 Complex Edward A. Berrya, Dong-Woo Leeb, Li-Shar Huanga and Fevzi Daldalb* a
Lawrence Berkeley National Laboratory, 1 Cyclotron Road, Berkeley, CA 94720 U.S.A.; b Department of Biology, University of Pennsylvania, Philadelphia, PA, 19104 U.S.A.
Summary ............................................................................................................................................................... 425 I. Introduction..................................................................................................................................................... 426 II. Structural and Mutational Studies .................................................................................................................. 427 A. The Overall Architecture of the Bacterial Cytochrome bc1 .............................................................. 427 B. The Cytochrome b Subunit .............................................................................................................. 431 1. N- and C-Termini and Internal Inserts of Bacterial Cytochrome b ......................................... 431 2. The Q/QH2 Binding Sites......................................................................................................... 433 3. Conformational Changes in Cytochrome b ............................................................................ 435 C. The Cytochrome c1 Subunit ............................................................................................................. 436 1. Structural Properties of Bacterial Cytochrome c1 ................................................................... 436 2. An Unusual Disulfide Bridge in Some Bacterial Cytochrome c1 ............................................. 438 D. The Fe/S Protein Subunit ................................................................................................................. 438 1. The Transmembrane Helix and the Linker Region of the Fe/S Protein .................................. 439 2. Additional Inserts in the Fe/S Protein Extrinsic Domain .......................................................... 439 E. Movement of the Fe/S Protein Extrinsic Domain and Electron Transfer .......................................... 441 1. The Flexible Linker Region does not Transmit, but Rather Permits Conformational Changes .................................................................................................................................. 441 2. Contact Surfaces Between the Fe/S Protein and Cytochrome b when the Extrinsic Domain is in the b or c1 Position .......................................................................................................... 442 3. Effects of Inhibitors on the Position, Em and Mobility of the Fe/S Protein ............................... 444 4. Mutations Affecting the Mobility of the Fe/S Protein .............................................................. 445 III. Conclusions and Perspectives ....................................................................................................................... 446 Acknowledgments ................................................................................................................................................. 447 References ............................................................................................................................................................ 447
Summary The ubihydroquinone:cytochrome c oxidoreductase, or the cytochrome bc1, is a widespread multi-subunit, multi-cofactor-bearing, membrane-integral enzyme complex of crucial importance both for photosynthesis and respiration. It is a major contributor of proton motive force that is subsequently used depending on the growth conditions for multiple cellular tasks, including ATP production via photo- or oxidative- phosphorylation. The simplest form of the cytochrome bc1 is generally found in prokaryotes, and that of the purple nonsulfur anoxygenic photosynthetic bacteria of Rhodobacter species is amenable to diverse multidisciplinary approaches. This chapter is focused on recent progress related to structural and mutational studies on the cytochrome bc1. Detailed information derived from the three-dimensional (3-D) structures of the bacterial enzyme, which con*Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 425–450. © 2009 Springer Science + Business Media B.V.
426
Edward A. Berry, Dong-Woo Lee, Li-Shar Huang and Fevzi Daldal
sists of the iron-sulfur protein, cytochrome b and cytochrome c1 subunits, is integrated with molecular genetic studies that yielded invaluable mutant variants. In particular, the roles of various structural components of the cytochrome bc1 that affect the unique mobility of the iron-sulfur subunit during catalysis are emphasized. Clearly, the nature of specific protein-protein interactions between the surface loops of the iron-sulfur and the cytochrome b subunits is of importance for this mobility and the steady-state activity of the enzyme. Recent progress illustrates that the cytochrome bc1 provides an invaluable model system for studying the mechanism of redox driven proton translocation across energy transducing membranes. I. Introduction The ubihydroquinone:cytochrome (Cyt) c oxidoreductases (or the Cyt bc1) are multi-subunit membranebound enzymes encountered in a broad variety of organisms, including the purple non-sulfur facultative photosynthetic bacteria (Robertson et al., 1993; Schagger et al., 1995; Cramer et al., 1996; Berry et al., 2000; Crofts, 2004). In the latter species, these enzymes are parts of both the cyclic photosynthetic (Ps) electron transport (ET) and the mitochondriallike linear respiratory (Res) ET chains (Crofts and Meinhardt, 1982; Gennis et al., 1993; Jenney et al., 1994; Hochkoeppler et al., 1996). In these pathways, the Cyt bc1 takes electrons from lipid soluble electron carriers, hydroquinone (QH2) derivatives forming the quinone (Q) pool, and donates them to various membrane-external electron carriers such as the ctype Cyts (Cyt c2 and Cyt cy) (Jenney and Daldal, 1993; Robertson et al., 1993; Jenney et al., 1994), and the high potential iron swulfur proteins (Brandt et al., 1991; Hochkoeppler et al., 1996; Berry et al., 2000). These proteins then convey electrons to the photochemical reaction center (RC) and to the Cyt c oxidase during photosynthesis and respiration, respectively. Indeed, absence of the Cyt bc1 is detrimental to Ps ET, but often not to the Res ET, due to the occurrence of alternative pathways independent of this enzyme (Zannoni and Daldal, 1993). The purple non-sulfur bacteria of Rhodobacter species, such as Rhodobacter (Rba.) capsulatus and Rba. sphaeroides, are widely used organisms for multidisciplinary studies aimed at understanding the structure and function of the Cyt bc1 (Crofts Abbreviations: 3-D – three-dimensional; Cyt bc1 – ubihydroquinone:Cyt c oxidoreductase; Cyt – cytochrome; ED – extrinsic domain of the Fe/S protein; EPR – electron paramagnetic resonance; ESEEM – electron spin echo envelope modulation; ET – electron transfer; MOA – β-methoxyacrylate; n side – negative side; p side – positive side; Q – quinone; QH2 – hydroquinone; UHDBT – 5-undecyl-6-hydroxy-4,7-dioxobenzothiazole; XXX. pdb – protein structure data bank access code
and Meinhardt, 1982; Robertson et al., 1993; Berry et al., 2004; Darrouzet et al., 2004). These species yield ample amounts of Cyt bc1 for purification and physicochemical analyses, including crystallographic studies. In bacterial genomes, structural genes for the Cyt bc1 are frequently grouped as a single cluster, and genetic manipulations to obtain modified variants are feasible. The Ps growth phenotype allows assessment of the functional effects of mutations on the Cyt bc1, and nonfunctional variants of the enzyme often accumulate gratuitously in membranes during Res growth. Using membrane vesicles (i.e., chromatophores) that Rhodobacter species produce, ET pathways internal to the Cyt bc1 complex can be photo-activated in situ and monitored in a time-resolved manner (Hochkoeppler et al., 1996; Crofts et al., 1999b; Darrouzet et al., 1999; Osyczka et al., 2004). In chromatophores, light activation of the RC rapidly oxidizes electron carrier c-type Cyts and reduces Q, providing the Cyt bc1 with both an electron donor (QH2) and an acceptor (oxidized Cyt c). Light-triggered primary redox events in the RC being extremely rapid (i.e., much faster than microsecond), reduction kinetics of the b and c-type Cyts of the Cyt bc1 can be monitored on the millisecond time scale to dissect individual ET steps during the catalytic cycle of this enzyme. For a net oxidation of one QH2 molecule by Cyt bc1, two protons (H+) are taken up from the n side of the membrane, four H+ are released to the p side, and two elemental charges are moved across the membrane (Hinkle, 2005). The mechanism of this redox coupled proton translocation by the Cyt bc1 is best described by the modified Q cycle mechanism (Mitchell, 1976; Crofts and Meinhardt, 1982; Crofts et al., 1983). This mechanism involves oxidation of two QH2 on the p side, and reduction of a quinone (Q) on the n side of the membrane, resulting in electrical charge separations (∆Ψ) in the lipid bilayer and proton translocations (∆pH) across the membrane. The ∆Ψ and ∆pH thus formed contribute together to the generation of an electrochemical potential ∆µH+, which is used subsequently for energy-requiring cel-
Chapter 22
Structure of the Cytochrome bc1 Complex
427
lular processes, including ATP production by the ATP synthase (Mitchell, 1961; Dutton et al., 1998; Saraste, 1999). This chapter is focused on the structural and mutational studies of the Cyt bc1 while Chapter 23 (Kramer et al.) deals with its mechanistic features, including bifurcated Qo site electron transfer. II. Structural and Mutational Studies The first 3-D structures for the Cyt bc1 were obtained some years ago using enzymes purified from mitochondria (Xia et al., 1997; Iwata et al., 1998, 1999; Zhang et al., 1998; Berry et al., 1999; Hunte et al., 2000; Lange and Hunte, 2002). This was followed by the chloroplast Cyt b6 f (Kurisu et al., 2003; Stroebel et al., 2003; Cramer et al., 2004) and then the bacterial enzymes. Preparation of high quality crystals from the latter proteins has proven to be more difficult than from their mitochondrial counterparts, perhaps due to their lesser stability (Elberry et al., 2006a) and the absence of large hydrophilic ‘core proteins’ which are present only in mitochondrial enzymes. Although crystals of Rhodobacter enzymes were obtained earlier, it is only during the last few years that bacterial Cyt bc1 structures became available. First, a low-resolution structure for Rba. capsulatus enzyme (Berry et al., 2004), and more recently, a higher resolution structure from a stable variant of Rba. sphaeroides Cyt bc1 (Table 1) were reported (Esser et al., 2006). The two structures are quite similar except in a few places where insertions in one protein relative to the other result in differences. Comparison with the Rba. sphaeroides structure allowed rebuilding the Rba. capsulatus structure at a few previously ambiguous locations (Berry, unpublished). As a result, we now have a clearer picture of the overall structure of the purple bacterial Cyt bc1 complex (Fig. 1A). Now, the availability of well-defined structures coupled with powerful directed mutagenesis approaches renders the bacterial enzyme an unparalleled experimental system. A. The Overall Architecture of the Bacterial Cytochrome bc1 The overall architecture of the bacterial Cyt bc1 is an intertwined homodimer with the two monomers organized around a twofold molecular axis, and is more similar to its mitochondrial than its chloroplast homologs (Fig. 1A). Each monomer of Rhodobacter
Fig. 1. Overall structure of the Rba. capsulatus Cyt bc1. The dimeric Cyt bc1 is depicted in the membrane, indicating the likely positions of the boundaries of the membrane lipid bilayer (A). The lower panels show a monomeric Cyt bc1 with the Fe/S protein ED in the b (B) or c1 (C) positions. The substrate Cyt c2 (1C2R. pdb) is uppermost in (B) and (C) and is modeled into its likely reactive position based on the yeast co-crystal structure (1KYO. pdb). The regions of the Cyt bc1 structure that differ significantly from the mitochondrial enzyme are in dark gray. The [2Fe2S] cluster is shown as a space-filled structure. The Qo site inhibitor stigmatellin locks the Fe/S protein ED in the b position. See also Color Plate 10, Fig. 16.
Cyt bc1 is constituted of three subunits, which are the high potential [2Fe2S] cluster containing iron-sulfur (Fe/S) protein, also called the Rieske protein (Rieske et al., 1964), the multi span integral membrane protein Cyt b, and the covalently attached heme containing c-type Cyt, called Cyt c1 (Berry et al., 2000; Darrouzet et al., 2004). The Cyt b subunit has two b-type hemes (named bL and bH, for low and high redox mid point potentials) that are located on the p and n sides, respectively, of the membrane. Ten transmembrane helices (one from the Fe/S protein, one from Cyt c1 and eight from Cyt b) form the central core of the monomer in the bacterial enzymes. This core por-
Edward A. Berry, Dong-Woo Lee, Li-Shar Huang and Fevzi Daldal
428
Table 1. Some recently described Rhodobacter cytochrome bc1 mutations Cyt bc1 subunit Cytochrome b
Mutation location/nature
Properties of the enzyme
Reference
Rba. capsulatus
G146A,V T163F G182S
Inactive but assembled Assembly defective Low activity; suppressor of assembly defect inflicted by T163F of Cyt b No major effect Cyt b split into Cyt b6 and SuIV homologs; highly decreased activity Additional single-base-pair substitution at the second codon of petBIV; increased enzyme amount and activity Suppressor of macro-mobility defect of ED inflicted by Ins(+1Ala46) in the Fe/S protein Inactive enzyme, mobility defect in T288S,N
(Saribas et al., 1997) (Saribas et al., 1998) "
Del(231-237) Del(231-237)S Del(231-237)S*
L286F T288S,Y,N,K T288C,G,L,I,V T288S+G182S T288S+A185V E295A,V,F,H,K,Q I304M+R306K I304M+R306K+Y302C Y302C Rba. sphaeroides
V64C G89C M92C A185C D187N F195A,Y,H,W Y199A F203A S287R Y302F,L,G,Q Del(309-326) Del(309-326)A Del(309-311)A Del(312-314)A Del(315-317)A Del(318-321)A Del(322-323)A Del(324-326)A F323A S322A,T,Y,C I326C T386C
Decreased activity, mobility defect Suppressor of macro-mobility defect of Fe/S protein ED; decreased activity Suppressor of macro-mobility defect of Fe/S protein ED, decreased activity Decreased activity Decreased activity Decreased activity Decreased activity Forms S-S bond with L34C of the Fe/S protein ED Forms S-S bond with P33C or N36C of the Fe/S protein ED Forms S-S bond with P33C of the Fe/S protein ED Forms S-S bond with K70C of the Fe/S protein ED DCCD effect? No major effect, increased superoxide production? No major effect No major effect Increases stability with V135 of the Fe/S protein ED Decreased activity Decreased activity Decreased activity Decreased activity Decreased activity Decreased activity Decreased activity Decreased activity Decreased activity Decreased activity Decreased activity, modified Qo site, increased superoxide production? Forms S-S bond with G165C of the Fe/S protein ED Forms S-S bond with K164C of the Fe/S protein ED
(Saribas et al., 1999) " "
(Darrouzet and Daldal, 2002) (Darrouzet and Daldal, 2003) " " " (Osyczka et al., 2006) (Mather et al., 2005) " " (Xiao et al., 2001) " " " (Shinkarev et al., 2000) (Gong et al., 2005) " " (Elberry et al., 2006a) (Crofts et al., 2000) (Gong et al., 2005) " " " " " " " " " (Xiao et al., 2000) "
Chapter 22
Structure of the Cytochrome bc1 Complex
429
Table 1. Continued. Cyt bc1 subunit
Mutation location/nature Del(433-445) Del(427-445) Del(425-445) Del(421-445)
Properties of the enzyme No major effect Decreased activity Decreased activity Decreased activity
Reference (Liu et al., 2004) " " "
R46C
Low activity; suppressor of assembly defect inflicted by T163F of Cyt b Decreased activity Decreased Em, inactive enzyme, native S-S bond with C167 No major effect (has Factor Xa cleavage site) Decreased Em, inactive enzyme Decreased Em, inactive enzyme Inactive enzyme (has Factor Xa cleavage site) Re-increased Em and active enzyme
(Saribas et al., 1998)
Decreased Em, inactive enzyme Decreased Em, inactive enzyme No major effect No major effect No major effect No major effect No major effect No major effect No major effect
(Darrouzet et al., 1999) (Li et al., 2002) (Li et al., 2003) " " " " " "
Decreased Em, decreased activity, forms native S-S bond with C169 Decreased Em, decreased activity, forms native S-S bond with C145 Decreased Em, decreased activity Decreased Em, decreased activity Decreased Em, decreased activity Decreased Em, inactive enzyme
(Elberry et al., 2006b)
Increased Em, suppressor of macro-mobility of the Fe/S protein ED defect inflicted by T288N of Cyt b or by L136G,H of Fe/S protein ED Increased Em, suppressor of macro-mobility of the Fe/S protein ED defect inflicted by T288S of Cyt b; suppressor of assembly defect inflicted by T163F of Cyt b Slowed macro-mobility of the Fe/S protein ED, increased Em Very slow macro-mobility of the Fe/S protein ED, drastically decreased activity, increased Em Both macro- and micro-mobility defective Fe/S subunit ED, increased Em, inactive Decreased activity No major effect No major effect No major effect No major effect Decreased activity
(Darrouzet and Daldal, 2003)
Cytochrome c1 Rba. capsulatus
F138A,V,Y C144A,S Y152R C167A C144A+C167A C167A+Y152R C144A+C167A +A181T,V M183H,K,L M183H,K Y194A,V,F E195K,Q D196K D189N+D190N D189E+D190E+D196E D189N+D190N+D196N D189K+D190K+D196K Rba. sphaeroides
C145A,S C169A,S C145A+C169A C145S+C169S M185K,L M185H
(Li et al., 2003) (Osyczka et al., 2001) " " " " "
" " " (Zhang et al., 2006) "
Fe/S Protein Rba. capsulatus
V44A,L,F
A46T,V
Ins(1Ala46) Ins(2Ala46) Ins(3Ala46) D43E,G,H,N,S K45A A46T,V M47A S49A Del1(46)
" (Saribas et al., 1998)
(Darrouzet et al., 2000b) " (Cooley et al., 2004) (Darrouzet et al., 2000a) " " " " "
Edward A. Berry, Dong-Woo Lee, Li-Shar Huang and Fevzi Daldal
430 Table 1. Continued Cyt bc1 subunit
Rba. sphaeroides
Mutation location/nature Del3(45-47) Del5(44-48) Del7(43-49) 6Pro(44-49) 6Gly 3Pro-Del3 3Gly-Del3
Properties of the enzyme Decreased activity Decreased activity Drastically decreased activity Inactive enzyme, increased Em Drastically decreased activity Drastically decreased activity Drastically decreased activity
Reference " " " " " " "
T134R,H,G T134D,N L136R,H,G,D L136G L136H L136A,Y L136H+V44L L136G+V44F L135G+A46T V139A
Decreased activity, decreased Em Decreased activity Decreased activity, decreased Em Inactive enzyme, decreased Em Inactive enzyme Decreased activity, decreased Em Decreased activity, decreased Em Decreased activity, decreased Em Decreased activity, decreased Em Suppressor of macro-mobility defect of Fe/S protein ED defect inflicted by T288S
(Liebl et al., 1997) " " (Brasseur et al., 1997) " " " " " (Darrouzet and Daldal, 2003)
P33C
Forms S-S bond with G89C or M92C of Cyt b Forms S-S bond with V64C of Cyt b Forms S-S bond with G89C of Cyt b Forms S-S bond with A185C of Cyt b Affects assembly of the enzyme Affects assembly of the enzyme Affects assembly of the enzyme Decreased enzyme activity Increased stability with S287R of Cyt b Forms S-S bond with I326C of Cyt b Forms S-S bond with T386C of Cyt b
(Xiao et al., 2001)
L34C N36C K70C Del(96-107) Del(96-107)A Del(104-107)A Del(96-99)A V135S G165C K164C
tion is smaller (a total of 886 and 895 amino acid residues in Rba. capsulatus and Rba. sphaeroides, respectively) than the organellar counterparts, which contain additional non-catalytic subunits. Some purified bacterial Cyt bc1 like that of Rba. sphaeroides also contain a cofactor-less fourth subunit that is neither required for enzymatic activity (Chen et al., 1994) nor present in the 3-D structure (PDB access code, 2FYN.pdb). A large portion of the bacterial Cyt bc1, constituted mainly of the hydrophilic parts of the Fe/S protein and the Cyt c1 subunits, protrudes on the p side of the membrane, leaving the n side of the structure quasi-devoid of proteins. The two monomers of the Cyt bc1 are assembled together in an unusual way. Of the three catalytic subunits, the transmembrane helix of the Fe/S protein is together with the Cyt b and Cyt c1 subunits of one monomer, while its extrinsic domain (ED) lies on top of the other monomer of the dimeric enzyme. Therefore, in the dimeric Cyt bc1 each Fe/S protein is associated with
" " " (Xiao et al., 2004) " " " (Elberry et al., 2006a) (Xiao et al., 2000) "
both monomers, but the significance of this unusual architecture is unknown. A Cyt bc1 monomer has two Q/QH2 binding sites that are separated roughly by 30 Å, and referred to as Qo (QH2 oxidation, H+ output) and Qi (Q reduction, H+ input) sites, according to the reactions that they catalyze. As seen in Fig. 1A, the Qo site, where QH2 is oxidized to Q and protons are released, is at the interface between the Fe/S protein and Cyt b on the p side of the membrane. The Qi site, where Q is reduced to QH2 and protons are taken up, is confined solely to Cyt b and closer to the n side of the membrane. In each monomer, the distances between the bH–bL and bL–c1 hemes are about 20 and 34 Å (Fe to Fe), whereas in the dimer the two bL-bL and bH-bH hemes are separated by roughly 20 and 34 Å, respectively (e.g., Rba. capsulatus structure 1ZRT.pdb). Resolution of the 3-D structure of the Cyt bc1 provided precise description of many of its known features, including shapes of the subunits, location of
Chapter 22
Structure of the Cytochrome bc1 Complex
the cofactors and their amino acid ligands, and yielded startling discoveries. It was found that the ED of the Fe/S protein undergoes a large-scale domain movement (roughly 56° rotation, resulting in movement of the cluster by about 13 Å; i.e., macro-movement) between the surface of Cyt b (b position, where it resides in the resting enzyme with the Fe/S protein reduced) and its immediate electron acceptor Cyt c1 heme (c1 position) (Fig. 1B and C, respectively). This macro-movement is essential for the steady-state turnover of the Cyt bc1, as mutants unable to perform it are inactive (Tian et al., 1999; Darrouzet et al., 2000b). The features of the Fe/S protein macro-movement and its role in the bifurcated electron transfer during QH2 oxidation will be discussed later. Here it suffices to state that the bifurcated electron transfer pathway at the Qo site is a unique feature of the Cyt bc1 as the basis of transmembrane charge separation, assuring the energetic efficiency of this enzyme. All amino acid residue numbers in the text refer to those of Rba. capsulatus Cyt bc1 subunits, unless indicated specifically.
431
such as their redox mid point potentials (Em(bL) = –90 mV and Em(bH) = 50 mV) (Saribas et al., 1997; Crofts et al., 1999a), and their optical absorption maxima (λbL = 558 and 566 nm and λbH = 560 nm) are distinct. The remaining four helices (helix E to H) have no prosthetic group, and helix E (residue 247 to 270), which is in close proximity to that of the Fe/S protein from the opposite monomer, runs anti parallel to the transmembrane helical anchor of Cyt c1. The α-helical surface domains ab, cd1, cd2 are located on the p side, and α-a on the n side, of the membrane. Bacterial Cyt b subunits are longer than organellar homologs, and sequence alignment shows that the extra residues are often distributed between the N- and C- terminal extensions and the three internal insertions. Omitting these five regions, which will be reviewed below in detail, the remaining 344 residues of Rba. capsulatus Cyt b encompassing amino acid residues 40 to 122, 125 to 227, 241 to 309 and 329 to 417 can be superimposed with their homologs from mitochondrial structures to serve as a central core onto which other subunits of the Cyt bc1 are attached.
B. The Cytochrome b Subunit The Cyt b subunit is the structural core of the Cyt bc1 in the lipid bilayer. A Cyt b monomer is composed of eight (A to H) membrane-spanning α-helices with the amino and carboxyl terminal ends located on the n side of the membrane (Iwata et al., 1998; Berry et al., 1999; Iwata et al., 1999; Lange et al., 2001) (Figs. 2A, and 3A and B). As a dimer, it has dimensions of 42 Å in length and a maximal diameter of 100 Å versus 70 by 100 Å for all three catalytic subunits together. Rba. capsulatus Cyt b has 437 residues of which those located between the positions 3 and 428 are represented in the 3-D structure. A schematic diagram of the secondary structure elements is shown in Fig. 2A. The first four helices of Rba. capsulatus Cyt b [A (residue 44 to 67), B (residue 89 to 119), C (residue 125 to 149) and D (187 to 218)] form a four-helix bundle, bear the heme bL and bH cofactors, and define the interface between the two monomers. The hemes bL and bH are axially coordinated between the His residues 97 and 111 of helix B, and 198 and 212 of helix D, respectively, with the individual tetrapyrrole planes being parallel to the axis of the four helix bundle, but rotated by an angle of about 50° one relative to the other. The overall protein environments of the two heme groups are different enough from each other so that their physicochemical properties
1. N- and C-Termini and Internal Inserts of Bacterial Cytochrome b The N-terminal amphipathic helix α-a is made up of residues 23 to 34. It runs parallel to the surface of the membrane going towards the other monomer of Cyt b to make a dimer contact, with Pro 24 closely approaching Trp 214 near the end of helix D of the other Cyt b monomer. The last turn of the helix α-a forms one wall of the Qi site, and can be seen as a rod connecting it and His 217 of one monomer with the end of the helix D around Trp 214 on the other monomer (Berry et al., 2004) (Fig. 3A and B). In comparison with vertebrate Cyt b sequences, the first internal insert of Rba. capsulatus Cyt b corresponds to the residues 123 to 124 in a linker region between helices B and C, in the sequence 121YxxPRE. In yeast Cyt b (Hunte et al., 2000) the two additional residues form a cis-peptide linkage, and they have been modeled as such in the Rba. capsulatus (1ZRT.pdb) and Rba. sphaeroides (2FYN. pdb) structures. The second internal insert in Rba. capsulatus Cyt b extends from residues 230 to 237 in the de linker on the n side of the membrane, near the Qi site (Figs. 2 and 3). Due to this insertion, the five-residue linker in the mitochondrial Cyt b is replaced in the Rba. capsulatus Cyt b by 13 residues
432
Edward A. Berry, Dong-Woo Lee, Li-Shar Huang and Fevzi Daldal
Fig. 2. Topologies of the Cyt b, Cyt c1 and the Fe/S protein subunits of the Cyt bc1. Panels A, B, and C represent topological drawings of the Cyt b, Cyt c1 and the Fe/S protein subunits of the Cyt bc1, respectively. In all cases, cylinders and boxes depict the transmembrane helices and β-sheets, and numbers correspond to the amino acid residues. Shaded areas in A indicate the Qo (top) and Qi (bottom) sites, respectively. The heme binding motif and the disulfide bond found in Cyt c1 are shown in B, and the [2Fe2S] cluster ligands, the disulfide bond and various regions of the Fe/S protein discussed in the text are shown in C (see the text for more details).
(positions 228 to 240) of which several are positively or negatively charged. This linker is even longer in some other bacterial species (e.g., by five residues in Chromatium (C.) vinosum), and also in the Cyt b6 f where Cyt b6 and subunit IV are split to form two distinct polypeptides. Deletion of these residues in Rba. capsulatus (Saribas et al., 1999) (Table 1) or Rba. sphaeroides Cyt b has no drastic effect on the bacterial Cyt bc1 (Chen et al., 1994). The high degree of structural conservation surrounding this region explains a posteriori why Rhodobacter Cyt b can be split into two polypeptides (Kuras et al., 1998; Saribas et al., 1999) to mimic the Cyt b6 f. The third internal insertion in Rba. capsulatus Cyt b is on the p side of the membrane, near the Qo site. It spans from the residues 308 to 325 and is located between the ef amphipathic helix and the
transmembrane helix F. The region extending from residue 309 to 328 includes a 3-turn α-helix. Similar to mitochondria, the residues 309 and 328 of Rba. capsulatus Cyt b (residues 285 and 286 of yeast/284 and 285 of the bovine sequence) form a portion of the docking site for the Fe/S protein ED. This region is in van der Waals contacts with the third loop of the Fe/S protein ED when it is located in the b position. In the bacterial enzyme, this third insert provides a gap on the rim of the Fe/S protein docking site on Cyt b, possibly facilitating its macro-movement towards the c1 position (Fig. 1B and C). Remarkably, this region is highly variable among the bacterial species, being 11 residues shorter in some α-proteobacteria, and 28 residues longer in other β-proteobacteria (e.g., some Burkholderiales). It is even absent in Rickettsia and Magnetospirillum relative to the mitochondrial Cyt b,
Chapter 22
Structure of the Cytochrome bc1 Complex
433
Fig. 3. Structural details of the Cyt b, Cyt c1 and the Fe/S protein. Panels A, C and E depict the secondary structure of the three subunits, Cyt b, Cyt c1 and the Fe/S protein, respectively, while B, D and F show the same views of the subunits in an open-coil format with key residues depicted as ball-and-sticks models. See the text for more details.
suggesting that the diversity of this insert might have arisen before endosymbiosis (Schutz et al., 2000) (Figs. 2 and 3). The C-terminus of Rba. capsulatus Cyt b is extended by 16 amino acid residues compared to the vertebrate homologs, but its last helix (H) ends early at Lys 416. The extra C-terminal residues plus the shorter H helix result in a 20-residue extended coil from position 417 to the end of the protein. This coil initially passes over the end of helix G, and makes β-bridges (between residues 360 and 418) holding together the n side ends of helices G and H. Deletion of these segments in Rba. sphaeroides Cyt bc1 does
not affect the enzyme activity, although the protein stability appears to be decreased slightly (Liu et al., 2004). Lastly, the Rba. sphaeroides structure indicates that the C-terminal eight residues of Cyt b form a short helix, which is consistent with the electron density in the corresponding portion of Rba. capsulatus structure. 2. The Q/QH2 Binding Sites The Qi site is made up of a pocket formed by the transmembrane helices A, D, E, the surface helix α-a, and the heme bH of Cyt b (Tokito and Daldal,
434
Edward A. Berry, Dong-Woo Lee, Li-Shar Huang and Fevzi Daldal
Fig. 4. Stereoview of structural details of Q sites in the Rba. capsulatus Cyt bc1. Panel A, helices α-a, A, D, and E making the Qi pocket are labeled. Heme bH and quinone bound to the Qi site are shown as ball-and-stick models. Panels B and C, the surface of Cyt b, showing interaction of the Fe/S protein ED with the ‘binding crater’ at the Qo site when the ED is in the ‘b’ (B) or ‘c1’ (C) positions. Cyt b is depicted using space-filling representation with different regions as described below. The white outline demarks the position of the Fe/S protein ED. Four regions surrounding the binding crater, the cd1-cd2 helices, ef-loop, gh-loop and α-ef2 portion are indicated. The side chains of the H135 and H156 residues of the Fe/S protein subunit are shown as ball-and-stick models. Atoms of Cyt b that are in contact (i.e., less than 3.9 Å) with the Fe/S protein ED are colored in gray, and Y302 in the α-ef helix and K329 between the α-ef2 and F helices are indicated by arrows. See also Color Plate 10, Fig. 17.
1993; Saribas et al., 1998; Berry et al., 2000) (Fig. 4A). It is shielded by the de loop and several residues located near the N-terminus of Cyt b. Similar to the mitochondrial structures (e.g., 1EZV.pdb or 2BCC. pdb), the Rba. capsulatus structure also contains
density attributable to a Q molecule at the Qi site. His 217 of Cyt b is located very close to the head group of this Q, as proposed earlier (Gray et al., 1994). The location of the Qi site, recognized by the entrapped Q molecule in the 3-D structures of the Cyt bc1, coin-
Chapter 22
Structure of the Cytochrome bc1 Complex
cides well with that defined previously by the analysis of mutants conferring resistance to Qi site inhibitors (like antimycin) or inactivating this site (Brasseur et al., 1996; Lange et al., 2001) (Table 1). As expected, the amino acid residues Ala 52, Ile 213, His 217, Gly 220, Asn 223, Pro 243 and Asp 252 of Cyt b, known to perturb the Qi site kinetics (Brasseur et al., 1996), are in the vicinity of the bound Q molecule. Unlike the Qi site, the true physical location of the Qo site is less certain due to the absence, or highly disordered state, of Q/QH2 molecule(s) in all known structures. Delimiting the exact boundaries of the Qo site is further complicated by the fact that this site involves not only Cyt b but also the Fe/S protein and, indirectly, the Cyt c1 subunits (Saribas et al., 1998) (Fig. 4B and C). In all bacterial structures available at this time, the Qo site is occupied by stigmatellin, which facilitates crystallization possibly by stabilizing the Fe/S protein ED at the b position, and no bacterial structure with any other Qo site inhibitor is yet available. Early genetic analyses of Qo site inhibitor resistant or defective mutants (see for a review Brasseur et al., 1996) indicated that Qo site mutations are grouped in two portions of Cyt b (QoI and QoII, Daldal et al., 1989). Later on, structural information derived from the mitochondrial Cyt bc1 crystals containing Qo site inhibitors also defined two overlapping but distinct regions, where either the bound stigmatellin or the MOA-type inhibitors (including myxothiazol) were found (Crofts and Berry, 1998). Both types of inhibitors occupy a region located between two ‘elbows’ of the Cyt b protein (light gray shaded ovals in Fig. 2): the junction of transmembrane helix C and the αCD1 transverse helix on one side, and the junction of the ef loop (including conserved sequence PEWY) and the helix αEF on the other. More specifically, all Qo inhibitors pass between the ring of Pro 294 residue of the 294PEWY motif and the conserved Ile 162 of the cd1 helix of Cyt b, but the two classes of inhibitors extend in different directions. The MOA-type inhibitors stretch toward heme bL, and form a H-bond to the backbone nitrogen of Glu 295 of the PEWY motif via the oxygen atom of the acrylate carbonyl of the MOA group (Table 1). On the other hand, stigmatellin extends farther in the direction away from heme bL, reaches the surface of the protein, and H-bonds to the Fe/S protein ED, located in position b (Fig. 4B) As stigmatellin lacks the acrylate group, in its presence the carboxylate group of Glu 295 of the PEWY motif occupies its space, with H-bonds bridging the inhibitors and backbone nitrogen of Glu295. These two Qo
435
site subdomains, defined by combined structural and genetic information, have been attributed the ability to accommodate different Q/QH2 intermediates during catalysis (Crofts et al., 1999b; Crofts et al., 2006). They are also intimately involved in controlling the multiple positions (i.e., micro-movement) of the Fe/S protein ED at the Cyt b surface (Darrouzet et al., 2002) as well as its macro-movement away from the b position (Esser et al., 2006) (Fig. 4B and C). Earlier, the Rba. capsulatus Cyt b residues Met 140, Tyr 147, Phe 144, Gly 152, and Gly 158 have been shown to affect either the electron transfer from QH2 to heme bL or the binding affinity of Q/QH2 to the Qo site (Daldal et al., 1989; Saribas et al., 1998). The transverse helix cd1 (which includes residues Gly 152 and Gly 158) is also located near the highly conserved 294PEWY motif of the ef loop of Cyt b. Glu 295 residue of this motif has been shown to affect the electron transfer from QH2 to heme bL (Crofts et al., 2000). Interestingly, Glu 295 residue is not universally conserved as it is absent in C. vinosum (Chen et al., 1998) or Rubrivivax gelatinosus (Ouchane et al., 2002). Recent work with Rba. capsulatus indicated that an acidic side chain at this position, although important for stigmatellin binding and QH2 oxidation, appears to be not essential for Cyt bc1 activity (Osyczka et al., 2006) (Table 1). 3. Conformational Changes in Cytochrome b Structural studies to date suggest that the transmembrane helices and the surface loops of Cyt b on the n side of the membrane are rather rigid (Fig. 4A). Binding of antimycin A to the Qi site shows no large effect in the presence of stigmatellin at the Qo site (Huang et al., 2005), or in available structures of low resolution (Berry et al., 1999; Gao et al., 2003), where the Qo site is not occupied with stigmatellin. On the other hand, long-range conformational changes in the hemes bH and bL were detected by electron paramagnetic resonance (EPR) studies using oriented membrane preparations (Cooley et al., 2005). More remarkably, in the absence of stigmatellin, major changes in the environment of the Fe/S protein [2Fe/2S] are seen readily upon binding of Qi site inhibitors to Cyt bc1 (Cooley et al., 2005). Especially antimycin A reveals two distinct subpopulations of Fe/S proteins. These changes are thought to reflect the micro-movement of the Fe/S protein ED in response to the occupancy state of the Qi site on the other side of the membrane (Fig. 4B and C). The α-a transverse helix near the
436
Edward A. Berry, Dong-Woo Lee, Li-Shar Huang and Fevzi Daldal
N-terminus of Cyt b, forming part of the surface of the Qi site in its own monomer, is poorly ordered in the orthorhombic bovine crystals (Huang et al., 2005). This helix reaches with its N-terminal end to contact Cyt b in the other monomer, and it has been proposed that it may transmit information between the Qi sites in the two monomers (Berry et al., 2004; Covian and Trumpower, 2006). However, a different role for this helix has also been proposed as the brace maintaining the separation between the Cyt b monomers on the n side of the membrane (Berry et al., 2004). Available structures show a conformational change in this region, with various structures differing by being out of register by one residue before position 31 of Rba. capsulatus Cyt b. The first turn of helix α-a is expanded in a sort of π bulge, taking up the extra residue, in all but the Rba. sphaeroides and bovine tetragonal crystal structures. The significance, if any, of the apparent rotation of this helix remains to be seen. The conformational changes in Cyt b on the p side of the membrane, involving the positions of the cd1 and cd2 surface helices and the ef loop are elaborate. Their correlations with the positions of the Fe/S protein ED, the Cyt b surface residues and the occupancy of the Qo site will be discussed in a later section. C. The Cytochrome c1 Subunit The Cyt c1, which is the most variable of the three catalytic subunits, is the exit gate for electrons from the Cyt bc1 to the electron acceptor (Cyt c2, cy or c8 in purple bacteria depending on the species) (Jenney et al., 1994; Kerfeld et al., 1996). The closest homologs of mitochondrial Cyt c1 are found in the α-proteobacteria, including purple non-sulfur bacteria. The nearest sequence match to the bovine Cyt c1 among the bacteria and archaea currently in the sequence databases is that from Rickettsia bellii, supporting the notion that mitochondria arose from an endosymbiotic event apparently linked to intracellular parasites related to Rickettsia. In bacteria, Cyt c1 is the nucleation subunit for the assembly of the Cyt bc1, as in its absence no other subunit is found in the membranes (Konishi et al., 1991; Davidson et al., 1992b). Cyt c1 is synthesized as a precursor protein, translocated across the membrane by the general secretion (SecYEG) system and processed to its mature form. It remains anchored to the membrane by a C-terminal transmembrane helix with a globular
N-terminal part protruding into the periplasm. Its prosthetic group is a protoporphyrin IX-Fe (heme b) covalently attached to the conserved heme-binding motif (34CXXCH) via the c-type Cyt maturation (Ccm) apparatus (Thony-Meyer, 2002; Turkarslan et al., 2006). The heme group is almost perpendicular to the membrane, with its axial ligands being His 38 and Met 183 in Rba. capsulatus (Fig. 3C and 3D). The membrane extrinsic domain of Cyt c1 sits atop the A, B, C, D helices and their connecting loops, while its membrane anchor runs parallel to the helix E of Cyt b (Berry et al., 2004). Some of the salient features of bacterial Cyt c1 are described below. 1. Structural Properties of Bacterial Cytochrome c1 The 3-D structure of Cyt c1 is that of a typical member of Ambler’s class 1 Cyts (Ambler, 1982; Zhang et al., 1998), differing from the electron carrier Cyts c, such as Cyt c2, in the lengths of the loops connecting the conserved features characterizing this class. The common core-folding pattern of Rba. capsulatus Cyt c1 consists of the helix α-1 with the heme binding motif at its end, helix α-5 near the Cterminus, and the highly conserved heme-supporting 98P(N/D)L(x)6R sequence. Helix α-3 is also present in nearly the same location as in Cyts c (Darrouzet et al., 2004). The α- and γ-proteobacterial Cyts c1, like those of mitochondria, are characterized by a long loop between the heme-binding CXXCH sequence and the heme-supporting P[D/N]L(X)6R motif, which is absent in class I Cyts. Based on Rhodobacter structures this long loop is branched like in the mitochondrial counterparts (loops 1a and 1b in Figs. 2B, 3C and 3D, and 5). Interestingly, the δ-and ε-proteobacterial Cyt bc1 homologs tend to have diheme Cyts c1 counterparts, and similar subunits can be artificially engineered in Rba. capsulatus as well (Lee et al., 2006). While no structure is available for such diheme Cyts c1, it was noted (Baymann et al., 2004) that the sequences are consistent with two class I Cyt c heme-binding domains joined together as seen in the Cyts c4, for which structures are available from Pseudomonas stutzeri and Acidothiobacillus ferrooxidans (Kadziola and Larsen, 1997; Abergel et al., 2003). In this respect, an unusual situation was noted in the case of the Aquifex aeolicus Cyt c1 counterpart, which is structurally closer to the monoheme Cyts c1 from purple bacteria, but sequence-wise more homologous
Chapter 22
Structure of the Cytochrome bc1 Complex
to diheme Cyts c. The N-terminal half of the Aquifex aeolicus Cyt c1 is similar to the N-terminal part of Cyts c1, while its C-terminal part aligns better with the C-terminal part of the second heme-binding domain of diheme Cyts c. This finding led to the proposal that Cyt c1 may have arisen by an internal deletion in a diheme Cyt c in the progenitor cell, fusing distal halves of the two heme-binding domains and expelling the intervening sequence from the resulting single globular heme-binding domain to yield the long branched loop characteristic of monoheme Cyts c1 (Baymann et al., 2004). Like in the ‘small’ Cyts c from Pseudomonas, algal Cyt c-553 and each domain of Cyt c4 (Kadziola and Larsen, 1997; Abergel et al., 2003), the long loop before the helix α-3 of class I c-type Cyts, which covers the propionate-bearing edge of the heme (tetrapyrrole rings A and D) and includes helix α-2, is absent in the mitochondrial Cyts c1. This exposes the heme propionate edge and allows close proximity of the Fe/S protein to heme on this side of the protein (Iwata et al., 1998; Zhang et al., 1998). Indeed, this ‘back-door’ access to heme, in addition to the exposure of the C ring on the other side of the
437
protein, to contact its physiological electron acceptor as confirmed in the yeast Cyt bc1–Cyt c co-crystal structure (Lange and Hunte, 2002), allows both the electron donor (Fe/S protein) and acceptor (Cyt c2) to bind simultaneously to the mitochondrial Cyt c1 to conduct electron transfer. In Rba. capsulatus and certain other purple bacteria, this loop is about the same length as in the ‘large’ Cyts c like Cyt c2, suggesting that this ‘back door’ may be closed (Fig. 5). This segment was poorly ordered in the Rba. capsulatus structure and was modeled as an extended coil looping out parallel to the membrane (Berry et al., 2004). The Rba. sphaeroides structure confirmed that indeed this loop does not cover the heme propionates, but depicted it in a rather different position (Fig. 3C, loop 2), going down towards the membrane and forming a transverse helix probably at the level of the lipid head groups. Thus, Cyt c1 interacts with the lipid bilayer via its transmembrane helix, via its earlier described second anchor at its N-terminus (Berry et al., 2004) (labeled 0 in Fig. 3C) and via this transverse helix in loop 2. Other structural features of bacterial Cyt c1 were described previously, and can be found in Berry et al. (2004).
Fig. 5. Top view of the Cyt c1 extrinsic domain. The heme crevice is seen edge-on from the viewpoint of Cyt c1. The Cys residues (positions 144 and 167) forming a disulfide bond are shown, along the residue Tyr 152 where a protease cleavage site was introduced to demonstrate the occurrence of the disulfide bridge (see Osyczka et al., 2001). Also shown are the heme ligands as well as the conserved aromatic residues Phe 138 and Tyr 194 and the acidic residues Asp 189, Asp 190, Glu 195, and Asp 196. The loops are numbered as in Fig. 3 panel C.
438
Edward A. Berry, Dong-Woo Lee, Li-Shar Huang and Fevzi Daldal
2. An Unusual Disulfide Bridge in Some Bacterial Cytochrome c1 A number of purple bacterial Cyt c1, including in Rhodobacter, have an insert extending from the residue 160 to 178, at positions corresponding to 154–155 in the bovine Cyt c1. This loop (loop 4 in Figs. 3 and 5) contains a conserved cysteine (Cys 167 in Rba. capsulatus) and those species containing this insert also have another cysteine in the more universally conserved loop 3, corresponding to the positions 140 to 152 in the bovine Cyt c1 (Cys 144 in Rba. capsulatus). Both Rba. capsulatus and Rba. sphaeroides structures show the presence of a disulfide bridge between these cysteines (Fig. 5) whose presence had been predicted based on genetic and biochemical data prior to structural studies (Osyczka et al., 2001) (Table 1). Loop 3 is present in all Cyts c1 and is involved in binding the acidic ‘hinge’ protein in mitochondrial enzymes. The disulfide thus holds the inserted loop 4 in roughly the same position as the hinge protein, suggesting that loop 4 might have a function similar to that of the hinge protein. However, this loop 4 does not appear to be essential, as many bacterial Cyts c1 have neither the hinge protein nor the loop 4 with the disulfide bridge. In Rba. capsulatus and Rba. sphaeroides Cyts c1 this disulfide bridge is critical for modulating the Em of the heme, as mutants lacking it have much lower Em values (Em7 of about –50 mV). The position of the sixth axial ligand Met 183 is stabilized by β-bridges between residues 179–183 and residues 153–156, forming an imperfect anti parallel two-strand β-sheet. In the mitochondrial Cyt c1 the corresponding strands are 148–150 and 156–160, connected by a relatively short linker between the residues 150 to 155. Rba. capsulatus Cyt c1 has an inserted loop 4 in this linker region, spanning the residues 160 to 177 (Fig. 5). This long loop is likely to reduce the probability of the above-mentioned two strands coming together to form a sheet, affecting the position of Met 183 as a ligand. In addition, the tetra-peptide 180WARM residues have high propensity to form an α-helix. Possibly, initiation of such a helix in this region, and its propagation into the 18-residue insert might generate a misfolded structure that could hinder Met 183 to act as a heme ligand. It appears that evolution has reduced the occurrence of this dead-end product by either providing a short linker promoting the formation of the anti parallel β-sheet, and also destabilizing the helical seed region by replacing Ala 181 with Val or Thr (of low helical propensity), or by
using a disulfide bridge to pin the 18-residue insert in an extended conformation, preventing helix extension and favoring β-sheet formation (Fig. 5). Remarkably, Rba. capsulatus Cyt c1 mutants lacking the disulfide bridge acquire readily additional mutations positioning a β-branched amino acid residue (Val or Thr) near the heme macrocycle (Table 1). These double mutants regain a functional Em7 (~ 230 mV), but like the isolated mitochondrial Cyt c1 are auto-oxidized (Osyczka et al., 2001). It is likely that in these suppressors the insert might still form an α-helix but the Thr or Val residue prevents helix initiation allowing the β-sheet to form, and Met to act as a heme ligand as in the native complex (Fig. 5). Apparently, the Em of Cyt c1 is controlled by both the occurrence of the disulfide bridge, and the nature of the amino acid side chains in the vicinity of its heme group. Mutagenesis studies similar to those described above have also been conducted on Rba. sphaeroides by two different groups, but yielded conflicting results (Table 1). The study by Elberry et al. (2006b) found that elimination of the cysteine residues forming the disulfide bond decreased the Em of Rba. sphaeroides Cyt c1 similar to that seen with Rba. capsulatus mutants. However, similar studies conducted by Zhang et al. (2006) found that the Em of such Rba. sphaeroides Cyt c1 mutants remained unchanged (Table 1). Considering that the native form of Rba. sphaeroides Cyt c1 already bears a β-branched residue close to its heme moiety elimination of the disulfide bond would be expected to have little effect on its Em, as shown previously with Rba. capsulatus suppressors (Osyczka et al., 2001) (Table 1). This interpretation is consistent with the slightly lower Em of the native Rba. sphaeroides Cyt c1 as compared to that of Rba. capsulatus (Elberry et al., 2006b). D. The Fe/S Protein Subunit The overall structure of the Rba. capsulatus Fe/S protein is very similar to its organellar counterparts in other organisms, with its globular ED bearing a [2Fe2S] cluster at one tip, connected to a single, Nterminal, transmembrane helix by a flexible linker region. In recent years, it has become clear that the Fe/S protein ED moves within the Cyt bc1 to shuttle electrons from the Qo site in Cyt b to the heme of Cyt c1, conferring onto this three-part structure the role of a periplasmic electron carrier (i.e., ED) tied to the membrane (transmembrane helix) by a flexible tether (linker region) (Fig. 3E and F).
Chapter 22
Structure of the Cytochrome bc1 Complex
Unlike the Cyt c1 subunit, the Fe/S protein does not have a processed signal sequence, and its high potential [2Fe2S] cluster is incorporated into the apoprotein prior to its membrane insertion through the twin arginine translocation (TAT) system (Bachmann et al., 2006). High-resolution 3-D structures have been obtained for the Fe/S protein ED from bovine mitochondria (Iwata et al., 1996) chloroplasts (Carrell et al., 1997), and more recently, from Rba. sphaeroides (Kolling et al., 2007), making this part of the Cyt bc1 structurally best defined. The Fe/S protein ED is formed of a base fold and four antiparallel β sheets (1 to 4), with a rubredoxin-like metal binding fold (Iwata et al., 1996; Carrell et al., 1997) (Fig. 3E and F). The [2Fe2S] cluster is coordinated by two of the four Cys and two His residues located in the two universally conserved 133C[T/K]HLGC and 153CPCHGSxY motifs (also called Box I and II, respectively). The identity of the liganding residues was postulated based on electron spin echo envelope modulation (ESEEM) measurements, indicating that two of the four ligands of the cluster come from nitrogen atoms attributed to His residues (Britt et al., 1991; Gurbiel et al., 1991). Site directed mutagenesis studies demonstrated that the conserved four Cys residues were essential (Davidson et al., 1992a; Denke et al., 1998), with two of them acting as cluster ligands, and the remaining two forming a disulfide bridge near the [2Fe2S] cluster to convey stability (Liebl et al., 1997) (Table 1). These assignments were confirmed by the structure of the Fe/S protein ED (Iwata et al., 1996) and by that of the intact Cyt bc1 (Zhang et al., 1998). In Rba. capsulatus, the solvent exposed His 156 ligand forms a H-bond with the Qo site inhibitor stigmatellin when the Fe/S protein ED is in the b position (Zhang et al., 1998), or with the electron acceptor Cyt c1 heme when in the c1 position (Iwata et al., 1998). The [2Fe2S] cluster of the Fe/S protein has several unique properties with respect to other binuclear ironsulfur clusters. Although it does not have a prominent optical spectrum, its reduced form is visible by EPR spectroscopy, with an asymmetric EPR spectrum that is centered around gav of 1.9. The high field (or gx = 1.80) region of the spectrum is variable in position and shape, depending upon the environment of the cluster domain, and is invaluable in monitoring the changes in the environment of the [2Fe2S] cluster (Von Jagow and Ohnishi, 1985; Von Jagow et al., 1986; Ohnishi et al., 1988; Berry et al., 2000; Darrouzet et al., 2001). The Em7 value of the [2Fe2S]
439
cluster is around 320 mV in intact Cyt bc1, but is lower in Q-depleted membranes or in the isolated ED (~ 260–275 mV) (Robertson et al., 1993; Saribas et al., 1998). It also exhibits pH dependence, with an apparent pKa of around 7.6 (Link and Iwata, 1996). The Em of the cluster in the intact complex is higher in the presence of Q or certain inhibitors, especially when stigmatellin is bound at the Qo site, or in some macro-movement defective mutants (450–500 mV) (Darrouzet et al., 2001) (Table 1). These Em variations will be discussed in a later section. In the following part, salient structural differences between the purple bacterial and mitochondrial Fe/S proteins, pertinent to the transmembrane helical anchor and the two specific insertions at the cluster-bearing ED, are described. 1. The Transmembrane Helix and the Linker Region of the Fe/S Protein In the bacterial structures, the first eight to ten residues on the n side of the membrane are disordered, so the Rba. sphaeroides and Rba. capsulatus structures start at residues 9 and 11, respectively (Fig. 3E and F). In Rba. capsulatus, a cluster of basic and aromatic residues, as a part of the TAT specific signal sequence, starts at position 11 (11R(R/K)xFxY) (Bachmann et al., 2006) and forms an amphipathic section to fix the Fe/S protein to the membrane surface and to lock its rotational orientation. The transmembrane helices of bacterial and chloroplast Fe/S proteins are three residues longer than their mitochondrial counterparts, and consequently form a larger angle to the membrane normal. In Rba. capsulatus, the helical structure ends at residue 37, and the protein continues with a well conserved motif, 38MxxSxDVx(L/M)xx, called a ‘linker’ or ‘hinge’ (Fig. 3E and F). Some species (Bradyrhizobium, Blastochloris, and Chromatium) have a linker one residue longer, which is inserted near position 46 occupied by a Leu or Met. 2. Additional Inserts in the Fe/S Protein Extrinsic Domain The Fe/S protein ED is formed of four β sheets and connecting loops (Fig. 2C). Sheet 1 contains the attachment of the linker region as well as the C-terminus of the protein, with the ED starting and ending in this region. The cluster-binding fold consists of β-sheet 3 (which has 4 strands) together with the loops before and after it. These latter loops and the loop between
440
Edward A. Berry, Dong-Woo Lee, Li-Shar Huang and Fevzi Daldal
the central two strands come together like three fingers at the tip of the ED to surround the cluster, which is ligated by residues in the central loop and the loop preceding the sheet (Figs. 2C and 3E and F). The bacterial Fe/S protein ED is highly superimposable with counterparts from organellar Cyt bc1, with a few modified regions. These modified portions are on the side of the Fe/S protein ED facing away from the rest of the Cyt bc1 (Hunsicker-Wang et al., 2003), keeping the side contacting the Cyt b highly conserved. The major differences are two variable inserts in the long ‘crest’ loop (Darrouzet et al., 2004) between the second and third strands of β-sheet 2, which in both bacterial and mitochondrial proteins reach out across the distal side of the ED and H-bonds with the backbone of the third cluster-surrounding loop in the cluster-binding fold (Fig. 3E and F). In the returning part of this loop is a single turn of 3-10 helix, with the highly conserved sequence motif in bacteria and mitochondria, corresponding to 110DxxR in Rba. capsulatus. The conserved Asp and Arg form salt-bridges and H-bonds with different parts of the Fe/S protein ED. Asp 110 ion-pairs with a conserved Arg 77 at the back end of the middle strand of β-sheet 2, holding these parts of the structure together, and the guanidino group of Arg 113 H-bonds to backbone carbonyl oxygens at the back end of the cluster-binding fold (Fig. 2C). Two variable regions flank the conserved 3-10 helix containing the 110DxxR motif: in most bacteria, an 11-residue and a two- or three-residue insert (except in Rba. capsulatus and a few other species like Jannaschia, Oceanicola, Roseovarius that have seven) are found before and after this 3-10 helix, respectively (Figs. 2C and 3E and F). In the case of Rba. capsulatus, the first insert preceding 110DxxR (residues 96–107) loops out towards the ‘crossover’ linker between the cluster-bearing fold and sheet 1. The very high-resolution 3-D structure of the Fe/S protein ED from Rba. sphaeroides (Kolling et al., 2007) confirms that the residues 99–101 make parallel β-bridges with residues 173 (corresponding to 177 in Rba. capsulatus) just before the linker becomes sheet β-9. Aligning the sequences corresponding to this region from bacterial Cyt bc1 shows that the Fe/S protein from Rickettsia, Rba. capsulatus, Rhodospirillum rubrum, and Blastochloris has the same number of residues as mitochondria, while in Rba. sphaeroides this insert is one residue longer (i.e., 13 residues instead of 12). Superimposition of the structures from Rba. capsulatus (1ZRT) and Rba. sphaeroides
Fe/S protein ED (Kolling et al., 2007) indicates that Ala 105 of the latter species is the inserted residue, as its neighbors Asp 104 and Gly 106 coincide with Pro 104 and Gly 105 of Rba. capsulatus (Fig. 3E and F). Experiments on Rba. sphaeroides showed that deleting this first insert (residue 96–107), or mutating all of its residues to alanine, was detrimental to the function of the Cyt bc1 (Xiao et al., 2004). Further narrowing down the critical residues, mutating either Asp 104 or Gly 106 to Ala had no effect, while mutating both together yielded a nonfunctional Cyt bc1. The structure indicates that Gly 106 is a conserved residue in a turn, with φ:ψ values of –69:–13, which are on the edge of the allowed region for non-glycine residues. Furthermore, Asp 104 reaches out into the solvent both in the high-resolution Rba. sphaeroides ED structure and in the intact Cyt bc1 structure (Fig. 3E) where the ED is in the b position. Modeling the Fe/S protein ED in the c1 position still leaves it not contacting other subunits of the Cyt bc1, thus the essential role of these residues is unclear. In Rba. sphaeroides, where the position 105 is Ala, a possibility is that mutating both positions 104 and 106 to Ala renders this stretch too hydrophobic to face the solvent, thereby causing misfolding with additional destabilization due to the strained backbone induced by substituting Gly 106 (Table 1). The second insert of the Rba. capsulatus Fe/S protein, which follows the conserved 3-10 helix with the 110DxxR motif, is seven residues (Pro 116 to Asn 122) longer relative to the mitochondrial subunits (Fig. 3E and F), while that of Rba. sphaeroides has only two (Glu 118–Ala 119) additional residues. Aligning these structures shows that in the mitochondrial Fe/S protein the Glu 118 and Ala 119 of Rba. sphaeroides are lacking, while the Pro 116 to Asn 122 stretch of Rba. capsulatus replaces Asp 117 of Rba. sphaeroides. Thus while Rba sphaeroides has two residues (118-119) inserted between 129 and 130 of the equivalent bovine sequence, Rba capsulatus has seven residues (116-122) inserted between 128 and 129 of the bovine sequence. Residues 116–118 seem to form β-bridges with residues 181–183, adding a fourth strand to sheet 1. Clearly, the two inserts flanking the conserved 110RxxD region serve to further link sheet 2 to sheets 3 and 1, and are on the distal surface of the Fe/S protein ED, away from Cyt b (Fig. 3). In addition, Rba. capsulatus and Rba. sphaeroides Fe/S proteins have one additional residue in the linker between strands β-5 and β-6 in sheet 3, as compared with the bovine counterpart. This seems
Chapter 22
Structure of the Cytochrome bc1 Complex
to be a variable region, as some other bacteria like Rhodospirillum rubrum and C. vinosum have longer (five and nine residues, respectively) inserts. E. Movement of the Fe/S Protein Extrinsic Domain and Electron Transfer The first X-ray structures of the Cyt bc1 were derived from tetragonal crystals of the bovine enzyme (Yu et al., 1996; Xia et al., 1997). In these crystals, the Fe/S protein ED could not be modeled due to disorder, but the predominant position of its [2Fe2S] cluster could be located using anomalous scattering of the iron atoms. This cluster was found close to the surface of Cyt b, near the binding site of myxothiazol, taken to be the Qo site of the Cyt bc1. Disorder of the Fe/S protein ED together with a large distance separating the [2Fe2S] cluster position and its acceptor, the Cyt c1 heme, suggested mobility. Later structures obtained using the chicken and mammalian enzymes exhibited the Fe/S protein ED in different positions (Fig. 1B and C). In these structures, the [2Fe2S] cluster was found near Cyt c1 in the absence of the Qo site inhibitors, and in a position similar to that in the bovine tetragonal crystals when the chicken enzyme was treated with stigmatellin before crystallization (Zhang et al., 1998). Crystallographic data indicated that the macro-movement of the Fe/S protein ED is approximately a rotation, with ED never leaving the surface of Cyt b. During the macro-movement between the b and c1 positions, while Fe-2 of the [2Fe2S] cluster moves by 20 Å the center of gravity of ED moves only by 5.8 Å, keeping extensive contacts with the p side surface residues of Cyt b. The mechanism(s) governing the macro-movement are complicated and still being worked out. Currently, the majority of the structural evidence derives from mitochondrial Cyt bc1 crystals, while the bacterial enzymes allow probing the movement through generation of mutations and characterization of appropriate mutants and their suppressors (Table 1). This ongoing topic of research is considered in detail below. 1. The Flexible Linker Region does not Transmit, but Rather Permits Conformational Changes Currently, no evidence for any motion of the transmembrane helix of the Fe/S protein is available. The flexible nature of the linker region makes it unlikely that a conformational change transmitted from the
441
transmembrane helix of the Fe/S protein through its linker is responsible for the macro-movement of the Fe/S protein. Thus the linker region seems to allow the Fe/S protein ED to migrate between the b and c1 positions on the surface of the Cyt b. Comparison of mitochondrial structures with the Fe/S protein ED in different positions indicates that the flexibility of the linker region resides mainly in three residues, centered around position 47 (Leu or Met) of a well conserved motif corresponding to 38MxxSxDVxx(L/M)xx in Rba. capsulatus (Fig. 6). Starting with residue 48, the Fe/S protein ED moves nearly in a rigid fashion. The first strand of β-sheet 1 of the ED begins with residue 50, and the residues 48 and 49 have backbone H-bonds with the partially conserved Trp 67–Arg 68 pair. The portion spanning the residues 37–42 of the linker sequence appears unstructured when the subunit is viewed in isolation (Fig. 6). In reality, it is tightly packed with the Cyt b subunits from both monomers. This ‘clamp’ or ‘vise’ region (Zhang et al., 1998) consists of the conserved di-glycine residues in the angle between the cd2 and D helices of Cyt b at positions 182–183 in Rba. capsulatus of the monomer with which the Fe/S protein ED interacts, and the residues 86–88, in the angle between the ab and B helices of Cyt b of the other monomer. These regions of Cyt b are indicated by small, gray shaded ovals in Fig. 2A. Although multiple van der Waals contacts and H-bonds hold this part of the linker immobile, the clamp region has a gap through which the linker could pop out. This gap explains the easy dissociation and re-association of the Fe/S protein and the Cyt bc1, via its intact transmembrane helix (Trumpower et al., 1980; Valkova-Valchanova et al., 1998). In the Rba. capsulatus linker, side chains of the conserved Ser 41 and Asp 43 residues are H-bonded, holding the intervening sequence in a non-helical loop (Fig. 6). The helical region of the linker starts with the residue 43, and continues as far as 48, depending on the position of the Fe/S protein ED. When the Fe/S protein ED is at the c1 position this entire stretch is helical, although irregular with mixed αand 3-10 H-bonding. Conversely, when the ED is at the b position, this helix becomes extended, leaving a single turn, pinned by 3-10 helical H-bonds between the residues 42 to 45 and 43 to 46 (Fig. 6). In the same region of the chicken Cyt bc1 structure with stigmatellin, only a single H-bond (stretched out to 3.7 Å) remains between the residues occupying the positions 41 and 44, but the H-bond connecting the side chains of Asp 43 and Ser 41 is still retained.
442
Edward A. Berry, Dong-Woo Lee, Li-Shar Huang and Fevzi Daldal
Fig. 6. Residues in the ‘neck’ region of the Fe/S protein. The left side of the figure shows the structure of the stigmatellin-inhibited Cyt bc1 of Rba. capsulatus (1ZRT.pdb), with the Fe/S protein ED located in the Cyt b position and the neck region maximally extended. The right side of the figures is a model of the same region with Fe/S protein ED in the Cyt c1 state, constructed by orienting the bacterial Fe/S protein ED (structure 1ZRT.pdb) after residue 48 relative to Cyt b according to bovine structure 1BE3.pdb. For clarity, the structure is truncated after the first strand of sheet 1, which is shown to indicate the reorientation of the head domain by comparison with the same strand in the left-side of the figure. The three residues undergoing major conformational change, Ala 46, Met 47 and Ala 48, were taken from the bovine structure. Note that while Ser 49 and Lys 45 have nearly the same orientation in the two structures, Met 47 is facing in the opposite direction. On the left side with ED in the b position and the neck extended, Met 47 is located toward the viewer, while with the ED in the c1 position and the neck contracted it faces away from the viewer. This results from converting one turn of helix into extended beta-like structure. It is interesting that the residue undergoing the large rotation is always a large hydrophobic amino acid like Met or Leu. Note that the conserved Asp 43 H-bonds with the conserved Ser 41 (hidden behind the backbone) of the Fe/S protein, and also with His 68 in Cyt b.
Interestingly, in Rhizobiales this Ser–Asp pair is conserved, but reversed, underlining the importance of these interactions. 2. Contact Surfaces Between the Fe/S Protein and Cytochrome b when the Extrinsic Domain is in the b or c1 Position Besides the ‘vise’ region of Cyt b, clamping the linker of the Fe/S protein between the two monomers, many presumably transient, interfering or modulating interactions occur between the Fe/S protein and the Cyt b surface. As can be seen in Fig. 1 and Fig. 4B and C, the Fe/S protein ED interacts with Cyt b when it is located in both the b (e.g., Rba. capsulatus Cyt bc1 in the presence of stigmatellin) and c1 (e.g., chicken Cyt bc1 in the presence of azoxystobin, a myxothiazol like Qo site inhibitor) positions. Thus, during the macro-
movement, which is mainly a rotation of the Fe/S protein ED, its [2Fe2S] cluster moves from near Cyt b to near Cyt c1. Through this large conformational change, the ED contacts the Cyt b surface with its three cluster-surrounding loops, and the loop between the first and middle strands of sheet 2 (strands β2 and β3). Two of the cluster-surrounding loops provide the His and Cys ligands for the cluster. The importance of the conserved residues was recognized early on (Brasseur et al., 1997) and named Box I and Box II (see above) (Table 1). Specific residues involved in the contact with Cyt b are residues 134 to 139 (cluster ligands from Box I and disulfide bridge region), 154 to 156 (cluster ligands from Box II), 169 to 172 (third cluster-surrounding loop with no cluster ligands), and residues 70 and 71 (from the turn between strands β2 and β3 in sheet 2). When the ED is in the b position, the middle cluster
Chapter 22
Structure of the Cytochrome bc1 Complex
443
Fig. 7. Locations of mutations affecting the interactions between the Cyt b and the Fe/S protein. The left panel depicts the inter subunit disulfide bridges with the tail of the Fe/S protein of the same monomer: Mutating Gly 89, Met 92 or Val 64 to Cys led to the formation of disulfide bridges with the Fe/S protein residues Pro 33, Asn 36 or Leu 34, respectively, which were also replaced by Cys residues. These residues are located in one ‘jaw’ of the Cyt b ‘vise’ portion, which clamps the region between the trans-membrane helix and the neck of the Fe/S protein. The crystal structure is consistent with formation of these disulfides. Residues Ala 185 in Cyt b and Lys 70 in the Fe/S protein also form a disulfide bridge when mutated to Cys, consistent with their locations in the structure. Note that when Ser 287 of Cyt b is mutated to Arg and Val 135 of the Fe/S protein is mutated to Ser, the stability of the Cyt bc1 is increased. While this looks reasonable in view of their locations in the structure, the structure (1FYN) shows that these two residues do not interact in the crystal. The right panel shows that Thr 386 forms a disulfide bond with the residue 164 of the Fe/S protein upon their substitutions with Cys, consistent with their structural positions. Also Ile 326 of Cyt b forms a disulfide with the position 165 of the Fe/S protein, which looks less plausible. However, this region is poorly ordered and may be flexible enough to allow the formation of a disulfide bond. It appears that the residues Ser 322 in Cyt b of Rba. sphaeroides, corresponding to Thr 322 in Rba. capsulatus Cyt b, is especially important for enzyme activity.
loop of the Fe/S protein, bearing the cluster ligands His 156 and Cys 153, inserts loosely into the Qo site, with its only close contact being the Qo site occupant stigmatellin and an H-bond between the Tyr 302 in Cyt b and the carbonyl of Cys 155. When the ED is in the c1 position, the middle cluster loop becomes the closest region of the Fe/S protein to Cyt c1, approaching the loop 4, the N-terminus, and the D propionate of heme c1 of the latter subunit, with no contact in Cyt b closer than 6 Å (Fig. 7). The first cluster loop, bearing the cluster ligands His 135 and Cys 133, is the closest of the three cluster loops to Cyt b and to the linker portion of the Fe/S protein. Extending the ‘head’ and ‘neck’ analogies to the Fe/S protein ED and the linker, this loop can be called the ‘chin.’ In the b position, when the middle cluster loop is inserted into the Qo pocket, the chin rests on the cd1 helix of Cyt b, with the Leu 136, Cys 138 and Val 139 of the ED being in van der Waals contacts with the Trp 157 and Val 161 of Cyt b (Fig. 7). In addition, the side chain of Thr 288 of the ef loop of Cyt b H-bonds to the backbone nitrogen of
Val 139 on the chin loop of the Fe/S protein. On the other hand, when the ED is in the c1 position, the head of the Fe/S protein tilts up to face Cyt c1, its chin is raised, breaking these contacts and coming to rest on the pinnacle of the Cyt b ef loop, a loose turn pinned off by a H-bond between backbone atoms of positions 284 and 287, and supported by H-bonds from the side chains of Gln 153 and Trp 157 in the cd1 helix of Cyt b. The H-bonds formed between the backbone atoms of the residues 136 to 138 of the Fe/S protein ED and the residues Leu 286 to Thr 288 of Cyt b stabilize this position. In this conformation, an additional H-bond from the side chain of Cyt b Thr 288 to the nitrogen of position 137 in the Fe/S protein is also plausible. Presumably docking the ‘chin’ loop of the ED on the pinnacle of the Cyt b ef loop in this way holds the [2Fe2S] cluster in a position allowing its favorable interaction with heme c1 (Fig. 7). In this contact region, clearly these interactions are mainly between the backbone nitrogen and oxygen atoms, and remarkably, a number of mutants substituting the residues Leu 136 or Thr 134 of the Fe/S protein
444
Edward A. Berry, Dong-Woo Lee, Li-Shar Huang and Fevzi Daldal
(Brasseur et al., 1996; Liebl et al., 1997) and Leu 286 and Thr 288 of the Cyt b (Darrouzet and Daldal, 2002, 2003) either inactivate the Cyt bc1, or restore its activity upon the occurrence of second site suppressors (Darrouzet and Daldal, 2003) (Table 1). Next, in the third cluster loop (with no cluster ligands) Gly 169 and Pro 170 make contact with the linker between helices ef and F in Cyt b and the residues 327 to 330 at the start of helix F. In addition, a H-bond from the nitrogen of Lys 329 of Cyt b to the oxygen of Pro 170 in the ED of the Fe/S protein is seen when it is in the b position, while no contact is made in the c1 position (Fig. 7). The final region of the Fe/S protein ED involved in the contacts with Cyt b during its macro-movement is the linker between strands β2 and β3 in sheet 3. The residues Trp 67 and Arg 68 make H-bonds with the backbone in the neck region, not contacting Cyt b, except when the ED is in the c position with the neck region contracted into a helix. In vertebrate Cyt bc1 (the only example in which the c position of the Fe/S protein ED is available), this contraction brings these two residues down onto Phe (168 in the bovine sequence) of Cyt b, with Arg 68 sandwiched between Trp 67 of the Fe/S protein and the Phe of Cyt b in a cation-π interaction. When the neck region is extended in the b position, Trp 67 and Arg 68 lift off, releasing the Phe 168. Although interesting, the significance of this interaction in bacteria is unknown. While these three residues are highly conserved in mitochondrial complexes, the Phe 168 (bovine number) is replaced by Pro 184 in bacteria, and although Arg 68 is common in bacteria, it is replaced by Leu in Rba. sphaeroides. On the other hand, the next three residues (Gly 69, Lys 70, Pro 71) of the Fe/S ED are highly conserved. Gly 69 is in a reverse turn between the strands, Lys 70 contacts Cyt b in both b and c1 positions, and Pro 71 does so only in the b position. Specifically, when the ED is in the b position, the nitrogen of its Lys 70 H-bonds with the oxygen of Pro 184 of Cyt b (via a water molecule according to the high resolution bovine structure 1PPJ), and Pro 71 contacts Pro 285 of Cyt b. When the ED is in the c1 position, the nitrogen of its Lys 70 H-bonds with the Pro 285 of Cyt b while the contact with Pro 71 is abolished. It is noteworthy that as these residues are near the axis of rotation of the Fe/S ED they move very little, allowing this contact to involve Pro 285 in both positions.
3. Effects of Inhibitors on the Position, Em and Mobility of the Fe/S Protein Usually, changes in the EPR spectra reflect the changes in the environment of paramagnetic metal clusters. In the case of the Fe/S protein this feature has been used extensively to monitor the micro-mobility of the ED on the Cyt b surface, and its liganding interactions with the Qo site occupants (Von Jagow et al., 1984; Brandt et al., 1991; Darrouzet et al., 2001). Binding of specific Qo site inhibitors has pronounced, correlated effects on EPR lineshape, the Em of the Fe/S protein cluster and the position of the ED seen in crystal structures. One class of inhibitor typified by stigmatellin raises the Em, and sharpens and shifts the EPR spectrum of the cluster to higher gz value relative to those of the isolated ED fragment. These inhibitors induce the ‘fixed’ or ‘Cyt b’ position of the ED in crystal structures. Another class of inhibitors, exemplified by myxothiazol, results in a broadened gx signal and a low Em similar to those of the isolated fragment, and crystal structures show the ED in the ‘released’ or ‘Cyt c1’ position. As described above the interaction of the cluster-bearing tip of the ED with Cyt b is much greater in the Cyt b position. From this it might be inferred that interaction with the protein surrounding the Qo site and the occupant of the Qo site are responsible for the changes, and when in the released (Cyt c1) position, the cluster is in an environment similar to that in the isolated fragment. The slight decrease in Em seen on binding myxothiazol can be explained as due to displacement of endogenous ubiquinone if it is assumed the latter has an effect like stigmatellin, only weaker (Crofts et al., 2002). Thus inhibitor binding, properties of the cluster, and position of the ED show strong correlation. However, going from correlation to causation and mechanism (that is, what causes what, and how?) is still not straightforward. Classically, it has been assumed that ligands like stigmatellin raise the Em by binding more tightly when the cluster is in the reduced form. However, the fact that the binding site involves Cyt b, and the resulting requirement of a specific position of the ED for tight binding to the reduced form to occur, introduces complicating factors, considered in Crofts et al. (2002) and Darrouzet and Daldal (2002). An alternative explanation is that the Em is directly related to the position of the cluster, and thus its environment. Of course the two mechanisms are not mutually exclusive, and the rela-
Chapter 22
Structure of the Cytochrome bc1 Complex
tive importance of each may depend on the inhibitor being considered. Early attempts to explain the position of the ED in the presence of various inhibitors (Crofts et al., 1999c) centered on the observation of a strong Hbond between the inhibitor and the cluster-ligand His 156 in the Fe/S protein, first reported for stigmatellin (Zhang et al., 1998), but later observed with 5-undecyl-6-hydroxy-4,7-dioxobenzothiazole (UHDBT) (Palsdottir et al., 2003; Esser et al., 2004), 2-n-heptyl-4-hydroxyquinoline-N-oxide (HQNO) (Esser et al., 2004), atovaquone and a crocacin analog (Berry, unpublished). It was assumed that this strong H-bond provided the extra energy to shift the equilibrium toward the Cyt b position. In addition, since the bond involves the imidazole ring of histidine which is part of the extended π-orbital system of the cluster, it is reasonable to suppose that the strength of the bond would be sensitive to the redox state of the cluster (in fact it is well known that the pKa of this histidine is redox-linked), which would account for the tighter binding of the inhibitor when the cluster is reduced (and hence the effect of inhibitor on Em) as well as the different position for reduced and oxidized Fe/S protein inferred from the directional dependence of the EPR spectra of oriented membranes (Brugna et al., 2000). At the same time an alternative mechanism was proposed to explain the position of the ED, in which the affinity of the ED for the fixed position is altered by conformational changes in the surface of Cyt b induced by the inhibitor (Kim et al., 1998). This mechanism was later supported by the discovery that the inhibitors famoxadone (Gao et al., 2002) and JG144 (Esser et al., 2006) induce a fixed position of the ED but do not form the H-bond to the cluster-ligand histidine. Famoxadone induces a barely-significant 26 mV increase in Em of the cluster (Gao et al., 2002). This is in line with the observation that there is no significant structural change in the ED with oxidation state of the cluster (Kolling et al., 2007), and hence no way for the binding of the inhibitor in Cyt b to be affected by redox state of the cluster. However it is also consistent with the microstate environment explanation, as the position of the ED in the presence of famoxadone is distinct from that in the presence of stigmatellin. Comparing the structure with JG-144 (2FYU.pdb, Esser et al., 2006) or chicken structures with famoxadone or fenamidone (Berry et al., unpublished) with structures containing stigmatellin, there is a 10° rotation of the
445
ED. The cluster-bearing tip moves away from the Qo site by about 1 Å allowing the cluster-ligand His 161 (His 156 in Rba. capsulatus) to H-bond with Tyr 279 of Cyt b (Tyr 302 in Rba. capsulatus), which in the presence of stigmatellin H-bonds the backbone O of the preceding residue in the ED. In conclusion it seems that neither the H-bond nor the surface conformational changes can account for the changes in affinity for the ED induced by all the inhibitors. The H-bond cannot be responsible for fixing the ED in the presence of famoxadone, fenamidone, or JG144; as they do not form the H-bond. The surface conformational changes cannot be responsible for the extremely tight binding of specifically the reduced Fe/S protein ED in the presence of stigmatellin, since the ED protein structure does not change with redox state. It seems likely that both factors are involved to varying degrees. 4. Mutations Affecting the Mobility of the Fe/S Protein Currently there are a fair number of Cyt bc1 mutations that affect the movement of the Fe/S protein ED, deserving a full review (Table 1). However, only a few of them will be covered here due to space limitation. First, the linker region of the Fe/S protein (Fig. 6) has been heavily mutagenized since the discovery of the ED mobility. The role of the linker region length and flexibility was probed by using low-helical propensity and rotation-constrained Pro or highly flexible Gly substitutions, several residue long deletions or extensions as well as one to three Ala (+nALa) insertions (Darrouzet et al., 2000b) (Table 1). The substitution or insertion mutations severely inhibited the activity of the Cyt bc1. The EPR spectra of these mutants indicated that the ED is docked at the Qo site and the Em of the [2Fe2S] cluster increased (Darrouzet et al., 2002). Shortening the linker region by up to five residues was better tolerated, and even partially counteracted the inhibitory effect of Pro substitutions (Darrouzet et al., 2000a). However, a deletion of seven residues abolished the Cyt bc1 activity, and EPR spectroscopy showed no interaction of the [2Fe2S] cluster with the Qo site (Table 1). Thus, the length of the linker region is important as a short one could not allow the Fe/S protein ED to reach the b position properly. In contrast, considering that the Gly and Pro substitutions had similar effects, and that a common property of these amino acids is their low helix-forming propensities, these mutants
446
Edward A. Berry, Dong-Woo Lee, Li-Shar Huang and Fevzi Daldal
suggested that the helical conformation of the linker region might be more important than its flexibility (Fig. 6 and Table 1). While rotating the Fe/S protein ED, the linker region apparently provides some tension to pull the cluster-bearing tip away from the Qo site (Berry et al., 2000). In the native Cyt bc1, as the linker is attached to the opposite side of the ED in respect to the cluster, movement into the Qo site results in stretching out the helix in the linker domain, dissolving one turn of it. The propensity for this helix to re-form maintains a tension on the linker. In the Gly or Pro substitutions, or in the +nAla insertions (Table 1), this tension could be decreased, enhancing the Fe/S protein ED– Cyt b surface interactions. This would then slow the macro-movement of the Fe/S protein away from the Qo site, as observed either by the slowed re-reduction of flash-oxidized Cyt c1, or resistance to conformation-dependent cleavage by the protease thermolysin (Darrouzet and Daldal, 2002). Moreover, in these mutants, the Fe/S protein ED would be located mainly in the b position, would interact with Q at the Qo site, and increase the cluster Em (Darrouzet et al., 2002). The severity of these interactions would eventually immobilize completely the ED as seen with the multiple Pro substitution or the +3Ala insertion mutants where no re-reduction of flash-oxidized Cyt c1 can be observed even in the presence of myxothiazol (Table 1). In these mutants the Fe/S protein ED might be ‘trapped’ at the Qo site so tightly that even myxothiazol-induced conformational changes are insufficient to release it. Another unusual class of Fe/S protein mutations arose in the linker region as second-site suppressors of Leu 136 to Gly or His substitutions, located immediately next to the [2Fe2S] cluster, at the tip of the Fe/S protein ED (Brasseur et al., 1997). The initial Rba. capsulatus mutant Leu 136 to Gly or His substitutions yielded an inactive Cyt bc1, with weak interaction of the cluster with the Qo site occupants and lower cluster Em values (Liebl et al., 1997) (Fig. 7 and Table 1). As Leu 136 interacts with the ef loop of Cyt b when the Fe/S protein ED is in the c position, or with the cd1 helix when it is in the b position, these mutations decreased the occupancy of the b position by the Fe/S protein ED. Remarkably, the second-site suppressors of these mutants were located at positions 44 and 46 of the linker region of the Fe/S protein, more than 30 Å from the cluster region (Brasseur et al., 1997) (Fig. 6 and Table 1). Mutation of Ala 46 to Val, Thr or Pro, or mutation of Val 44 to Leu or Phe restored partial Cyt bc1 activity and raised the Em to
more native-like values. Moreover, in the absence of the initial Leu 136 substitutions, the Em values in these mutants were greatly elevated, and the EPR spectra indicated strong interactions with the Qo site occupants. These mutations apparently favored the location of the Fe/S protein ED in the b position, and in doing so partially overcame its inability to bind the Qo site, rendering the double mutants kinetically competent (Table 1). Considering that the region around position Ala 46 is helical when the Fe/S protein ED is in the c1, and not when in the b position, substitution of Ala 46 of high helical propensity by Val or Thr of lower helical propensities would favor the b position. This would then compensate for the decreased interactions between the ED and the Cyt b upon mutation of Leu 146 of the Fe/S protein. In contrast, a structural explanation for the mode of action of the second site revertants Leu or Phe substituting Val 44 at the linker region is less obvious (Table 1). As these substitutions are bulky residues, a possibility is that they may perturb the local structure by preventing the H-bond between Asp 43 and Ser 45 (Fig. 6), thereby extending the linker and favoring the b position. A better understanding of the effects of these mutations should await the resolution of the mutant structures. III. Conclusions and Perspectives The Cyt bc1 continues to remain a well-studied model enzyme for defining the atomic basis of vectorial charge separation across, and charge transfer within, the biomembrane, as a fundamental tenet of all living organisms. Indeed, multidisciplinary approaches combining physiological, molecular genetic, biochemical and biophysical techniques have been extremely successful for fruitful studies of bacterial Cyt bc1 as a guiding light for all organisms and their organelles. The progress in this field has been, and continues to be, at a fast pace with novel findings allowing us to unravel structural and functional intricacies of the Cyt bc1, which is of utmost importance for some diseases. These discoveries lay down the foundations for future inventions and engineering of non-native versions of the Cyt bc1 with novel properties. The exquisite experimental properties of the bacterial systems have now been supplemented with a family of X-ray structures including Rhodobacter species, rendering the bacterial systems experimentally unequalled. Clearly, progress in this
Chapter 22
Structure of the Cytochrome bc1 Complex
area will continue at a rapid pace in the future, with the conception and production of unusual, often nonnative and possibly more robust variants of the Cyt bc1 with significance to important aspects of energy conservation and human health. Acknowledgments Work in the laboratories of the authors was supported by grants from NIH (GM 38237) and DOE (91ER20052) to FD and NIH (DK44842) to EAB. The authors thank A. R. Crofts for providing the coordinates for the high resolution structures of bacterial Fe/S protein EDs prior to release. References Abergel C, Nitschke W, Malarte G, Bruschi M, Claverie JM and Giudici-Orticoni MT (2003) The structure of Acidithiobacillus ferrooxidans c4-cytochrome: A model for complex-induced electron transfer tuning. Structure 11: 547–555 Ambler RP (1982) The structure and classification of cytochrome c. In: Kaplan NO and Robinson A (eds) From Cyclotrons to Cytochromes, pp 263–282. Academic Press, New York Bachmann J, Bauer B, Zwicker K, Ludwig B and Anderka O (2006) The Rieske protein from Paracoccus denitrificans is inserted into the cytoplasmic membrane by the twin-arginine translocase. FEBS J 273: 4817–4830 Baymann F, Lebrun E and Nitschke W (2004) Mitochondrial cytochrome c1 is a collapsed di-heme cytochrome. Proc Natl Acad Sci USA 101: 17737–17740 Berry EA, Huang LS and DeRose VJ (1991) Ubiquinol-cytochrome c oxidoreductase of higher plants. Isolation and characterization of the bc1 complex from potato tuber mitochondria. J Biol Chem 266: 9064–9077 Berry EA, Huang LS, Zhang Z and Kim SH (1999) Structure of the avian mitochondrial cytochrome bc1 complex. J Bioenerg Biomembr 31: 177–190 Berry EA, Guergova-Kuras M, Huang LS and Crofts AR (2000) Structure and function of cytochrome bc complexes. Annu Rev Biochem 69: 1005–1075 Berry EA, Huang LS, Saechao LK, Pon NG, Valkova-Valchanova M and Daldal F (2004) X-Ray structure of Rhodobacter capsulatus cytochrome bc1: Comparison with its mitochondrial and chloroplast counterparts. Photosynth Res 81: 251–275 Brandt U, Haase U, Schagger H and von Jagow G (1991) Significance of the ‘Rieske’ iron-sulfur protein for formation and function of the ubiquinol-oxidation pocket of mitochondrial cytochrome c reductase (bc1 complex). J Biol Chem 266: 19958–19964 Brasseur G, Saribas AS and Daldal F (1996) A compilation of mutations located in the cytochrome b subunit of the bacterial and mitochondrial bc1 complex. Biochim Biophys Acta 1275: 61–69 Brasseur G, Sled V, Liebl U, Ohnishi T and Daldal F (1997)
447
The amino-terminal portion of the Rieske iron-sulfur protein contributes to the ubihydroquinone oxidation site catalysis of the Rhodobacter capsulatus bc1 complex. Biochemistry 36: 11685–11696 Britt RD, Sauer K, Klein MP, Knaff DB, Kriauciunas A, Yu CA, Yu L and Malkin R (1991) Electron spin echo envelope modulation spectroscopy supports the suggested coordination of two histidine ligands to the Rieske Fe-S centers of the cytochrome b6 f complex of spinach and the cytochrome bc1 complexes of Rhodospirillum rubrum, Rhodobacter sphaeroides R-26, and bovine heart mitochondria. Biochemistry 30: 1892–1901 Brugna M, Rodgers S, Schricker A, Montoya G, Kazmeier M, Nitschke W and Sinning I (2000) A spectroscopic method for observing the domain movement of the Rieske iron-sulfur protein. Proc Natl Acad Sci USA 97: 2069–2074 Carrell CJ, Zhang H, Cramer WA and Smith JL (1997) Biological identity and diversity in photosynthesis and respiration: Structure of the lumen-side domain of the chloroplast Rieske protein. Structure 5: 1613–1625 Chen YL, Dincturk HB, Qin H and Knaff DB (1998) The pet operon, encoding the prosthetic group-containing subunits of the cytochrome bc1 complex, of the purple sulfur bacterium Chromatium vinosum Photosynth Res 57: 139–158 Chen YR, Usui S, Yu CA and Yu L (1994) Role of subunit IV in the cytochrome bc1 complex from Rhodobacter sphaeroides. Biochemistry 33: 10207–10214 Cooley JW, Roberts AG, Bowman MK, Kramer DM and Daldal F (2004) The raised midpoint potential of the [2Fe2S] cluster of cytochrome bc1 is mediated by both the Qo site occupants and the head domain position of the Fe-S protein subunit. Biochemistry 43: 2217–2227 Cooley JW, Ohnishi T and Daldal F (2005) Binding dynamics at the quinone reduction (Qi) site influence the equilibrium interactions of the iron sulfur protein and hydroquinone oxidation (Qo) site of the cytochrome bc1 complex. Biochemistry 44: 10520–10532 Covian R, and Trumpower BL (2006) Regulatory interactions between ubiquinol oxidation and ubiquinone reduction sites in the dimeric cytochrome bc1 complex. J Biol Chem 281: 30925–30932 Cramer WA, Soriano GM, Ponomarev M, Huang D, Zhang H, Martinez SE and Smith JL (1996) Some new structural aspects and old controversies concerning the cytochrome b6 f complex of oxygenic photosynthesis. Annu Rev Plant Phys Plant Mol Biol 47: 477–508 Cramer WA, Zhang H, Yan J, Kurisu G and Smith JL (2004) Evolution of photosynthesis: Time-independent structure of the cytochrome b6 f complex. Biochemistry 43: 5921–5929 Crofts AR (2004) The cytochrome bc1 complex: Function in the context of structure. Ann Rev Physiol 66: 689–733 Crofts AR and Berry EA (1998) Structure and function of the cytochrome bc1 complex of mitochondria and photosynthetic bacteria. Curr Opin Struct Biol 8: 501–509 Crofts AR and Meinhardt SW (1982) A Q-cycle mechanism for the cyclic electron-transfer chain of Rhodopseudomonas sphaeroides. Biochem Soc Trans 10: 201–203 Crofts AR, Meinhardt SW, Jones KR and Snozzi M (1983) The role of the quinone pool in the cyclic electron-transfer chain of Rhodopseudomonas sphaeroides. A modified Q-cycle mechanism. Biochim Biophys Acta 723: 202–218 Crofts AR, Barquera B, Gennis RB, Kuras R, Guergova-Kuras
448
Edward A. Berry, Dong-Woo Lee, Li-Shar Huang and Fevzi Daldal
M and Berry EA (1999a) Mechanism of ubiquinol oxidation by the bc1 complex: different domains of the quinol binding pocket and their role in the mechanism and binding of inhibitors. Biochemistry 38: 15807–15826 Crofts AR, Hong S, Ugulava N, Barquera B, Gennis R, GuergovaKuras M and Berry EA (1999b) Pathways for proton release during ubihydroquinone oxidation by the bc1 complex. Proc Natl Acad Sci USA 96: 10021–10026 Crofts AR, Hong S, Zhang Z and Berry EA (1999c) Physicochemical aspects of the movement of the rieske iron sulfur protein during quinol oxidation by the bc1 complex from mitochondria and photosynthetic bacteria. Biochemistry 38: 15827–15839 Crofts AR, Guergova-Kuras M, Kuras R, Ugulava N, Li J and Hong S (2000) Proton-coupled electron transfer at the Qo site: what type of mechanism can account for the high activation barrier? Biochim Biophys Acta 1459: 456–466 Crofts AR, Shinkarev VP, Dikanov SA, Samoilova RI and Kolling D (2002) Interactions of quinone with the iron-sulfur protein of the bc1 complex: Is the mechanism spring-loaded? Biochim Biophys Acta 1555: 48–53 Crofts AR, Lhee S, Crofts SB, Cheng J and Rose S (2006) Proton pumping in the bc1 complex: A new gating mechanism that prevents short circuits. Biochim Biophys Acta 1757: 1019–1034 Daldal F, Tokito MK, Davidson E and Faham M (1989) Mutations conferring resistance to quinol oxidation (Qz) inhibitors of the Cyt bc1 complex of Rhodobacter capsulatus. EMBO J 8: 3951–3961 Darrouzet E and Daldal F (2002) Movement of the iron-sulfur subunit beyond the ef loop of cytochrome b is required for multiple turnovers of the bc1 complex but not for single turnover Qo site catalysis. J Biol Chem 277: 3471–3476 Darrouzet E and Daldal F (2003) Protein-protein interactions between cytochrome b and the Fe-S protein subunits during QH2 oxidation and large-scale domain movement in the bc1 complex. Biochemistry 42: 1499–1507 Darrouzet E, Mandaci S, Li J, Qin H, Knaff DB and Daldal F (1999) Substitution of the sixth axial ligand of Rhodobacter capsulatus cytochrome c1 heme yields novel cytochrome c1 variants with unusual properties. Biochemistry 38: 7908–7917 Darrouzet E, Valkova-Valchanova M and Daldal F (2000a) Probing the role of the Fe-S subunit hinge region during Qo site catalysis in Rhodobacter capsulatus bc1 complex. Biochemistry 39: 15475–15483 Darrouzet E, Valkova-Valchanova M, Moser CC, Dutton PL and Daldal F (2000b) Uncovering the [2Fe2S] domain movement in cytochrome bc1 and its implications for energy conversion. Proc Natl Acad Sci USA 97: 4567–4572 Darrouzet E, Moser CC, Dutton PL and Daldal F (2001) Large scale domain movement in cytochrome bc1: A new device for electron transfer in proteins. Trends Biochem Sci 26: 445–451 Darrouzet E, Valkova-Valchanova M and Daldal F (2002) The [2Fe-2S] cluster Em as an indicator of the iron-sulfur subunit position in the ubihydroquinone oxidation site of the cytochrome bc1 complex. J Biol Chem 277: 3464–3470 Darrouzet E, Cooley JW and Daldal F (2004) The cytochrome bc1 complex and its homologue the b6 f complex: Similarities and differences. Photosynth Res 79: 25–44 Davidson E, Ohnishi T, Atta-Asafo-Adjei E and Daldal F (1992a) Potential ligands to the [2Fe-2S] Rieske cluster of the cytochrome bc1 complex of Rhodobacter capsulatus probed by
site-directed mutagenesis. Biochemistry 31: 3342–3351 Davidson E, Ohnishi T, Tokito M and Daldal F (1992b) Rhodobacter capsulatus mutants lacking the Rieske FeS protein form a stable cytochrome bc1 subcomplex with an intact quinone reduction site. Biochemistry 31: 3351–3358 Denke E, Merbitz-Zahradnik T, Hatzfeld OM, Snyder CH, Link TA and Trumpower BL (1998) Alteration of the midpoint potential and catalytic activity of the Rieske iron-sulfur protein by changes of amino acids forming hydrogen bonds to the iron-sulfur cluster. J Biol Chem 273: 9085–9093 Dutton PL, Moser CC, Sled VD, Daldal F and Ohnishi T (1998) A reductant-induced oxidation mechanism for complex I. Biochim Biophys Acta 1364: 245–257 Elberry M, Xiao K, Esser L, Xia D, Yu L and Yu CA (2006a) Generation, characterization and crystallization of a highly active and stable cytochrome bc1 complex mutant from Rhodobacter sphaeroides. Biochim Biophys Acta 1757: 835–840 Elberry M, Yu L and Yu CA (2006b) The disulfide bridge in the head domain of Rhodobacter sphaeroides cytochrome c1 is needed to maintain its structural integrity. Biochemistry 45: 4991–4997 Esser L, Quinn B, Li YF, Zhang M, Elberry M, Yu L, Yu CA and Xia D (2004) Crystallographic studies of quinol oxidation site inhibitors: A modified classification of inhibitors for the cytochrome bc1 complex. J Mol Biol 341: 281–302 Esser L, Gong X, Yang S, Yu L, Yu CA and Xia D (2006) Surfacemodulated motion switch: Capture and release of iron-sulfur protein in the cytochrome bc1 complex. Proc Natl Acad Sci USA 103: 13045–13050 Gao X, Wen X, Yu C, Esser L, Tsao S, Quinn B, Zhang L, Yu L and Xia D (2002) The crystal structure of mitochondrial cytochrome bc1 in complex with famoxadone: The role of aromatic-aromatic interaction in inhibition. Biochemistry 41: 11692–11702 Gao X, Wen X, Esser L, Quinn B, Yu L, Yu CA and Xia D (2003) Structural basis for the quinone reduction in the bc1 complex: A comparative analysis of crystal structures of mitochondrial cytochrome bc1 with bound substrate and inhibitors at the Qi site. Biochemistry 42: 9067–9080 Gennis RB, Barquera B, Hacker B, Van Doren SR, Arnaud S, Crofts AR, Davidson E, Gray KA and Daldal F (1993) The bc1 complexes of Rhodobacter sphaeroides and Rhodobacter capsulatus. J Bioenerg Biomembr 25: 195–209 Gong X, Yu L, Xia D and Yu CA (2005) Evidence for electron equilibrium between the two hemes bL in the dimeric cytochrome bc1 complex. J Biol Chem 280: 9251–9257 Gray KA, Dutton PL and Daldal F (1994) Requirement of histidine 217 for ubiquinone reductase activity (Qi site) in the cytochrome bc1 complex. Biochemistry 33: 723–733 Gurbiel RJ, Ohnishi T, Robertson DE, Daldal F and Hoffman BM (1991) Q-band ENDOR spectra of the Rieske protein from Rhodobactor capsulatus ubiquinol-cytochrome c oxidoreductase show two histidines coordinated to the [2Fe-2S] cluster. Biochemistry 30: 11579–11584 Hinkle PC (2005) P/O ratios of mitochondrial oxidative phosphorylation. Biochim Biophys Acta 1706: 1–11 Hochkoeppler A, Zannoni D, Ciurli S, Meyer TE, Cusanovich MA and Tollin G (1996) Kinetics of photo-induced electron transfer from high-potential iron-sulfur protein to the photosynthetic reaction center of the purple phototroph Rhodoferax fermentans. Proc Natl Acad Sci USA 93: 6998–7002
Chapter 22
Structure of the Cytochrome bc1 Complex
Huang LS, Cobessi D, Tung EY and Berry EA (2005) Binding of the respiratory chain inhibitor antimycin to the mitochondrial bc1 complex: A new crystal structure reveals an altered intramolecular hydrogen-bonding pattern. J Mol Biol 351: 573–597 Hunsicker-Wang LM, Heine A, Chen Y, Luna EP, Todaro T, Zhang YM, Williams PA, McRee DE, Hirst J, Stout CD and Fee JA (2003) High-resolution structure of the soluble, respiratorytype Rieske protein from Thermus thermophilus: Analysis and comparison. Biochemistry 42: 7303–7317 Hunte C, Koepke J, Lange C, Rossmanith T and Michel H (2000) Structure at 2.3 Å resolution of the cytochrome bc1 complex from the yeast Saccharomyces cerevisiae co-crystallized with an antibody Fv fragment. Structure 8: 669–684 Iwata M, Bjorkman J and Iwata S (1999) Conformational change of the Rieske [2Fe-2S] protein in cytochrome bc1 complex. J Bioenerg Biomembr 31: 169–175 Iwata S, Saynovits M, Link TA and Michel H (1996) Structure of a water soluble fragment of the ‘Rieske’ iron-sulfur protein of the bovine heart mitochondrial cytochrome bc1 complex determined by MAD phasing at 1.5 Å resolution. Structure 4: 567–-579 Iwata S, Lee JW, Okada K, Lee JK, Iwata M, Rasmussen B, Link TA, Ramaswamy S and Jap BK (1998) Complete structure of the 11-subunit bovine mitochondrial cytochrome bc1 complex. Science 281: 64–71 Jenney FE Jr. and Daldal F (1993) A novel membrane-associated c-type cytochrome, Cyt cy, can mediate the photosynthetic growth of Rhodobacter capsulatus and Rhodobacter sphaeroides. EMBO J 12: 1283–1292 Jenney FE Jr., Prince RC and Daldal F (1994) Roles of the soluble cytochrome c2 and membrane-associated cytochrome cy of Rhodobacter capsulatus in photosynthetic electron transfer. Biochemistry 33: 2496–2502 Kadziola A and Larsen S (1997) Crystal structure of the dihaem cytochrome c4 from Pseudomonas stutzeri determined at 2.2 Å resolution. Structure 5: 203–216 Kerfeld CA, Chan C, Hirasawa M, Kleis-SanFrancisco S, Yeates TO and Knaff DB (1996) Isolation and characterization of soluble electron transfer proteins from Chromatium purpuratum. Biochemistry 35: 7812–7818 Kim H, Xia D, Yu CA, Xia JZ, Kachurin AM, Zhang L, Yu L and Deisenhofer J (1998) Inhibitor binding changes domain mobility in the iron-sulfur protein of the mitochondrial bc1 complex from bovine heart. Proc Natl Acad Sci USA 95: 8026–8033 Kolling DJ, Brunzelle JS, Lhee S, Crofts AR and Nair SK (2007) Atomic resolution structures of Rieske iron-sulfur protein: Role of hydrogen bonds in tuning the redox potential of iron-sulfur clusters. Structure 15: 29–38 Konishi K, Van Doren SR, Kramer DM, Crofts AR and Gennis RB (1991) Preparation and characterization of the water-soluble heme-binding domain of cytochrome c1 from the Rhodobacter sphaeroides bc1 complex. J Biol Chem 266: 14270–14276 Kuras R, Guergova-Kuras M and Crofts AR (1998) Steps toward constructing a cytochrome b6 f complex in the purple bacterium Rhodobacter sphaeroides: An example of the structural plasticity of a membrane cytochrome. Biochemistry 37: 16280–16288 Kurisu G, Zhang H, Smith JL and Cramer WA (2003) Structure of the cytochrome b6f complex of oxygenic photosynthesis: Tuning the cavity. Science 302: 1009–1014 Lange C and Hunte C (2002) Crystal structure of the yeast cy-
449
tochrome bc1 complex with its bound substrate cytochrome c. Proc Natl Acad Sci USA 99: 2800–2805 Lange C, Nett JH, Trumpower BL and Hunte C (2001) Specific roles of protein-phospholipid interactions in the yeast cytochrome bc1 complex structure. EMBO J 20: 6591–6600 Lee DW, Ozturk Y, Mamedova A, Osyczka A, Cooley JW and Daldal F (2006) A functional hybrid between the cytochrome bc1 complex and its physiological membrane-anchored electron acceptor cytochrome cy in Rhodobacter capsulatus. Biochim Biophys Acta 1757: 346–352 Li J, Darrouzet E, Dhawan IK, Johnson MK, Osyczka A, Daldal F and Knaff DB (2002) Spectroscopic and oxidation-reduction properties of Rhodobacter capsulatus cytochrome c1 and its M183K and M183H variants. Biochim Biophys Acta 1556: 175–186 Li J, Osyczka A, Conover RC, Johnson MK, Qin H, Daldal F and Knaff DB (2003) Role of acidic and aromatic amino acids in Rhodobacter capsulatus cytochrome c1. A site-directed mutagenesis study. Biochemistry 42: 8818–8830 Liebl U, Sled V, Brasseur G, Ohnishi T and Daldal F (1997) Conserved nonliganding residues of the Rhodobacter capsulatus Rieske iron-sulfur protein of the bc1 complex are essential for protein structure, properties of the [2Fe-2S] cluster, and communication with the quinone pool. Biochemistry 36: 11675–11684 Link TA and Iwata S (1996) Functional implications of the structure of the ‘Rieske’ iron-sulfur protein of bovine heart mitochondrial cytochrome bc1 complex. Biochim Biophys Acta 1275: 54–60 Liu X, Yu C-A and Yu L (2004) The role of extra fragment at the C-terminal of cytochrome b (residues 421-445) in the cytochrome bc1 complex from Rhodobacter sphaeroides. J Biol Chem 279: 47363-47371 Mather MW, Darrouzet E, Valkova-Valchanova M, Cooley JW, McIntosh MT, Daldal F and Vaidya AB (2005) Uncovering the molecular mode of action of the antimalarial drug atovaquone using a bacterial system. J Biol Chem 280: 27458–27465 Mitchell P (1961) Coupling of phosphorylation to electron and hydrogen transfer by a chemi-osmotic type of mechanism. Nature 191: 144–148 Mitchell P (1976) Possible molecular mechanisms of the protonmotive function of cytochrome systems. J Theor Biol 62: 327–367 Ohnishi T, Brandt U and von Jagow G (1988) Studies on the effect of stigmatellin derivatives on cytochrome b and the Rieske iron-sulfur cluster of cytochrome c reductase from bovine heart mitochondria. Eur J Biochem 176: 385–389 Osyczka A, Dutton PL, Moser CC, Darrouzet E and Daldal F (2001) Controlling the functionality of cytochrome c1 redox potentials in the Rhodobacter capsulatus bc1 complex through disulfide anchoring of a loop and a beta-branched amino acid near the heme-ligating methionine. Biochemistry 40: 14547–14556 Osyczka A, Moser CC, Daldal F and Dutton PL (2004) Reversible redox energy coupling in electron transfer chains. Nature 427: 607–612 Osyczka A, Zhang H, Mathe C, Rich PR, Moser CC and Dutton PL (2006) Role of the PEWY glutamate in hydroquinone-quinone oxidation-reduction catalysis in the Qo Site of cytochrome bc1. Biochemistry 45: 10492–10503 Ouchane S, Agalidis I and Astier C (2002) Natural resistance to
450
Edward A. Berry, Dong-Woo Lee, Li-Shar Huang and Fevzi Daldal
inhibitors of the ubiquinol cytochrome c oxidoreductase of Rubrivivax gelatinosus: Sequence and functional analysis of the cytochrome bc1 complex. J Bacteriol 184: 3815–3822 Palsdottir H, Lojero CG, Trumpower BL and Hunte C (2003) Structure of the yeast cytochrome bc1 complex with a hydroxyquinone anion Qo site inhibitor bound. J Biol Chem 278: 31303–31311 Rieske JS, MacLennan DH and Coleman R (1964) Isolation and properties of an iron-protein from the (reduced coenzyme Q)-cytochrome c reductase complex of the respiratory chain. Biochem Biophys Res Commun 15: 338–344 Robertson DE, Ding H, Chelminski PR, Slaughter C, Hsu J, Moomaw C, Tokito M, Daldal F and Dutton PL (1993) Hydroubiquinone-cytochrome c2 oxidoreductase from Rhodobacter capsulatus: Definition of a minimal, functional isolated preparation. Biochemistry 32: 1310–1317 Saraste M (1999) Oxidative phosphorylation at the fin de siècle. Science 283: 1488–1493 Saribas AS, Ding H, Dutton PL and Daldal F (1997) Substitutions at position 146 of cytochrome b affect drastically the properties of heme bL and the Qo site of Rhodobacter capsulatus cytochrome bc1 complex. Biochim Biophys Acta 1319: 99–108 Saribas AS, Valkova-Valchanova M, Tokito MK, Zhang Z, Berry EA and Daldal F (1998) Interactions between the cytochrome b, cytochrome c1, and Fe-S protein subunits at the ubihydroquinone oxidation site of the bc1 complex of Rhodobacter capsulatus. Biochemistry 37: 8105–8114 Saribas AS, Mandaci S and Daldal F (1999) An engineered cytochrome b6 c1 complex with a split cytochrome b is able to support photosynthetic growth of Rhodobacter capsulatus. J Bacteriol 181: 5365–5372 Schagger H, Brandt U, Gencic S and von Jagow G (1995) Ubiquinol-cytochrome c reductase from human and bovine mitochondria. Methods Enzymol 260: 82–96 Schutz M, Brugna M, Lebrun E, Baymann F, Huber R, Stetter KO, Hauska G, Toci R, Lemesle-Meunier D, Tron P, Schmidt C and Nitschke W (2000) Early evolution of cytochrome bc complexes. J Mol Biol 300: 663–675 Shinkarev VP, Ugulava NB, Takahashi E, Crofts AR and Wraight CA (2000) Aspartate-187 of cytochrome b is not needed for DCCD inhibition of ubiquinol: cytochrome c oxidoreductase in Rhodobacter sphaeroides chromatophores. Biochemistry 39: 14232–14237 Stroebel D, Choquet Y, Popot JL and Picot D (2003) An atypical haem in the cytochrome b6f complex. Nature 426: 413–418 Thony-Meyer L (2002) Cytochrome c maturation: A complex pathway for a simple task? Biochem Soc Trans 30: 633–638 Tian H, White S, Yu L and Yu CA (1999) Evidence for the head domain movement of the Rieske iron-sulfur protein in electron transfer reaction of the cytochrome bc1 complex. J Biol Chem 274: 7146–7152 Tokito MK and Daldal F (1993) Roles in inhibitor recognition and quinol oxidation of the amino acid side chains at positions of Cyt b providing resistance to Qo-inhibitors of the bc1 complex from Rhodobacter capsulatus. Mol Microbiol 9: 965–978 Trumpower BL, Edwards CA and Ohnishi T (1980) Reconstitution
of the iron-sulfur protein responsible for the g = 1.90 electron paramagetic resonance signal and associated cytochrome c reductase activities to depleted succinate-cytochrome c reductase complex. J Biol Chem 255: 7487–7489 Turkarslan S, Sanders C and Daldal F (2006) Extracytoplasmic prosthetic group ligation to apoproteins: Maturation of c-type cytochromes. Mol Microbiol 60: 537–541 Valkova-Valchanova MB, Saribas AS, Gibney BR, Dutton PL and Daldal F (1998) Isolation and characterization of a two-subunit cytochrome bc1 subcomplex from Rhodobacter capsulatus and reconstitution of its ubihydroquinone oxidation (Qo) site with purified Fe-S protein subunit. Biochemistry 37: 16242–16251 Von Jagow G and Ohnishi T (1985) The chromone inhibitor stigmatellin–binding to the ubiquinol oxidation center at the C-side of the mitochondrial membrane. FEBS Lett 185: 311–315 Von Jagow G, Ljungdahl PO, Graf P, Ohnishi T and Trumpower BL (1984) An inhibitor of mitochondrial respiration which binds to cytochrome b and displaces quinone from the ironsulfur protein of the cytochrome bc1 complex. J Biol Chem 259: 6318–6326 Von Jagow G, Gribble GW and Trumpower BL (1986) Mucidin and strobilurin A are identical and inhibit electron transfer in the cytochrome bc1 complex of the mitochondrial respiratory chain at the same site as myxothiazol. Biochemistry 25: 775–780 Xia D, Yu CA, Kim H, Xia JZ, Kachurin AM, Zhang L, Yu L and Deisenhofer J (1997) Crystal structure of the cytochrome bc1 complex from bovine heart mitochondria. Science 277: 60–66 Xiao K, Yu L and Yu CA (2000) Confirmation of the involvement of protein domain movement during the catalytic cycle of the cytochrome bc1 complex by the formation of an intersubunit disulfide bond between cytochrome b and the iron-sulfur protein. J Biol Chem 275: 38597–38604 Xiao K, Chandrasekaran A, Yu L and Yu CA (2001) Evidence for the intertwined dimer of the cytochrome bc1 complex in solution. J Biol Chem 276: 46125–46131 Xiao K, Liu X, Yu CA and Yu L (2004) The extra fragment of the iron-sulfur protein (residues 96-107) of Rhodobacter sphaeroides cytochrome bc1 complex is required for protein stability. Biochemistry 43: 1488–1495 Yu CA, Xia JZ, Kachurin AM, Yu L, Xia D, Kim H and Deisenhofer J (1996) Crystallization and preliminary structure of beef heart mitochondrial cytochrome bc1 complex. Biochim Biophys Acta 1275: 47–53 Zannoni D and Daldal F (1993) The role of c-type cytochromes in catalyzing oxidative and photosynthetic electron transport in the dual functional plasmamembrane of facultative phototrophs. Arch Microbiol 160: 413–423 Zhang H, Osyczka A, Moser CC and Dutton PL (2006) Resilience of Rhodobacter sphaeroides cytochrome bc1 to heme c1 ligation changes. Biochemistry 45: 14247–14255 Zhang Z, Huang L, Shulmeister VM, Chi YI, Kim KK, Hung LW, Crofts AR, Berry EA and Kim SH (1998) Electron transfer by domain movement in cytochrome bc1. Nature 392: 677–684
Chapter 23 The Cytochrome bc1 and Related bc Complexes: The Rieske/Cytochrome b Complex as the Functional Core of a Central Electron/Proton Transfer Complex David M. Kramer* Institute of Biological Chemistry, Washington State University, Pullman WA 99164-6340, U.S.A.
Wolfgang Nitschke Laboratoire de Bioénergétique et Ingénierie des Protéines, Institut de Biologie Structurale et Microbiologie, CNRS, 13402 Marseille, France
Jason W. Cooley Department of Chemistry, University of Missouri-Columbia, Columbia MO 65211, U.S.A.
Summary ............................................................................................................................................................... 452 I. Introduction..................................................................................................................................................... 452 II. Structures of the Cytochrome bc1 and Related Rieske/Cytochrome b Complexes ....................................... 453 A. The Electron Carriers of the Rieske/Cytochrome b Complexes ...................................................... 453 B. The Quinone Binding Sites of the Cytochrome bc1 Complexes ....................................................... 453 III. Catalysis in the Rieske/Cytochrome b Complexes: The General Q-cycle Framework .................................. 455 IV. Phylogeny and Evolution ................................................................................................................................ 456 V. The ‘Third’ Redox Subunit is a Phylogenetic Marker ..................................................................................... 458 VI. The Rieske/Cytochrome b Complex, a Primordial Enzyme ........................................................................... 459 VII. The Molecular Mechanism of the Qo Site: Avoiding Q-Cycle Short Circuits................................................... 460 A. Kinetic Models .................................................................................................................................. 460 B. Thermodynamic Control of Semiquinone Intermediates .................................................................. 462 C. Gating and Shielding ........................................................................................................................ 464 D. Inter-monomer Electron Transfer .................................................................................................... 464 E. Interactions Between Quinone Binding Sites .................................................................................. 465 F. The Rieske Pivot as a Gating Mechanism........................................................................................ 465 G. Coulombic Gating Mechanisms ....................................................................................................... 466 H. ‘Proton Stripping’ .............................................................................................................................. 466 I. Complete Avoidance of Semiquinone Intermediates ........................................................................ 466 VIII. The Quinone Reduction Site, Qi of the Cytochrome bc1 Complexes ............................................................. 467 IX. The Quinone Reduction Site,Qi of the Cytochrome b6 f and Related Complexes .......................................... 468 X. The Functional Mechanism of the Rieske/Cytochrome b Complexes is Conserved...................................... 469 Acknowledgments ................................................................................................................................................. 469 References ............................................................................................................................................................ 469
*Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 451–473. © 2009 Springer Science + Business Media B.V.
452
David M. Kramer, Wolfgang Nitschke and Jason W. Cooley
Summary The cytochrome (Cyt) bc1 and related complexes play a central role in purple bacterial photosynthesis, transferring electrons between electron carriers reduced and oxidized by the photochemical reaction centers, oxidizing quinol (QH2) and reducing Cyt c while translocating protons via some variation of the Q-cycle mechanism. In this chapter, we discuss recent advances in the biochemical, biophysical and evolutionary understanding of these complexes. The mechanistic core of these complexes, conserved over billions of years, contains the Cyt b protein (with its two associated b-type hemes) and the Rieske iron-sulfur center. Together, this central core performs the central (and well-conserved) reaction of the Q-cycle, that is the ‘bifurcated’ oxidation of QH2 at the quinol oxidation (Qo) site, with two electrons sent to different acceptors, one to the Rieske iron-sulfur center and the other to the Cyt b chain. The subsequent reactions of the Q-cycle, involving the reduction of secondary carriers (a high potential Cyt c in the case of purple bacteria) and quinone at the quinone reduction (Qi) site are less well conserved both in terms of structure and mechanism. We thus use the term Rieske/Cytochrome b (RB) complexes for these enzymes. Key issues surrounding the mechanisms of the RB complexes are discussed, including a series of currently debated models for the avoidance of deleterious side reactions within the Qo site, the mechanism of stabilization of semiquinone intermediates within the Qi site, and the role of the pivoting iron-sulfur protein subunit. I. Introduction The cytochrome (Cyt) bc1, b6 f and related complexes are essential components of the energy transduction machinery of the vast majority of photosynthetic bacteria including all purple photosynthetic bacteria. From an evolutionary perspective, these complexes have in common a pair of Cyt b hemes and a [2Fe-2S] Rieske iron-sulfur cluster, and we thus will refer to them collectively as Rieske/Cyt b (RB) complexes and, specifically, by their historical names. Our reasoning for adopting this nomenclature is further justified in Sections IV and V. In addition to this catalytic core, most complexes possess structurally variable (non-conserved) secondary electron carriers Abbreviations: [2Fe-2S] – the ‘Rieske’ iron-sulfur center of the iron sulfur protein; AA – antimycin A; C – high potential electron carriers that accept electrons from the [2Fe-2S] center; CW-EPR – continuous wave-electron paramagnetic resonance; Cyt – cytochrome; Em – redox midpoint potential; EPR – electron paramagnetic resonance; HQNO – 2-n-heptyl-4-hydroxyquinoline-N-oxide; ISP – iron-sulfur protein subunit; kSOP – the rate constant for reduction of O2 by SQ; KSQ – the equilibrium constant for oxidation of QH2 to SQ; LUCA – last universal common ancestor; MOA-stilbene – E-β-methoxyacrylate-stilbene; pmf – proton motive force; Q – quinone; QB – the quinone reductase site of type II reaction centers; QH2 – quinol or hydroquinone; Qi – quinone reductase site of the RB complexes; Qo – quinol oxidase site of the RB complexes; RB – Rieske/Cytochrome b ; SC – secondary electron carriers; SQ – semiquinone (the term ‘SQ’is used to indicate both the anion (Q·– ) and neutral (Q·H) forms); UQ – ubiquinone; UQH2 – ubiquinol; vSOP – velocity of superoxide production
(SC), typically a c-type Cyt or functionally equivalent redox carriers, while some contain an additional ctype heme bound to the Cyt b protein. The primary roles of the RB complexes are 1) to make an ‘electronic’ connection between the two-electron chemistry of the quinone pool and the one-electron chemistry of downstream electron transfer chains; and 2) to translocate protons across the bioenergetic membrane, thus storing a portion of the potential energy from the two electron / two proton oxidation reaction in the electrochemical proton gradient, or proton motive force (pmf) (reviewed in Gennis et al., 1993; Trumpower and Gennis, 1994; Gray and Daldal, 1995; Crofts, 2004; Cape et al., 2006). The pmf in turn drives the synthesis of ATP at the FO-F1-ATP synthase, or other coupled transport processes. A general equation for the overall reaction catalyzed by the RB complexes can be written: QH2 + 2C(ox) + H+(n) ⇔ Q + 2C(r) + 2H+(p) where QH2 and Q are the reduced and oxidized forms of the native quinone, C(ox) and C(r) are oxidized and reduced downstream electron carriers, and H+(p) and H+(n) are aqueous protons on the positively and negatively charged sides of the energy transducing membrane. In the case of purple photosynthetic bacteria, the QH2 and C substrates for the Cyt bc1 complex are reduced ubiquinol (UQH2) and an oxidized Cyt c, either the soluble c2 or the membrane anchored
Chapter 23
Function and Evolution of Rieske/Cytochrome b Complexes
Cyt cy (Jenney et al., 1996), whose redox states are held out of equilibrium by photochemistry within the reaction center (see Chapters 16 to 20 in this volume). The chemical make-up of the analogous substrates in other species can be quite varied (see Sections IV and V). The RB field has made dramatic advances, spurred on by high resolution structural information as well as new biophysical techniques and new ways of thinking, as extensively reviewed recently ( Crofts, 2004; Osyczka et al., 2005; Cape et al., 2006; Cramer and Zhang, 2006; Crofts et al., 2006; Chapter 22, Berry et al.). The purpose of this chapter is to present a unified view of the entire range of RB complexes, to give insights into the core mechanisms. Historically, work in the RB field has been carried out using both photosynthetic (especially Rhodobacter species, cyanobacteria and chloroplast) and non-photosynthetic models (e.g., yeast or mammalian mitochondria and bacterial) and thus, where there is no evidence to the contrary, we make the simplifying assumption that all RB complexes function similarly. We use the betterunderstood Cyt bc1 complexes as a model system, but also discuss clear cases of divergent function (e.g., the Qi site of the b6 f and related complexes). II. Structures of the Cytochrome bc1 and Related Rieske/Cytochrome b Complexes Crystal structures of a number of Cyt bc1 and Cyt b6 f complexes have been published with and without bound inhibitors (Xia, 1997; Iwata et al., 1998; Hunte et al., 2000; Gao et al., 2002, 2003; Lange and Hunte, 2002; Kessl et al., 2003; Kurisu et al., 2003; Palsdottir et al., 2003; Stroebel et al., 2003). Figure 1 shows representations of ‘functional monomers’ of the Rhodobacter (Rba.) capsulatus Cyt bc1 complex (Berry et al., 2004) and Mastigocladus laminosum b6 f complex (Huang et al., 2004), showing important structural features. See Chapter 22 by Berry et al. for an extensive discussion of the structure of the Cyt bc1 complex. The Cyt bc1 complexes are dimeric integral membrane complexes with each monomer composed of as few as three subunits in some prokaryotes and up to eleven subunits in mitochondria. The Cyt b6 f complexes have from 8–11 subunits (Smith et al., 2004), notably with the homologous Cyt b protein ‘split’ into two subunits, termed ‘Cyt b’ and ‘subunit IV.’
453
A. The Electron Carriers of the Rieske/Cytochrome b Complexes As illustrated in Fig. 1, the Cyt bc1 and b6 f complexes each contain the same four essential redox-active metal centers in two distinct electron transport chains. For clarity, this arrangement is also illustrated in the absence of the protein components (Fig. 1, lower panels) and in cartoon form (Fig. 2), comparing functional components of both Cyt bc1 and b6f complexes. The Cyt b6 f complexes also contain an additional, unusually ligated, heme cofactor in the membrane portion (Stroebel et al., 2003). The role of this extra cofactor is currently being debated, as discussed in sections IV, V, VI and IX. In addition, the Cyt b6 f complexes contain bound chlorophyll (Fig. 1B) and carotenoid molecules, which might play structural roles (Smith et al., 2004). The basic ‘wiring diagram’ of two spatially diverging electron chains is essentially unchanged for all RB complexes, though (again) with some variations. The ‘low potential chain,’ so named for its redox potential versus the Q/QH2 couple, consists of two b-type hemes housed within the Cyt b protein. The two b hemes are often termed Cyt bH and Cyt bL due to their relatively higher and lower redox potentials. Some authors prefer the terms Cyt bn and bp to indicate the relative positions of the hemes with respect to the membrane polarity, with the former being located towards the negative face of the membrane. The ‘high potential chains’ always contain a ‘Rieske’ iron-sulfur cluster ([2Fe-2S]), and this is often, but not always (see section V) followed by a c-type heme; in the Cyt bc1 and b6 f complexes these hemes are bound within the Cyt c1 or Cyt f subunits, respectively. During catalysis, the ISP ‘head’ domain, containing [2Fe2S], undergoes a rotational displacement from its ‘b’ position at the Cyt b surface to its ‘c’ position near the Cyt c/f surface (as illustrated in Fig. 2), facilitating and perhaps also gating electron transfer (Zhang et al., 1998; Berry et al., 2000). There are also indications that the Cyt bc1 complexes function as dimers with inter-monomer electron transfer processes, as discussed in Sections VII.E and VII.F (Covian et al., 2004; Covian and Trumpower, 2005). B. The Quinone Binding Sites of the Cytochrome bc1 Complexes The RB complexes possess two Q/QH2 binding sites.
454
David M. Kramer, Wolfgang Nitschke and Jason W. Cooley
Fig. 1. Similarities and differences in the electron transfer chains and core structures of the Cyt bc1 and Cyt b6 f complexes. The gross functional monomeric structures of the Cyt b, Cyt c and the ISP coordinating portions of the Cyt bc1- (left) or Cyt b6 f-type (right) complexes are illustrated side-by-side as matching ribbon structures with the cofactors coordinated in each drawn as space filled spheres. The glaring differences of the presence of an extra c-type heme (heme ci) adjacent to the Qi (or QN) site of the Cyt b6 f, as well as the presence of the non-redox active chlorophyll molecule in this same complex, are readily seen in these side-by-side illustrations. However, in each case, a similar high and low potential chain of cofactors facilitating bifurcated electron flow following QH2 oxidation is easily envisioned. Both types of complex are shown with the inhibitor, stigmatellin bound at the Qo (or Qp) site as well as amino acids, pictured as stick models, thought to be important for Qi (or QN) site substrate binding to aid in visually identifying the two spatially distinct Q binding sites. The illustrations have been worked up from the Rba. capsulatus and Mastigocladus laminosus derived atomic coordinate files 1ZRT.pdb and 1VF5.pdb, respectively. See also Color Plate 11, Fig. 18.
The first of these, the quinol oxidase (Qo, sometimes called Qp) site, is located on the positively charged side (p-side) of the membrane (i.e., the periplasm of Rba. sphaeroides or Rba. capsulatus), at the interface between Cyt bL and the ISP. This site acts during normal turnover to oxidize QH2 to Q. The Qo site is large compared to the substrate QH2 headgroup, and is thought to contain two distinct ‘niches’ that bind distinct types of inhibitors and that are thought by some to have direct functional significance (see below). The ‘distal’ niche is close to where the ISP [2Fe-2S] cluster is nearest the surface of the Cyt b protein surface and is defined by the binding of qui-
none analog or ‘type I inhibitors’ such as stigmatellin (Fig. 2A and B) (reviewed in Muller et al., 2003). The ‘distal niche’ has been proposed by some to be the site of initial or primary oxidation of QH2 (reviewed in Berry and Huang, 2003). The type I inhibitors tend to form strong hydrogen bonds with one of the [2Fe-2S] cluster liganding histidine residues of the ISP, fixing the head domain in the ‘b’ position (near the Qo site). It is important to note that, while nearly all highresolution structures have been solved with inhibitors at the Qo site, none have been solved with the native substrate QH2 present, leading to difficulties
Chapter 23
Function and Evolution of Rieske/Cytochrome b Complexes
455
the Cyt b protein, together with the propionates and an edge of the Cyt bH heme. Key residues making up the Qi site cavity are highlighted in Fig. 1 as ball and stick representations. The niche functions as the site where Q is reduced to QH2 by the low potential chain. A stable SQ has been trapped (and observed via EPR) at this site as a function of reduction at raised pH values, implying that the temporally stable SQ inhabitant is in the form of a radical anion (De Vries et al., 1983). Binding of antimycin A (AA), HQNO, funiculosin and various other exotic Q analogs to the Qi site, discussed in more detail in section VII.E, inhibits the Q-cycle by displacing Q and preventing the re-oxidation of the Cyt b hemes. The Cyt b6 f (and related) version of the Qi site differs dramatically in structure and probably function, as will be covered in Section VIII. Fig. 2. The general Q-cycle framework. Shown is a cartoon depicting the common ‘wiring diagram’ of the RB complexes, known as the Q-cycle. Key functional components of both the Cyt bc1 and the Cyt b6 f complex are included. The Q-cycle mechanism is described in detail in the text. Sequence alignments suggest that the Cyt b subunit of ethylbenzene dehydrogenase (2IVF.pdb) may represent a structural model for the archaeal Cyt b558/566 subunit (Lebrun, Felle, Brugna & Nitschke, unpublished).
in narrowing the number of mechanistic models for catalysis at the Qo site. However, there is compelling evidence from electron paramagnetic resonance (EPR) work on the Rba. sphaeroides Cyt bc1 complex that the product, Q, forms a H-bond to the ε-N of His161 within the Qo pocket; this residue is also a ligand to [2Fe-2S] and to stigmatellin (Dikanov et al., 2006). Most recently, evidence has been presented for a SQ intermediate bound within the Qo site (Cape et al., 2007). The ‘proximal niche’ of the Qo site lies closer to the Cyt bL heme and thus is thought by some to receive semiquinone (SQ) produced by the distal niche (Berry et al., 2000). The proximal niche binds ‘type II inhibitors’ such as myxothiazol, mucidin, funiculosin, MOA-stilbene and other type II inhibitors which all lack the ability to hydrogen bond with the ε-N of His161 imidazole nitrogen associated with the ISP metal center as Q analogs (see review in Muller et al., 2003). The second of the Q/QH2 binding sites in the RB complexes is the Qi site, located towards the negatively charged side (n-side) of the membrane. It is formed by a cavity in the Cyt b structure defined by three helices (A, D and E) and their connecting loops of
III. Catalysis in the Rieske/Cytochrome b Complexes: The General Q-cycle Framework In this section, we summarize the generally agreed Q-cycle framework (Mitchell, 1975; Trumpower, 1990; Ding et al., 1995; Brandt, 1996; Crofts and Berry, 1998). The more controversial aspects of catalysis will be discussed in separate sections. The Q-cycle is initiated when a QH2 molecule is bound at the Qo site with the [2Fe-2S] cluster and the Cyt bL heme in their oxidized forms, allowing a ‘bifurcated’ oxidation of QH2, with one electron going to the high potential chain, and the other going to the low potential chain (Fig. 2). In most models (reviewed in Gennis et al., 1993; Trumpower and Gennis, 1994; Gray and Daldal, 1995; Berry et al., 2000; Kramer et al., 2003; Crofts, 2004; Cape et al., 2006) the first electron is transferred to the [2Fe-2S] cluster, transiently forming a semiquinone (SQ)1 species at the Qo site, which in turn reduces Cyt bL, a process that involves proton transfer from QH2 and SQ species to protein residues, and perhaps movement of the SQ from the distal to proximal niches (Crofts, 2004). In some other models, two quinone species are simultaneously bound to Qo, allowing electrons to move from one Q/SQ couple to another (Ding et al., 1995). The electron on the [2Fe-2S] cluster is transferred to the secondary high potential chain carrier, Cyt c1 1 The term SQ represents both the neutral and anionic forms of semiquinone
456
David M. Kramer, Wolfgang Nitschke and Jason W. Cooley
in the Cyt bc1 complex, Cyt f in the b6 f complex. This electron transfer step is facilitated by a largescale pivoting of the ISP from the ‘b’ position at Cyt b surface to the ‘c’ position in contact with the Cyt c1 subunit (Zhang, 1998). From Cyt c1, electrons are transferred to a downstream, tertiary carrier and then to a source of oxidant (e.g., the photosynthetic reaction center). The tertiary carriers are clearly not conserved, though their redox properties are matched to those of their respective RB complexes. In the case of the purple bacterial Cyt bc1 complex, the tertiary carrier is either the soluble Cyt c2 or the membrane anchored Cyt cy (Jenney et al., 1996). The electron transferred from SQ to Cyt bL is passed to Cyt bH, driven by a substantial (~100 mV) difference of potentials between the two hemes which is conserved across all RB complexes thus characterized (Kramer and Crofts, 1994; Nitschke et al., 1995; Crofts, 2004). In the case of the Cyt bc1 complex, electrons from Cyt bH are transferred to a Q bound at the Qi site forming a stabilized SQ. The process of oxidizing the Cyt b hemes is clearly different in the case of the Cyt b6 f and related complexes, and likely involves a third heme between Cyt bH and a Q species bound at the Qi site, as discussed in Section VIII. In either case, oxidation of a second QH2 molecule at the Qo site is required to introduce a second electron into the low potential chain (Stroebel et al., 2003), which, together with the first, reduces a Q to a QH2 at the Qi site with an uptake of two protons from the n-side of the membrane. For all RB complexes, the bifurcated oxidation of QH2 at the Qo site results in the net release of two protons into the p-side (positively charged) aqueous phase. It is the engineering of this basic step that ensures the storage of pmf and the taming of an otherwise deleterious reactive intermediate. The remainder of this chapter will focus on three current issues: 1) the evolutionary, structural and functional differences among distantly related versions of the RB complexes; 2) the mechanism of the Qo site; and 3) the mechanism of the Qi site. IV. Phylogeny and Evolution Enzymes homologous to the Cyt bc1 and b6 f complexes are found in almost all major phyla of the tree of life. These enzymes function in diverse types of energy conserving chains such as photosynthetic (both oxygenic and anoxygenic) electron transport,
and aerobic as well as numerous anaerobic kinds of respiration. The subunit composition of the enzyme family is somewhat variable with respect to the subunits analogous to Cyt c1 or Cyt f but strictly conserved regarding the functional core, that is Cyt b or Cyt b6 /subunit IV and the ISP protein. Since the ISP and Cyt b subunits are well conserved, whereas the c-type Cyt containing subunit is not, we propose the term Rieske/Cyt b (RB) complexes when dealing with the family as a whole rather than ‘bc-complexes,’ previously proposed by Schütz et al., (2000). As we will discuss throughout the review, the nomenclature better reflects the true catalytic core of the complexes. Phylogenetic trees based on Cyt b sequences were reconstructed several decades ago to elucidate species relationships among eukaryotes (Avise, 1986). With the advent of high throughput genome sequencing, this approach was more recently extended to the entire prokaryotic world (Castresana et al., 1995; Schütz et al., 2000; Xiong et al., 2000; Hiller et al., 2003; Lebrun et al., 2006). As a rule, going from close relatives towards representatives encompassing phylogenetically very distant species entails a significantly decreased reliability of phylogenies. This is due to the high probability of multiple substitutions in a given sequence position and the difficulty in deducing the ‘correct’ alignment of sequences when similarities become low. Fortunately, the Cyt b protein sequence has turned out to be extremely well suited for phylogenetic approaches as it transcends the entire tree of life. The sequence length is of sufficient size to provide a good number of phylogenetically informative sites, required for the resolution of a divergent set of sequences. More importantly, it features a set of fully conserved positions almost uniformly distributed over the whole sequence length, which serve as landmarks for multiple alignments (Lebrun et al., 2006). Figure 3 shows a phylogenetic tree of the Cyt b subunit covering the entire tree of life. The topology of this tree is astonishingly close to that of the phylogeny of the parent species, with the obvious exception of the eukaryotic mitochondria, which are rooted within the α-proteobacteria reflecting the endosymbiotic origin of mitochondria. Examples of lateral gene transfer are provided by the two Aquificales (Aquifex aeolicus and Sulfurihydrogenibium azorense), which cluster with the ε-proteobacteria rather than being a stem-group phylum. Analysis of the operons encoding the enzyme
Chapter 23
Function and Evolution of Rieske/Cytochrome b Complexes
457
Fig. 3. Diversity of RB complexes on the Tree of Life. The phylogram represented to the left has been reconstructed from sequences of Cyt b using the neighbor-joining algorithm. Bootstrap values correspond to the frequency of nodes in 1000 bootstrap replicates. In the middle of the figure, operon structures for the respective RB complexes are schematically depicted. The arrowhead indicates the splitting within the cytochrome b gene yielding cytochrome b6 and subunit IV. The structures on the right hand side represent the functional core consisting of Cyt b and the ISP protein (top, based on 1EZV.pdb) and diverse cases of the terminal electron accepting subunit of the high potential chain (Cyt f: 1CTM.pdb; Cyt c3: 1GMB.pdb; Cyt c4: 1H1O.pdb; Cyt c1: 1EZV.pdb). The empty circle denotes the space taken by the diverse 3rd subunits (see corresponding structures) present in the different phyla.
shows that Cyt b and the ISP are present under all circumstances. In all cases except cyanobacteria and plastids (as well as an isolated case of a γ-proteobacterium), the genes encoding these two proteins form a tandem with Cyt b downstream of the ISP gene (Fig. 3). In cyanobacteria, the contiguousness of the ISP/Cyt b pair is disrupted, and the two gene loci are not part of a common operon. The split of the Cyt b protein into two separate subunits, i.e., Cyt b6 (N-terminal half of Cyt b) and subunit IV (C-terminal half), recognized several decades ago (Widger et al., 1984) for the cyanobacterial and plastidic enzyme, extends to the RB complexes present in low GC-Gram-positive and in heliobacteria (grouped together into the phylum of the firmicutes). As demonstrated by the tree shown in Fig. 3, firmicutes and cyanobacteria form a common clade indicating that the gene split took place in their common ancestor. Another feature setting the RB complexes from
firmicutes, heliobacteria and cyanobacteria aside from all other members of the family is the presence of an additional redox cofactor, heme ci. Heme ci was first discovered in the X-ray structures of Cyt b6 f complexes from Mastigocladus (Kurisu et al., 2003) and Chlamydomonas (Stroebel et al., 2003), though it was alluded to by observation of a covalently attached heme from biochemical work with the Bacillus RB complex (Yu and Lebrun, 1998). This cofactor was found to be covalently attached to the cytochrome b6 subunit via a single cysteine residue. This cysteine residue is conserved in the RB complexes from firmicutes and heliobacteria, and spectroscopic evidence indicates the presence of heme ci also in the enzymes from these two clades (A. Kanazawa and F. Baymann, personal communication). The tree topology thus suggests that heme ci was introduced into the RB complexes after the node characterizing the divergence of the
458
David M. Kramer, Wolfgang Nitschke and Jason W. Cooley
green sulfur bacteria and prior to the radiation of the firmicutes/heliobacteria/cyanobacteria cluster. Rather than being a mere structural detail, the emergence of heme ci most likely had far-reaching consequences for the functional mechanism of Qi-site turnover, as will be discussed in Section VIII. It is noteworthy that screening of emerging genomes suggests that two other clades of bacteria may possess a split Cyt b. The draft versions of the genomes from the acidobacterium Solibacter usitatus as well as the planctomycetes Candidatus Kuenenia stuttgartiensis and Blastopirellula marina contain operons coding for RB complexes, which feature split Cyt b subunits. Our preliminary phylogenetic analysis shows that these three cases of RB complexes indeed cluster with the firmicutes/heliobacteria/cyanobacteria branches. Interestingly, the cysteine residue ligating heme ci in firmicutes, heliobacteria and cyanobacteria (including chloroplasts) is conserved in the RB complexes from the two planctomycetes but not in that from the acidobacterium. Since the respective genome sequences are in a preliminary state, no truly reliable phylogenies can be constructed at the present time. We, however, tend to predict that in the near future, acidobacteria and planctomycetes will become promising targets for both evolutionary and functional studies of RB complexes. In addition to this conspicuous modification in the cyanobacterial and firmicutes Cyt b protein, several ‘minor’ structural changes are suggested by the available sequences. Green sulfur bacterial, firmicutes and cyanobacterial Cyt b proteins all lack the eighth (C-terminal) transmembrane helix. Since all these species are on a common branch of the tree, it appears that loss of this helix must have happened somewhere down the line from the stem of the tree to the node common to these three phylogenetic groups. Many archaeal Cyt b proteins, by contrast, have several transmembrane helices added to their C-terminal end, the functional or structural role of which is not yet understood. This is also true for the apparent anchoring of the actinobacterial ISP protein by three or four (rather than only one N-terminal, as in many other cases) membrane spanning stretches (Fig. 3). These additional helices of the actinobacterial ISP protein possibly correspond to the three extra helices provided by the small non-redox protein encoded by the gene fbcX observed in members of the Deinococcus/Thermus group (Mooser et al., 2005; Mooser et al., 2006). The position of fbcX right upstream of the ISP suggests the possibility that in actinobacteria, this
open reading frame became fused to the N-terminal end of the ISP genes. V. The ‘Third’ Redox Subunit is a Phylogenetic Marker A closer inspection of the genes present in RB operons from all examined species shows that, in addition to the possible presence of non-redox subunits, the nature of the protein analogous to Cyt c1 and Cyt f strongly varies between phyla. Mono-, di- and tetraheme c-type as well as b-type Cyts from different structural families are observed. Almost every phylum contains its proper Cyt c1 or f equivalent giving this subunit the status of a phylogenetic marker. Figure 3 schematically shows the types of Cyts present in various phyla defined by the tree. Cyt c1 only exists in α-, β- and γ-proteobacteria. The ε-proteobacteria contain a diheme Cyt c belonging to the Cyt c4 family (Abergel et al., 2003), whereas members of the δ-subgroup apparently have recruited their ‘favorite’ type of heme protein, i.e., tetrahemic Cyt c3, into the RB complexes. Cyt c3 is a major component of many δ-proteobacterial species (amounting to several percent of the total cell mass in sulfate reducers) and a few dozen genes encoding c3-related Cyts have been found in the respective genomes. In contrast to many previous phylogenies of the proteobacterial phylum, the Cyt b-based tree places the δ-subgroup at positions corresponding to the earliest branching cluster, indicating a common node for ε–proteobacteria and the subsequently diverging γ-, β- and α-subgroups. Similar topologies have been reported in more recent 16S rRNA based studies (von Mering et al., 2007). In line with this phylogeny, Cyt c1 appears to be derived from the ε-proteobacterial di-heme Cyt (Baymann et al., 2004). The δ-proteobacterial tetraheme Cyt c3, by contrast, is structurally unrelated to Cyt c1. Cyanobacteria and plastids contain Cyt f, the structure of which (Martinez et al., 1994) is again completely unrelated to Cyt c1, Cyt c3 or Cyt c4. Low GC Gram-positive bacteria with a low guanine/cytosine content in their DNA feature a monoheme whose sequence indicates it belongs to the class I c-type Cyts. This c-type Cyt is fused to the C-terminal end of subunit IV. In heliobacteria, a diheme c-type Cyt plays the role of the terminal electron acceptor in the high potential chain of the enzyme. The actinobacteria (formerly called the high GC Gram-positive bacteria)
Chapter 23
Function and Evolution of Rieske/Cytochrome b Complexes
again contain a RB complex featuring a diheme Cyt as the third redox subunit of the complex. In contrast to all cases mentioned so far, the gene for this protein is located upstream of those corresponding to the ISP/Cyt b tandem (Fig. 3). As the primary sequences of these diheme proteins do not allow their classification unambiguously into the Cyt c4 group, it cannot be excluded that the actinobacterial third redox subunit represents yet another structural motif in the heterogeneous group of Cyt c1/f analogs. This may also be true for the monoheme Cyts present in the Deinococcus/Thermus group (Mooser et al., 2005; Mooser et al., 2006), whose sequences bear very little resemblance to a typical (class I) Cyt c and may contain very surprising structural elements such as a transmembrane helix predicted to be situated roughly in the centre of the sequence. RB complexes in Archaea are not well characterized by biochemical and biophysical methods. The only case where some data are available is that of the enzyme most probably containing the so-called Cyt b558/566 as third redox subunit (Hiller et al., 2003). In contrast to all other Cyt c1/f analogs discussed above, this monoheme protein contains a b-type heme (Hettmann et al., 1998). Its structure is predicted to be composed of mainly β-sheet elements as another striking difference to the other members of the group except Cyt f. The gene for Cyt b558/566 is located upstream of those corresponding to the ISP/Cyt b tandem, just as in actinobacteria and in the Deinococcus/Thermus group. VI. The Rieske/Cytochrome b Complex, a Primordial Enzyme The deep branching between archaeal and bacterial subtrees indicates that the origin of the RB complexes pre-dates the Archaea/Bacteria divergence more than 3 billion years ago. This conclusion is corroborated by phylogenetic analyses of the ISP subunit. The ISP is a basic building block of several distinct enzymes and also found, for example, in arsenite oxidases or dioxygenases. The composite phylogeny of RB complexes and arsenite oxidases shows well-separated subtrees for the two enzymes with a pronounced Archaea/Bacteria divergence within each subtree (Lebrun et al., 2006). This suggests that ISP proteins have been integrated into these two enzymes within the last universal common ancestor (LUCA) of Bacteria and Archaea, thus well before
459
the divergence of these major domains of the tree of life. Similar ‘pre-LUCA’ origins have been observed for several enzymes the substrates of which were probably abundant in habitats of the early Earth, such as hydrogenases (Vignais et al., 2001; BrugnaGuiral et al., 2003) or arsenite oxidase (Lebrun et al., 2003). In contrast, many enzymes working with substrates that probably appeared only later, i.e., due to the increase of oxidation state of the environment, do not show Archaea/Bacteria splits and often show signs of extensive lateral gene transfer. Prominent examples are provided by the enzymes involved in sulfate (Stahl et al., 2002) or arsenate reduction (S. Duval, W. Nitschke and B. Schoepp-Cothenet, unpublished) (see also Chapter 27, Zannoni et al.). Sulfate and arsenate certainly were virtually absent under the environmental conditions prevailing in the archaean era of our planet. RB complexes play a pivotal role in aerobic and anaerobic respiration, which use oxidized substrates with fairly high electrochemical potentials as terminal electron acceptors. Since these substrates are unlikely to have been abundantly present in their oxidized form, the early (pre-LUCA) origin of this enzyme is puzzling, and the chemical nature of the terminal electron sink in such an ancestral respiratory chain remains an open question. Similarly ‘crucial’ enzymes, such as for example [NiFe] hydrogenases, frequently seem to have been duplicated prior to the Archaea/ Bacteria cleavage (Vignais et al., 2001; Brugna-Guiral et al., 2003). This does not seem to have happened to the RB complexes. Only one type of enzyme can be discerned within the Bacteria, and the only case of a bacterium (Acidithiobacillus ferrooxidans) possessing two distinct Cyt bc1 complexes turned out to be the result of a relatively late operon duplication within the Acidithiobacillus lineage (Fig. 3). By contrast, the enzyme was apparently duplicated early on in the evolutionary history of the archaeal kingdom (prior to the Euryarchaota/Crenarchaota split) giving rise to two distinct classes of archaeal RB complexes, i.e., the F/G and the L/N-groups. The names of these two groups are derived from the gene labels SoxF/SoxL and SoxG/SoxN, denoting the Rieske protein and cytochrome b respectively, in each of the two gene clusters in the Archaeon Sulfolobus acidocaldarius. These two operons were vertically inherited into many extant Archaea, and only a few species lost one or the other of the two enzymes (Fig. 3). Whereas for Acidithiobacillus, function of the individual Cyt bc1 complexes in distinct electron transfer directions was
460
David M. Kramer, Wolfgang Nitschke and Jason W. Cooley
proposed (Brasseur et al., 2002), the significance of the coexisting F/G- and L/N-type RB enzymes in many Archaea is enigmatic. VII. The Molecular Mechanism of the Qo Site: Avoiding Q-Cycle Short Circuits The bifurcated oxidation of QH2 at the Qo site is remarkable, perhaps even unique, in biology, and possibly one of the oldest extant reactions in biology (see section VI); it is certainly one of the most debated, with numerous, radically different models, which have resulted partly from of a lack of information about any of the intermediates in the process. The fundamental biochemical problem faced by the Qo site is in itself profound. Free in solution, the oxidation of QH2 by a one-electron oxidant would result in reactive free radical intermediates (SQ species), which could lead to loss of energy or damage of cellular components, if uncontrolled. The Qo site not only engineers the taming of these intermediates, but it does so in a way that it stores additional energy for the cell. In general, the bifurcated reaction competes with several thermodynamically favored ‘bypass’ or ‘short-circuit’ reactions (Kramer and Crofts, 1993; Muller et al., 2002; Kramer et al., 2003; Sun and Trumpower, 2003; Osyczka et al., 2004; Cape et al., 2006), some of which lead to deleterious superoxide production (Naqui et al., 1986; Muller, 2000; Muller et al., 2002; Gong et al., 2004). To understand catalysis at the Qo site, several research groups have begun to focus on what the RB complexes do not do during catalysis, i.e., how these side reactions are avoided (Cape et al., 2006). We discuss here four of these bypass reactions that are particularly important for a mechanistic understanding because they differ in the fate of the SQ formed within the Qo site, the key intermediate in the entire Q-cycle. These reactions are depicted in Fig. 4 with key steps numbered as described in the following. The first bypass reaction (Fig. 4A) occurs when [2Fe-2S] oxidizes both QH2 electrons, upon two sequential one-electron transfers to [2Fe-2S]. This requires the oxidation of bound QH2 by the oxidized [2Fe-2S] (step 1), leading to formation of SQ. Oxidation of the second electron by the high potential chain requires the pivoting of the ISP head domain to the ‘c’ position (step 2), transfer of an electron to C (step 3), return of the ISP head domain to the ‘b’ position (step 4) followed by transfer of an electron from SQ
to [2Fe-2S] (step 5). The electron on [2Fe-2S] can then be transferred to C (steps 6 and 7). Two protons are released into the aqueous phase on the p-side of the membrane for each two electrons transferred to C. The second bypass reaction (Fig. 4B) involves the production of superoxide. Bound QH2 is oxidized by [2Fe-2S] (step 1), forming SQ, which reduces O2 (step 2) releasing Q (step 3). The electron which was passed to [2Fe-2S] is eventually transferred in the usual way to C. In the third bypass reaction (Fig. 4C), the SQ, once formed (step 1), can also escape from the Qo pocket (step 2) allowing it to react with other redox carriers, in particular other SQ species, leading to disproportionation (formation of Q and QH2 from two SQ, step 3). The fourth bypass reaction occurs when SQ is formed within Qo with the Cyt b chain partly or fully reduced (step 1). Since SQs are both good oxidants (forming QH2) and reductants (forming Q), SQ can oxidize reduced Cyt bL (step 2), producing QH2 (step 3). The bypass routes are far more favored by thermodynamics than the concerted Q-cycle. For example, in the Cyt b6 f complex the complete oxidation of plastoquinol by two Cyt f hemes is about 104 ~ 105 times more favorable than the bifurcated reaction, leading to reduction of one Cyt f and one Cyt bH heme (Kramer and Crofts, 1993). An open question is, how does the Qo site prevent high rates of superoxide formation under conditions where they would be most favored? Models in the literature to explain the relative lack of bypass reactions can be categorized into several classes (Cape et al., 2006), as briefly reviewed in the following section. A. Kinetic Models Recently, Cape et al. (2007) reported that an O2sensitive SQ is generated by the Qo site. This species is currently the best candidate for the intermediates involved in all of the above bypass reactions, especially the immediate reductant for O2. However, it remains unclear whether this species is involved in the Q-cycle, or is only accumulated under conditions where the enzyme is inhibited. In any case, an obvious way to prevent superoxide formation would be to ensure that this SQ, or its precursor, is rapidly oxidized by Cyt b, so that it never accumulates to a large extent. In principle, SQ accumulation can be avoided by maintaining the flow of electrons through the entire Q-cycle, i.e., by keeping the complexes in active states. However, this is only possible under ideal
Chapter 23
Function and Evolution of Rieske/Cytochrome b Complexes
461
Fig. 4. Side reactions that bypass the Q-cycle. Reactions shown are those that differ from the Q-cycle in the fate of the SQ generated at the Qo site. Details are given in the text. See also Color Plate 12, Fig. 19.
conditions (Muller, 2000; Gong et al., 2004; Osyczka et al., 2004), and in reality, under in vivo conditions, the RB complexes are likely to be at least partially inhibited, e.g., whenever reduced Cyt b accumulates, and prone to superoxide production. Involvement of some type of kinetic competition is supported by a large body of work, which shows that the yield of the Q-cycle is very high under uninhibited conditions. However, any process which prevents the normal electron transfer from the Qo site to the Cyt b chain appears to accelerate superoxide production, most likely by allowing the buildup of Qo site SQ intermediates (Muller et al., 2002, 2003; Kramer
et al., 2003; Sun and Trumpower, 2003), which are highly reactive in solution and which readily reduce O2 to superoxide (Afanas’ev, 2004). For example, the buildup of ‘backpressure’ from high pmf can hinder electron transfer from Cyt bL to Cyt bH, and has been shown to induce superoxide production at the Cyt bc1 complex in mitochondria as may occur under stress or disease conditions when respiration outstrips cellular ATP consumption (Guidot et al., 1993; Lebovitz et al., 1996; Zhang et al., 1998; Longo et al., 1999; Muller, 2000). Similar effects are also seen upon addition of specific inhibitors (reviewed in Muller, 2000). Blocking the Qi site with AA prevents the re-
462
David M. Kramer, Wolfgang Nitschke and Jason W. Cooley
oxidation of Cyt b hemes, resulting in the buildup of reduced Cyt b hemes, preventing further oxidation of SQ and subsequent superoxide production (reviewed in Kramer et al., 2003). Similarly, the ‘proximal niche’ (or type II) inhibitors prevent Cyt b reduction, but still allow QH2 to bind at the ‘distal niche’ of the Qo site, where it can be partially oxidized to SQ by [2Fe-2S] (Muller et al., 2003). There is good evidence that the ‘processing’ of SQ after it is formed is important for avoiding bypass reactions. Such processing might involve specific SQ binding or movements as well as proton transfer reactions. However, exploring these interactions at a molecular level has been made difficult by the lack of information about where the substrate, product and intermediates bind within the Qo site. One possibility is that QH2 bound in the same location as stigmatellin, a Q analog that strongly H-bonds between H161 of the ISP (one of the ligands to its [2Fe-2S] cluster) and Glu295/272 (Rba. sphaeroides/yeast numbering) on the Cyt b protein (Crofts et al., 1999) Functionally, this position makes sense if the first electron transfer to the oxidized [2Fe-2S] was coupled to or gated by proton transfer to His161 while the remaining proton could be transferred to another residue, perhaps Glu295/272 (Rba. sphaeroides/yeast numbering) or water molecule in the site. The remaining SQ anion would be at some distance from the Cyt bL heme, and secondary SQ movements within the Qo pocket could gate electron transfer. This model could also explain other known features of catalysis. For example, an apparent lack of QH2 binding to the Qo would be expected, as effective H-bonding to the site can only occur after the [2Fe-2S] is oxidized, allowing deprotonation of H161 under physiological conditions. The stigmatellin binding position of QH2 also reasonably explains the partial inhibition of catalysis by type II Qo site inhibitors that bind towards the Cyt bL heme (Muller et al., 2003). Type II inhibitors prevent full Q-cycle turnover but still allow QH2 binding. Subsequent partial oxidation leads to essentially 100% yield of superoxide production (Muller et al., 2003). The smaller inhibitors, like MOA-stilbene, do not significantly interfere with QH2 binding, whereas larger ones, such as myxothiazol, increase the apparent KM substantially. Modeling of the Qo site suggests that UQH2 can bind in the stigmatellin site without steric clashes with MOA-stilbene, but not myxothiazol. Thus it may not be possible to characterize inhibitors as strictly type I or II. On the other hand, the effects of mutation of
Glu295/272 on Qo site catalysis have been interpreted as major (Crofts et al., 1999; Wenz et al., 2007), or relatively minor (Osyczka et al., 2006). More importantly, however, Wenz et al. (2007) and Crofts et al. (2006) showed that residual QH2 oxidation activity of Glu295/272 mutants resulted in substantial yields of superoxide rather than Q-cycle. Intriguingly, mutations at this residue show only a small effect on the KM for UQH2 at the Qo site (Crofts et al., 1999; Wenz et al., 2007), suggesting that Glu272 might not (after all) be involved in the initial binding of substrate to Qo. Wenz et al. (2007) further showed that mutation of other Cyt b residues within the Qo site, notably F129 and Y132 and Y279 (yeast numbering), also inhibit the Q-cycle while increasing the yield of superoxide production. These mutants exhibit altered KM values for UQH2, as well as Qo site inhibitor resistance, while F129K also lowered the redox mid-point potential of Cyt bL. In our view, these mutations (and perhaps many others yet to be fully characterized) likely induce behavior similar to that of Qo site proximal inhibitors, preventing the processing rather than the formation of SQ, consistent with a competitive kinetic view of superoxide production (Muller et al., 2002; Kramer et al., 2003; Muller et al., 2003). It seems likely that these specific residues are involved in SQ processing at the levels of SQ motion or proton extraction (as in Crofts et al., 2006), or perhaps indirectly by changing the coordination of water molecules within the pocket (as in Osyczka et al., 2006; Wenz et al., 2007). B. Thermodynamic Control of Semiquinone Intermediates It is possible to limit side reactions by controlling the thermodynamic properties of the reactive intermediates of QH2 oxidation. In many ways, these mechanisms are the simpler to conceive, as similar mechanisms are known to operate in other systems, e.g., the highly stabilized SQ species in the photochemical reaction centers, nitrate reductase, complex II etc., and require few additional assumptions. Models invoking thermodynamic control of SQ are not mutually exclusive with most other models, like those invoking gating or kinetics. The general concept for this type of control is illustrated in Fig. 5, as discussed in Cape et al. (2006; 2007), highlighting factors that should govern the rate of superoxide production (vSOP). Assuming that the steps leading up to the formation of SQ in Fig. 5 are
Chapter 23
Function and Evolution of Rieske/Cytochrome b Complexes
463
Fig. 5. Kinetic model for competition between the Q-cycle and superoxide production.
in rapid equilibrium, vSOP will be determined by three factors: i) how much SQ is present in the steady-state, related to KSQ, the equilibrium constant for oxidation of QH2 to SQ and in turn to the difference in redox potentials of the SQ/QH2 couple (Em(SQ/QH2) and the [2Fe-2S] cluster (Em([2Fe-2S]); ii) the rate constant for reduction of O2 by SQ, kSOP, which is a function of SQ redox properties; iii) the rate of O2 diffusion into and/or the rate of escape of SQ from the Qo site. In general, the more stable the SQ the higher will be its concentration in the steady-state, but the less reactive it will be to O2. Data compiled from Afanas’ev (1989) for organic free radicals in aqueous solution shows that the rate constant for superoxide production (kSOP) increases as the SQ becomes more reducing, until it becomes diffusion-limited (to about 108 M–1s–1 at about –400 mV). The situation in the Qo site should be qualitatively similar, but the maximal (diffusion limited) rate will depend on the effective rate of collision of SQ with O2, and will likely be slower than in water, and thus reaching a diffusion limit at a less reducing potential. The interplay of kSOP and KSQ predicts an interesting dependence of superoxide production rate on the redox properties of Q/SQ/QH2 species (Cape et al., 2006). As SQ becomes more stable, its concentration will increase but its reactivity to O2 will decrease. On the other hand, as SQ becomes less stable, its reactivity to O2 will increase but its steady state concentration will decrease. The vSOP, being the product of [SQ], [O2] and kSOP, will reach a maximum at moderate SQ stability constant, falling off steeply as SQ is either stabilized or destabilized. It follows that the Qo site could limit superoxide
production by thermodynamic control of the redox properties of bound SQ/QH2 in two distinct ways. The Qo site can stabilize the SQ by specific binding interactions (as in models proposed by Link, 1997; Berry and Huang, 2003), rendering it less reactive to O2 or to other bypass routes. This type of mechanism is apparent in the QB site of type II reaction centers (Rinyu et al., 2004; Zhu and Gunner, 2005) or within Complexes I and II (Ohnishi et al., 2005; Rothery et al., 2005; Yap et al., 2006) where SQ intermediates are stable on a time scale of minutes, and relatively slow to react with O2. In stabilized SQ models for the Qo site, the overall bifurcated reaction must be pulled forward via the oxidation of SQ by Cyt bL, predicting the accumulation in the steady-state of significant amounts of SQ, which is clearly not observed either directly (Cape et al., 2007), or via their expected effects on Cyt b and c reduction kinetics (Crofts et al., 2003). Alternatively, the Qo site can de-stabilize (or simply not stabilize given its inherent instability) the SQ. As long as the effective rate constant for SQ reduction of O2 is lower than that for its reduction of Cyt bL, destabilization of SQ will result in decreased superoxide production. A very unstable SQ intermediate is implied in several Q-cycle models (Crofts and Wang, 1989; Crofts, 2004), and an unstable Qo site SQ is strongly supported by a number of experimental results. The rate-limiting step (i.e., that probed by Arrhenius plots of the reaction) in the Qo site reaction involves the transfer of an electron from QH2 to the [2Fe-2S] cluster (Snyder et al., 1999; Guergova-Kuras et al., 2000; Forquer et al., 2006), and is modulated by changing the redox properties
464
David M. Kramer, Wolfgang Nitschke and Jason W. Cooley
of the [2Fe-2S]. By contrast, in models invoking a highly stable SQ intermediate, the oxidation of the SQ by Cyt b should be rate-determining (Forquer et al., 2006). A low stability SQ is consistent with the high Eact for Qo site turnover (Crofts and Wang, 1989; Crofts et al., 2000) as well as the low concentration of SQ detected in recent freeze quench experiments (Cape et al., 2007). It thus seems likely that the Qo site SQ intermediate is highly unstable (perhaps destabilized by the Qo site), and that this instability is a key property which controls the rate of O2 reduction. Indeed, when the redox properties of the substrate QH2 were modulated, so that SQ formation was more favored by increasing the reducing potential of the Q moiety, the complex was transformed from an energy-storing system to one that produced superoxide at rates exceeding normal Q-cycle turnover (Cape et al., 2005). Since the SQ is likely an intermediate in both superoxide production and the Q-cycle (Forquer et al., 2005, 2006), its destabilization represents a compromise between avoidance of deleterious side reactions on one hand, and the overall rate of the Qcycle on the other. C. Gating and Shielding The Qo site may also prevent side reactions by sequestering reactive intermediates from potential oxidants or reductants. Essentially, gating would allow the Qo site to act as a ‘reaction chamber’ shielding the dangerous intermediates from potential reaction partners. This type of shielding would be more important with more unstable (and thus more reactive) SQ intermediates, and it would be critical to prevent O2 leakage into the site (leading to superoxide production), as well as SQ escape from the site (leading to disproportionation, Fig. 4C). Unstable SQ species are, by thermodynamic principle, less tightly bound to the Qo site than stable ones (Crofts and Wang, 1989), but the Qo site can prevent SQ release by ‘kinetic trapping’, where the SQ is only loosely bound within the site, but trapped by a protein barrier for a significant time. D. Inter-monomer Electron Transfer The architecture of the RB complexes strongly suggests that they operate as functional as well as structural dimers. The ISP hydrophobic anchors extend from the Qo pocket of one monomer to Cyt b protein of the opposite monomer, so that monomerization of
the complex would render it inactive (Huang et al., 1994; Yu et al., 1998; Zhang, 1998; Gong et al., 2004; Covian and Trumpower, 2005). In addition, the Cyt b hemes are in close proximity with their counterparts found in the opposite monomer, providing an easy route for electron transfer (Crofts and Berry, 1998; Osyczka et al., 2004, 2005; Covian and Trumpower, 2005; Gong et al., 2005), as illustrated in Fig. 1. Accordingly, several groups have suggested models where such interactions may be involved in the function of the enzyme during its normal turnover (Snyder et al., 2000; Covian et al., 2004; Osyczka et al., 2004; Cooley et al., 2005; Covian and Trumpower, 2005; Gong et al., 2005; Klishin and Mulkidjanian, 2005; Mulkidjanian, 2005; Gong et al., 2006). These interactions fall roughly into two categories, one involving straightforward inter-monomer electron transfer, and the other involving more elaborate conformation changes or substrate exchange that gate various processes. It has been suggested that inter-dimer electron transfer could serve as a ‘release valve’ to limit buildup of reduced Cyt b, and thus limiting bypass reactions (Yu et al., 1998; Osyczka et al., 2004). Gong et al. (2005) found that mutating a phenylalanine residue lying between the two bL hemes at the monomer: monomer interface had only a moderate effect on the kinetics of Rba. sphaeroides Cyt bc1 complex turnover, but a large effect on the yield of superoxide produced by the complex, leading to the proposal that electron sharing between the cofactors of the two monomers is important for preventing side reactions. However, as no changes in inter-monomer electron transfer rates were measured in this work, effects on other processes in the enzyme are not ruled out. More recently, Shinkarev et al. (2007) demonstrated inter-monomer electron transfer from the two Cyt bL hemes, but found that the normal Cyt bL to bH were considerably faster so that intermonomer electron flow should only occur to a great extent when Cyt bH in a monomer is pre-reduced before QH2 oxidation. Covian et al. (2005) probed the effects of substoichiometric concentrations of the Qi site inhibitor AA on the kinetics of Cyt b redox reactions, suggesting that electrons derived from one Qi site QH2 molecule can reach the heme bL of the alternate monomer adjacent to the inhibitor. A concern with these measurements is that they are carried under pre-steady state conditions, and the sub-stoichiometric amount of the inhibitor is free to equilibrate with each of the two monomers Qi sites in
Chapter 23
Function and Evolution of Rieske/Cytochrome b Complexes
the time-scale of the measurement. The authors have attempted to address this issue by mathematically fitting the electron transfer phenomena in such a way that the rates appear to be faster than what would be expected from simple binding equilibria and diffusion of the inhibitor molecule (Covian et al. 2006). However, Crofts and coworkers (Crofts et al., 2003; Crofts, 2005) argue that these data can be equally explained within the context of a simple Q-cycle if one takes into account the random distribution of inhibitors and other kinetic factors. E. Interactions Between Quinone Binding Sites A number of models have been proposed where one monomer is prevented from turning over until the opposing monomer has completed a part of its reaction cycle. Inspiration for these models comes from unexpected effects of inhibitors. Over 30 years ago, Berden (1972) presented evidence for differential binding of the first or second AA molecules at the Qi site. Trumpower and coworkers (Gutierrez-Cirlos and Trumpower, 2002) found evidence that sub-stoichiometric amounts of AA per Cyt bc1 dimer were sufficient to fully arrest yeast enzyme activity, leading these authors to propose that only half of the sites of the Cyt bc1 are active at any one time (Trumpower, 2002; Trumpower et al., 2004). The Trumpower group has also proposed that AA binding at the Qi site influences the binding of stigmatellin to the opposite monomer’s Qo site (Covian et al., 2004). In a similar vein, Valkova-Valchanova et al. (2000) found that when AA is bound at the Qi site of the purple bacterial Cyt bc1 complex, the ISP tether (or hinge) region at the opposite side of the membrane is more susceptible to proteolysis by exogenously added protease. More recently, Cooley et al. (2005) used CW-EPR of ordered membranes to show that mutation of the Cyt b protein at the Qi site and binding of Qi site inhibitors each alter the environment of the [2Fe-2S] cluster. These effects have been interpreted as reflecting functionally important interactions between the Qi and Qo sites that are transmitted via specific transmembrane helices of the Cyt b protein. The observed inter-monomer effects have suggested models in which the dimer of the Cyt bc1 complex operates in some alternating fashion, with one Qo site operating at a time in a manner controlled by the status of the other monomer’s quinone/quinol
465
binding or redox state (Covian et al., 2004; Cooley et al., 2005). These interactions could function as a valve mechanism, to prevent introduction of substrate or primary oxidant until the enzyme is in an optimal state for catalysis, thus preventing the buildup of reactive intermediates. Such a mechanism could be analogous to the valve system in an internal combustion engine that times the introduction of fuel, or the release of exhaust, preventing uncontrolled explosions. However, the structural bases and consequent energetic costs of any of these effects have yet to be sorted out. Furthermore, evidence at the atomic scale has not yet been correlative of these spectroscopic observations, as available crystal structures do not show consistent long-range conformational changes that could transmit such information between the various Q binding sites. However, it should be noted that the published, well-resolved atomic level structures have been prepared in the presence of Qo site inhibitors, rendering the ISP less dynamic to yield better diffraction data. One of these Qo site inhibitors, stigmatellin, was found to inhibit the structural interactions between the Qo and Qi sites observed by CW-EPR (Cooley et al., 2005), possibly masking inter-monomer or intersite structural interactions. F. The Rieske Pivot as a Gating Mechanism The striking >16 Å displacement of the soluble ‘head’ domain of the ISP was first observed in crystal structures of the mitochondrial Cyt bc1 complex with different Qo site occupants (Xiao et al., 2000), and since then demonstrated to be an essential catalytic feature of a range of RB complexes (Zhang, 1998; Brugna et al., 2000; Darrouzet et al., 2000; Roberts et al., 2001). The pivoting is necessary for catalysis because with the ISP in the ‘b’ position, the [2Fe-2S] cluster is too far (~ 31 Å) from its oxidant (heme c/f) to allow rapid electron transfer. Likewise, when the ISP is in the ‘c’ position, the [2Fe-2S] cluster is too distant from the Qo pocket to electronically interact with the substrate, QH2. The ISP pivot has been an attractive means by which to explain how the second and seemingly energetically uphill electron transfer may be facilitated, i.e., the low potential chain cannot complete the oxidation of the QH2 as long as the [2Fe-2S] cluster is prevented from swinging to the ‘c’ position (Zhang, 1998; Brugna et al., 2000; Darrouzet et al., 2000; Roberts et al., 2001). Darrouzet and coworkers (Darrouzet et al., 2000; Darrouzet et al., 2001) showed that the mobility of the ISP head
466
David M. Kramer, Wolfgang Nitschke and Jason W. Cooley
domain is not required for the initial oxidation of the QH2 by either the high or low potential chains, achieving both reduction of the [2Fe-2S] cluster and Cyt b, but ISP pivoting is required for the re-oxidation of [2Fe-2S] to complete with Q-cycle. It remains unclear how such a gate might function as the pivot motion appears to be more rapid than overall enzymatic turnover under many conditions (Darrouzet et al., 2000). On the other hand, interactions between ISP conformation and the occupancy of the Qo site by inhibitors and the redox state of [2Fe2S] cluster have been found. The redox mid potential (Em) of the [2Fe-2S] center depends dramatically on the occupant at the Qo site, ranging from 180 mV in the presence of DBMIB (Malkin, 1981), to ~ 300 mV with native Q present in the Cyt b6 f complex. Whereas in the Cyt bc1 Em values of the [2Fe-2S] cluster range from over 500 mV with stigmatellin bound tightly at the Qo site (Ohnishi et al., 1988) to 300–320 mV with native Q and are further decreased by 30 or 50 mV when myxothiazol is bound or the site is emptied of Q, respectively, prior to titration of the [2Fe-2S] cluster. Brugna et al. (2000) found that the positions of the ISP head group depended upon the redox state of the [2Fe-2S] cluster, while Daldal and coworkers (Darrouzet et al., 2002; Cooley et al., 2004) have shown that the Em of the [2Fe-2S] cluster was dramatically increased when the head domain was limited to positions nearer the b position by mutation of the tether domain (Cooley et al., 2004). The occupants of the Qo site can thus affect the redox properties of the [2Fe-2S], either by direct interaction via H-bonds (Shinkarev et al., 2002), or by altering the conformation of the ISP head domain via its interactions with the Cyt b surface (Darrouzet et al., 2002; Cooley et al., 2004). How interactions between substrate occupancy, ISP position and Em changes might affect gating at the Qo site remains untested. G. Coulombic Gating Mechanisms Gating mechanisms have been proposed that control the fate of SQ by electrostatic or Coulombic interactions. An example of this type of model was presented by Crofts (Crofts et al., 2006), positing that the mobility of SQ in the Qo site is affected by the redox state of the Cyt bL heme. In its oxidized (Fe3+) state SQ can migrate towards Cyt bL, facilitating electron transfer. In its reduced state, however, electrostatic repulsion prevents SQ from approaching Cyt bL,
thus hindering electron transfer. The overall effect is proposed to be the gating off of electron transfer when the Cyt b chain is reduced, preventing SQ from oxidizing the Cyt b chain. H. ‘Proton Stripping’ The oxidation of Cyt bL by the neutral form of SQ, forming QH–, is thermodynamically favored, but that by the anionic form (Q• –), forming the double anion (Q2–) is not. Kramer et al. (2005) thus proposed that Cyt b Glu272 (yeast numbering) acts to ‘strip’ off the SQ proton, leaving exclusively the anionic form, thus preventing the oxidation of reduced Cyt b by SQ (as depicted in Fig. 4D). The anionic character of the SQ will also act to prevent its release through the hydrophobic entrance to the Qo site, thus preventing the disproportionation reactions depicted in Fig. 4C (Kramer et al., 2005; Shinkarev, 2006). The proton stripping model is in line with the anion form of the SQ trapped in recent freeze quench experiments (Cape et al., 2007), but it remains unclear what residues are essential for the deprotonation to occur. Also critical for this model is the absence of proton donors to SQ when it is close to Cyt bL, as these would allow proton coupled electron transfer and the formation of QH– or QH2. In this regard, the existing high-resolution structures of Qo do not (yet) show water molecules in the Qo pocket (E. Berry, personal communication). I. Complete Avoidance of Semiquinone Intermediates A more exotic model was recently proposed by Osyczka and coworkers (2004) wherein QH2 at the Qo site is oxidized by a ‘simultaneous’ two-electron transfer, without the appearance of a SQ intermediate. The attractiveness of this model is that it would prevent O2 reduction, or other short-circuiting bypass reactions, by eliminating the reactive intermediate. A ‘truly concerted’ electron transfer reaction of this type is theoretically possible (Zusman and Beratan, 1996), but to operate with high yield, it must compete with the ‘normal’ sequential reaction by forming a lower energy activated intermediate. These processes must thus have very different reaction coordinates, activation energies and reorganizational energies. Recent work by Forquer et al. (2005, 2006) tested this prediction by comparing the rates and activation energies for normal Q-cycle turnover and superoxide
Chapter 23
Function and Evolution of Rieske/Cytochrome b Complexes
467
production (which must proceed by formation of a SQ via a sequential electron transfer mechanism) in wild type and a series of mutant Cyt bc1 complexes with altered driving forces for QH2 oxidation (by changing the Em of the [2Fe-2S] cluster). The results indicated that the uninhibited Q-cycle and superoxide production involve very similar transition intermediates, and dependencies on driving force, in conflict with a truly concerted electron transfer. The results are in support of models where both Q-cycle and superoxide production share the early steps of the reaction, splitting after the formation of activated intermediate, which involves the formation of the Qo site SQ. VIII. The Quinone Reduction Site, Qi of the Cytochrome bc1 Complexes The reduction of Q at the Qi site involves two, temporally-separated n = 1 redox events, which require the site to stabilize the SQ intermediate. A stable SQ anion, observed by EPR and studied in some detail (Ding et al., 1992; Gray et al., 1994; Kolling et al., 2003; Dikanov et al., 2004), is coupled to specific binding of SQ over Q or QH2. This lowers the in situ Em of the Q/SQ couple to about 150 mV (vs. SHE), similar to that of the Q/QH2 couple (~ 100 mV), while raising the energetic barrier for the escape of the electron to potential deleterious acceptors like molecular oxygen. Several specific residues have been suggested to play a role in Qi site function based on mutations conferring resistance to Qi inhibitors, in both purple bacterial systems Rba. sphaeroides and Rba. capsulatus and in Saccharomyces cerevisiae, as concisely reviewed by Brasseur et al. (1996). Recent high-resolution structures place all but a few of these residues outside of the Qi site, and only a select few are reasonable candidates for hydrogen bond donors, as expected for stabilization of SQ. In particular, the E helix adjacent pair of Asp252/Lys251 (Rba. capsulatus numbering) appear from all of the structures to be well positioned to hydrogen bond with one oxygen of the Q species, while a conserved histidine residue located at the base of the D helix His217 is the most probable candidate for hydrogen bonding (Berry et al., 2000; Berry et al., 2002), either directly or through a water molecule, to the Q-borne oxygen (Fig. 6). Both of these atomic structure based observations are borne out by experimental data in that mutations located at both the Asp/Lys pair and
Fig. 6. Architecture of the quinone reduction (Qi) site of the Cyt b subunit. A cartoon depicting a subset of residues thought to be prominent players in the binding of the Q and/or stabilization of the SQ is shown overlapping the antimycin A containing Qi site from the chicken derived Cyt bc1 structure (1PPJ.pdb). The D helix residue His217 (202 in chicken numbering), A helix Ile46, E helix Asp252/Lys251 and the de loop Asn221 residue are depicted with proposed hydrogen bonds to the Q (darker dotted lines), or H20 (lighter dotted lines) and alternate positions of the His and Lys residues (various shaded dashed stick side chains) are shown.
the His confer not just inhibitor resistance, but they also significantly effect the ability of the enzyme to reduce/bind Q or even stabilize SQ (Gray et al. 1992; Hacker et al. 1993). Other, less well conserved, residues (particularly Asn221 in Rba. sphaeroides) located in the de loop region at the base of the site, have also been postulated to be interacting via hydrogen bonding with the Q molecule residing at the Qi site (Kolling et al., 2003). The burning question that remains is how the SQ is stabilized at this site. A direct question is whether the electron resides on the E helix or the D helix adjacent oxygens of the SQ, or is delocalized throughout the Q. This very question has been addressed using EPR methodologies, where the authors (Kolling et al., 2003; Dikanov et al., 2004) have noted that the unpaired electron interacts with a nitrogen resembling that of imidazole, and hence assumed to be the D helix His 217 adjacent to the Qi site occupant. Lastly, the conformations of both the His217 and Lys251 residues appear to be flexible, leading to proposals of structural gating and proton release mechanisms (Gao et al., 2003; Palsdottir et al., 2003). For example, His217 has been observed to be within hydrogen bonding distance of the Qi site inhibitor
468
David M. Kramer, Wolfgang Nitschke and Jason W. Cooley
antimycin A (Berry et al., 1999), while, when NQNO is present, the imidazole ring is clearly rotated and a crystallographically visible water molecule bridges the residues nitrogen and the oxygen of the inhibitor (Gao et al., 2003). However, it remains unclear whether the observed conformational changes are induced by the inhibitor molecules, or whether they reflect important functional roles. In another set of structures Xia and coworkers have pointed out that the Lys251 (corresponding to Lys278 in bovine numbering) is rotated away from the site when no occupant is found at the site (Berry et al., 2000, 2002; Gao et al., 2003; Palsdottir et al., 2003). From this observation they hypothesized that this lysine may act to shuttle protons into the site upon the reduction events. IX. The Quinone Reduction Site,Qi of the Cytochrome b6 f and Related Complexes Even before the availability of the high-resolution structures, there were strong indications that the Qi site of the Cyt b6 f complexes is distinct from that of the Cyt bc1 complexes. In the latter case, there has been no evidence for a stabilized Qi site SQ, while electrons passed to the low potential chain remained on Cyt b for extended periods at high potential, despite the availability of Q as an oxidant (Joliot and Joliot, 1984; Rich et al., 1987; Kramer and Crofts, 1992; Kramer and Crofts, 1993). In addition, the inhibitor sensitivity of the Cyt b6 f complex Qi site is very different from that of the Cyt bc1 complex. For instance, antimycin A does not appear to bind at the Qi site of Cyt b6 f complexes, whereas it is a potent competitive inhibitor of the Cyt bc1 complexes). Furthermore, the protein structure in the Qi region of the Cyt b6 f complex is surprisingly distinct from that of the Cyt bc1 complex. (Cramer and Crofts, 1982; Cramer et al., 1997; Soriano et al., 1999; Smith et al., 2004). The high-resolution structures for the Chlamydomonas reinhardtii (Stroebel et al., 2003) and Mastigocladus laminosus (Kurisu et al., 2003) demonstrated clearly why these differences existed by revealing a previously unsuspected ‘extra’ heme at the Qi site. This same heme group is very likely present in a range of photosynthetic (e.g., heliobacterial) and non-photosynthetic (bacilli) b6 f-type complexes. The heme is covalently bound to the Cyt b polypeptide by a cysteine residue conserved among firmicutes and cyanobacterial/plastidic RB complexes and has unusual axial ligands, with one oxygen ligand most
likely from H2O or OH– (Zatsman et al., 2006; Baymann et al., 2007). EPR spectroscopy further revealed spin exchange with Cyt bH (Baymann et al., 2007), implying very close interactions between these two redox carriers. Furthermore, dramatic changes in the EPR spectrum of the Cyt ci upon binding of inhibitors like NQNO to the Qi site have been interpreted as reflecting changes in its axial liganding (Zatsman et al., 2006; Baymann et al., 2007), suggesting that the inhibitors, and by extrapolation, the substrate Q, bind to heme ci itself. These results are consistent with some of the unusual redox properties of Cyt ci, which also demonstrate strong interactions with the Qi site ligands (Alric et al., 2005). In our opinion, the most likely function of the Cyt ci is in the reduction of Q at the Qi site. The Cyt bc1 complexes appear to stabilize the SQ intermediate at this site by specific interactions with conserved amino acid residues (i.e., the aforementioned His and Asp residues) that are not well conserved in the Cyt b6 f. It seems likely that the Cyt b6 f type complexes deal with this problem in a different way. Several recent papers (Zhang et al., 2004; Alric et al., 2005; Zatsman et al., 2006; Baymann et al., 2007) have implied that the low potential chain acts as a charge storage device, preventing the buildup of potentially reactive semiquinone at the Qo site. Electron transfer is ‘gated’ by heme ci, so that the Q is reduced by an essentially n = 2 process only when two electrons have accumulated in the low potential chain. This behavior was previously suggested to account for the slow oxidation of Cyt b in intact thylakoids under oxidizing conditions (Kramer and Crofts, 1993) but without any mechanistic basis. The ‘tunable’ redox properties of the Cyt ci could act as a redox gate in this process (Baymann et al., 2007). Alternatively, it has been suggested for chloroplasts and cyanobacteria that the Cyt ci can act as a redox conduit in ‘cyclic’ electron transfer, allowing ferredoxin to reduce the Q pool. Indeed, Cyt ci seems well placed to carry electrons from the n-side aqueous phase to the Qi site cavity. On the other hand, cyclic electron flow cannot explain the function of Cyt ci in non-photosynthetic organisms possessing Cyt ci containing RB complexes. Furthermore, earlier work failed to observe rapid reduction of Cyt bH that would be expected if the Cyt ci was reduced by added ferredoxin (reviewed in Zhang et al., 2004), and at present, the issue appears unresolved.
Chapter 23
Function and Evolution of Rieske/Cytochrome b Complexes
X. The Functional Mechanism of the Rieske/Cytochrome b Complexes is Conserved Judging from what we know about representatives of the various phyla, the basic mode of functioning of the RB complexes family seems to have been conserved through three billion or more years of evolution. The basic biochemical/biophysical problem they resolve is the same: taming the otherwise dangerous oxidation of QH2. They all appear to do this with the benefit of added energy storage via a Q-cycle mechanism. The central reaction of the Q-cycle, the bifurcated oxidation of QH2 at Qo, thus appears to be highly conserved. The conservation includes the chemical nature of the principal oxidants, relatively high and low potential [2Fe-2S] cluster and Cyt bL heme respectively. Furthermore, the domain mobility of the Rieske protein has been observed in all species (bacterial and archaeal) where it was looked for (Brugna et al., 1998, 1999), indicating comparable modes of Qo-site mechanism in all these species. Over evolutionary time scales, the chemical nature (and hence the redox potential) of the substrate QH2 vary substantially over the tree. The archaeal caldariellaquinone (CQ), for example, is structurally quite different from ubi- (UQ) or plastoquinone (PQ) while featuring virtually the same redox potential as the latter molecules. Most menaquinones (MKs), the prevalent chemical type of quinone found on the tree of species, by contrast, are some 150 mV more reducing than ubi-, plasto- or caldariella-quinone (Schütz et al., 2000). Intriguingly, the Em values of the redox cofactors in the enzyme follow quite tightly that of their Q substrate (Kramer et al. 1997; Schütz et al. 2003). Therefore, microscopic driving forces and redox equilibria are astonishingly well conserved for the respective reaction steps of enzyme turnover between the low (MK-based) and high (UQ-, PQ- or CQ-based) potential groups. This conservation argues strongly for mechanisms that require specific driving forces for the oxidation of QH2 and subsequent reactions, such as the destabilized SQ models (Cape et al., 2006; Cape et al., 2007). The further downstream components are from the Qo site, the less well conserved their structural features, as evidenced by the diversity of solutions for the secondary high potential carrier, and the large divergence in function of the Qi site. Here, we are left with two possibilities; either these large differences represent evolutionary drift (where the different
469
versions are equally effective because of the lack of selection pressure), or they represent adaptations to additional specific physiological demands (e.g., accommodating a cyclic electron transfer pathway in the b6 f-type complexes). Acknowledgments The authors would like to thank Drs. Edward Berry, Barbara Schoepp-Cothenet, Jonathan Cape, Atsuko Kanazawa, Michael Bowman and Fevzi Daldal for stimulating discussions. Work in DMK’s laboratory was supported by NIH (2 RO1 GM061904). References Abergel C, Nitschke W, Malarte G, Bruschi M, Claverie JM and Giudici-Orticoni MT (2003) The structure of Acidithiobacillus ferrooxidans c4-cytochrome: A model for complex-induced electron transfer tuning. Structure 11: 547–555 Afanas’ev I (2004) Interplay between superoxide and nitric oxide in aging and diseases. Biogerontology 5: 267–270 Afanas’ev IB (1989) Superoxide ion: Chemistry and Biological Implications. CRC Press, Boca Raton Alric J, Pierre Y, Picot D, Lavergne J and Rappaport F (2005) Spectral and redox characterization of the heme ci of the cytochrome b6 f complex. Proc Natl Acad Sci USA 102: 15860–15865 Avise JC (1986) Mitochondrial DNA and the evolutionary genetics of higher animals. Philos Trans R Soc London B312: 325–342 Baymann F, Giusti F, Picot D and Nitschke W (2007) The ci/bH moiety in the b6 f complex studied by EPR: A pair of strongly interacting hemes. Proc Natl Acad Sci USA 104: 519–524 Baymann F, Lebrun E and Nitschke W (2004) Mitochondrial cytochrome c1 is a collapsed di-heme cytochrome. Proc Natl Acad Sci USA 101: 17737–17740 Berden JA and Slater EC (1972) The allosteric binding of antimycin to cytochrome b in the mitochondrial membrane. Biochim Biophys Acta 256: 199–215 Berry EA and Huang LS (2003) Observations concerning the quinol oxidation site of the cytochrome bc1 complex. FEBS Lett 555: 13–20 Berry EA, Zhang Z, Huang LS and Kim SH (1999) Structures of quinone-binding sites in bc complexes: Functional implications. Biochem Soc Trans 27: 565–572 Berry EA, Guergova-Kuras M, Huang LS and Crofts AR (2000) Structure and function of cytochrome bc complexes. Annu Rev Biochem 69: 1005–1075 Berry EA, Huang LS, Saechao LK, Pon NG, Valkova-Valchanova M and Daldal F (2004) X-ray structure of Rhodobacter capsulatus cytochrome bc1: Comparison with its mitochondrial and chloroplast counterparts. Photosynth Res. 81: 251–275 Brandt U (1996) Bifurcated ubihydroquinone oxidation in the cytochrome bc1 complex by proton-gated charge transfer. FEBS Lett 387: 1–6
470
David M. Kramer, Wolfgang Nitschke and Jason W. Cooley
Brasseur G, Saribas AS and Daldal F (1996) A compilation of mutations located in the cytochrome b subunit of the bacterial and mitochondrial bc1 complex. Biochim Biophys Acta 1275: 61–69 Brasseur G, Bruscella P, Bonnefoy V and Lemesle-Meunier D (2002) The bc1 complex of the iron-grown acidophilic chemolithotrophic bacterium Acidithiobacillus ferrooxidans functions in the reverse but not in the forward direction. Is there a second bc1 complex? Biochim Biophys Acta 1555: 37–43 Brugna-Guiral M, Tron P, Nitschke W, Stetter KO, Burlat B, Guigliarelli B, Bruschi M and Giudici-Orticoni MT (2003) [NiFe] hydrogenases from the hyperthermophilic bacterium Aquifex aeolicus: Properties, function, and phylogenetics. Extremophiles 7: 145–157 Brugna M, Albouy D and Nitschke W (1998) Diversity of cytochrome bc complexes: Example of the Rieske protein in green sulfur bacteria. J Bacteriol 180: 3719–3723 Brugna M, Nitschke W, Asso M, Guigliarelli B, Lemesle-Meunier D and Schmidt C (1999) Redox components of cytochrome bc-type enzymes in acidophilic prokaryotes. II. The Rieske protein of phylogenetically distant acidophilic organisms. J Biol Chem 274: 16766–16772 Brugna M, Rogers S, Schricker A, Montoya G, Kazmeier M, Nitschke W and Sinning I (2000) A spectroscopic method for observing the domain movement of the Rieske iron-sulfur protein. Proc Natl Acad Sci USA 97: 2069–2074 Cape JL, Strahan JR, Lenaeus MJ, Yuknis BA, Le TT, Shepherd JN, Bowman MK and Kramer DM (2005) The respiratory substrate rhodoquinol induces Q-cycle bypass reactions in the yeast cytochrome bc1 complex. J Biol Chem 280: 34654–34660 Cape JL, Bowman MK and Kramer DM (2006) Understanding the cytochrome bc complexes by what they don’t do. The Qcycle at 30. Trends Plant Sci 11: 46–55 Cape JL, Bowman MK and Kramer DM (2007) A semiquinone intermediate generated at the Qo site of the cytochrome bc1 complex. Importance for the Q-cycle and superoxide production. Proc Natl Acad Sci USA 104 7887–7892 Castresana J, Lubben M and Saraste M (1995) New archaebacterial genes coding for redox proteins: Implications for the evolution of aerobic metabolism. J Mol Biol 250: 202–210 Cooley JW, Roberts AG, Bowman MK, Kramer DM and Daldal F (2004) The raised midpoint potential of the [2Fe-2S] cluster of cytochrome bc1 is mediated by both the Qo site occupants and the head domain position of the Fe-S protein subunit. Biochemistry 43: 2217–2227 Cooley JW, Ohnishi T and Daldal F (2005) Binding dynamics at the quinone reduction (Qi) site influence the equilibrium interactions of the iron sulfur protein and hydroquinone oxidation (Qo) site of the cytochrome bc1 complex. Biochemistry 44: 10520–10532 Covian R and Trumpower BL (2005) Rapid electron transfer between monomers when the cytochrome bc1 complex dimer is reduced through center N. J Biol Chem 280: 22732–22740 Covian R, Gutierrez-Cirlos EB and Trumpower BL (2004) Anticooperative oxidation of ubiquinol by the yeast cytochrome bc1 complex. J Biol Chem 279: 15040–15049 Cramer WA and Crofts AR (1982) Electron and proton transport. In: [editors?] Photosynthesis: Energy Conversion By Plants and Bacteria, Vol I, pp 387–466. Academic Press, New York Cramer WA and Zhang H (2006) Consequences of the structure of the cytochrome b6 f complex for its charge transfer pathways.
Biochim Biophys Acta 1757: 339–345 Cramer WA, Soriano GM, Zhang H, Ponamarev MV and Smith JL (1997) The cytochrome b6 f complex. Novel aspects. Physiol Plant 100: 852–862 Crofts AR (2004a) The cytochrome bc1 complex: Function in the context of structure. Annu Rev Physiol 66: 689–733 Crofts AR (2004b) Proton-coupled electron transfer at the Qosite of the bc1 complex controls the rate of ubihydroquinone oxidation. Biochim Biophys Acta 1655: 77–92 Crofts AR (2005) The bc1 complex: What is there left to argue about? In: M Wikström M (ed) Biophysical and Structural Aspects of Bioenergetics, pp 123–155. Royal Society of Chemistry, Cambridge Crofts AR and Berry EA (1998) Structure and function of the cytochrome bc1 complex of mitochondria and photosynthetic bacteria. Curr Op Struct Biol 8: 501–509 Crofts AR, Wang Z (1989) How rapid are the internal reactions of the ubiquinol:cytochrome c2 oxidoreductase? Photosynth. Res. 22: 69–87 Crofts A, Hong S, Ugulava N, Barquera B, Gennis R, GuergovaKuras M and Berry E (1999) Pathways for proton release during ubihydroquinone oxidation by the bc1 complex. Proc Natl Acad Sci USA 96: 10021–10026 Crofts AR, Guergova-Kuras M, Kuras R, Ugulava N, Li J and Hong S (2000) Proton-coupled electron transfer at the Qo site: What type of mechanism can account for the high activation barrier? Biochim Biophys Acta 1459: 456–466 Crofts AR, Shinkarev VP, Kolling DR, Hong S (2003) The modified Q-cycle explains the apparent mismatch between the kinetics of reduction of cytochromes c1 and bH in the bc1 complex. J Biol Chem 278: 36191–36201 Crofts AR, Lhee S, Crofts SB, Cheng J and Rose S (2006) Proton pumping in the bc1 complex: A new gating mechanism that prevents short circuits. Biochim Biophys Acta 1757: 1019–103 Darrouzet E, Valkova-Valchanova M and Daldal F (2000a) Probing the role of the Fe-S subunit hinge region during Qo site catalysis in Rhodobacter capsulatus bc1 complex. Biochemistry 39: 15475–15483 Darrouzet E, Valkova-Valchanova M, Moser C, Dutton PL and Daldal F (2000b) Uncovering the [2Fe-2S] domain movement in cytochrome bc1 and its implications for energy conversion. Proc Natl Acad Sci USA 97: 4567–4572 Darrouzet E, Moser CC, Dutton PL and Daldal F (2001) Large scale domain movement in cytochrome bc1: A new device for electron transfer in proteins. Trends Biochem Sci 26: 445–451 Darrouzet E, Valkova-Valchanova M and Daldal F (2002) The [2Fe-2S] cluster Em as an indicator of the iron-sulfur subunit position in the ubihydroquinone oxidase site of the cytochrome bc1 complex. J Biol Chem 277: 3464–3470 De Vries S, Albracht SPJ, Berden JA, Marres CAM and Slater EC (1983) The Effect of pH, ubiquinone depletion and myxothiazol on the reduction kinetics of the prosthetic groups of ubiquinol: cytochrome c oxidoreductase. Biochim Biophys Acta 723: 91–103 Dikanov SA, Kolling DR, Endeward B, Samoilova RI, Prisner TF, Nair SK and Crofts AR (2006) Identification of hydrogen bonds to the Rieske cluster through the weakly coupled nitrogens detected by electron spin echo envelope modulation spectroscopy. J Biol Chem 281: 27416–27425 Dikanov SA, Samoilova RI, Kolling DRJ, Holland JT and Crofts
Chapter 23
Function and Evolution of Rieske/Cytochrome b Complexes
AR (2004) Hydrogen bonds involved in binding the Qi-site semiquinone in the bc1 complex, identified through deuterium exchange using pulsed EPR. J Biol Chem 279: 15814–15823 Ding HG, Robertson DE, Daldal F and Dutton PL (1992) Cytochrome-bc1 complex [2Fe-2S] cluster and its interaction with ubiquinone and ubihydroquinone at the Qo site — a double-occupancy Qo site model. Biochemistry 31: 3144–3158 Ding H, Moser CC, Robertson DE, Tokito MK, Daldal F and Dutton PL (1995) Ubiquinone pair in the Qo site central to the primary energy conversion reactions of cytochrome bc1 complex. Biochemistry 34: 15979–15996 Forquer I, Cape J, Bowman MK and Kramer DM (2005) Sequential two-electron oxidation of ubiquinol by the cytochrome bc1 complex. In: A van der Est and D Bruce (eds), Photosynthesis: Fundamental Aspects to Global Perspectives, Vol 1, pp 430-432. ACG Publishing, Lawrence Forquer I, Covian R, Bowman MK, Trumpower B and Kramer DM (2006) Similar transition states mediate the Q-cycle and superoxide production by the cytochrome bc1 complex. J Biol Chem 281: 38459–38465 Gao X, Wen X, Esser L, Quinn B, Yu L, Yu CA and Xia D (2003) Structural basis for the quinone reduction in the bc1 complex: a comparative analysis of crystal structures of mitochondrial cytochrome bc1 with bound substrate and inhibitors at the Qi site. Biochemistry 42: 9067–9080 Gao X, Wen X, Yu C, Esser L, Tsao S, Quinn B, Zhang L, Yu L and Xia D (2002) The crystal structure of mitochondrial cytochrome bc1 in complex with famoxadone: The role of aromatic-aromatic interaction in inhibition. Biochemistry 41: 11692–11702 Gennis RB, Barquera B, Hacker B, Van Doren SR, Arnaud S, Crofts AR, Davidson E, Gray KA and Daldal F (1993) The bc1 complex of Rhodobacter sphaeroides and Rhodobacter capsulatus. J Bioenerg Biomembr 25: 195–209 Gong X, Yu L, Xia D and Yu CA (2004) Evidence for electron equilibrium between the two hemes bL in the dimeric cytochrome bc1 complex. J Biol Chem 280: 9251–9257 Gong X, Yu L and Yu CA (2006) The role of an extra fragment of cytochrome b (residues 309-326) in the cytochrome bc1 complex from Rhodobacter sphaeroides. Biochemistry 45: 11122–11129 Gray KA and Daldal F (1995) Mutational studies of the cytochrome bc1 complexes. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 747–774. Kluwer Academic Publishers, Dordrecht Gray KA, Dutton PL and Daldal F (1994) Requirement of histidine-217 for ubiquinone reductase-activity (Qi-Site) in the cytochrome-bc1 complex. Biochemistry 33: 723–733 Guergova-Kuras M, Kuras R, Ugulava N, Hadad I and Crofts AR (2000) Specific mutagenesis of the Rieske iron-sulfur protein in Rhodobacter sphaeroides shows that both the thermodynamic gradient and the pK of the oxidized form determine the rate of quinol oxidation by the bc1 complex. Biochemistry 39: 7436–7444 Guidot DM, McCord JM, Wright RM and Repine JE (1993) Absence of electron transport (Rho 0 state) restores growth of a manganese-superoxide dismutase-deficient Saccharomyces cerevisiae in hyperoxia. Evidence for electron transport as a major source of superoxide generation in vivo. J Biol Chem 268: 26699–26703
471
Gutierrez-Cirlos EB and Trumpower BL (2002) Inhibitory analogs of ubiquinol act anti-cooperatively on the yeast cytochrome bc1 complex — Evidence for an alternating, halfof-the-sites mechanism of ubiquinol oxidation. J Biol Chem 277: 1195–1202 Hettmann T, Schmidt CL, Anemüller S, Zähringer U, Moll H, Petersen A and Schäfer G (1998) Cytochrome b558/566 from the archaeon Sulfolobus acidocaldarius. A novel highly glycosylated, membrane-bound b-type hemoprotein. J Biol Chem 273: 12032–12040 Hiller A, Henninger T, Schäfer G and Schmidt CL (2003) New genes encoding subunits of a cytochrome bc1-analogous complex in the respiratory chain of the hyperthermoacidophilic crenarchaeon Sulfolobus acidocaldarius. J Bioenerg Biomembr 35: 121–131 Huang D, Everly RM, Cheng RH, Heymann JB, Schägger H, Sled V, Ohnishi T, Baker TS and Cramer WA (1994) Characterization of the chloroplast cytochrome b6 f complex as a structural and functional dimer. Biochemistry 33: 4401–4409 Huang LS, Cobessi D, Tung EY and Berry EA (2005) Binding of the respiratory chain inhibitor antimycin to the mitochondrial bc1 complex: A new crystal structure reveals an altered intramolecular hydrogen-bonding pattern. J Mol Biol 351: 573–597 Hunte C, Koepke J, Lange C, Rossmanith T and Michel H (2000) Structure at 2.3 Å resolution of the cytochrome bc1 complex from the yeast Saccharomyces cerevisiae co-crystallized with an antibody Fv fragment. Structure with Folding and Design 8: 669–684 Iwata S, Lee JW, Okada K, Lee JK, Iwata M, Rasmussen B, Link TA, Ramaswamy S and Jap BK (1998) Complete structure of the 11-subunit bovine mitochondrial cytochrome bc1 complex. Science 281: 64–71 Jenny Jr. FE, Prince RC and Daldal F (1996) The membrane-bound cytochrome cy of Rhodobacter capsulatus can serve as an electron donor to the photosynthetic reaction center of Rhodobacter sphaeroides. Biochim Biophys Acta 1273: 159–164 Joliot P and Joliot A (1984) Electron transfer between the two photosystems. II Equilibrium constants. Biochim Biophys Acta 765: 219–226 Kessl JJ, Lange BB, Merbitz-Zahradnik T, Zwicker K, Hill P, Meunier B, Palsdottir H, Hunte C, Meshnick S and Trumpower BL (2003) Molecular basis for atovaquone binding to the cytochrome bc1 complex. J Biol Chem 278: 31312–31318 Klishin SS and Mulkidjanian AY (2005) Electron/proton coupling in the cytochrome bc1 complex of Rhodobacter capsulatus. Biophys J 88: 326A Kolling DRJ, Samoilova RI, Holland JT, Berry EA, Dikanov SA and Crofts AR (2003) Exploration of ligands to the Qi site semiquinone in the bc1 complex using high-resolution EPR. J Biol Chem 278: 39747–39754 Kramer D and Crofts AR (1992) The concerted reduction of cytochromes b and f of the b/f complex by plastoquinol. EBEC Short Reports 7: 4 Kramer DM and Crofts AR (1993) The concerted reduction of the high- and low-potential chains of the bf complex by plastoquinol. Biochim Biophys Acta 1183: 72–84 Kramer DM and Crofts AR (1994) Re-examination of the properties and function of the b cytochromes of the thylakoid cytochrome bf complex. Biochim Biophys Acta 1184: 193–201 Kramer DM, Roberts AG, Muller F, Cape J and Bowman MK (2003) Q-cycle bypass reactions at the Qo site of the cyto-
472
David M. Kramer, Wolfgang Nitschke and Jason W. Cooley
chrome bc1 (and related) complexes. Methods in Enzymology 382: 21–45 Kramer DM, Cape JL, Forquer I and Bowman MK (2005) Kinetic steering of quinol oxidation by ‘proton stripping’ at the cytochrome bc1 complex Qo site. In: A van der Est and D Bruce (eds) Photosynthesis: Fundamental Aspects to Global Perspectives, Vol 1, pp 424-427. ACG Publishing, Lawrence Kurisu G, Zhang H, Smith JL and Cramer WA (2003) Structure of the cytochrome b6 f complex of oxygenic photosynthesis: Tuning the cavity. Science 302: 1009–1014 Lange C and Hunte C (2002) Crystal structure of the yeast cytochrome bc1 complex with its bound substrate cytochrome c. Proc Natl Acad Sci USA 99: 2800–2805 Lebovitz RM, Zhang H, Vogel H, Cartwright J, Dionne L, Lu N, Huang S and Matzuk MM (1996) Neurodegeneration, myocardial injury, and perinatal death in mitochondrial superoxide dismutase-deficient mice. Proc Natl Acad Sci USA 93: 9782–9787. Lebrun E, Brugna M, Baymann F, Muller D, Lievremont D, Lett MC and Nitschke W (2003) Arsenite oxidase, an ancient bioenergetic enzyme. Mol Biol Evol 20: 686–693 Lebrun E, Santini JM, Brugna M, Ducluzeau AL, Ouchane S, Schoepp-Cothenet B, Baymann F and Nitschke W (2006) The Rieske protein: a case study on the pitfalls of multiple sequence alignments and phylogenetic reconstruction. Mol Biol Evol 23: 1180–1191 Link TA (1997) The role of the ‘Rieske’ iron sulfur protein in the hydroquinone oxidation (QP) site of the cytochrome bc1 complex — The ‘proton-gated affinity change’ mechanism. FEBS Lett 412: 257–264 Longo VD, Liou LL, Valentine JS and Gralla EB (1999) Mitochondrial superoxide decreases yeast survival in stationary phase. Arch Biochem Biophys 365: 131–142 Malkin R (1981) Interaction of the quinone analogue, DBMIB, with the photosynthetic Rieske iron-sulfur center. Israel J Chem 21: 301–305 Martinez SE, Huang D, Szczepaniak A, Cramer WA and Smith JL (1994) Crystal structure of chloroplast cytochrome f reveals a novel cytochrome fold and unexpected heme ligation. Structure 2: 95–105 Mitchell P (1975) The protonmotive Q cycle: A general formulation. FEBS Lett 59: 137-139 Mooser D, Maneg O, Corvey C, Steiner T, Malatesta F, Karas M, Soulimane T and Ludwig B (2005) A four-subunit cytochrome bc1 complex complements the respiratory chain of Thermus thermophilus. Biochim Biophys Acta 1708: 262–274 Mooser D, Maneg O, MacMillan F, Malatesta F, Soulimane T and Ludwig B (2006) The menaquinol-oxidizing cytochrome bc complex from Thermus thermophilus: Protein domains and subunits. Biochim Biophys Acta 1757: 1084–1095 Mulkidjanian AY (2005) Ubiquinol oxidation in the cytochrome bc1 complex: Reaction mechanism and prevention of shortcircuiting. Biochim Biophys Acta 1709: 5–34 Muller F (2000) The nature and mechanism of superoxide production by the electron transport chain: Its relevance to aging. J Amer Aging Assoc 23: 227–256 Muller F, Crofts AR and Kramer DM (2002) Multiple Q-cycle bypass reactions at the Qo-site of the cytochrome bc1 complex. Biochemistry 41: 7866–7874 Muller F, Roberts AG, Bowman MK and Kramer DM (2003) Architecture of the Qo site of the cytochrome bc1 complex probed
by superoxide production. Biochemistry 42: 6493–6499 Naqui A, Chance B and Cadenas E (1986) Reactive oxygen intermediates in biochemistry. Annu Rev Biochem 55: 137–166 Nitschke W, Kramer DM, Riedel A and Liebl U (1995) From Naphtho- to benzoquinones — (r)evolutionary reorganisations of electron transfer chains. In: Mathis P (ed) Photosynthesis: From Light to Biosphere, Vol I, pp 945–950. Kluwer Academic Press, Dordrecht Ohnishi T, Brandt U and von Jagow G (1988) Studies on the effect of stigmatellin derivatives on cytochrome b and the Rieske iron-sulfur cluster of cytochrome c reductase from bovine heart mitochondria. FEBS Lett 176: 385–389 Ohnishi T, Johnson JE, Jr., Yano T, LoBrutto R and Widger WR (2005) Thermodynamic and EPR studies of slowly relaxing ubisemiquinone species in the isolated bovine heart complex I. FEBS Lett 579: 500–506 Osyczka A, Moser CC, Daldal F and Dutton PL (2004) Reversible redox energy coupling in electron transfer chains. Nature 427: 607–612 Osyczka A, Moser CC and Dutton PL (2005) Fixing the Q cycle. Trends Biochem Sci 30: 176–182 Osyczka A, Zhang H, Mathe C, Rich PR, Moser CC and Dutton PL (2006) Role of the PEWY glutamate in hydroquinone-quinone oxidation-reduction catalysis in the Qo site of cytochrome bc1. Biochemistry 45: 10492–10503 Palsdottir H, Lojero CG, Trumpower BL and Hunte C (2003) Structure of the yeast cytochrome bc1 complex with a hydroxyquinone anion Qo site inhibitor bound. J Biol Chem 278: 31303–31311 Rich P, Heathcote P and Moss DA (1987) Electron and proton transfer in the cytochrome bf complex. In: Biggins J (ed) Progress in Photosynthesis Research, Vol II, pp 453–460. Martinus Nijhoff, Dordrecht Rinyu L, Martin EW, Takahashi E, Maroti P and Wraight CA (2004) Modulation of the free energy of the primary quinone acceptor (QA) in reaction centers from Rhodobacter sphaeroides: Contributions from the protein and protein-lipid(cardiolipin) interactions. Biochim Biophys Acta 1655: 93–101 Roberts A, Bowman MK and Kramer DM (2001) Certain metal ions are inhibitors of cytochrome b6 f ‘Rieske’ iron-sulfur protein domain movements. Biochemistry 41: 4070–4079 Rothery RA, Seime AM, Spiers AM, Maklashina E, Schroder I, Gunsalus RP, Cecchini G and Weiner JH (2005) Defining the Q-site of Escherichia coli fumarate reductase by site-directed mutagenesis, fluorescence quench titrations and EPR spectroscopy. FEBS J 272: 313–326 Schütz M, Brugna M, Lebrun E, Baymann F, Huber R, Stetter KO, Hauska G, Toci R, Lemesle-Meunier D, Tron P, Schmidt C and Nitschke W (2000) Early evolution of cytochrome bc complexes. J Mol Biol 300: 663–675 Shinkarev VP (2006) Ubiquinone (coenzyme Q10) binding sites: low dielectric constant of the gate prevents the escape of the semiquinone. FEBS Lett 580: 2534–2539 Shinkarev VP and Wraight CA (2007) Intermonomer electron transfer in the bc1 complex dimer is controlled by the energized state and by impaired electron transfer between low and high potential hemes. FEBS Lett 581: 1535–1541 Shinkarev VP, Kolling DR, Miller TJ and Crofts AR (2002) Modulation of the midpoint potential of the [2Fe-2S] Rieske iron sulfur center by Qo occupants in the bc1 complex. Biochemistry 41: 14372–14382
Chapter 23
Function and Evolution of Rieske/Cytochrome b Complexes
Smith JL, Zhang H, Yan J, Kurisu G and Cramer WA (2004) Cytochrome bc complexes: A common core of structure and function surrounded by diversity in the outlying provinces. Curr Opin Struct Biol 14: 432–439 Snyder CH, Gutierrez-Cirlos EB and Trumpower BL (2000) Evidence for a concerted mechanism of ubiquinol oxidation by the cytochrome bc1 complex. J Biol Chem 275: 13535–13541 Snyder CH, Merbitz-Zahradnik T, Link TA and Trumpower BL (1999) Role of the Rieske iron-sulfur protein midpoint potential in the protonmotive Q-cycle mechanism of the cytochrome bc1 complex. J Bioenerg Biomembr 31: 235–242 Soriano GM, Ponamarev MV, Carrell CJ, Xia D, Smith JL and Cramer WA (1999) Comparison of the cytochrome bc1 complex with the anticipated structure of the cytochrome b6 f complex: De plus ça change de plus c’est la même chose. J Bioenerg Biomemb 31, 201–213 Stahl DA, Fishbain S, Klein M, Baker BJ and Wagner M (2002) Origins and diversification of sulfate-respiring microorganisms. Antoine Van Leeuwenhoek 81: 189–195 Stroebel D, Choquet Y, Popot JL and Picot D (2003) An atypical haem in the cytochrome b6 f complex. Nature 426: 413–418 Sun H and Trumpower BL (2003) Superoxide anion generation by the cytochrome bc1 complex. Arch Biochem Biophys 419: 198–206 Trumpower BL (1990) The protonmotive Q cycle. J Biol Chem 265: 11409–11412 Trumpower BL (2002) A concerted, alternating sites mechanism of ubiquinol oxidation by the dimeric cytochrome bc1 complex. Biochim Biophys Acta 1555: 166–173 Trumpower BL and Gennis RB (1994) Energy transduction by cytochrome complexes in mitochondrial and bacterial respiration: The enzymology of coupling electron transfer reactions to transmembrane proton translocation. Annu Rev Biochem 63: 675–716 Trumpower BL, Covian R and Gutierrez-Cirlos EB (2004) Halfof-the-sites reactivity of the yeast cytochrome bc1 complex. Biochim Biophys Acta1658: 33–33 Valkova-Valchanova M, Darrouzet E, Moomaw C, Slaughter C and Daldal F (2000) Proteolytic cleavage of the Fe-S subunit hinge region of Rhodobacter capsulatus bc1 complex: Effects of inhibitors and mutations. Biochemistry 39: 15484–15492 Vignais PM, Billoud B and Meyer J (2001) Classification and phylogeny of hydrogenases. FEMS Microbiol Rev 25: 455–501 von Mering C, Hugenholtz P, Raes J, Tringe SG, Doerks T, Jensen LJ, Ward N and Bork P (2007) Quantitative phylogenetic assessment of microbial communities in diverse environments. Science 315: 1126-1130 Wenz T, Covian R, Hellwig P, Macmillan F, Meunier B, Trumpower BL and Hunte C (2007) Mutational analysis of cytochrome b at the ubiquinol oxidation site of yeast complex III. J Biol Chem 282: 3977–3988 Widger WR, Cramer WA, Herrmann RG and Trebst A (1984)
473
Sequence homology and structural similarity between cytochrome b of mitochondrial complex III and the chloroplast b6-f complex: Position of the cytochrome b hemes in the membrane. Proc Natl Acad Sci USA 81: 674–678 Xia D, Yu C-A, Kim H, Xia J-Z, Kachurin AM, Zhang L, Yu L and Deisenhofer J (1997) Crystal structure of the cytochrome bc1 complex from bovine heart mitochondria. Science 277: 60–66 Xiao KH, Yu L and Yu CA (2000) Confirmation of the involvement of protein domain movement during the catalytic cycle of the cytochrome bc1 complex by the formation of an intersubunit disulfide bond between cytochrome b and the iron-sulfur protein. J Biol Chem 275: 38597–38604 Xiong J, Fischer WM, Inoue K, Nakahara M and Bauer CE (2000) Molecular evidence for the early evolution of photosynthesis. Science 289: 1724–1730 Yap LL, Samoilova RI, Gennis RB and Dikanov SA (2006) Characterization of the exchangeable protons in the immediate vicinity of the semiquinone radical at the QH site of the cytochrome bo3 from Escherichia coli. J Biol Chem 281: 16879–16887 Yu J, and Le Brun (1998) Studies of the cytochrome subunits of menaquinone:cytochrome c reductase (bc complex). Evidence of the covalent attachment of the cytochrome b subunit. J Biol Chem 273: 8860–8866 Yu CA, Xia D, Kim H, Deisenhofer J, Kachurin AM, Zhang L, Deng KP and Yu L (1998a) Three-dimensional structure and functions of bovine heart mitochondrial cytochrome bc1 complex. Biofactors 8: 187–189 Yu CA, Xia D, Kim H, Deisenhofer J, Zhang L, Kachurin AM and Yu L (1998b) Structural basis of functions of the mitochondrial cytochrome bc1 complex. Biochim Biophys Acta 1365: 151–158 Zatsman AI, Zhang H, Gunderson WA, Cramer WA and Hendrich MP (2006) Heme-heme interactions in the cytochrome b6 f complex: EPR spectroscopy and correlation with structure. J Am Chem Soc 128: 14246–14247 Zhang H, Primak A, Cape J, Bowman MK, Kramer DM and Cramer WA (2004) Characterization of the high-spin heme x in the cytochrome b6 f complex of oxygenic photosynthesis. Biochemistry 43: 16329–16336 Zhang L, Yu L and Yu CA (1998) Generation of superoxide anion by succinate-cytochrome c reductase from bovine heart mitochondria. J Biol Chem 273: 33972–33976 Zhang Z, Huang L, Shulmeister V, Chi Y, Kim K, Hung L, Crofts A, Berry E and Kim S (1998) Electron transfer by domain movement in cytochrome bc1. Nature 392: 677–684 Zhu Z and Gunner MR (2005) Energetics of quinone-dependent electron and proton transfers in Rhodobacter sphaeroides photosynthetic reaction centers. Biochemistry 44: 82–96 Zusman LD and Beratan DN (1996) Two-electron transfer reactions in polar solvents. J Chem Phys 165–176
Chapter 24 Proton Translocation and ATP Synthesis by the FoF1-ATPase of Purple Bacteria Boris A. Feniouk Chemical Resources Laboratory, Tokyo Institute of Technology, Nagatsuta 4259, Midori-ku, Yokohama 226-8503 Japan
Wolfgang Junge* Department of Biophysics, University of Osnabrück, DE-49069 Osnabrück, Germany
Summary ............................................................................................................................................................... 475 I. Introduction..................................................................................................................................................... 476 II. Structure and Rotary Catalysis....................................................................................................................... 476 III. Proton Translocation and its Coupling to ATP Synthesis/Hydrolysis ............................................................. 478 A. The Structure of FO, the Rotary Electromotor ................................................................................... 478 1. Subunit c and its Oligomers .................................................................................................... 478 2. Subunit a ................................................................................................................................ 480 3. Subunit b ................................................................................................................................ 481 B. Proton Transport through FO : Functional Aspects ............................................................................ 481 1. Hypothetical Rotary Mechanism of Proton Translocation ....................................................... 481 2. The Unitary Proton Conductance of FO and FOF1 as Determined in Membrane Vesicles of Purple Bacteria........................................................................................................................ 482 3. Mechanistic Implications ......................................................................................................... 484 C. Coupling of Proton Transport to ATP Synthesis/Hydrolysis ............................................................. 485 1. Elastic Coupling Hypothesis.................................................................................................... 485 2. Coupling Efficiency.................................................................................................................. 486 III. Role of Proton Translocation in the Regulation of ATP Synthase .................................................................. 486 Acknowledgments ................................................................................................................................................. 487 References ............................................................................................................................................................ 488
Summary In purple bacteria both the light driven and respiratory electron transfers serve the sole purpose of generating a proton-motive force across their inner membrane. The backflow of protons is monopolized by the enzyme FOF1-ATP synthase producing ATP from ADP and Pi. Almost all the useful work derived from absorbed sunlight is delivered to the cell in form of the ATP/ADP-couple in purple bacteria, whereas this accounts for only about 20% of the total in green plants and cyanobacteria. FOF1 is composed of two rotary motors; FO uses a transmembrane proton motive force to generate torque and F1 uses the torque to synthesize ATP. When operat*Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 475–493. © 2009 Springer Science + Business Media B.V.
476
Boris A. Feniouk and Wolfgang Junge
ing in reverse the enzyme generates proton motive force at the expense of ATP hydrolysis. Both portions are mechanically coupled by a central rotary shaft and held together by a peripheral stalk. Elasticity of mechanical power transmission is a prerequisite for high kinetic efficiency of the two motor/generators. Although being elastically coupled, they do not slip against each other, and under physiological conditions there is no proton leak without concomitant ATP synthesis. If the membrane is de-energized, the enzyme is inactivated. Re-activation requires a protonmotive force, and it is presumably achieved via proton driven rotation of FO transmitted to F1. Therefore, proton translocation by the FOF1 has the dual functions of driving and regulating ATP synthesis. The subunit compositions and construction principles of FOF1-ATP synthases of various prokaryotic and eukaryotic organisms are very similar. The enzymes from photosynthetic organisms have proven to be particularly useful for elucidating the electrochemical aspects, as emphasized in this article. I. Introduction The FOF1-ATP synthase (also known as F-type ATPase, or simply as FOF1) is a membrane protein of bacteria, chloroplasts and mitochondria. In all cases, it synthesizes ATP from ADP and inorganic phosphate (Pi). By translocating protons (H+) it derives the free energy for ATP synthesis from the electrochemical potential difference of the proton across the respective coupling membrane (∆µ~H+). Some bacteria reverse this operation, generating ~ ∆µ H+ at the expense of ATP hydrolysis to power motility and secondary transport. The unique ion translocation and chemical synthesis activities of the ATP synthase are performed by two rotary machines, named FO and F1, which are mechanically coupled by a central rotary shaft and mounted one to the other by an outer bearing. Today’s rather comprehensive view of this enzyme has emerged from studies carried out on several different organisms offering different advantages, e.g., Escherichia (E.) coli being superior for molecular biology, mitochondria for classical enzymology and X-ray crystal structure analysis, and chloroplast and purple bacteria for rapid kinetic studies. This raises a question: is it acceptable to pool these data together and try to fit them in a coherent common model? Clearly, there are many distinctions between mitochondrial, chloroplast, and bacterial FOF1, such as the subunit composition and structure, kinetic properties, and regulatory features. Moreover, the coupling ion is not always the proton, but it can be Na+ in certain bacteria (Dimroth, 1994). However, the overall topology, the catalytic core, the primary Abbreviations: AFM – atomic force microscopy; E. – Escherichia; FOF1 – F-type ion translocating ATP synthase; NMR – nuclear magnetic resonance; Pi – inorganic phosphate; Rba. – Rhodo~ bacter TMH – transmembrane helix; ∆ µH+ – transmembrane electrochemical potential difference of proton
structure of subunits directly involved in catalysis and ion transport, and the main features of the catalytic mechanism are all strikingly similar in enzymes from various organisms. This accounts for the fact that the enzyme must have emerged before the evolutionary separation of archaea and eubacteria, and that it has been conserved to an astounding degree among bacteria, chloroplasts and mitochondria. In this review we briefly survey the general properties of the enzyme, and emphasize what has been done on purple bacteria. This deliberately biases the reference list. In doing so we chose to focus on the otherwise least studied side of FOF1: the mechanism of proton translocation and its coupling to ATP synthesis/hydrolysis. As will be described below, chromatophores from purple bacteria have proved particularly useful for experiments in this line. In reading this chapter, it might be helpful to view the animations of enzyme operation, which can be found at http://www.biologie.uni-osnabrueck.de/biophysik/ Junge/pics.htm. II. Structure and Rotary Catalysis The enzyme consists of two distinct portions, both of which are multi-subunit complexes: the hydrophobic proton-translocating FO-portion that is embedded in the membrane, and the hydrophilic F1-portion that protrudes by ~100 Å from the plane of the membrane and bears the nucleotide binding sites (Fig. 1). The F1 portion can be detached from the membrane-embedded FO by removal of magnesium ions from the medium (Baccarini-Melandri et al., 1970), or by addition of chaotropic agents (Ponomarenko et al., 1999). The isolated water-soluble F1-portion maintains the ability to hydrolyze ATP, and is often referred to as ‘F1-ATPase.’ However, in some cases (e.g., in chloroplasts) the ATPase activity of F1 is latent
Chapter 24
Proton Translocation in ATP Synthase
Fig. 1. Cartoon representation of bacterial ATP synthase (crosssection; only one of the two b subunits, and only one half of the α3β3 hexamer and of the c-ring are shown). Rotor subunits are dotted.
because of several inhibitory mechanisms (ADP-inhibition, inhibition by subunit ε, thiol modulation). The isolated FO-portion operates as a passive proton transporter that facilitates the dissipation of ∆µ~H+. It is possible to reconstitute a fully active enzyme from isolated FO and F1 (Melandri et al., 1971; Saphon et al., 1975b; Feng and McCarty, 1990). The catalytic core of the F1-portion is composed of three types of subunit, named by Greek letters according to their molecular weights (from highest to lowest) and with a stoichiometry of α3β3γ1. Three αβ heterodimers form an ‘orange’-shaped hexamer with a long central cavity filled by the elongated portion of subunit γ which comprises a N-terminal coiled coil structure and an α helical domain close to its C-terminus. The isolated (αβ)3γ -complex is capable of ATP hydrolysis at a high rate (up to 300 s–1), but cannot catalyze ATP synthesis without being coupled to the FO-portion and to the other F1 subunits. There are six nucleotide-binding sites on the α3β3 hexamer, located in the six clefts between adjacent subunits. Only three of these six sites participate in catalysis, as nucleotide binding to the other three is not essential for activity (Weber et al., 1994, 1995). Whether the role of the other three sites is regulatory (Milgrom et al., 1990, 1991; Jault and Allison, 1993; Jault et al., 1995; Matsui et al., 1997), or just structural is an open question. The catalytic sites are formed mainly by the subunit β, with the subunit α contributing an essential arginine close to the γ-phosphate of ATP. All three sites are identical, but at any given time they
477
have different conformations and different affinities to nucleotides. Changes in affinity to nucleotides are coordinated between the sites, so that nucleotide binding to one site triggers the nucleotide release from another site, the so-called ‘binding change mechanism’ (Kayalar et al., 1977). This coordination of the catalytic sites’ affinity changes is intimately linked to the rotation of the subunit γ inside the α3β3 hexamer; see Yoshida et al. (2001), Senior et al. (2002), Weber and Senior (2003), Kinosita, Jr. et al. (2004) and Gao et al. (2005) for recent reviews on the rotary catalysis in F1. Under ATP hydrolysis conditions, such rotation, initially experimentally supported by cross-linking data (Duncan et al., 1995) and time resolved by polarized absorption after photobleaching (Sabbert et al., 1996), was directly visualized in single-molecule experiments (Noji et al., 1997; Yasuda et al., 1998; Hisabori et al., 1999; Noji et al., 1999; Omote et al., 1999; Panke et al., 2000). It became apparent that it is the rotation of the subunit γ relative to (αβ)3 that serves as the power transmission from the ion-driven FO. Due to this amazing feature, it is convenient to classify the subunits into ‘rotor’ (γ and ε of F1 and the C-ring of FO) and ‘stator’ (α3β3δ of F1 and b2a of FO). Because the enzyme carries out rapid Brownian rotation in the membrane, correlation time 200 µs (Sabbert et al., 1997), the terms stator and rotor are arbitrary, and what really matters for the function of the enzyme is their relative rotation against each other. The subunit ε (in chloroplasts and bacteria) is attached to the subunit γ in the region where the latter protrudes from the α3β3 hexamer (from the side facing the membrane), hence it functionally belongs to the rotor (Schulenberg et al., 1997; Häsler et al., 1998; Kato-Yamada et al., 1998). The subunit δ is attached to the periphery of the α3β3 hexamer from the other side and belongs to the stator (Fig. 1). Its binding to the (αβ)3 hexamer is very strong with an affinity in the sub-nanomolar range (Häsler et al., 1999; Weber et al., 2002). The mitochondrial F1-portion has some additional subunits and a slightly different nomenclature. For instance, the homolog of the bacterial subunit ε is named δ, while the homolog of bacterial δ is named OSCP (oligomycin sensitivity conferring protein) in mitochondria. Detailed information on the structure of the mitochondrial F1 can be found elsewhere (Gibbons et al., 2000; Dickson et al., 2006). FO is bi-partite: its rotor portion is made up of a ring of very hydrophobic c subunits, and the interior of the
478 ring is filled with lipid (Meier et al., 2002; Oberfeld et al., 2006). The stator portion of FO includes the subunits a and b, both of which are located at the periphery of the c-ring (Birkenhager et al., 1995; Singh et al., 1996; Takeyasu et al., 1996). Subunit a is a highly hydrophobic protein, and is also involved in proton translocation (Zhang and Vik, 2003), whereas subunit b is elongated, largely α helical and anchored to the membrane by its N-terminal hydrophobic domain. It is present as a single copy in the mitochondrial enzyme, as a homodimer in E. coli and many other bacterial enzymes, and as a heterodimer of similar b and b-prime (b´) subunits in chloroplast and purple bacterial enzymes. The elongated, hydrophilic and α-helical domain of subunit b is bound to the subunits α and δ (Häsler et al., 1999; Dunn et al., 2000a; Weber et al., 2004). Besides the subunits a and b, a few additional FO subunits are also present in the mitochondrial ATP synthase (Collinson et al., 1994; Velours et al., 2000). There is little doubt that during catalysis the c-ring rotates together with the γε-complex relative to the rest of the enzyme. In E. coli FOF1, an actin filament mono-specifically linked to the c-ring rotates when F1 is driven by ATP hydrolysis (Sambongi et al., 1999; Panke et al., 2000). It has been demonstrated that covalent cross-linking of the subunits γ, ε and c has little effect on the activity (Tsunoda et al., 2001a). ATP-driven rotation of the c-ring that was sensitive to tributyltin chloride, a specific inhibitor of proton translocation through FO, was observed in Bacillus PS3 FOF1 (Ueno et al., 2005). Many additional pieces of experimental evidence also support the rotation of the c-ring during catalysis (Jones et al., 2000; Hutcheon et al., 2001; Tanabe et al., 2001; Gumbiowski et al., 2002; Kaim et al., 2002; Suzuki et al., 2002). It should be noted that rotation of the γ ε cn-complex relative to the stator part of the enzyme, as directly observed in the above-mentioned experiments, was powered by ATP hydrolysis. Demonstration of the rotation of the γ ε cn complex during ATP synthesis was technically much more challenging. The most convincing evidence in favor of rotary ATP synthesis has been the external, magnetically driven rotation of the subunit γ resulting in ATP synthesis (Itoh et al., 2004; Rondelez et al., 2005). Moreover, single-molecule fluorescence resonance energy transfer (FRET) experiments with the E. coli FOF1 incorporated into liposomes suggested that during ATP synthesis powered by ∆µ~H+ a complex formed from the γ and ε subunits (and presumably from the
Boris A. Feniouk and Wolfgang Junge c-ring) rotates relative to the subunit b (Diez et al., 2004; Zimmermann et al., 2005). It is thus established that proton transport coupled to ATP synthesis also involves rotation of the c-ring oligomer relative to the rest of FO, and that FO operates as a rotary proton transporter. III. Proton Translocation and its Coupling to ATP Synthesis/Hydrolysis A. The Structure of FO, the Rotary Electromotor 1. Subunit c and its Oligomers Photo-affinity labeling has revealed that the highly hydrophobic c-subunit, named the proteolipid, is made up of two transmembrane helices interconnected by a polar loop (Hoppe and Sebald, 1984). This gross conformation does not change between the proton translocating activity and inactivity (Weber et al., 1986). Nuclear magnetic resonance (NMR) studies with this subunit dissolved in chloroform/methanol/ water mixture indicated a hairpin structure of two extended helices (Girvin and Fillingame, 1994; Matthey et al., 1999; Rastogi and Girvin, 1999; Dmitriev and Fillingame, 2001). It was assumed that several copies of the c-subunit were arranged as a homooligomeric ring. This was first established by crystal structure analyses of the yeast enzyme at medium resolution (Stock et al., 1999), and corroborated for the Na+-translocating Ilyobacter tartaricus at high resolution (Meier et al., 2005a). It was also established for several organisms by atomic force microscopy (AFM) studies (Seelert et al., 2000; Stahlberg et al., 2001; Pogoryelov et al., 2005). In the oligomer, the c-subunits are arranged ‘face-to-back’, with their Cterminal transmembrane helices (cTMH2) being at the outer side of the ring; see Fillingame and Dmitriev (2002) and the references therein. The number of copies of the c-subunits in this ring varies between organisms. The partial X-ray structure of the yeast enzyme has revealed 10 copies of c subunits (Stock et al., 1999). The same number was then determined in E. coli by genetic fusion of the c-subunits (Jiang et al., 2001), and also in the thermophilic Bacillus PS3 (Mitome et al., 2004). However, the decameric c-ring structure is not universal. AFM studies have revealed a ring with 11 copies of the c-subunit in the sodium-translocating enzymes from Ilyobacter
Chapter 24
Proton Translocation in ATP Synthase
tartaricus (Stahlberg et al., 2001) and Propionigenium modestum (Meier et al., 2003), 14 copies in the chloroplast (Seelert et al., 2000), and 15 copies in the alkaliphilic cyanobacterium Spirulina platensis enzymes (Pogoryelov et al., 2005). It is likely that the ring size is determined mainly by pair-wise neighborto-neighbor interactions between the c-subunits. AFM studies show that rings with gaps where one or more c-subunits are missing still have the same diameter as intact rings (Meier et al., 2005b; Pogoryelov et al., 2005). This organism-specific variation implies a gearshift between the ion-driven electromotor (FO) and the chemical generator (F1). The physiological necessity for this gearshift and its structural determinants still await characterization. Primary structure analysis of the c-subunits from various organisms reveals that there are only a few
479
highly conserved residues (Fig. 2). An acidic residue (usually glutamate, but aspartate in E. coli, i.e., cAsp61) that is located approximately in the middle of cTMH2 is essential for proton translocation, and is absolutely conserved. It could not be displaced up or down without loss of function, although in E. coli a second-site revertant was found, cD61N/cA24D, where the essential acid residue was swapped to the other helix, keeping it at the same depth in the membrane, (Miller et al., 1990). The residue three positions down, either alanine (if the acidic residue is a glutamic acid as in chloroplasts) or glycine (if the acidic residue is an aspartic acid, cD61, as in E. coli) is also conserved (Fig. 2). A highly conserved Arg-(Gln/Asn)-Pro trio is located in the polar loop between the cTMH1 and cTMH2 along with several glycines in the cTMH1. One possible function of the
Fig. 2. Multiple sequence alignments for ATP synthase subunit a (TMH4 and TMH5 only) and subunit c. Non-purple bacteria: Escherichia coli, Propionigenium modestum, Bacillus PS3, Ilyobacter tartaricus, Paracoccus denitrificans, Silicibacter pomeroyi DSS-3, Rhizobium leguminosarum bv viciae 3841. Purple bacteria: Rhodobacter capsulatus, Rhodopseudomonas palustris CGA009, Rhodospirillum rubrum ATCC 11170, Rhodobacter sphaeroides. Mitochondria: Saccharamyces cerevisiae, Bos taurus, Homo sapiens. Chloroplasts: Spinacia oleracea. Shaded in grey are regions where significant homology for all sequences was found by MACAW software with a pairwise score cutoff of 35. Bold and underlined are highly conserved residues.
480 latter glycines is to keep the cross-section diameter of the cTMH1 (i.e., the inner TMH in the c-ring) smaller, so that the ring oligomer could be properly formed with a larger cTMH2 facing outwards. The Arg-(Gln/Asn)-Pro trio in the polar loop is involved in binding of the F1 subunits γ and ε (Zhang and Fillingame, 1995; Watts et al., 1996; Watts and Capaldi, 1997). The functional role of the conserved Ala/Gly in the cTMH2 (Gly58 in E. coli) is unknown. Finally, the conserved acidic residue in the middle of the cTMH2 is directly involved in proton translocation. This residue can be covalently labeled by DCCD with complete loss of the proton translocating function; hence, subunit c is sometimes referred to as the ‘DCCD-binding protein’ (Altendorf and Zitzmann, 1975; Fillingame, 1975; Sone et al., 1979). The c-ring itself is incapable of transmembrane proton translocation, and requires other FO subunits for its proper function (Schneider and Altendorf, 1985). 2. Subunit a Electron microscopy (Birkenhager et al., 1995) and AFM studies (Singh et al., 1996; Takeyasu et al., 1996) revealed that the a and b subunits are attached to the outer side of the c-ring. The subunits ab2 alone can form a stable complex (Stalz et al., 2003), and only one such complex is attached to the c-ring (Schneider and Altendorf, 1985). As the ring is symmetrical, stable, and in principle large enough to accommodate at least two ab2-complexes, the stoichiometry of one ab2 per FOF1 is probably dictated by the binding of b2 via δ to the N-terminal collar of (αβ)3 at the top of F1. A moderate binding strength between the c-ring and the ab2 complex is an obvious pre-requisite for rotation of the c-ring relative to ab2. Still, this binding is strong enough to assemble, in the absence of F1-components, a proton conducting FO from the a, b, and c subunits (Schneider and Altendorf, 1985; Steffens et al., 1988), even though the presence of F1 during assembly significantly increases the activity (Pati et al., 1991). To account for both the sufficient strength of binding and the mobility of the c-ring relative to ab2, it is tempting to speculate that the subunits a and b alternate in binding when they process along the surface of the c-ring during the rotary proton translocation. No crystal structure is yet available for the subunit a, but quite a bit of biochemical evidence exists (see Vik et al., 2000 for a review). The E. coli subunit a probably has five transmembrane helices (Long et al., 1998; Valiyaveetil and Fillingame, 1998; Wada et al.,
Boris A. Feniouk and Wolfgang Junge 1999), of which the last two, aTMH4 and aTMH5 (see Fig. 2), host the amino acid residues essential for proton transport. For this function, the most important residue is likely to be an arginine (aArg210 in E. coli, see Fig. 2) located in aTMH4, in the center of the hydrophobic span with a slight shift to the cytoplasmic side (Lightowlers et al., 1987). Substitution of this arginine by alanine results in uncoupling of proton transport and ATP hydrolysis. The mutant is incapable of ATP synthesis, no ATP-driven proton pumping is observed, and passive proton flow through FO is no longer blocked by addition of F1, in contrast to wild type FO (Hatch et al., 1995; Valiyaveetil and Fillingame, 1997). Other conserved residues in aTMHs 4–5 include a glutamine three positions down from the above mentioned arginine, another glutamine approximately in the middle of aTMH5 (Gln252 in E. coli), a histidine and a glutamate. In the E. coli enzyme, the latter two residues are aHis245 on aTMH5 and aGlu219 on aTMH4 (underlined in Fig. 2) However, in the majority of other organisms, including purple bacteria, the locations of these residues are switched, with the histidine being in aTMH4 in the position corresponding to the E. coli aGlu219, while glutamate is in aTMH5 in the position of E. coli His245. Experiments on E. coli mutants demonstrated that both the histidine and the glutamate residues are involved in (although not essential for) proton translocation, and can be swapped with retention of FO function (Lightowlers et al., 1987, 1988; Cain and Simoni, 1988; Vik and Antonio, 1994; Hatch et al., 1998). It is noteworthy that the E. coli-like arrangement is found in organisms having aspartate, but not glutamate, as a conserved acidic residue in the cTMH2 (cAsp61 in E. coli). There are more residues in aTMHs 1–3 which affect the rate of proton translocation (Vik et al., 1988; Patterson et al., 1999; Long et al., 2002; DeLeon-Rangel et al., 2003), but none of them seems to be indispensable. Similar to what was observed with the subunit c, second-site revertants have been reported for subunit a, aR210X/aQ252R (Hatch et al., 1995) and aR210K/aG252R (Howitt and Cox, 1992), with the essential basic residue swapped over to another helix, but kept at the same depth in the membrane. These observations support the notion that the mechanism of proton transport through FOF1-ATP synthase occurs at the interface of the c-ring and the subunit-a, and the electrostatic interaction (rather than the van-der-Waals contact) between the charged residue on the c-subunit with the basic residue on the a-subunit is essential.
Chapter 24
Proton Translocation in ATP Synthase
3. Subunit b Subunit b is generally considered to have a purely structural role, and it connects as a homo- (b2) or heterodimer (b,b´) the subunits a and c with α3β3δ (Fig. 1; see Dunn et al. (2000b) and Weber (2006) for reviews). It is a stretched out, highly α-helical protein with its hydrophobic N-terminus anchored in the membrane, and with its hydrophilic C-terminal domain reaching up to the top of the (αβ)3 binding to δ. In E. coli the fully stretched b is 190 Å, far longer than required to reach the ‘top’ of F1 (Weber, 2006). In line with this ‘excessive’ length, the E. coli FOF1 was demonstrated to tolerate a deletion of up to 11 amino acid residues (shortening by 16 Å) or an insertion of up to 14 residues (lengthening by 20 Å) with retention of activity (Sorgen et al., 1998; Sorgen et al., 1999). A recent structural study in the laboratory of Stanley Dunn (Del Rizzo et al., 2006) revealed that in E. coli the dimerization domain of the ATP synthase subunit b forms an atypical parallel two-stranded coiled-coil. This right-handed coiled-coil structure has intrinsic asymmetry: the two helices are offset rather than in register. It is conceivable that other bacterial (and probably chloroplast) FOF1 have a similar structure for the peripheral stalk. In E. coli FOF1 some mutations in the b subunit, for example bG9D (Jans et al., 1984; Porter et al., 1985), impair proton translocation, despite the fact that binding of F1 to FO was unaffected. Second site revertants of the bG9D mutation were found in position a240 (Kumamoto and Simoni, 1986), implying that subunit b might influence the proper functioning of subunit a. B. Proton Transport through FO : Functional Aspects 1. Hypothetical Rotary Mechanism of Proton Translocation As the rotary catalysis in F1 was gaining experimental support, several hypothetical mechanisms of rotary proton translocation were suggested. The common feature in all mechanisms proposed is the rotation of the c-ring oligomer relative to the a-subunit (Vik and Antonio, 1994; Junge et al., 1997; Elston et al., 1998; Dimroth et al., 1999). A fundamental question is: how can ion flow perpendicular to the membrane plane drive the rotation of the c-ring in this plane? An answer based on two assumptions was proposed
481
by one of us (WJ) in Junge et al. (1997). The first assumption was that the c-ring performs stochastic rotational diffusion relative to the a-subunit. This diffusion is restricted by electrostatic constraint; the carboxyl of the conserved acidic residue in cTMH2 may be deprotonated (i.e., negatively charged) when facing the a-subunit, but must be protonated (i.e., electrically neutral) when facing the lipid core because of electrostatic penalty for exposure of a charged group to the hydrophobic environment. The second assumption was that there are two non-collinear proton access half-channels from each side of the membrane leading to the carboxyl in cTMH2 facing the a subunit, giving rise to chirality. In such a model the c-subunit with a deprotonated carboxyl cannot leave the interface of a-subunit, and the rotational diffusion of the c-ring is constrained to a narrow angular domain. When a proton comes from one of the half-channels and neutralizes the negative charge on the respective c-subunit, this constraint is relieved from one side, so that the cring can eventually rotate one step forward during its rotational fluctuations. The net direction of rotation is then determined by the probabilities of the cTMH2 carboxyl protonation/deprotonation in each halfchannel. In turn, these probabilities are dependent on µ~H+ on each side of the membrane. Thus FO acts as an entropic (i.e., proton counting) rotary machine; see Junge et al. (1997), Elston et al. (1998) and Dimroth et al. (1999), for details. Two modifications to the above simple model of rotary ion translocation have been proposed (see Junge and Nelson (2005) and references therein). (i) Fillingame and colleagues proposed a model where the c-ring was not rotating as a stiff entity but with obligatory swiveling of the hairpin-helices when in contact with subunit a (Fillingame et al., 2000a). This model was based on seemingly conflicting structural data obtained on the one hand by the solution-NMR of the c-subunit monomer and on the other by cysteinemapping of the c-ring (Fillingame et al., 2000b). The recent crystal structure of the c-ring from I. tartaricus (Meier et al., 2005a), however, deviates from the one inferred from solution NMR of E. coli subunit c. Because the arrangement of helices is now compatible with that deduced from cysteine-mapping there is no more need to invoke a swiveling mechanism. Moreover, the above mentioned experiments on the relocation of the conserved Arg from aTMH4 and Glu/Asp from cTMH2 to other transmembrane helices of the respective subunits are compatible with the
482 simple rotary model rather than with those requiring swiveling helices. (ii) Another modification was proposed by Dimroth and colleagues, who have postulated a ‘one-channel model’ (Dimroth et al., 2000), as opposed to the above described ‘two-channel model.’ The experimental evidence in favor of the ‘one-channel model’ came from the NMR structure of the c monomer from P. modestum in detergent micelles (Matthey et al., 1999), and from several other lines of circumstantial evidence suggesting that on the c-ring the cTMH2 conserved acidic residues are located in the proximity to the cytoplasmatic surface of the membrane. Again this topology is not supported by the crystal structure (Meier et al., 2005a). As already mentioned, the high-resolution structure of the I. tartaricus c-ring (Meier et al., 2005a), as well as the structural model of the c-ring from yeast (Stock et al., 1999), indicate that the acidic residue in the cTMH2 is located approximately in the middle of the lipid bilayer. Cysteines have been introduced into TMHs of the E. coli a subunit in order to estimate the accessibility of the respective residues from the aqueous phases by their chemical modification with Ag+ and N-ethylmaleimide (Angevine and Fillingame, 2003; Angevine et al., 2003). The Ag+-sensitive residues in aTMH2, –4, and –5 were found to form a continuum extending from the periplasmic to the cytoplasmic side of the membrane. These residues cluster at the interior of a four-helix bundle formed by aTMH2-5 and probably account for the postulated two proton half-channels. The presence of highly conserved bulky hydrophobic residues in subunit a (Fig. 2B) might reflect the necessity to insulate the proton half-channels against each other. Such insulation might be universal and therefore highly conserved, while the nature of polar residues along the proton pathway is likely to be variable in enzymes optimized for different conditions (e.g., periplasmic pH, coupling ion, temperature). It is noteworthy that in the simple rotary model proton translocation can be driven with both the chemical and the electrical components of ∆µ~H+. If the half-channels were mono-specific for the proton, a voltage of 59 mV was expected to be equivalent to a pH-difference of one unit. Their non-equivalence, in other words a strict requirement of a certain transmembrane voltage (∆ψ ) for ATP synthesis, has been claimed by Dimroth for FOF1 from E. coli (Kaim and Dimroth, 1998a), chloroplasts (Kaim and Dimroth, 1999) and also for the Na+ transporting enzyme from
Boris A. Feniouk and Wolfgang Junge P. modestum (Kaim and Dimroth, 1998b; Kaim and Dimroth, 1999). However, the experiments performed by Fischer and Gräber on FOF1 incorporated in liposomes demonstrated that this ∆ψ is necessary for ATP synthesis only for the E. coli, but not for the chloroplast enzyme. The authors suggested that ∆ψ might be required not for the catalysis of ATP synthesis, but for activation of the enzyme (Fischer and Graber, 1999). Experiments using Rhodobacter (Rba.) capsulatus chromatophores, where the fraction of active FOF1 was measured as a function of ∆pH and ∆ψ, revealed that the magnitude of this active fraction depends on both ∆µ~H+ components. However, at high ∆pH the active FOF1 fraction cannot be increased further by increasing ∆pH, whereas elevation of ∆ψ increases this fraction (Turina et al., 1992). Subsequent experiments on Rba. capsulatus FOF1 pointed to a role of the conserved glutamate E219 in the aTMH5, in the position of E. coli aHis245 being important for discrimination between ∆ψ and ∆pH (Turina and Melandri, 2002). The regulation of the activity of the chloroplast enzyme by ∆µ~H+ was discovered long ago (Junge, 1970; Junge et al., 1970; Bakker-Grunwald and Van Dam, 1973; Graber et al., 1977), and in this chapter we devote the entire section III to this phenomenon. At this point we would like to emphasize that it is unclear whether ∆ψ is critically important for the ATP synthesis reaction or for the activation of FOF1. It is noteworthy that studies on the electric conductance of FO from Rba. capsulatus, when F1 was stripped off, gave no indication of any voltage threshold down to the level of 5 mV (Feniouk et al., 2004). The implications of this finding are two-fold: (i) the requirement for a certain transmembrane voltage for activity, or activation, may be a regulatory feature in certain organisms, and not a universal property of FOF1; (ii) if there is such a requirement it seems to be due to FOF1, but not to the FO-portion per se. 2. The Unitary Proton Conductance of FO and FOF1 as Determined in Membrane Vesicles of Purple Bacteria The unit conductance of FO for the proton is expected to be in the range of femto-Siemens (fS), that is outside of the range accessible by patch-clamp techniques (suited for detecting pico-Siemens levels of conductance). The photosynthetic membranes of purple bacteria and of plants have two unique properties for quantitative studies on the electric conductance of FO
Chapter 24
Proton Translocation in ATP Synthase
483
Fig. 3. A. Chromatophores prepared from Rba. capsulatus by sonication have an average diameter of ~30 nm; only one out of three contains one molecule of ATP synthase per vesicle. In response to a short flash of light photosynthetic reaction centers generate transmembrane voltage and release quinol (QH2); subsequent oxidation of QH2 by the cytochrome bc1 complex is coupled to proton pumping ~ + drives protons through the ATP synthase and powers ATP synthesis. B. Electrochromic into the chromatophore lumen. The resultant ∆µ H signals recorded at 522 nm reflect changes in the transmembrane voltage (∆ψ). Flash excitation leads to a sharp step in ∆ψ due to charge separation in the reaction centers, followed by further increase due to electrogenic proton pumping by the cytochrome bc1 complex. Simultaneously membrane discharge due to various ion flows (including proton transport through ATP synthase) starts. Addition of ADP and Pi to intact chromatophores markedly increases the rate of this ∆ψ decay. The difference between the trace recorded before and after addition of ADP and Pi reflects proton transfer through ATP synthase coupled to ATP synthesis. C. Same experimental approach as applied to chromatophores with F1 removed by EDTA treatment. In this case, the chromatophore membrane is initially leaky due to futile proton flow via FO lacking F1. This leak can be specifically blocked by FO inhibitors (e.g., oligomycin); the difference between the trace before and after inhibitor addition reflects proton transport through FO. See Feniouk et al. (2002) and Feniouk et al. (2004) for experimental details. See also Color Plate 13, Fig. 21.
and FOF1. They intrinsically contain both a molecular voltmeter with sub-nanosecond time resolution,
owing to the electrochromism of intrinsic pigments, and a voltage generator with sub-nanosecond rise
484 time, which are the photochemical reaction centers (Junge and Witt, 1968; Emrich et al., 1969; Jackson and Crofts, 1971). A jump of the transmembrane voltage can be generated in less than one nanosecond by excitation with a short laser flash, and the relaxation of the transmembrane voltage can be followed spectroscopically. These properties of photosynthetic membranes are superior to those of the coupling membranes of mitochondria or E. coli that lack the inbuilt voltmeter, and where a voltage jump can be induced only by invasive techniques like mixing in substrate, or inducing a salt jump in the suspension. Thus, most quantitative studies were carried out with chromatophores of purple bacteria and with thylakoid membranes of chloroplasts. Because the leak conductance for protons is low, and with specific inhibitors of FO (e.g., venturicidin or DCCD) at hand, the specific conductance of chloroplast CFO (Althoff et al., 1989; Schönknecht et al., 1986) and CFOF1 (Junge, 1987) has been inferred from the observed voltage relaxation (see Junge, 2004 for a review) using information on the area-specific density of reaction centers and the membrane capacitance. The electric conductance of CFO was extremely proton specific (a rejection factor of >107 against Na+ and K+). Despite this specificity, the conductance was only slightly pH-dependent (Althoff et al., 1989), and was greater than 10 fS or about 6000 protons per second at a driving force of 100 mV (Schönknecht et al., 1986). There was, however, one serious shortcoming in these experiments, which was that the proportion of exposed FO that was actually conducting was not determined rigorously. If this proportion was low, then the unitary conductance must be higher. The above ambiguity was eventually resolved by using membrane vesicles of purple bacteria (Fig. 3). The preparation and the physical properties of chromatophores are well characterized: the mean diameter of the vesicles is 30–60 nm (Saphon et al., 1975a; Packham et al., 1978; Feniouk et al., 2002); the electric capacitance is 1.1 µF cm–2 (Packham et al., 1978). The proton permeability is 0.2 µm s–1, and the pH buffer properties of the inner chromatophore volume can be modeled with 2 buffer groups with pK values of 3.6 and 6.7 at concentrations of 250 and 24 mM respectively (Turina et al., 1990). The number of bacteriochlorophylls and reaction centers per vesicle is 1000–5000 and 10–50, respectively (Saphon et al., 1975a; Packham et al., 1978). Sub-reactions of ∆µ~H+ generation by the photosynthetic reaction centers and the cytochrome bc1 complexes are also described in
Boris A. Feniouk and Wolfgang Junge detail (Crofts and Wraight, 1983; Jackson, 1988). An important advantage of the chromatophore preparations, which is lacking in preparations from chloroplast membranes, is their stability. In the absence of oxygen, Rba. capsulatus chromatophores were shown to perform photophosphorylation at a constant rate for 10 days (Larreta-Garde and Thomas, 1985). To solve the above-mentioned ambiguity about the number of active FO molecules per vesicle, chromatophores were prepared to such a small size (diameter 30 nm) that they contained only about 0.3 copies of FOF1 on the average (Fig. 3A). This allowed the determination of the unitary conductance in the subset of vesicles containing a single copy of FO (Feniouk et al., 2004). These mini-chromatophores contained the full complement of the primary processes of photosynthesis, i.e., about 10 copies of the bacterial reaction centers and the cytochrome bc1 complexes. A jump of the transmembrane voltage (magnitude about 70 mV) was generated by flashing the chromatophores with a xenon flash, and the decay of the voltage was monitored spectrophotometrically by the above-mentioned electrochromic absorption changes of intrinsic pigments (Fig. 3B, C). The relaxation time of the voltage was about 2.2 ms in vesicles with FO, and by several orders of magnitude longer after the proton conduction by FO was specifically blocked. The calculated unitary proton conductance of Rba. capsulatus FO is 10 fS which, at a voltage of 100 mV, implies 6240 protons transported per second. This figure matches the lower limit of the values previously determined for FO from chloroplasts (Schönknecht et al., 1986), implying that all CFO that are exposed by removal of CF1 retain their ability to conduct protons. 3. Mechanistic Implications Experiments with mini-chromatophores reveal an Ohmic conductance in the voltage domain between 5 and 70 mV. The proton specificity is extremely high; at pH 8 and against a background of 100 mM of K+ or Na+, the apparent rejection factor is 107. Despite the very high specificity, the pH-dependence is low. In a wide range of pH from 6.5 to 10, the conductance varies only by a factor of two (Feniouk et al., 2004). How does the conductance of FO compare with the conductance of gramicidin A, which has been determined at low pH, ranging from zero to four units (Cukierman, 2000)? Extrapolated to pH 8, the expected proton conductance of gramicidin A is only
Chapter 24
Proton Translocation in ATP Synthase
1 fS, which is tenfold lower than that of FO, which is astounding in the light of the required rotation in the latter. How does the conductance of FO compare with the conductance of the coupled enzyme, FOF1? The typical rate of ATP synthesis under high protonmotive force (e.g., at 200 mV) is 300 s–1 or 100 rps. With a C10-symmetrical c-ring the rate of proton transport is 1000 s–1, that is an order of magnitude less than the expected rate of proton transport by free FO (>10000 s–1 at 200mV). Thus FO operates as a low-impedance electro-motor for the chemical generator, F1. How do these data compare with the above described Brownian ratchet model for proton conduction by a rotary mechanism? A simulation of the observed behavior by statistical mechanics in terms of the rotary transport model required only three free parameters, two different pK-values in the respective access channels (6.1 and 10) and a parameter to describe the energy profile across the membrane. The least to say is that the rotary model is compatible with the observed conductance of FO (Feniouk et al., 2004). The weak pH-dependence between these two pK values follows from this simulation. The proton transfer rate is determined by the probability of proton binding to a relay group at the opening of one halfchannel and proton release from a relay group in the other half-channel. When the pH is the same on both sides of the membrane (as was in our experiments), the product probability is constant in the interval between the calculated pK values. In energized membranes, on the other hand, the pK values of the relay groups tend to match the ambient pH at the cytoplasm and periplasm sides of the coupling membrane. This allows higher turnover numbers of FO even against back pressure (Cherepanov et al., 1999). It is tempting to attribute these two relay groups to particular amino acids. The group with a pK of 10 is likely to be the conserved arginine of aTMH4 itself (i.e., the E. coli Arg210), which indeed is located close to the cytoplasmic side of the membrane. It was demonstrated in E. coli FOF1 that Arg210 is readily accessible to hydrophilic molecules such as N-ethylmaleimide (Angevine and Fillingame, 2003). This suggestion is indirectly supported by the loss of rotary proton translocation upon mutation of Arg210 to alanine (Valiyaveetil and Fillingame, 1997). For the group with a pK of 6, the conceivable candidates are the highly conserved Glu or His of subunit a (i.e., in E. coli Glu219 and His245). It is noteworthy that in FOF1 from alkaliphilic bacteria there is a lysine that
485
precedes the conservative glutamine corresponding to E. coli Glu219, and a glycine instead of the conservative histidine, corresponding to the E. coli His245 (Wang et al., 2004). Since such a combination was demonstrated to be a special feature required for non-fermentative growth and oxidative phosphorylation of alkaliphilic bacteria (Wang et al., 2004), it is probable that these residues serve to elevate the pK of the proton relay group A to match the high periplasmic pH. In line with the speculation above, in acidophilic bacteria from Heliobacter and Campylobacter genera there is a pair of Asp residues in positions corresponding to the E. coli a218-219, and a Leu instead of the conserved E. coli His245. C. Coupling of Proton Transport to ATP Synthesis/Hydrolysis 1. Elastic Coupling Hypothesis The structure of FOF1 with 10-11-14 and 15 copies of subunit c in FO and a three-fold symmetry in F1 implies a variable proton per ATP ratio ranging from 3.33 to 5. An obvious solution for the mechanism to tolerate this variability with either symmetry-match or -mismatch is an elastic power transmission between the two rotary stepper motors, FO and F1 (Junge et al., 1997; Cherepanov et al., 1999; Pänke and Rumberg, 1999). The contour composed of subunit γ, α3β3-hexamer, subunits b and a, and the c-ring might function as a torsional spring, storing the energy provided by sequential proton translocation trough FO by its compliance. Experimental evidence for elastic energy transmission was obtained in single-molecule rotation experiments on E. coli F1 (Panke et al., 2001). The normal operation of this double engine is such that FO represents a load for F1 which is driven by ATP hydrolysis, or vice versa. A simulation of the turnover of the behavior of F1 driving a heavy load has revealed that the turnover rate increases with increasing compliance of the power transmission (Cherepanov and Junge, 2001; Panke et al., 2001). Increasing the compliance straightens the free energy profile of the driving stepper motor as seen by the driven one. This straightening increases the kinetic efficiency, which is defined as the turnover rate over the maximum possible one; see Fig. 5 in Cherepanov and Junge (2001) and Panke et al. (2001). Ongoing studies by one of us (WJ) aiming at the torsional spring constant of the E. coli FOF1 by fluctuation analysis revealed a number of about 20 pN.nm.rad–1, serving
486 the purpose of high kinetic efficiency quite well. 2. Coupling Efficiency If the coupling efficiency between proton transport and ATP synthesis was not perfect, a greater proton per ATP ratio than that expected for a perfect coupling would result. In the case of the chloroplast enzyme for which the most extended set of studies has been presented, structural AFM data with 14 copies of subunit c in the ring suggest a H+/ATP ratio of 4.67 under perfect coupling conditions. Instead, a ratio of 4 was the ‘résumé’ of extensive studies by three groups, see van Walraven et al., (1996), recently joined by a fourth one (Turina et al., 2003). This is an open discrepancy. The least to conclude is that the chloroplast enzyme, if offered saturating nucleotide concentration, is perfectly coupled and it does not slip. The same result was obtained in single-molecule studies with bacterial F1 and FOF1 where the enzyme drove actin filaments against a viscous drag. Here, the energy output (torque times angle) matched the free energy input of ATP hydrolysis if the torque was determined by using the compliance of a rotating actin filament as in Panke et al. (2001b). In related studies the torque was calculated using the viscosity of the bulk fluid; this perfect match (although being claimed as such) was not obtained because the rotating filament of a few µM length, moving 10 nm or less over ground, experiences a greater viscosity than in the bulk due to flow-coupling to the surface, and this was neglected (Noji et al., 1997; Yasuda et al., 1998). It is noteworthy that even in the absence of nucleotides and phosphate the enzyme is well coupled. In Rba. capsulatus chromatophores there is no detectable proton translocation through FOF1 in response to excitation with flashing light, indicating that the enzyme does not ‘leak’ under these conditions (Feniouk et al., 2001). This implies that when the γεcn complex (n being the copy-number of subunit c in the ring) is fixed in the F1 portion, no proton translocation occurs even when a pulse of protonmotive force over 100 mV is applied. The physiologically tight coupling between FO and F1 can be loosened by several factors. The chloroplast enzyme, for instance, leaks protons if the nucleotide concentration drops below 1 µM (Groth and Junge, 1993). This has been interpreted to indicate that the lack of nucleotide binding loosens the grip of (αβ)3 on the central rotary shaft, so that the c-ring of FO is
Boris A. Feniouk and Wolfgang Junge free-wheeling. In Rba. capsulatus chromatophores the absence of nucleotides per se does not induce a leak (Feniouk et al., 2001), and application of high ∆µ~H+ for several seconds is additionally required for its induction (Feniouk et al., 2005). It is noteworthy that this induction is accompanied with release of the tightly bound ADP from the enzyme. Once such a ‘slip’ is induced the enzyme is no longer tightly coupled, and passive proton translocation occurs at a high rate if ∆µ~H+ is applied. It is likely that the cause of ‘slip’ induction is ∆µ~H+-driven rotation of the γεcn complex that forces all the catalytic sites on the α3β3 hexamer into an ‘open’ conformation; thus, subunit γ is no longer fixed in the F1, as in the structure of the nucleotide-free α3β3 hexamer from Bacillus PS3 (Shirakihara et al., 1997), and therefore no longer limits the rotation of the c-ring and proton transport. An obvious pre-requisite for tight coupling between ATP hydrolysis and proton translocation is sufficiently strong binding between subunits both within the rotor and the stator that excludes ‘slipping’ during rotary catalysis. This includes interactions between α3β3 and the other stator subunits (Weber, 2006) as well as between the c-ring and the γε complex. Studies on bacteria indicated that these interactions are affected by Mg2+ (Weber et al., 2002; Weber et al., 2004), Ca2+ (Pick and Weiss, 1988; Casadio and Melandri, 1996), sulfite (Cappellini et al., 1997) and pH (Weber et al., 2002). In Rba. capsulatus chromatophores it was demonstrated that uncoupling between ATP hydrolysis and ATP-driven proton pumping is induced if ADP or Pi are absent from the medium (Turina et al., 2004). Under physiological conditions, however, the coupling seems to be perfect in all organisms studied. III. Role of Proton Translocation in the Regulation of ATP Synthase An interesting regulatory mechanism found in ATP synthase is the so-called activation of ATP hydrolysis by ∆µ~H+, or increase in ATPase activity of FOF1 after short membrane energization (Carmeli and Lifshitz, 1972; Baltscheffsky and Lundin, 1979; Turina et al., 1992; Galkin and Vinogradov, 1999; Fischer et al., 2000; Zharova and Vinogradov, 2004). As reviewed in Feniouk and Junge (2005), the activation by ∆µ~H+ occurs due to release of tightly bound Mg-ADP that inhibits ATPase activity of FOF1 when bound in a high-affinity catalytic site in the absence of phos-
Chapter 24
Proton Translocation in ATP Synthase
phate (Smith et al., 1983; Drobinskaya et al., 1985; Milgrom and Boyer, 1990; Hyndman et al., 1994). Single-molecule experiments on Bacillus PS3 α3β3γ complex revealed that the mechanical rotation of subunit γ by external force relieves inhibition by ADP (Hirono-Hara et al., 2005). Therefore, the cause of ~ ~ ∆µ H+ induced ADP release is probably the ∆µH+ powered rotation of the γεcn complex. It is noteworthy that in the chloroplast ATP synthase at least, the ∆µ~H+ necessary for the activation is higher than the ∆µ~H+ thermodynamically necessary for phosphorylation (Junge, 1970; Hangarter et al., 1987). Recent experiments on FOF1 from Paracoccus denitrificans (Zharova and Vinogradov, 2004) and from Bacillus PS3 (Feniouk et al., 2007) demonstrated another regulatory role of ∆µ~H+. It was revealed that ~ ∆µ H+ not only relieved the enzyme from inhibition by ADP, but also supported the steady-state ATPase activity in the presence of ADP. The latter effect was observed only in the presence of Pi (in contrast to the described above activation by ∆µ~H+). We have proposed earlier that such Pi dependent stimulation of the steady-state ATPase activity might be due to ~ ∆µ H+- induced increase in the affinity of FOF1 to Pi (Feniouk and Junge, 2005). The ADP-inhibition of ATPase activity occurs only if ADP is bound at the catalytic site without Pi. Therefore, one could expect that the increase in Pi concentration, or in the affinity of FOF1 to Pi, would diminish inhibition by ADP and thereby stimulate the ATPase activity. This suggestion is supported by the results obtained on bacterial, chloroplast and mitochondrial enzymes (Carmeli and Lifshitz, 1972; Melandri et al., 1975; Moyle and Mitchell, 1975; Dunham and Selman, 1981; Drobinskaya et al., 1985; Turina et al., 1992; Mitome et al., 2002; Zharova and Vinogradov, 2006; Feniouk et al., 2007). We have recently put forward a hypothesis that in bacterial FOF1 ∆µ~H+-driven rotation of the γεcn complex might have another regulatory role influencing the conformation of subunit ε (Feniouk and Junge, 2005). The latter small subunit is composed of two domains, the N-terminal β-barrel necessary for tight binding of subunit γ to the c-ring and the C-terminal α-helical domain that plays a regulatory role (see Feniouk et al., 2006 for a review). In FOF1 from E. coli and Bacillus PS3 the C-terminal domain was shown to undergo large conformational changes (Tsunoda et al., 2001b; Suzuki et al., 2003). When extended along the subunit γ towards α3β3, the Cterminal domain selectively inhibits ATP hydrolysis,
487
but not synthesis (Suzuki et al., 2003). When this domain is contracted in a hairpin structure near FO, neither activity is affected. Studies on Bacillus PS3 FOF1 revealed that the inhibitory effect of subunit ε is at least in part due to the stabilization of the ADP inhibited state of the enzyme (Feniouk et al., 2007), in line with our earlier speculation (Feniouk and Junge, 2005). It was proposed earlier that rotation of subunit γ in the hydrolysis direction induces the contracted conformation of subunit ε, while if rotation is stopped or reversed, the transition to the extended conformation occurs (Feniouk and Junge, 2005). In view of the recent experimental evidence, reviewed in Feniouk et al., (2006), it seems more probable that the conformation of the C-terminal domain of subunit ε is determined by the overall F1 conformation and by the angular position of subunit γ rather than by the direction of its rotation. In turn, this angular position is defined by the nucleotide composition of the medium, presence of inhibitors and ∆µ~H+. The probable ∆µ~H+ dependent regulatory mechanisms described above suggest that rotary proton translocation through FO serves not only as energy input for ATP synthesis, but also plays a regulatory role, relieving F1 from ADP inhibition and maintaining the active state of the F1 by preventing the re-inhibition by ADP. This helps bacteria to distinguish between a short-term decrease in the activity of the primary generators of ∆µ~H+ (e.g., due to decrease in the light intensity or in oxygen concentration) and severe de-energization of the cell due to damage of the coupling membrane. In the former case ATP-driven proton pumping by FOF1 is advantageous because it will restore ∆µ~H+ and cell motility so that the bacteria can move to a more favorable environment. In the latter case, on the contrary, ATP hydrolysis will merely deplete the ATP pool without any use, so an inhibitory mechanism is needed to prevent such a waste. Acknowledgments Studies of Rba. capsulatus FOF1 conducted by BF were supported by the Alexander von Humboldt Foundation. Stimulating discussions with Professors Andrei D. Vinogradov, Masasuke Yoshida and Toru Hisabori, and Drs. Dmitry Cherepanov, Armen Mulkidjanian, Paola Turina, Toshiharu Suzuki, Hiroki Konno and Yasuyuki Kato-Yamada are gratefully acknowledged. WJ acknowledges financial support
488 by the Volkswagen Foundation, the European Union and the Fonds der Chemie and wishes to thank his coworkers, in particular Prof. Siegfried Engelbrecht, for continuing discussion and support. References Altendorf K and Zitzmann W (1975) Identification of the DCCDreactive protein of the energy transducing adenosinetriphosphatase complex from Escherichia coli. FEBS Lett 59: 268–272 Althoff G, Lill H and Junge W (1989) Proton channel of the chloroplast ATP synthase, CFo: Its time-averaged single-channel conductance as function of pH, temperature, isotopic and ionic medium composition. J Membr Biol 108: 263–271 Angevine CM and Fillingame RH (2003) Aqueous access channels in subunit a of rotary ATP synthase. J Biol Chem 278: 6066–6074 Angevine CM, Herold KA and Fillingame RH (2003) Aqueous access pathways in subunit a of rotary ATP synthase extend to both sides of the membrane. Proc Natl Acad Sci USA 100: 13179–13183 Baccarini-Melandri A, Gest H and Pietro AS (1970) A coupling factor in bacterial photophosphorylation. J Biol Chem 245: 1224–1226 Bakker-Grunwald T and Van Dam K (1973) The energy level associated with the light-triggered Mg2+ -dependent ATPase in spinach chloroplasts. Biochim Biophys Acta 292: 808–814 Baltscheffsky M and Lundin A (1979) Flash-induced increase of ATPase activity in Rhodospirillum rubrum chromatophores. In: Mukohata Y, Packer L (eds) Cation Flux Across Biomembranes, pp 209-218. Academic Press, New York Birkenhager R, Hoppert M, Deckers-Hebestreit G, Mayer F and Altendorf K (1995) The F0 complex of the Escherichia coli ATP synthase. Investigation by electron spectroscopic imaging and immunoelectron microscopy. Eur J Biochem 230: 58–67 Cain BD and Simoni RD (1988) Interaction between Glu-219 and His-245 within the a-subunit of FOF1-ATPase in Escherichia coli. J Biol Chem 263: 6606–6612 Cappellini P, Turina P, Fregni V and Melandri BA (1997) Sulfite stimulates the ATP hydrolysis activity of but not proton translocation by the ATP synthase of Rhodobacter capsulatus ~ +. Eur J Biochem 248: and interferes with its activation by ∆µ H 496–506 Carmeli C and Lifshitz Y (1972) Effects of P 1 and ADP on ATPase activity in chloroplasts. Biochim Biophys Acta 267: 86–95 Casadio R and Melandri BA (1996) CaATP inhibition of the MgATP-dependent proton pump (H+-ATPase) in bacterial photosynthetic membranes with a mechanism of alternate substrate inhibition. J Bioinorg Chem 1: 284–291 Cherepanov DA and Junge W (2001) Viscoelastic dynamics of actin filaments coupled to rotary F-ATPase: Curvature as an indicator of the torque. Biophys J 81: 1234–1244 Cherepanov DA, Mulkidjanian AY and Junge W (1999) Transient accumulation of elastic energy in proton translocating ATP synthase. FEBS Lett 449: 1–6 Collinson IR, Runswick MJ, Buchanan SK, Fearnley IM, Skehel JM, van Raaij MJ, Griffiths DE and Walker JE (1994) Fo membrane domain of ATP synthase from bovine heart mitochondria:
Boris A. Feniouk and Wolfgang Junge Purification, subunit composition, and reconstitution with F1-ATPase. Biochemistry 33: 7971–7978 Crofts AR and Wraight CA (1983) The electrochemical domain of photosynthesis. Biochim Biophys Acta 726: 149–185 Cukierman S (2000) Proton mobilities in water and in different stereoisomers of covalently linked gramicidin A channels. Biophys J 78: 1825–1834 Del Rizzo PA, Bi Y and Dunn SD (2006) ATP synthase b subunit dimerization domain: A right-handed coiled coil with offset helices. J Mol Biol 364: 735–746 DeLeon-Rangel J, Zhang D and Vik SB (2003) The role of transmembrane span 2 in the structure and function of subunit a of the ATP synthase from Escherichia coli. Arch Biochem Biophys 418: 55–62 Dickson VK, Silvester JA, Fearnley IM, Leslie AG and Walker JE (2006) On the structure of the stator of the mitochondrial ATP synthase. EMBO J 25: 2911–2918 Diez M, Zimmermann B, Borsch M, Konig M, Schweinberger E, Steigmiller S, Reuter R, Felekyan S, Kudryavtsev V, Seidel CA and Graber P (2004) Proton-powered subunit rotation in single membrane-bound FOF1-ATP synthase. Nat Struct Mol Biol 11: 135–141 Dimroth P (1994) Bacterial sodium ion-coupled energetics. Antonie Van Leeuwenhoek 65: 381–395 Dimroth P, Wang H, Grabe M and Oster G (1999) Energy transduction in the sodium F-ATPase of Propionigenium modestum. Proc Natl Acad Sci USA 96: 4924–4929 Dimroth P, Matthey U, and Kaim G (2000) Critical evaluation of the one- versus the two-channel model for the operation of the ATP synthase’s F(o) motor. Biochim Biophys Acta 1459: 506–513 Dmitriev OY and Fillingame RH (2001) Structure of Ala(20) → Pro/Pro(64) → Ala substituted subunit c of Escherichia coli ATP synthase in which the essential proline is switched between transmembrane helices. J Biol Chem 276: 27449–27454 Drobinskaya IY, Kozlov IA, Murataliev MB and Vulfson EN (1985) Tightly bound adenosine diphosphate, which inhibits the activity of mitochondrial F1-ATPase, is located at the catalytic site of the enzyme. FEBS Lett 182: 419–424 Duncan TM, Bulygin VV, Zhou Y, Hutcheon ML and Cross RL (1995) Rotation of subunits during catalysis by Escherichia coli F1-ATPase. Proc Natl Acad Sci USA 92: 10964–10968 Dunham KR and Selman BR (1981) Interactions of inorganic phosphate with spinach coupling factor 1. Effects on ATPase and ADP binding activities. J Biol Chem 256: 10044–10049 Dunn SD, McLachlin DT and Revington M (2000a) The second stalk of Escherichia coli ATP synthase. Biochim Biophys Acta 1458: 356–363 Dunn SD, Revington M, Cipriano DJ and Shilton BH (2000b) The b subunit of Escherichia coli ATP synthase. J Bioenerg Biomembr 32: 347–355 Elston T, Wang H, and Oster G (1998) Energy transduction in ATP synthase. Nature 391: 510–513 Emrich HM, Junge W and Witt HT (1969) Further evidence for an optical response of chloroplast bulk pigments to a light induced electrical field in photosynthesis. Z Naturforsch B 24: 1144–1146 Feng Y and McCarty RE (1990) Purification and reconstitution of active chloroplast F0. J Biol Chem 265: 5104–5109 Feniouk BA and Junge W (2005) Regulation of the F0F1-ATP synthase: The conformation of subunit ε might be deter-
Chapter 24
Proton Translocation in ATP Synthase
mined by directionality of subunit γ rotation. FEBS Lett 579: 5114–5118 Feniouk BA, Cherepanov DA, Junge W and Mulkidjanian AY (2001) Coupling of proton flow to ATP synthesis in Rhodobacter capsulatus: FOF1-ATP synthase is absent from about half of chromatophores. Biochim Biophys Acta 1506: 189–203 Feniouk BA, Cherepanov DA, Voskoboynikova NE, Mulkidjanian AY and Junge W (2002) Chromatophore vesicles of Rhodobacter capsulatus contain on average one FOF1-ATP synthase each. Biophys J 82: 1115–1122 Feniouk BA, Kozlova MA, Knorre DA, Cherepanov DA, Mulkidjanian AY and Junge W (2004) The proton driven rotor of ATP synthase: Ohmic conductance (10 fS), and absence of voltage gating. Biophys J 86: 4094–4109 Feniouk BA, Mulkidjanian AY and Junge W (2005) Proton slip in the ATP synthase of Rhodobacter capsulatus: induction, proton conduction, and nucleotide dependence. Biochim Biophys Acta 1706: 184–194 Feniouk BA, Suzuki T and Yoshida M (2006) The role of subunit epsilon in the catalysis and regulation of F0F1-ATP synthase. Biochim Biophys Acta 1757: 326–338 Feniouk BA, Suzuki T and Yoshida M (2007) Regulatory Interplay between proton motive force, ADP, phosphate, and subunit-epsilon in bacterial ATP synthase. J Biol Chem 282: 764–772 Fillingame RH (1975) Identification of the dicyclohexylcarbodiimide-reactive protein component of the adenosine 5´triphosphate energy-transducing system of Escherichia coli. J Bacteriol 124: 870–883 Fillingame RH and Dmitriev OY (2002) Structural model of the transmembrane Fo rotary sector of H+-transporting ATP synthase derived by solution NMR and intersubunit cross-linking in situ. Biochim Biophys Acta 1565: 232–245 Fillingame RH, Jiang W and Dmitriev OY (2000a) Coupling H(+) transport to rotary catalysis in F-type ATP synthases: Structure and organization of the transmembrane rotary motor. J Exp Biol 203: 9–17 Fillingame RH, Jiang W, Dmitriev OY and Jones PC (2000b) Structural interpretations of FO rotary function in the Escherichia coli FOF1 ATP synthase. Biochim Biophys Acta 1458: 387–403 Fischer S and Graber P (1999) Comparison of ∆pH- and ∆φ-driven ATP synthesis catalyzed by the H+-ATPases from Escherichia coli or chloroplasts reconstituted into liposomes. FEBS Letters 457: 327–332 Fischer S, Graber P and Turina P (2000) The activity of the ATP synthase from Escherichia coli is regulated by the transmembrane proton motive force. J Biol Chem 275: 30157–30162 Galkin MA and Vinogradov AD (1999) Energy-dependent transformation of the catalytic activities of the mitochondrial FOF1-ATP synthase. FEBS Lett 448: 123–126 Gao YQ, Yang W and Karplus M (2005) A structure-based model for the synthesis and hydrolysis of ATP by F1-ATPase. Cell 123: 195–205 Gibbons C, Montgomery MG, Leslie AG, and Walker JE (2000) The structure of the central stalk in bovine F1-ATPase at 2.4 Å resolution. Nat Struct Biol 7: 1055–1061 Girvin ME and Fillingame RH (1994) Hairpin folding of subunit c of F1Fo ATP synthase: 1H distance measurements to nitroxidederivatized aspartyl-61. Biochemistry 33: 665–674 Graber P, Schlodder E and Witt HT (1977) Conformational change of the chloroplast ATPase induced by a transmembrane electric
489
field and its correlation to phosphorylation. Biochim Biophys Acta 461: 426–440 Groth G and Junge W (1993) Proton slip of chloroplast ATPase: Its nucleotide dependence, energetic threshold and relation to an alternating site mechanism of catalysis. Biochemistry 32: 8103–8111 Gumbiowski K, Panke O, Junge W and Engelbrecht S (2002) Rotation of the c subunit oligomer in EF0EF1 mutant cD61N. J Biol Chem 277: 31287–31290 Hangarter RP, Grandoni P and Ort DR (1987) The effects of chloroplast coupling factor reduction on the energetics of activation and on the energetics and efficiency of ATP formation. J Biol Chem 262: 13513–13519 Häsler K, Engelbrecht S and Junge W (1998) Three-stepped rotation of subunits gamma and epsilon in single molecules of F-ATPase as revealed by polarized, confocal fluorometry. FEBS Lett 426: 301–304 Häsler K, Pänke O and Junge W (1999) On the stator of rotary ATP synthase: the binding strength of subunit δ to (αβ)3 as determined by fluorescence correlation spectroscopy. Biochemistry 38: 13759–13765 Hatch LP, Cox GB and Howitt SM (1995) The essential arginine residue at position 210 in the a subunit of the Escherichia coli ATP synthase can be transferred to position 252 with partial retention of activity. J Biol Chem 270: 29407–29412 Hatch LP, Cox GB and Howitt SM (1998) Glutamate residues at positions 219 and 252 in the a-subunit of the Escherichia coli ATP synthase are not functionally equivalent. Biochim Biophys Acta 1363: 217-223 Hirono-Hara Y, Ishuzuka K, Kinosita K, Yoshida M and Noji H (2005) Activation of pausing F-1 motor by external force. Proc Natl Acad Sci USA 102: 4288–4293 Hisabori T, Kondoh A and Yoshida M (1999) The gamma subunit in chloroplast F1-ATPase can rotate in a unidirectional and counter-clockwise manner. FEBS Lett 463: 35–38 Hoppe J and Sebald W (1984) The proton conducting F0-part of bacterial ATP synthases. Biochim Biophys Acta 768: 1-27 Howitt SM and Cox GB (1992) Second-site revertants of an arginine-210 to lysine mutation in the a subunit of the FOF1ATPase from Escherichia coli: Implications for structure. Proc Natl Acad Sci USA 89: 9799–9803 Hutcheon ML, Duncan TM, Ngai H and Cross RL (2001) Energy-driven subunit rotation at the interface between subunit a and the c oligomer in the F0 sector of Escherichia coli ATP synthase. Proc Natl Acad Sci USA 98: 8519–8524 Hyndman DJ, Milgrom YM, Bramhall EA and Cross RL (1994) Nucleotide-binding sites on Escherichia coli F1-ATPase. Specificity of noncatalytic sites and inhibition at catalytic sites by MgADP. J Biol Chem 269: 28871–28877 Itoh H, Takahashi A, Adachi K, Noji H, Yasuda R, Yoshida M and Kinosita K, Jr. (2004) Mechanically driven ATP synthesis by F1-ATPase. Nature 427: 465–468 Jackson JB (1988) Bacterial photosynthesis. In: Anthony C (ed) Bacterial Energy Transduction, pp 317–376. Academic Press, London Jackson JB and Crofts AR (1971) The kinetics of light induced carotenoid changes in Rhodopseudomonas sphaeroides and their relation to electrical field generation across the chromatophore membrane. Eur J Biochem 18: 120–130 Jans DA, Fimmel AL, Hatch L, Gibson F and Cox GB (1984) An additional acidic residue in the membrane portion of the
490 b-subunit of the energy-transducing adenosine triphosphatase of Escherichia coli affects both assembly and function. Biochem J 221: 43–51 Jault JM and Allison WS (1993) Slow binding of ATP to noncatalytic nucleotide binding sites which accelerates catalysis is responsible for apparent negative cooperativity exhibited by the bovine mitochondrial F1-ATPase. J Biol Chem 268: 1558–1566 Jault JM, Matsui T, Jault FM, Kaibara C, Muneyuki E, Yoshida M, Kagawa Y and Allison WS (1995) The α3β3γ complex of the F1-ATPase from thermophilic Bacillus PS3 containing the αD261N substitution fails to dissociate inhibitory MgADP from a catalytic site when ATP binds to noncatalytic sites. Biochemistry 34: 16412–16418 Jiang W, Hermolin J and Fillingame RH (2001) The preferred stoichiometry of c subunits in the rotary motor sector of Escherichia coli ATP synthase is 10. Proc Natl Acad Sci USA 98: 4966–4971 Jones PC, Hermolin J, Jiang W, and Fillingame RH (2000) Insights into the rotary catalytic mechanism of FOF1 ATP synthase from the cross-linking of subunits b and c in the Escherichia coli enzyme. J Biol Chem 275: 31340–31346 Junge W (1970) Critical electric potential difference for photophosphorylation. Its relation to the chemiosomotic hypothesis and to the triggering requirements of the ATPase system. Eur J Biochem 14: 582–592 Junge W (1987) Complete tracking of transient proton flow through active chloroplast ATP synthase. Proc Natl Acad Sci USA 84: 7084–7088 Junge W (2004) Protons, proteins and ATP. Photosynth Res 80: 197–221 Junge W and Nelson N (2005) Structural biology. Nature’s rotary electromotors. Science 308: 642–644 Junge W and Witt HT (1968) On the ion transport system of photosynthesis–investigations on a molecular level. Z Naturforsch 23: 244–254 Junge W, Rumberg B and Schroder H (1970) The necessity of an electric potential difference and its use for photophosphorylation in short flash groups. Eur J Biochem 14: 575–581 Junge W, Lill H and Engelbrecht S (1997) ATP synthase: An electrochemical transducer with rotatory mechanics. Trends Biochem Sci 22: 420–423 Kaim G and Dimroth P (1998a) ATP synthesis by the F1Fo ATP synthase of Escherichia coli is obligatorily dependent on the electric potential. FEBS Lett 434: 57–60 Kaim G and Dimroth P (1998b) Voltage-generated torque drives the motor of the ATP synthase. EMBO J 17: 5887–5895 Kaim G and Dimroth P (1999) ATP synthesis by F-type ATP synthase is obligatorily dependent on the transmembrane voltage. EMBO J 18: 4118–4127 Kaim G, Prummer M, Sick B, Zumofen G, Renn A, Wild UP, and Dimroth P (2002) Coupled rotation within single F0F1 enzyme complexes during ATP synthesis or hydrolysis. FEBS Lett 525: 156–163 Kato-Yamada Y, Noji H, Yasuda R, Kinosita K, Jr. and Yoshida M (1998) Direct observation of the rotation of epsilon subunit in F1-ATPase. J Biol Chem 273: 19375–19377 Kayalar C, Rosing J and Boyer PD (1977) An alternating site sequence for oxidative phosphorylation suggested by measurement of substrate binding patterns and exchange reaction inhibitions. J Biol Chem 252: 2486–2491
Boris A. Feniouk and Wolfgang Junge Kinosita K, Jr., Adachi K and Itoh H (2004) Rotation of F1-ATPase: How an ATP-driven molecular machine may work. Annu Rev Biophys Biomol Struct 33: 245–268 Kumamoto CA and Simoni RD (1986) Genetic evidence for interaction between the a and b subunits of the F0 portion of the Escherichia coli proton translocating ATPase. J Biol Chem 261: 10037–10042 Larreta-Garde V and Thomas D (1985) Influence of light on longterm ADP phosphorylation. J Biol Chem 260: 6060–6062 Lightowlers RN, Howitt SM, Hatch L, Gibson F and Cox GB (1987) The proton pore in the Escherichia coli FOF1-ATPase: A requirement for arginine at position 210 of the a-subunit. Biochim Biophys Acta 894: 399–406 Lightowlers RN, Howitt SM, Hatch L, Gibson F and Cox G (1988) The proton pore in the Escherichia coli FOF1-ATPase: Substitution of glutamate by glutamine at position 219 of the alpha-subunit prevents FO-mediated proton permeability. Biochim Biophys Acta 933: 241–248 Long JC, Wang S and Vik SB (1998) Membrane topology of subunit a of the FOF1 ATP synthase as determined by labeling of unique cysteine residues. J Biol Chem 273: 16235–16240 Long JC, DeLeon-Rangel J and Vik SB (2002) Characterization of the first cytoplasmic loop of subunit a of the Escherichia coli ATP synthase by surface labeling, cross-linking, and mutagenesis. J Biol Chem 277: 27288–27293 Matsui T, Muneyuki E, Honda M, Allison WS, Dou C and Yoshida M (1997) Catalytic activity of the alpha3beta3gamma complex of F1-ATPase without noncatalytic nucleotide binding site. J Biol Chem 272: 8215–8221 Matthey U, Kaim G, Braun D, Wuthrich K and Dimroth P (1999) NMR studies of subunit c of the ATP synthase from Propionigenium modestum in dodecylsulphate micelles. Eur J Biochem 261: 459–467 Meier T, Matthey U, Henzen F, Dimroth P and Müller DJ (2002) The central plug in the reconstituted undecameric c cylinder of a bacterial ATP synthase consists of phospholipids. FEBS Lett 505: 353–356 Meier T, Matthey U, von Ballmoos C, Vonck J, Krug vN, Kühlbrandt W and Dimroth P (2003) Evidence for structural integrity in the undecameric c-rings isolated from sodium ATP synthases. J Mol Biol 325: 389–397 Meier T, Polzer P, Diederichs K, Welte W and Dimroth P (2005a) Structure of the rotor ring of F-Type Na+-ATPase from Ilyobacter tartaricus. Science 308: 659–662 Meier T, Yu J, Raschle T, Henzen F, Dimroth P and Müller DJ (2005b) Structural evidence for a constant c11 ring stoichiometry in the sodium F-ATP synthase. FEBS J 272: 5474–5483 Melandri AB, Fabbri E, and Melandri BA (1975) Energy transduction in photosynthetic bacteria. VIII. Activation of the energy-transducing ATPase by inorganic phosphate. Biochim Biophys Acta 376: 82–88 Melandri BA, Baccarini-Melandri A, San Pietro A and Gest H (1971) Interchangeability of phosphorylation coupling factors in photosynthetic and respiratory energy conversion. Science 174: 514–516 Milgrom YM and Boyer PD (1990) The ADP that binds tightly to nucleotide-depleted mitochondrial F1- ATPase and inhibits catalysis is bound at a catalytic site. Biochim Biophys Acta 1020: 43–48 Milgrom YM, Ehler LL and Boyer PD (1990) ATP binding at noncatalytic sites of soluble chloroplast F1-ATPase is required
Chapter 24
Proton Translocation in ATP Synthase
for expression of the enzyme activity. J Biol Chem 265: 18725–18728 Milgrom YM, Ehler LL and Boyer PD (1991) The characteristics and effect on catalysis of nucleotide binding to noncatalytic sites of chloroplast F1-ATPase. J Biol Chem 266: 11551–11558 Miller MJ, Oldenburg M and Fillingame RH (1990) The essential carboxyl group in subunit c of the F1F0 ATP synthase can be moved and H+-translocating function retained. Proc Natl Acad Sci USA 87: 4900–4904 Mitome N, Ono S, Suzuki T, Shimabukuro K, Muneyuki E and Yoshida M (2002) The presence of phosphate at a catalytic site suppresses the formation of the MgADP-inhibited form of F1-ATPase. Eur J Biochem 269: 53–60 Mitome N, Suzuki T, Hayashi S and Yoshida M (2004) Thermophilic ATP synthase has a decamer c-ring: indication of noninteger 10:3 H+/ATP ratio and permissive elastic coupling. Proc Natl Acad Sci USA 101: 12159–12164 Moyle J and Mitchell P (1975) Active/inactive state transitions of mitochondrial ATPase molecules influenced by Mg2+, anions and aurovertin. FEBS Lett 56: 55–61 Noji H, Yasuda R, Yoshida M and Kinosita K, Jr. (1997) Direct observation of the rotation of F1-ATPase. Nature 386: 299–302 Noji H, Häsler K, Junge W, Kinosita K, Jr., Yoshida M and Engelbrecht S (1999) Rotation of Escherichia coli F1-ATPase. Biochem Biophys Res Commun 260: 597–599 Oberfeld B, Brunner J and Dimroth P (2006) Phospholipids occupy the internal lumen of the c ring of the ATP synthase of Escherichia coli. Biochemistry 45: 1841–1851 Omote H, Sambonmatsu N, Saito K, Sambongi Y, Iwamoto-Kihara A, Yanagida T, Wada Y and Futai M (1999) The γ-subunit rotation and torque generation in F1-ATPase from wild- type or uncoupled mutant Escherichia coli. Proc Natl Acad Sci USA 96: 7780–7784 Packham NK, Berriman JA and Jackson JB (1978) The charging capacitance of the chromatophore membrane. FEBS Lett 89: 205–210 Pänke O and Rumberg B (1999) Kinetic modeling of rotary CF0F1-ATP synthase: storage of elastic energy during energy transduction. Biochim Biophys Acta 1412: 118–128 Pänke O, Gumbiowski K, Junge W and Engelbrecht S (2000) FATPase: Specific observation of the rotating c subunit oligomer of EF0EF1. FEBS Lett 472: 34–38 Pänke O, Cherepanov DA, Gumbiowski K, Engelbrecht S and Junge W (2001) Viscoelastic dynamics of actin filaments coupled to rotary F-ATPase: Angular torque profile of the enzyme. Biophys J 81: 1220–1233 Pati S, Brusilow WS, Deckers-Hebestreit G and Altendorf K (1991) Assembly of the F0 proton channel of the Escherichia coli F1F0 ATPase: Low proton conductance of reconstituted F0 sectors synthesized and assembled in the absence of F1. Biochemistry 30: 4710–4714 Patterson AR, Wada T and Vik SB (1999) His(15) of subunit a of the Escherichia coli ATP synthase is important for the structure or assembly of the membrane sector FO. Arch Biochem Biophys 368: 193–197 Pick U and Weiss M (1988) A light-dependent dicyclohexylcarbodiimide-sensitive Ca-ATPase activity in chloroplasts which is not coupled to proton translocation. Eur J Biochem 173: 623–628 Pogoryelov D, Yu J, Meier T, Vonck J, Dimroth P and Muller DJ (2005) The c15 ring of the Spirulina platensis F-ATP synthase:
491
F1/F0 symmetry mismatch is not obligatory. EMBO Rep 6: 1040–1044 Ponomarenko S, Volfson I and Strotmann H (1999) Proton gradient-induced changes of the interaction between CFO and CF1 related to activation of the chloroplast ATP synthase. FEBS Lett 443: 136–138 Porter AC, Kumamoto C, Aldape K and Simoni RD (1985) Role of the b subunit of the Escherichia coli proton-translocating ATPase. A mutagenic analysis. J Biol Chem 260: 8182–8187 Rastogi VK and Girvin ME (1999) Structural changes linked to proton translocation by subunit c of the ATP synthase. Nature 402: 263–268 Rondelez Y, Tresset G, Nakashima T, Kato-Yamada Y, Fujita H, Takeuchi S and Noji H (2005) Highly coupled ATP synthesis by F1-ATPase single molecules. Nature 433: 773–777 Sabbert D, Engelbrecht S and Junge W (1996) Intersubunit rotation in active F-ATPase. Nature 381: 623–625 Sabbert D, Engelbrecht S and Junge W (1997) Functional and idling rotatory motion within F1-ATPase. Proc Natl Acad Sci USA 94: 4401–4405 Sambongi Y, Iko Y, Tanabe M, Omote H, Iwamoto-Kihara A, Ueda I, Yanagida T, Wada Y and Futai M (1999) Mechanical rotation of the c subunit oligomer in ATP synthase (F0F1): direct observation. Science 286: 1722–1724 Saphon S, Jackson JB, Lerbs V and Witt HT (1975a) The functional unit of electrical events and phosphorylation in chromatophores from Rhodopseudomonas sphaeroides. Biochim Biophys Acta 408: 58–66 Saphon S, Jackson JB and Witt HT (1975b) Electrical potential changes, H+ translocation and phosphorylation induced by short flash excitation in Rhodopseudomonas sphaeroides chromatophores. Biochim Biophys Acta 408: 67–82 Schneider E and Altendorf K (1985) All three subunits are required for the reconstitution of an active proton channel (F0) of Escherichia coli ATP synthase (F1F0). EMBO J 4: 515–518 Schönknecht G, Junge W, Lill H and Engelbrecht S (1986) Complete tracking of proton flow in thylakoids-the unit conductance of CF0 is greater than 10 fS. FEBS Lett 203: 289–294 Schulenberg B, Wellmer F, Lill H, Junge W and Engelbrecht S (1997) Cross-linking of chloroplast F0F1-ATPase subunit ε to γ without effect on activity. ε and γ are parts of the rotor. Eur J Biochem 249: 134–141 Seelert H, Poetsch A, Dencher NA, Engel A, Stahlberg H and Muller DJ (2000) Structural biology. Proton-powered turbine of a plant motor. Nature 405: 418–419 Senior AE, Nadanaciva S and Weber J (2002) The molecular mechanism of ATP synthesis by F1F0-ATP synthase. Biochim Biophys Acta 1553: 188–211 Shirakihara Y, Leslie AG, Abrahams JP, Walker JE, Ueda T, Sekimoto Y, Kambara M, Saika K, Kagawa Y and Yoshida M (1997) The crystal structure of the nucleotide-free α3β3 subcomplex of F1-ATPase from the thermophilic Bacillus PS3 is a symmetric trimer. Structure 5: 825–836 Singh S, Turina P, Bustamante CJ, Keller DJ and Capaldi RA (1996) Topographical structure of membrane-bound Escherichia coli F1F0 ATP synthase in aqueous buffer. FEBS Lett 397: 30–34 Smith LT, Rosen G and Boyer PD (1983) Properties of ATP tightly bound to catalytic sites of chloroplast ATP synthase. J Biol Chem 258: 10887–10894 Sone N, Yoshida M, Hirata H and Kagawa Y (1979) Carbodiimidebinding protein of H+-translocating ATPase and inhibition
492 of H+ conduction by dicyclohexylcarbodiimide. J Biochem (Tokyo) 85: 503–509 Sorgen PL, Caviston TL, Perry RC and Cain BD (1998) Deletions in the second stalk of F1F0-ATP synthase in Escherichia coli. J Biol Chem 273: 27873–27878 Sorgen PL, Bubb MR and Cain BD (1999) Lengthening the second stalk of F1F0 ATP synthase in Escherichia coli. J Biol Chem 274: 36261–36266 Stahlberg H, Muller DJ, Suda K, Fotiadis D, Engel A, Meier T, Matthey U and Dimroth P (2001) Bacterial Na(+)-ATP synthase has an undecameric rotor. EMBO Rep 2: 229–233 Stalz WD, Greie JC, Deckers-Hebestreit G and Altendorf K (2003) Direct interaction of subunits a and b of the F0 complex of Escherichia coli ATP synthase by forming an ab2 subcomplex. J Biol Chem 278: 27068–27071 Steffens K, Hoppe J and Altendorf K (1988) F0 part of the ATP synthase from Escherichia coli. Influence of subunits a, and b, on the structure of subunit c. Eur J Biochem 170: 627–630 Stock D, Leslie AG and Walker JE (1999) Molecular architecture of the rotary motor in ATP synthase. Science 286: 1700–1705 Suzuki T, Ueno H, Mitome N, Suzuki J and Yoshida M (2002) F0 of ATP synthase is a rotary proton channel: Obligatory coupling of proton translocation with rotation of c-subunit ring. J Biol Chem 277: 13281–13285 Suzuki T, Murakami T, Iino R, Suzuki J, Ono S, Shirakihara Y and Yoshida M (2003) F0F1-ATPase/synthase is geared to the synthesis mode by conformational rearrangement of epsilon subunit in response to proton motive force and ADP/ATP balance. J Biol Chem 278: 46840–46846 Takeyasu K, Omote H, Nettikadan S, Tokumasu F, Iwamoto-Kihara A and Futai M (1996) Molecular imaging of Escherichia coli F0F1-ATPase in reconstituted membranes using atomic force microscopy. FEBS Lett 392: 110–113 Tanabe M, Nishio K, Iko Y, Sambongi Y, Iwamoto-Kihara A, Wada Y and Futai M (2001) Rotation of a complex of the γ subunit and c ring of Escherichia coli ATP synthase. The rotor and stator are interchangeable. J Biol Chem 276: 15269–15274 Tsunoda SP, Aggeler R, Yoshida M and Capaldi RA (2001a) Rotation of the c subunit oligomer in fully functional F1F0 ATP synthase. Proc Natl Acad Sci USA 98: 898–902 Tsunoda SP, Rodgers AJ, Aggeler R, Wilce MC, Yoshida M and Capaldi RA (2001b) Large conformational changes of the ε subunit in the bacterial F1F0 ATP synthase provide a ratchet action to regulate this rotary motor enzyme. Proc Natl Acad Sci USA 98: 6560–6564 Turina P, Giovannini D, Gubellini F and Melandri BA (2004) Physiological ligands ADP and Pi modulate the degree of intrinsic coupling in the ATP synthase of the photosynthetic bacterium Rhodobacter capsulatus. Biochemistry 43: 11126–11134 Turina P and Melandri BA (2002) A point mutation in the ATP synthase of Rhodobacter capsulatus results in differential contributions of ∆pH and ∆φ in driving the ATP synthesis reaction. Eur J Biochem 269: 1984–1992 Turina P, Rumberg B, Melandri BA and Graber P (1992) Activation of the H+-ATP synthase in the photosynthetic bacterium Rhodobacter capsulatus. J Biol Chem 267: 11057–11063 Turina P, Samoray D, and Graber P (2003) H+/ATP ratio of proton transport-coupled ATP synthesis and hydrolysis catalysed by CF0F1-liposomes. EMBO J 22: 418–426 Turina P, Venturoli G, and Melandri BA (1990) Evaluation of the buffer capacity and permeability constant for protons in
Boris A. Feniouk and Wolfgang Junge chromatophores from Rhodobacter capsulatus. Eur J Biochem 192: 39–47 Ueno H, Suzuki T, Kinosita K, Jr. and Yoshida M (2005) ATPdriven stepwise rotation of F0F1-ATP synthase. Proc Natl Acad Sci USA 102: 1333–1338 Valiyaveetil FI and Fillingame RH (1997) On the role of Arg210 and Glu-219 of subunit a in proton translocation by the Escherichia coli F0F1-ATP synthase. J Biol Chem 272: 32635–32641 Valiyaveetil FI and Fillingame RH (1998) Transmembrane topography of subunit a in the Escherichia coli F1F0 ATP synthase. J Biol Chem 273: 16241–16247 van Walraven HS, Strotmann H, Schwarz O, and Rumberg B (1996) The H+/ATP coupling ratio of the ATP synthase from thiol-modulated chloroplasts and two cyanobacterial strains is four. FEBS Lett 379: 309–313 Velours J, Paumard P, Soubannier V, Spannagel C, Vaillier J, Arselin G and Graves PV (2000) Organisation of the yeast ATP synthase F0: A study based on cysteine mutants, thiol modification and cross-linking reagents. Biochim Biophys Acta 1458: 443–456 Vik SB and Antonio BJ (1994) A mechanism of proton translocation by F1F0 ATP synthases suggested by double mutants of the a subunit. J Biol Chem 269: 30364–30369 Vik SB, Cain BD, Chun KT and Simoni RD (1988) Mutagenesis of the α subunit of the F1F0-ATPase from Escherichia coli. Mutations at Glu-196, Pro-190, and Ser-199. J Biol Chem 263: 6599–6605 Vik SB, Long JC, Wada T, and Zhang D (2000) A model for the structure of subunit a of the Escherichia coli ATP synthase and its role in proton translocation. Biochim Biophys Acta 1458: 457–466 Wada T, Long JC, Zhang D and Vik SB (1999) A novel labeling approach supports the five-transmembrane model of subunit a of the Escherichia coli ATP synthase. J Biol Chem 274: 17353–17357 Wang Z, Hicks DB, Guffanti AA, Baldwin K and Krulwich TA (2004) Replacement of amino acid sequence features of a- and c-subunits of ATP synthases of alkaliphilic Bacillus with the Bacillus consensus sequence results in defective oxidative phosphorylation and non-fermentative growth at pH 10.5. J Biol Chem 279: 26546–26554 Watts SD and Capaldi RA (1997) Interactions between the F1 and F0 parts in the Escherichia coli ATP synthase. Associations involving the loop region of c subunits. J Biol Chem 272: 15065–15068 Watts SD, Tang C and Capaldi RA (1996) The stalk region of the Escherichia coli ATP synthase. Tyrosine 205 of the subunit is in the interface between the F1 and F0 parts and can interact with both the ε and c oligomer. J Biol Chem 271: 28341–28347 Weber H, Junge W, Hoppe J and Sebald W (1986) Laser-activated carbene labels the same residues in the proteolipid subunit of the ATP synthase in energized and nonenergized chloroplasts and mitochondria. FEBS Lett 202: 23–26 Weber J (2006) ATP synthase: Subunit-subunit interactions in the stator stalk. Biochim Biophys Acta 1757: 1162–1170 Weber J, Bowman C, Wilke-Mounts S and Senior AE (1995) α-Aspartate 261 is a key residue in noncatalytic sites of Escherichia coli F1-ATPase. J Biol Chem 270: 21045–21049 Weber J and Senior AE (2003) ATP synthesis driven by proton transport in F1F0-ATP synthase. FEBS Lett 545: 61–70
Chapter 24
Proton Translocation in ATP Synthase
Weber J, Wilke-Mounts S, Grell E and Senior AE (1994) Tryptophan fluorescence provides a direct probe of nucleotide binding in the noncatalytic sites of Escherichia coli F1-ATPase. J Biol Chem 269: 11261–11268 Weber J, Wilke-Mounts S and Senior AE (2002) Quantitative determination of binding affinity of δ-subunit in Escherichia coli F1-ATPase: effects of mutation, Mg2+, and pH on Kd. J Biol Chem 277: 18390–18396 Weber J, Wilke-Mounts S, Nadanaciva S and Senior AE (2004) Quantitative determination of direct binding of b subunit to F1 in Escherichia coli F1F0-ATP synthase. J Biol Chem 279: 11253–11258. Yasuda R, Noji H, Kinosita K and Yoshida M (1998) F1-ATPase is a highly efficient molecular motor that rotates with discrete 120° steps. Cell 93: 1117–1124 Yoshida M, Muneyuki E and Hisabori T (2001) ATP synthase — a marvellous rotary engine of the cell. Nat Rev Mol Cell Biol 2: 669–677 Zhang D and Vik SB (2003) Helix packing in subunit a of the
493
Escherichia coli ATP synthase as determined by chemical labeling and proteolysis of the cysteine-substituted protein. Biochemistry 42: 331–337 Zhang Y and Fillingame RH (1995) Subunits coupling H+ transport and ATP synthesis in the Escherichia coli ATP synthase. Cys-Cys cross-linking of F1 subunit epsilon to the polar loop of F0 subunit c. J Biol Chem 270: 24609–24614 Zharova TV and Vinogradov AD (2004) Energy-dependent transformation of F0F1-ATPase in Paracoccus denitrificans plasma membranes. J Biol Chem 279: 12319–12324 Zharova TV and Vinogradov AD (2006) Energy-linked binding of Pi is required for continuous steady-state proton-translocating ATP hydrolysis catalyzed by FOF1 ATP synthase. Biochemistry 45: 14552–14558 Zimmermann B, Diez M, Zarrabi N, Graber P and Borsch M (2005) Movements of the ε-subunit during catalysis and activation in single membrane-bound H+-ATP synthase. EMBO J 24: 2053–2063
Chapter 25 Proton-Translocating Transhydrogenase in Photosynthetic Bacteria J. Baz Jackson* and U. Mirian Obiozo School of Biosciences, University of Birmingham, Edgbaston, Birmingham, B15 2TT, U.K.
Summary ............................................................................................................................................................... 495 I. Introduction..................................................................................................................................................... 496 II. An Overview of the Main Structural Features of Transhydrogenase.............................................................. 496 III. Distribution of Transhydrogenase Among Species ........................................................................................ 496 IV. Phylogenetic Relationships between Transhydrogenases from Different Species ........................................ 498 V. The Function of Transhydrogenase in Photosynthetic Bacteria ..................................................................... 500 VI. The Mechanism of Coupling Between Hydride Transfer and Proton Translocation in Transhydrogenase ... 501 A. Proton Translocation Through Transhydrogenase ........................................................................... 501 B. The Hydride Transfer Site of Transhydrogenase ............................................................................. 501 C. Gating at the Hydride-Transfer Site ................................................................................................. 503 1. Gating the Hydride-Transfer Reaction .................................................................................... 503 2. Gating the NADPH Release .................................................................................................... 503 D. The Mobile Loop of Transhydrogenase ........................................................................................... 504 E. The Alternating-Sites Mechanism for Transhydrogenase ................................................................ 504 VII. Conformational Changes in the Coupling Reactions of Transhydrogenase................................................... 504 Acknowledgments ................................................................................................................................................. 505 References ............................................................................................................................................................ 506
Summary Transhydrogenase uses the proton-motive force across a membrane to drive the reduction of NADP+ by NADH. It is found in the inner mitochondrial membrane of animal cells and in the cytoplasmic membrane of bacteria, including many photosynthetic bacteria. The enzyme has three components: dI, which binds NADH, and dIII, which binds NADP+, protrude from the membrane, whereas dII spans the membrane. Within this organization, the polypeptide arrangement varies among species. The Rhodospirillum (Rsp.) rubrum transhydrogenase was the first to be shown to have a dissociable dI component, a finding that opened up a new approach in studies on the kinetics and the structure of the enzyme. A complex of dI and dIII from Rsp. rubrum transhydrogenase catalyzes very fast hydride transfer from NADH to NADP+. Crystal structures of the complex show how the dihydronicotinamide ring of the NADH and the nicotinamide ring of the NADP+ are brought together to effect this redox reaction. They also indicate that the dihydronicotinamide ring of the NADH can move within its binding site. This is thought to gate the hydride-transfer reaction and thus prevent a slip in the coupling mechanism. Short polypeptide loops in dI and dIII are seen to bind and to position the (dihydro)nicotinamide rings. Longer loops regulate nucleotide and solvent access to the hydride-transfer site. Asymmetry in the structure of the dI-dIII complex suggests that the intact enzyme may operate by an alternating-site mechanism in which one monomer runs 180° out-of-phase with the other. *Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 495–508. © 2009 Springer Science + Business Media B.V.
496
J. Baz Jackson and U. Mirian Obiozo
I. Introduction Figure 1 shows that, like the ATP synthase, transhydrogenase is a consumer of the proton-motive force (∆p). It uses the energy made available from ‘downhill’ proton translocation to drive the reduction of NADP+ by NADH. Thus, as was first described some 40 years ago (Keister and Yike, 1967), we can observe a light-accelerated transhydrogenation reaction in chromatophore membranes from Rsp. rubrum (and other photosynthetic bacteria) that is abolished upon addition of protonophores (uncouplers). The Keister and Yike paper was published at about the same time as descriptions were appearing of a similar transhydrogenase driven by the ∆p generated by respiratory electron transport in the membranes of animal mitochondria and Escherichia (E.) coli. In the 1970s, Fisher and colleagues (Fisher and Guillory, 1971) discovered that, unlike other transhydrogenases known at the time, the enzyme from Rsp. rubrum could be resolved into a ‘water-soluble factor’ and an ‘insoluble membrane component.’ When the structural basis for this became clear (Cunningham et al., 1992; Williams et al., 1994) (see below), it established the Rsp. rubrum transhydrogenase as an excellent experimental system. There is now substantially more thermodynamic, kinetic and particularly structural information on the Rsp. rubrum enzyme than there is on the transhydrogenases from any other species. The purpose of this chapter is to review recent advances in our understanding of transhydrogenase, focusing mainly on the structural information that has emerged on the Rsp. rubrum protein. The proton-translocation reaction and the redox reaction of transhydrogenase are coupled together by conformational changes in the protein. The enzyme has to obey ‘specificity rules’ (Jenks, 1983) that gate the protonation and the redox steps. These features demand a highly intricate mechanism of action involving numerous intermediate steps, which, even now, are barely understood. Earlier reviews giving more detailed descriptions of transhydrogenase structure and mechanism are available (Bizouarn et al., 2000; Jackson et al., 2002, 2005; Jackson, 2003).
Abbreviations: E. – Eschericia; NAD(H) – both the oxidized, NAD+, and reduced, NADH, form of the nucleotide; Nic+ – nicotinamide; NicH – dihydronicotinamide; Rba. – Rhodobacter; Rsp. – Rhodospirillum
Fig. 1. Transhydrogenase in the chemiosmotic proton circuit of photosynthetic bacteria.
II. An Overview of the Main Structural Features of Transhydrogenase Gene sequences indicate that, although the polypeptide composition varies among species (Fig. 2), the overall structure of different transhydrogenases is always very similar. The enzyme has three components. The dI component (~350 amino acid residues), which binds NADH, and the dIII component (~200 residues), which binds NADP+, both protrude from the membrane into the bacterial cytoplasm. The proton-conducting dII component (~400 residues) spans the membrane. Transhydrogenase is thought to be a ‘dimer’ of two dI-dII-dIII ‘monomers.’ There is no high-resolution structure of the intact enzyme but Xray structures of a redox-active complex of the dI and dIII components of Rsp. rubrum transhydrogenase (described below) suggest the organization shown in Fig. 3 (Cotton et al., 2001). III. Distribution of Transhydrogenase Among Species The distribution of transhydrogenases in living organisms is really quite bewildering. Available genome sequences indicate that within both the Eubacterial and the Eukaryal kingdoms the enzyme is present
Chapter 25
Transhydrogenase
Fig. 2. Variation in the polypeptide composition of transhydrogenases from different species: representatives of the four groups. Vertical lines delineate the dI, dII and dIII components (see text and Fig. 3); vertical gaps delineate the polypeptides. The vertical alignment corresponds to sequence similarity. The gray shade indicates the transmembrane regions. The dashed line on the Entamoeba histolytica enzyme illustrates that the C-terminus of dIII and the N-terminus of dI are fused. PntAA, PntAB and PntB are the designated names of the three polypeptides of Rsp. rubrum transhydrogenase (Williams et al., 1994) and other members of the 3P group. The group notation, at the right, is defined in the text.
in some species but absent from others. It may be entirely absent from the Archaea. The transhydrogenase distribution in the photosynthetic bacteria does not yet form a recognizable pattern (Table 1). The current indications are that the enzyme is frequently, perhaps universally, present in the purple nonsulfur and the purple sulfur bacteria. There are insufficient data to draw conclusions about the Chloroflexi and
497 Chlorobi but clearly transhydrogenase is absent from some species within these two groups. Transhydrogenase is present in some species of the Cyanobacteria but absent from others. Remarkably, it is present in Synechococcus elongatus PCC7942 but absent from an organism in the same genus, Synechococcus sp. Ja-3-3B´Ab. Also of interest is the observation that transhydrogenase is present in at least some species of algae (including representatives from the Chlorophyta and Rhodophyta) but its gene is absent from the three currently available genome sequences (both nuclear and plastid) of plants. Table 1 also illustrates the fact that transhydrogenase is present in some fungi (Neurospora crassa) but not others (Saccharomyces cerevisiae), and present in some insects (mosquito) but not others (Drosophila). It is evident that, through evolution, some organisms have retained their transhydrogenase but other, often closely related organisms, have abandoned it. Transhydrogenase operates at the interface between the two major soluble redox cofactors of the cell (NAD and NADP) and the transmembrane protonmotive force, and it might, therefore, be thought of as an enzyme that is not to be trifled with. However, its patchy distribution shows that in some organisms the role performed by transhydrogenase can be
Fig. 3. Surface model of the dI2dIII1 complex of Rsp. rubrum transhydrogenase. The nomenclature of the domains and subunits is described in the text. The cleft separates dI.1 from dI.2. Note that dI.2(A) lies behind dI.2(B). The structure of the dI2dIII1 complex is shown with bound NADH and NADP+; X-ray structures, to date, have been obtained either as dead-end complexes or with analog nucleotides (for example, 2OO5.PDB). The dashed lines show the predicted disposition of other components in the intact enzyme.
498
J. Baz Jackson and U. Mirian Obiozo
Table 1. Distribution of transhydrogenases in photosynthetic organisms Archaea. TH genes are absent from all 4 of the complete genome sequences* of species in the Class, Halobacteria [and from all 28 of the complete genome sequences of non-photosynthetic archea]. α-Proteobacteria.
TH genes are present in all 10 of the complete genome sequences, and all 7 of the incomplete genome sequences, of the ‘Purple Non-Sulfur Bacteria.’ [Complete genome sequences from most, but not all, non-photosynthetic species in this Class have TH.] γ-Proteobacteria.
TH genes are present in the 2 complete genome sequences, and in the 2 incomplete genome sequences, of the ‘Purple Sulfur Bacteria.’ [Complete genome sequences from most, but not all, of the non-photosynthetic species in this Class have TH.] Chloroflexi. TH genes are absent from the 2 incomplete genome sequences of the ‘Green Non-Sulfur Bacteria.’ Chlorobi. TH genes are absent from the 3 complete genome sequences and from the 5 incomplete genome sequences. The incomplete genome sequence of Chlorobium phaeobacteroides, however, indicates part of a TH gene. Cyanobacteria. TH genes are absent from 3 of the complete genome sequences but present in all the 13 others and in 6 incomplete genome sequences. Algae. A TH gene is present in Acetabularia acetabulum and is indicated in ESTs from some other algae (Arkblad et al. 2001). Plantae. TH genes are absent from the 3 complete nuclear genome sequences, and from all incomplete nuclear genome sequences and ESTs. [TH genes are present in the genomes of some fungi (e.g. Neurospora crassa) and absent from others (e.g. Saccharomyces cerevisiae)] [TH genes are present in the genomes of nematodes, fish, amphibians and mammals. TH genes are present in some insects (Anopheles gambiae), and absent in others (Drosophila melanogaster).] TH = transhydrogenase; EST = expressed sequence tags; *completely sequenced genomes are those deposited in the NCBI genome database before September 2005 (non-photosynthetic organisms) and before September 2006 (photosynthetic organisms). Information in square brackets [ ] pertains to non-photosynthetic organisms.
satisfactorily carried out by other metabolic systems (see Section V). IV. Phylogenetic Relationships between Transhydrogenases from Different Species Figure 4 is a phylogenetic tree derived from transhydrogenase amino acid sequences which shows representative genera from both photosynthetic (underlined) and non-photosynthetic organisms (see Wilson et al. (2006) for a fuller picture of the non-photosynthetic species). As with many other proteins, the clade organization for the bacterial transhydrogenases differs greatly from that determined using 16S RNA sequences, indicating the importance of horizontal gene transfer in the species distribution of the enzyme. Thus, transhydrogenases from related genera do not group together; for example, the enzymes from the
two α-proteobacteria, Rhodobacter (Rba.) sphaeroides and Rsp. rubrum, reside in different clades, as do the enzymes from species in the cyanobacterial genera, Synechococcus and Synechocystis. Although the global structures of the transhydrogenases are similar (Fig. 3), there are four different types of polypeptide organization (Fig. 2). In some bacteria there are three polypeptides (the 3P group), in other bacteria there are two (the 2P group), and in the Eukaryota there is one (the 1P group). In the latter group, there are two ways in which the single polypeptide is constructed—there are the 1Pa transhydrogenases (found in the vertebrates, invertebrates, fungi and some protists) and the 1Pb transhydrogenases (found, thus far, in only a small group of protozoan parasites). In Fig. 4 the polypeptide group of the transhydrogenase from each of the representative genera is identified. Strikingly, the 1Pa, 1Pb, 2P and 3P
Chapter 25
Transhydrogenase
499
Fig. 4. An un-rooted phylogenetic tree of transhydrogenase amino acid sequences. Derived from 126 complete transhydrogenase sequences using methods described (Wilson et al., 2006). The sequences were from genomes of non-photosynthetic organisms available before September 2005, and from the genomes of photosynthetic organism available before September 2006. All genera for which there are photosynthetic species having the transhydrogenase gene are shown in the tree and are underlined. A few representative genera of non-photosynthetic species having transhydrogenase are also shown (more are illustrated in Wilson et al., (2006)). The assignment of a transhydrogenase to a 1Pa, 1Pb, 2P or 3P group (as defined in Fig. 2 and see text) is based on predictions of polypeptide composition from its gene sequence. Regions F1, F2 and F3 are defined in the text. Note that the length of a branch is not intended to correlate with the number of mutations along the branch.
groups precisely match the clades revealed by the phylogenetic analysis. To date, no transhydrogenase sequence has been found that violates this relationship (not shown) indicating that horizontal transfer of DNA has not disturbed the polypeptide groups, and suggesting an outline for the evolutionary path of the enzyme (Wilson et al., 2006). Thus, the ancestral transhydrogenase was a 3P enzyme. A mutation near F1 in Fig. 4 led to fusion of the C-terminus of
PntAA and the N-terminus of PntAB (see Fig. 2 for polypeptide nomenclature) and the consequent development of all members of the 2P group. Subsequent to this, two different fusions led to the emergence of the 1Pa group (at F2) and the 1Pb group (at F3) from ancestral 2P enzymes. This sequence of events would explain why analyses of amino acid sequences predict an extra transmembrane helix (TMH1) at the N-terminus of
500
J. Baz Jackson and U. Mirian Obiozo
Fig. 5. Predicted transmembrane helices in the dII component of transhydrogenase. The solid black lines show the predicted TM helices and their loops in dII of Rsp. rubrum transhydrogenase (Meuller and Rydstrom, 1999; Studley et al., 1999; Jackson et al., 2002). H1 and its linker are drawn in dashed lines to signify their absence from Rsp. rubrum and other members of the 3P group of transhydrogenases (as defined in Fig. 2). The helix numbering system is that of mammalian transhydrogenase (Yamaguchi et al., 1988): H5 is absent from all the bacterial enzymes. The ranges defined under the helix numbers correspond to the predictions for Rsp. rubrum. The disposition of invariant amino acid residues in helices and loops is indicated by the bold letters. The Rsp. rubrum equivalents of highly conserved residues are shown in gray. The figure was constructed using data from 126 complete transhydrogenase sequences (see legend to Fig. 4).
dII in 2P (and in 1Pa and 1Pb) transhydrogenases that is absent in the 3P enzymes (Wilson et al., 2006) (Fig. 5). Thus, the fusion at F1 (Fig. 4) between the PntAA and PntAB polypeptides of a 3P enzyme would have linked segments that lay on opposite sides of the membrane (cytoplasmic and periplasmic, respectively) and this, presumably, would have been favorable only if the TM linker were fairly hydrophobic. A similar argument explains the extra, predicted TM helix (TMH5) in 1Pa transhydrogenases that is absent in the 2P (and 3P) enzymes—again the polypeptide fusion (this time at F2) linked segments across the membrane. In contrast, the fusion (at F3) that led to the emergence of the 1Pb group did not produce an extra TMH because the linker is between dI and dIII which both lie on the cytoplasmic side of the membrane. V. The Function of Transhydrogenase in Photosynthetic Bacteria A long-held view is that the ∆p-driven transhydrogenase in bacteria provides NADPH for biosynthesis (Bragg et al., 1972; Gerolimatos and Hanson, 1978; Ambartsoumian et al., 1994). However, the genetic inactivation of transhydrogenase in Rba. sphaeroides leads to only a small decrease in the growth
rate of cells in succinate-based minimal medium in either aerobic or photosynthetic conditions (Hickman et al., 2002). Similar results are obtained with transhydrogenase-defective mutants of Rsp. rubrum (N. P. J. Cotton, T. Bizouarn and J. B. Jackson, unpublished). These observations probably reflect the fact that soluble metabolic enzymes such as glucose-6-phosphate dehydrogenase (G6PDH) and isocitrate dehydrogenase (ICDH) can also contribute to NADPH production for biosynthesis — see also Hanson and Rose (1980). An interesting recent finding is that the inactivation of transhydrogenase in Rba. sphaeroides causes an increased sensitivity to methanol poisoning in succinate-based aerobic cultures (Hickman et al., 2002). In the cell methanol is converted to formaldehyde, which is further metabolized by a reduced glutathione-dependent formaldehyde dehydrogenase. The increased methanol sensitivity in the transhydrogenase mutant was explained because the block in the NADPH supply would restrict the formation of reduced glutathione by glutathione reductase, and formaldehyde would then accumulate to toxic levels. The phenotype is not evident in glucose-grown cultures, presumably because the activity of G6PDH in the Entner-Doudoroff pathway provides an alternative means of NADPH production. The importance of transhydrogenase in providing NADPH for glu-
Chapter 25
Transhydrogenase
tathione reduction to minimize damage caused by reactive oxygen species has also been shown in the nematode, Caenorhabditis elegans (Arkblad et al., (2005) and see Oshino and Chance, (1977)). Equivalently perhaps, the inactivation of transhydrogenase in mice causes a defect in glucose-stimulated insulin release from pancreatic β-cells as a consequence of elevated levels of reactive oxygen species (Freeman et al., 2006). It remains puzzling as to why many cell types have two very different processes for reducing NADP+— through soluble dehydrogenases (like G6PDH and ICDH) and through transhydrogenase, an enzyme that is compulsorily linked to the proton-motive force. VI. The Mechanism of Coupling Between Hydride Transfer and Proton Translocation in Transhydrogenase A. Proton Translocation Through Transhydrogenase The organization of transmembrane helices in dII, based on biochemical experiments with the E. coli enzyme (Meuller and Rydstrom, 1999) and on predictive algorithms, is shown in Fig. 5. There is no high-resolution structure of dII from any species, and this model should probably be considered as fairly speculative. Despite a large number of mutagenesis experiments on E. coli transhydrogenase (Holmberg et al., 1994; Glavas et al., 1995; Yamaguchi and Hatefi, 1995a; Bragg and Hou, 1999; Hu et al., 1999; Yamaguchi et al., 2002; Yamaguchi and Stout, 2003), amino acid residues involved in the proton-transfer pathway through dII have not been convincingly identified. Some experiments have indicated the presence of a water cavity in dII (Bragg and Hou, 2000), and the importance of water molecules in proton transfer through proteins is now becoming widely recognized (Wraight, 2006). Given the global structure of the intact transhydrogenase (Fig. 3), it is significant that X-ray structures of dIII do not reveal a proton pathway between the dII-dIII interface and the hydride-transfer site (Prasad et al., 1999; White et al., 2000). There must be a transducer in dII, where proton-transfer reactions generate a conformational change that is then transmitted across dIII to the nucleotide-binding sites (some 30 Å). Conserved amino acid residues
501 are clustered in predicted TM helices 2 to 4, 9, 10, 13 and 14, mainly at their cytoplasmic ends and in their cytoplasmic loops (Fig. 5), suggesting perhaps that the transducer is located near the dII/dIII interface. Of particular interest is the fact that mutagenesis of several amino acid residues in these conserved regions of dII (of the E. coli enzyme) leads to changes in the binding affinity of NADP(H) (Holmberg et al., 1994; Glavas et al., 1995; Yamaguchi and Hatefi, 1995a; Bragg and Hou, 1999; Hu et al., 1999; Althage et al., 2001; Yamaguchi et al., 2002; Yamaguchi and Stout, 2003). It was suggested that protonation of the transducer from the outside aqueous phase induces a conformational change that blocks release of bound NADP+ and enables hydride transfer from bound NADH (Hutton et al., 1994; Jackson et al., 2002; Jackson, 2003; Jackson et al., 2005). Deprotonation of the transducer to the inside aqueous phase essentially reverses this conformational change, blocking hydride transfer and permitting nucleotide release. The switch in access of the transducer from the outside to the inside aqueous phase is effected by the redox state of the NADP(H). The nature of the conformational changes generated within the transducer is not known (Jackson et al., 2005). However, X-ray structures of the nucleotide-binding components of transhydrogenase have been revealing about the coupled conformational changes at the hydride-transfer site in dI-dIII (see below). A recent development which may help in the further investigation of the mechanism of action of the transducer is the observation that low concentrations of some metal ions, such as Zn2+ and Cd2+, strongly inhibit the binding/release of NADP+/NADPH, but have no effect either on the redox reaction of transhydrogenase or on the binding/release of NAD+/NADH (Whitehead et al., 2005). Since these metal ions specifically block proton-transfer steps in a number of other electron transport proteins (including bacterial reaction centers; Axelrod et al. (2000)), it was suggested that they might act on equivalent steps in transhydrogenase. This reinforces the view that proton translocation through the transducer is linked to changes in the binding energy of NADP+/NADPH. B. The Hydride Transfer Site of Transhydrogenase A simple mixture of recombinant dI and recombinant dIII from Rsp. rubrum transhydrogenase forms a tight dI2dIII1 complex which catalyses very fast hydride
502 transfer between bound NAD(H) and NADP(H) (Diggle et al., 1996; Venning et al., 1998). Recombinant dI and dIII from E. coli transhydrogenase form active complexes but with much lower affinity (Fjellstrom et al., 1997, 1999). Hybrid complexes of dI and dIII from different species have also been isolated and will catalyze the hydride-transfer reaction (Yamaguchi and Hatefi, 1995b; Peake et al., 1999; Fjellstrom et al., 1999; Wilson et al., 2006). The Rsp. rubrum dI2dIII1 complex has been crystallized in several nucleotide-bound states (Cotton et al., 2001; Singh et al., 2003; Sundaresan et al., 2003; Mather et al., 2004; Bhakta et al., 2006). The X-ray structure reveals the profound asymmetry of the complex (Fig. 3) — the dI(B) polypeptide is associated with a dIII to form a hydride-transfer site but the dI(A) does not have a dIII. NMR studies indicate that a second dIII can bind to the complex, presumably to dI(A), but only with very low affinity (Quirk et al., 1999a). A possible explanation for the asymmetry of the complex is discussed in section VI.E. Note, for now, that a second hydride-transfer site between dI(A) and a second dIII must be present in the intact enzyme. A recent structure with a bound, redox-inactive NADH analog in dI, and NADP+ in dIII, probably gives the best view of the hydride-transfer site at the dI/dIII interface of the complex (Bhakta et al., 2006) — see Fig. 3 and Fig. 8a. Hydride transfer is known to take place from C4 of the dihydro-
J. Baz Jackson and U. Mirian Obiozo nicotinamide (NicH) ring of the NADH to C4 of the nicotinamide (Nic+) ring of the NADP+. In the X-ray structure the planes of the rings are approximately parallel to one another; Fig. 6a shows how the rings are offset so that the 3-carboxamide group of the NADH stacks over the Nic+ ring of the NADP+ and the two C4 atoms are apposed at a distance of ~3.4 Å. Fig. 6b is a cartoon in which the rings are moved apart to illustrate the hydride-transfer reaction. The pro-R hydrogen on C4 of the dihydronicotinamide ring of NADH is transferred to the si-face of the nicotinamide ring of NADP+, corresponding to the ‘A-B’ stereochemistry that was experimentally determined many years ago (Lee et al., 1965). Quantum mechanical calculations with N-methyl derivatives of NicH and Nic+ show that the disposition of the two rings seen in the crystal structure of transhydrogenase is similar to that in a low-energy transition state for hydride transfer (Bhakta et al., 2006). The temperature dependence of the deuterium isotope effect on the first-order rate constant for hydride transfer in stopped-flow experiments suggested that there may be a tunneling contribution to the reaction but the possibility of conformational changes taking place during NADH binding (a ‘kinetic complexity’) make the analysis rather difficult (Venning et al., 1998).
Fig. 6. Hydride transfer in transhydrogenase. (a) The organization of the Nic+ ring of NADP+ and the tetrahydronicotinamide ring (H2NicH) of the NADH analogue, 1,4,5,6-tetrahydro-NAD, at the dI(B)-dIII interface of the 2.6 Å X-ray structure of the dI2dIII1 complex, 2OO5.pdb. C atoms, light gray; O atoms, black; N atoms, dark gray. The C4 atoms are labeled with asterisks. (b) A cartoon showing the hydride-transfer reaction between similarly organized (but slightly offset) Nic+ and NicH rings (thin and thick bonds, respectively). The pro-R H atom is transferred from C4 of the NicH to C4 on the si face of the Nic+.
Chapter 25
Transhydrogenase
Fig. 7. NADH in its proximal and distal conformations in the dI2dIII1 complex. The adenine rings of the tetrahydro-analog of NADH from the A and the B polypeptides of dI of 2OO5.pdb were superimposed. The nucleotide from B, shown in light gray, is in the proximal conformation — the C4 of its nicotinamide ring is 3.4 Å from that of the bound NADP+ (mid-gray and truncated at the pyrophosphate). The nucleotide from A, shown in dark gray, is in the distal conformation — when modelled against an equivalently docked dIII, the C4 of its nicotinamide ring would be too far from that of the bound NADP+ for hydride transfer. Similar results, but at a slightly lower resolution, are obtained by superimposing the NADH molecules from 1U2D.pdb.
C. Gating at the Hydride-Transfer Site Two specificity rules ensure that the redox reaction of transhydrogenase does not take place without proton translocation. (1) Hydride transfer must be blocked during the formation of the Michaelis complex of the enzyme, and then enabled at the appropriate point in the catalytic cycle. (2) At least one of the two nucleotide-substrates must be prevented from dissociating from the enzyme at the hydride-transfer step. The crystal structures of the Rsp. rubrum dI2dIII1 complex give clear indications as to how these two rules are obeyed. 1. Gating the Hydride-Transfer Reaction It appears that the hydride-transfer step is gated by movements of the NicH ring of NADH (Cotton et al., 2001; van Boxel et al., 2003; Mather et al., 2004; Bhakta et al., 2006; Brondijk et al., 2006). Each of the dI components of transhydrogenase has two protein domains, designated dI.1 and dI.2, and these are separated by a deep cleft (Fig. 3). The adenosine moiety of bound NADH is located in dI.2, the nucleotide pyrophosphate extends across the cleft and the NicH ring is located in dI.1 Thus, rotation of dI.1 against dI.2 causes a change in the conformation of the bound nucleotide. In dI(B) this brings the NicH of the NADH into a ‘proximal’ position relative to
503 the Nic+ ring of the NADP+ bound to dIII, forming the ground state for hydride transfer, as described above in Section VI.B. However, in dI(A) the two domains are oppositely rotated and the NicH of the NADH occupies a ‘distal’ position (Fig. 7). When this is modeled against the structure of dIII.NADP+, the C4 atoms of the NicH and Nic+ rings of the two nucleotides are too far apart (~ 4.6 Å) for hydride transfer to take place. It is proposed that adoption of the distal conformation in the Michaelis complex blocks premature hydride transfer. A cluster of invariant, charged/polar amino acid residues in the NicH binding loop of dI.1 (the socalled ‘RQD’ loop) appears to play an important role in stabilizing the distal and proximal conformations of the NADH. Mutation of these residues has a pronounced inhibitory effect on the rate of transhydrogenation (van Boxel et al., 2003; Brondijk et al., 2006). Figure 8a illustrates the extensive H-bond network between residues in the RQD loop and the NADH when the nucleotide occupies its proximal position, as in dI(B). In particular, the side chain of Arg127 lies deep in the cleft and forms H-bonds with the side chains of Asp135 and S138, and with one of the nucleotide pyrophosphates. However, when the nucleotide occupies its distal position, as in dI(A), the Arg127 side chain lies closer to the surface of the cleft and this set of H-bonds is broken. 2. Gating the NADPH Release NADP+ and NADPH are prevented from leaving their binding site in the enzyme intermediate that is responsible for the hydride-transfer step (Hutton et al., 1994; Diggle et al., 1996). Crystal structures indicate that this is the result of a loop in the polypeptide chain of dIII (loop E, the ‘lid’) closing down over, and occluding, the bound nucleotide from the solvent—Fig. 8b (White et al., 2000). In the occluded position a highly conserved Tyr residue (Tyr171) at the apex of the loop packs across the hydride transfersite. The conformational changes responsible for the closing and opening of loop E (i. e., before and after hydride transfer) are thought to result from vectorial protonation and deprotonation of the dII transducer as indicated in Section VI.A. A sequence of LysArg-Ser residues in the N-terminal segment of loop E determines the nucleotide-binding specificity of dIII by making hydrogen bonds with the 2´-phosphate group of NADP+/NADPH.
J. Baz Jackson and U. Mirian Obiozo
504 (a)
(b)
Fig. 8. (a) The hydride-transfer site in 2OO5.pdb. See Color Plate 12, Fig. 20 for further explanation. This view shows mainly the dI(B) component; the loop near NADP+ and the helix turn in the background are from dIII. The nucleotides are abeled and H-bonds are shown as gray dashed lines. (b) The loop E ‘lid’ of dIII and the ‘mobile loop’ of dI are closed down over the hydride transfer site of 2OO5.pdb. The nucleotides (NADP+ in dIII, and the NADH analog, H2NADH, in dI) are shown in a space-filling format. The dI(B) polypeptide forms the upper half of the figure, with its mobile loop labeled, and the dI(A) polypeptide in the background. The dIII polypeptide is shown in the lower part of the figure with its loop E labeled. Also shown are the RQD loop of dI and helix D/loop D of dIII (see text).
D. The Mobile Loop of Transhydrogenase The dI component of transhydrogenase has a loop (the ‘mobile loop’) which, like loop E of dIII, has a highly conserved Gly-Tyr-Ala motif at its apex. NMR experiments show that the mobile loop closes down upon the protein surface during either NAD+ or NADH binding (Diggle et al., 1995; Quirk et al., 1999b). However in contrast to the behavior of loop E, this seems not to be directly associated with proton-translocation steps in the enzyme because the protein motions are observed in isolated dI. X-ray structures show that, in the closed position of the mobile loop, the apical Tyr residue (Tyr235) is in van der Waals contact with the bound NADH, and that its OH group makes an H-bond with Arg127 in the RQD loop—see Figs. 3 and 8. It is thought that closure of the mobile loop is required to seal the site from the solvent water during hydride-transfer (Buckley et al., 2000; Mather et al., 2004). E. The Alternating-Sites Mechanism for Transhydrogenase In principle, the two monomers of the transhydrogenase dimer might operate independently, or they might interact during turnover. The latter was suggested by early experiments on the detergentdispersed mitochondrial enzyme in which it was
found that transhydrogenation is completely blocked when only 0.5 mol of covalent inhibitor is bound per protein monomer (Phelps and Hatefi, 1984; Phelps and Hatefi, 1985). The profound asymmetry of the dI2dIII1 complex suggested to us that the intact enzyme operates by an ‘alternating-sites mechanism’ (Cotton et al., 2001; Jackson et al., 2002). It was proposed that the conformational changes taking place in one dI-dII-dIII monomer run 180° out-ofphase with those taking place in the other monomer. While one monomer is in an open state (capable of exchanging NADP+/NADPH with the solvent but blocked in hydride transfer) the other is in an occluded state (capable of undergoing hydride transfer but blocked in solvent NADP+/NADPH exchange). As a protonation from the outside aqueous phase drives one monomer from the open to the occluded state, a deprotonation to the inside aqueous phase drives the other monomer from the occluded to the open state (Fig. 9). The advantage of an alternating-site mechanism might be that it equalizes the energy-generating (protonation/deprotonation) and energy-consuming (conformational change) steps of the enzyme. VII. Conformational Changes in the Coupling Reactions of Transhydrogenase Our current challenge is to understand better the
Chapter 25
Transhydrogenase
505
Fig. 9. The alternating-site mechanism of transhydrogenase. Re-drawn from Jackson et al. (2002). D is NAD+, DH is NADH, P is NADP+ and PH is NADPH. One dI-dII-dIII monomer is shown in gray, the other is stippled. In the top left cartoon the grey monomer is in the open state and is exchanging product nucleotides for fresh substrate nucleotides from the solvent, whereas the stippled monomer is in the occluded state catalyzing hydride transfer. In the top right cartoon protonation from the outside drives the conversion of the gray monomer into the occluded state, and deprotonation to the inside drives conversion of the stippled monomer into the open state. In the bottom right and bottom left cartoons the equivalent reactions take place in the complementary monomers. Note that the order of the protonation and deprotonation reactions remains to be established, but see Hutton et al. (1994).
character of the conformational changes that couple the redox reaction and the proton-translocation reaction of transhydrogenase. The emergence of X-ray structures of the nucleotide-binding components of the enzyme has led to improved descriptions of events at the hydride transfer site in dI2dIII1 complexes but much more information is required. One suggestion was that dIII might rotate 180° on an axis parallel to the membrane plane during catalysis (Sundaresan et al., 2005). However, we have argued that this conclusion, based on the observation of contacts between dI and dIII in X-ray structures, is probably in error—the contacts are more likely to result simply from crystal packing (Bhakta et al., 2006). More interesting is the finding of a crystal form of Rsp. rubrum dIII in which loop D can adopt two different conformations (Sundaresan et al., 2003). Loop D is well placed to participate in the conformational changes that couple the redox reaction to proton translocation—see Fig. 3 and Fig. 8b — because it contacts both the loop E lid of dIII and the RQD loop of dI, and because it buries the invariant dIII residue, Asp132. Mutation of this residue, which has an unusually high pKa (Jackson et al., 2002; Sundaresan et al., 2003; Pedersen et al.,
2005), inactivates intact transhydrogenase (Meuller et al., 1996) and lowers the apparent NADP(H)-binding affinity of isolated dIII from E. coli (Bergkvist et al., 2000). Several research groups are working towards a high-resolution structure of intact transhydrogenase. Such a structure will undoubtedly provide further important clues on the mechanism of coupling. We should then have an excellent platform from which to study conformationally coupled ion-transport. In comparison with ion-transporters that are coupled to, for example, ATP synthesis/hydrolysis, a crucial advantage of transhydrogenase is that it is accessible to real time-kinetics analysis. Acknowledgments We are grateful to Scott White, Nick Cotton, Harma Brondijk, Simon Whitehead and Tina Bhakta for discussion, and to Tina Bhakta for providing some of the figures.
506 References Althage M, Bizouarn T and Rydstrom J (2001) Identification of a region involved in the communication between the NADP(H)binding domain and the membrane domain in proton pumping E. coli transhydrogenase. Biochemistry 40: 9968–9976 Ambartsoumian G, Dari R, Lin RT and Newman EB (1994) Altered amino-acid metabolism in LRP mutants of Escherichia coli K12 and their derivatives. Microbiology 140: 1737–1744 Arkblad EL, Betsholtz C, Mandoli D and Rydstrom J (2001) Characterization of a nicotinamide nucleotide transhydrogenase gene from the green alga Acetabularia acetabulum and comparison of its structure with those of the corresponding genes in mouse and Caenorhabditis elegans. Biochim Biophy Acta 1520: 115–123 Arkblad EL, Tuck S, Pestov NB, Dmitriev RI, Kostina MB, Stenvall J, Tranberg M and Rydstrom J (2005) A Caenorhabditis elegans mutant lacking functional nicotinamide nucleotide transhydrogenase displays increased sensitivity to oxidative stress. Free Radical Biol Med 38: 1518–1525 Axelrod HL, Abresch EC, Paddock ML, Okamura MY and Feher G (2000) Determination of the binding sites of the proton transfer inhibitors Cd2+ and Zn2+ in bacterial reaction centers. Proc Nat Acad Sci USA 97: 1542–1547 Bergkvist A, Johansson C, Johansson T, Rydstrom J and Karlsson BG (2000) Interactions of the NADP(H)-binding domain III of proton-translocating transhydrogenase from Escherichia coli with NADP(H) and the NAD(H)-binding domain I studied by NMR and site-directed mutagenesis. Biochemistry 39: 12595–12605 Bhakta T, Whitehead SJ, Snaith JS, Wilkie J, Rajesh S, White SA and Jackson JB (2006) Structures of the dI2dIII1 complex of proton-translocating transhydrogenase with bound, inactive analogues of NADH and NADPH reveal active-site geometries. Biochemistry 46: 3304–3318 Bizouarn T, Fjellstrom O, Meuller J, Axelsson M, Bergkvist A, Johansson C, Karlsson G and Ry0dstrom J (2000) Proton translocating nicotinamide nucleotide transhydrogenase from E. coli. Mechanism of action deduced from its structural and catalytic properties. Biochim Biophys Acta 1457: 211–218 Bragg PD and Hou C (1999) Mutation of conserved polar residues in the transmembrane domain of the proton-pumping pyridine nucleotide transhydrogenase of Escherichia coli. Arch Biochem Biophys 363: 182–190 Bragg PD and Hou C (2000) The presence of an aqueous cavity in the proton-pumping pathway of the pyridine-nucleotide transhydrogenase of Escherichia coli is suggested by the reaction of the enzyme with sulphydryl inhibitors. Arch Biochem Biophys 380: 141–150 Bragg PD, Davies PL and Hou C (1972) Function of energy dependent transhydrogenase in Escherichia coli. Biochem Biophys Res Commun 47: 1248–1255 Brondijk THC, van Boxel GI, Singh A, Mather OM, White HA, Quirk PG, White SA and Jackson JB (2006) The role of invariant amino acid residues at the hydride-transfer site of proton-translocating transhydrogenase. J Biol Chem 281: 13345–13354 Buckley PA, Jackson JB, Schneider T, White SA, Rice DW and Baker PJ (2000) Protein-protein recognition, hydride transfer
J. Baz Jackson and U. Mirian Obiozo and proton pumping in the transhydrogenase complex. Structure 8: 809–815 Cotton NPJ, White SA, Peake SJ, McSweeney S and Jackson JB (2001) The crystal structure of an asymmetric complex of the two nucleotide-binding components of proton-translocating transhydrogenase. Structure 9: 165–176 Cunningham IJ, Williams R, Palmer T, Thomas CM and Jackson JB (1992) The relation between the soluble factor associated with H+-transhydrogenase of Rhodospirillum rubrum and the enzyme from mitochondria and Escherichia coli. Biochim Biophys Acta 1100: 332–338 Diggle C, Cotton NPJ, Grimley RL, Quirk PG, Thomas CM and Jackson JB (1995) Conformational dynamics of a mobile loop in the NAD(H)-binding subunit of proton-translocating transhydrogenases from Rhodospirillum rubrum and Escherichia coli. Eur J Biochem 232: 315–326 Diggle C, Bizouarn T, Cotton NPJ and Jackson JB (1996) Properties of the purified, recombinant, NADP(H)-binding domain III of the proton-translocating nicotinamide nucleotide transhydrogenase from Rhodospirillum rubrum. Eur J Biochem 241: 162–170 Fisher RR and Guillory RJ (1971) Resolution of enzymes catalyzing energy-linked transhydrogenation—Interaction of transhydrogenase factor with the Rhodospirillum rubrum chromatophore membrane. J Biol Chem 246: 4679–4686 Fjellstrom O, Johansson Cand Rydstrom J (1997) Structural and catalytic properties of the expressed and purified NAD(H)- and NADP(H)-binding domains of proton-pumping transhydrogenase from Escherichia coli. Biochemistry 36: 11331–11341 Fjellstrom O, Bizouarn T, Zhang J, Rydstrom J, Venning JD and Jackson JB (1999) Catalytic properties of hybrid complexes of the NAD(H)-binding and NADP(H)-binding domains of the proton-translocating transhydrogenases from Escherichia coli and Rhodospirillum rubrum. Biochemistry 38: 415–422 Freeman H, Shimomura K, Horner E, Cox RD and Ashcroft FM (2006) Nicotinamide nucleotide transhydrogenase: A key role in insulin secretion. Cell Metab 3: 35–45 Gerolimatos B and Hanson RL (1978) Repression of Escherichia coli pyridine nucleotide transhydrogenase by leucine. J Bacteriol 134: 394–400 Glavas NA, Hou C and Bragg PD (1995) Involvement of Histidine91 of the beta subunit in proton translocation by the pyridine nucleotide transhydrogenase of Escherichia coli. Biochemistry 34: 7694–7702 Hanson RL and Rose C (1980) Effects of an insertion mutation in a locus affecting pyridine nucleotide transhydrogenase (pnt::Tn5) on the growth of Escherichia coli. J Bacteriol 141: 401–404 Hickman JW, Barber RD, Skaar EP and Donohue TJ (2002) A link between the membrane-bound pyridine nucleotide transhydrogenase and glutathione-dependent processes in Rhodobacter sphaeroides. J Bacteriol 184: 400–409 Holmberg E, Olausson T, Hultman T, Rydstrom J, Ahmad S, Glavas NA and Bragg PD (1994) Prediction and site-specific mutagenesis of residues in transmembrane alpha-helices of proton-pumping nicotinamide nucleotide transhydrogenases from Escherichia coli and bovine heart mitochondria. Biochemistry 33: 7691–7700 Hu X, Zhang JW, Fjellstrom O, Bizouarn T and Rydstrom J (1999) Site directed mutagenesis of charged and potentially proton carrying residues in the beta subunit of the proton-translocating nicotinamide nucleotide transhydrogenase from Escherichia
Chapter 25
Transhydrogenase
coli. Characterisation of beta H91, beta D392 and beta K424. Biochemistry 38: 1652–1658 Hutton MN, Day JM, Bizouarn T and Jackson JB (1994) Kinetic resolution of the reaction catalysed by proton-translocating transhydrogenase from Escherichia coli as revealed by experiments with nucleotide substrate analogues. Eur J Biochem 219: 1041–1051 Jackson JB (2003) Proton translocation by transhydrogenase. FEBS Lett 545: 18–24 Jackson JB, White SA, Quirk PG and Venning JD (2002) The alternating site, binding change mechanism for proton translocation by transhydrogenase. Biochemistry 41: 4173–4185 Jackson JB, White SA and Brondijk THC (2005) Hydride transfer and proton translocation by nicotinamide nucleotide transhydrogenase. In: M Wikstrom (ed) Biophysical and Structural Aspects of Bioenergetics, pp 376–393. Royal Society of Chemistry, Cambridge, UK Jenks WP (1983) What is a coupled vectorial process? Current Topics in Membranes and Transport 19: 1–18 Keister DL and Yike NJ (1967) Energy-linked reactions in photosynthetic bacteria II—The energy-dependent reduction of oxidized nicotinamide-adenine dinucleotide phosphate by reduced nicotinamide-adenine dinucleotide in chromatophores of Rhodospirillum rubrum. Biochemistry 6: 3847–3857 Lee CP, Simard-Duquesne N, Ernster L and Hoberman HD (1965) Stereochemistry of hydrogen transfer in the energy-linked pyridine nucleotide transhydrogenase and related reactions. Biochim Biophys Acta 105: 397–409 Mather OM, van Boxel GI, White SA and Jackson JB (2004) Active-site conformational changes associated with hydride transfer in proton-translocating transhydrogenase. Biochemistry 43: 10952–10964 Meuller J and Rydstrom J (1999) The membrane topology of proton-pumping Escherichia coli transhydrogenase determined by cysteine labelling. J Biol Chem 274: 19072–19080 Meuller J, Hu X, Buntoff C, Olausson T and Rydstrom J (1996) Identification of an aspartic acid residue in the beta subunit which is essential for catalysis and proton pumping by transhydrogenase from Escherichia coli. Biochim Biophys Acta 1273: 191–194 Oshino N and Chance B (1977) Properties of glutathione release observed during reduction of organic hydroperoxide, demethylation of aminopyridine and oxidation of some substances in perfused rat liver, and their implications for the physiological function of catalase. Biochem J 162: 509–525 Peake SJ, Venning JD and Jackson JB (1999) A catalytically active complex formed from the recombinant dI protein of Rhodospirillum rubrum transhydrogenase and the recombinant dIII protein of the human enzyme. Biochim Biophys Acta 1411: 159–169 Pedersen A, Johansson T, Rydstrom J and Karlsson BG (2005) Titration of E. coli transhydrogenase domain III with bound NADP+ or NADPH studied by NMR reveals no pH-dependent conformational change in the physiological pH range. Biochim Biophy Acta 1707: 254–258 Phelps DC and Hatefi Y (1984) Interaction of purified nicotinamide nucleotide transhydrogenase with dicyclohexylcarbodiimide. Biochemistry 23: 4475–4480 Phelps DC and Hatefi Y (1985) Mitochondrial nicotinamide nucleotide transhydrogenase: active site modification by 5´-(p-(fluorosulfonyl)benzoyl) adenosine. Biochemistry 24:
507 3503–3507 Prasad GS, Sridhar V, Yamaguchi M, Hatefi Y and Stout CD (1999) Crystal structure of transhydrogenase domain III at 1.2 Å resolution. Nature Struct Biol 6: 1126–1131 Quirk PG, Jeeves M, Cotton NPJ, Smith KJ and Jackson JB (1999a) Structural changes in the recombinant, NADP(H)-binding component (dIII) of proton-translocating transhydrogenase revealed by NMR spectroscopy. FEBS Lett 446: 127–132 Quirk PG, Smith KJ, Thomas CM and Jackson JB (1999b) The mobile loop region of the NAD(H)-binding component (dI) of proton-translocating transhydrogenase from Rhodospirillum rubrum: Complete NMR assignment and effects of bound nucleotides. Biochim Biophys Acta 1412: 139–148 Singh A, Venning JD, Quirk P, van Boxel GI, Rodrigues DJ, White SA and Jackson JB (2003) Interactions between transhydrogenase and thio-nicotinamide analogues of NAD(H) and NADP(H) underline the importance of nucleotide conformational changes in coupling to proton translocation. J Biol Chem 278: 33208–33216 Studley WK, Yamaguchi M, Hatefi Y and Saier MH (1999) Phylogenetic analyses of proton-translocating transhydrogenases. Microbial Compar Genomics 4: 173–186 Sundaresan V, Yamaguchi M, Chartron J and Stout CD (2003) Conformational change in the NADP(H)-binding domain of transhydrogenase defines four states. Biochemistry 42: 12143–12153 Sundaresan V, Chartron J, Yamaguchi M and Stout CD (2005) Conformational diversity in NAD(H) and interacting transhydrogenase nicotinamide nucleotide binding domains. J Mol Biol 346: 617–629 van Boxel GI, Quirk P, Cotton NJP, White SA and Jackson JB (2003) Glutamine-132 in the NAD(H)-binding component of proton-translocating transhydrogenase tethers the nucleotides before hydride transfer. Biochemistry 42: 1217–1226 Venning JD, Bizouarn T, Cotton NPJ, Quirk PG and Jackson JB (1998) Stopped-flow kinetics of hydride transfer between nucleotides by recombinant domains of proton-translocating transhydrogenase. Eur J Biochem 257: 202–209 White SA, Peake SJ, McSweeney S, Leonard G, Cotton NNJ and Jackson JB (2000) The high resolution structure of the NADP(H)-binding component of proton-translocating transhydrogenase from human-heart mitochondria. Structure 8: 1–12 Whitehead SJ, Rossington KE, Hafiz A, Cotton NJP and Jackson JB (2005) Zinc ions selectively inhibit steps associated with binding and release of NADP(H) during turnover of protontranslocating transhydrogenase. FEBS Lett 579: 2863–2867 Williams R, Cotton NPJ, Thomas CM and Jackson JB (1994) Cloning and sequencing of the genes for the proton-translocating nicotinamide nucleotide transhydrogenase from Rhodospirillum rubrum and the implications for the domain structure of the enzyme. Microbiology 140: 1595–1604 Wilson R, Obiozo UM, Quirk PG, Besra GD and Jackson JB (2006) A hybrid of the transhydrogenases from Rhodospirillum rubrum and Mycobacterium tuberculosis catalyses rapid hydride transfer but not the complete, proton-translocating reaction. Biochim Biophy Acta 1537: 215–223 Wraight CA (2006) Chance and design—proton transfer in water, channels and bioenergetic proteins. Biochim Biophy Acta 1757: 886–912 Yamaguchi M and Hatefi Y (1995a) Proton-translocating
508 nicotinamide nucleotide transhydrogenase of Escherichia coli. Involvement of aspartate 213 in the membrane-intercalating domain of the beta subunit in energy transduction. J Biol Chem 270: 16653–16659 Yamaguchi M and Hatefi Y (1995b) Proton-translocating transhydrogenase. Reconstitution of the extramembranous nucleotidebinding domains. J Biol Chem 270: 28165–28168 Yamaguchi M and Stout CD (2003) Essential glycine residues in the proton channel of Escherichia coli transhydrogenase. J
J. Baz Jackson and U. Mirian Obiozo Biol Chem 278: 45333–45339 Yamaguchi M, Hatefi Y, Trach K and Hoch JA (1988) The primary structure of the mitochondrial energy-linked nicotinamide nucleotide transhydrogenase deduced from the sequence ofcDNA clones. J Biol Chem 263: 2761–2767 Yamaguchi M, Stout CD and Hatefi Y (2002) The proton channel of the energy-transducing nicotinamide nucleotide transhydrogenase of Escherichia coli. J Biol Chem 277: 33670–33675
Chapter 26 Functional Coupling Between Reaction Centers and Cytochrome bc1 Complexes Jérôme Lavergne* and André Verméglio CEA Cadarache, DSV/IBEB/SBVME/LBC, UMR 6191 CNRS/CEA/Univ Aix-Marseille, Saint-Paul-lez-Durance, F-13108 France
Pierre Joliot Institut de Biologie Physico-Chimique, UMR 7141 CNRS/Paris 6, 13 Rue Pierre et Marie Curie, 75005 Paris, France
Summary ............................................................................................................................................................... 509 I. Introduction..................................................................................................................................................... 510 II. Structure of the Protein Complexes ............................................................................................................... 512 III. The Electron Donors to the Reaction Center ................................................................................................. 514 A. Tetraheme Reaction Center Subunits .............................................................................................. 514 B. Triheme Reaction Center Subunits .................................................................................................. 515 C. Monoheme Electron Donor .............................................................................................................. 515 IV. Kinetics of P+ Reduction by Mobile Cytochromes .......................................................................................... 515 V. Donor Side Shuttling and Turnover of the Cytochrome bc1 Complex ............................................................ 517 VI. Quinone Reactions ......................................................................................................................................... 518 VII. Supramolecular Organization in Rhodobacter sphaeroides and Rhodobacter capsulatus ......................... 519 VIII. Quinone Confinement in Rhodobacter sphaeroides .................................................................................... 522 IX. Quinone Traffic in the PufX– Mutant of Rhodobacter sphaeroides ............................................................... 522 X. The Supercomplex Model: Difficulties and Alternative Possibilities ............................................................... 524 XI. Mitochondrial Supercomplexes ...................................................................................................................... 526 XII. Diffusion and Confinement of Cytochrome c2: Possible Mechanisms............................................................ 527 XIII. Diffusion and Confinement of Quinones: Possible Mechanisms .................................................................... 528 XIV. Conclusions .................................................................................................................................................... 529 References ............................................................................................................................................................ 530
Summary Light-induced cyclic electron transfer in purple bacteria involves two integral membrane protein complexes, the reaction center (RC) and the cytochrome bc1 complex, and two mobile carriers that shuttle between them. The mobile carrier on the acceptor side of the RC is a quinone molecule, confined to the hydrophobic membrane/protein regions. More diversity is found for the donor side, starting with the RC, which may or may not possess a multiheme donor subunit. Depending on species, the multiheme subunit includes three or four c-type hemes with alternating high and low midpoint potentials so that the electron transfer involves a succes*Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 509–536. © 2009 Springer Science + Business Media B.V.
510
Jérôme Lavergne, André Verméglio and Pierre Joliot
sion of uphill and downhill steps. The donor side mobile carrier, confined to the periplasmic space, is either a soluble cytochrome (c2 or c8), a membrane-anchored c-type cytochrome, or a high-potential iron-sulfur protein (HiPIP). The stoichiometric ratios between the components of the photosynthetic chain (generally in the order: quinones > RC > mobile cytochrome > bc1) are variable depending on species and growth conditions. In many species where the core antenna ring surrounds the RC completely, the quinone circuit requires the crossing of this barrier. For species endowed with a low bc1:RC ratio, the diffusion of the mobile carriers over relatively long ranges is mandatory for connecting the distant partners. In some Rhodobacter species, with a high bc1: RC ratio, evidence has been found for the formation of specific supercomplexes associating all the components of the cyclic transfer in largely independent functional units. The mechanisms that may be responsible for the confinement of the mobile carriers in such systems are discussed. I. Introduction The light-induced cyclic electron transfer in purple bacteria involves four different electron carrier components. Two of them, the RC and the cytochrome bc1 complex, are multi-subunit complexes, each encompassing several redox centers, embedded in the cytoplasmic membrane. Quinone molecules (Q) within the hydrophobic domain of the membrane and soluble carriers (C) in the periplasmic space act as electron transfer shuttles between the two membranebound protein complexes. The outcome of this light induced cyclic electron transfer is the translocation of protons from the cytoplasm to the periplasm (∆pH) and the formation of a transmembrane potential. Both components constitute the proton electrochemical potential difference ∆µ~H that ultimately drives ATP synthesis via the ATP synthase (Chapter 24, Feniouk and Junge). A schematic representation of the light-induced electron transfer in purple bacteria is shown in Fig. 1. Some of these electron carriers and components (quinone molecules, cytochrome bc1 complex, periplasmic carrier, ATP synthase) are shared with the respiratory chain that is also present in the cytoplasmic membrane of the purple facultative photosynthetic bacteria. Part of the energy formed during this process, using the ∆µ~H generated directly by the photochemical activity or from the hydrolysis of ATP, is used to produce Abbreviations: 3-D – three-dimensional; AFM – atomic force microscopy; bc1 – cytochrome bc1 complex; bc1:RC – stoichiometric ratio of cytochrome bc1 complex to reacton center; Blc. – Blastochloris; FeS – the Fe-S center in the Iron Sulfur Protein subunit of the cytochrome bc1 complex; HiPIP – high-potential iron-sulfur protein; LH1/LH2 – light-harvesting complex 1 (core) or 2 (peripheral); P – special pair of bacteriochlorophylls in the reaction center acting as the primary photochemical electron donor; Q – quinone; Rba. – Rhodobacter; RC – reaction center; Rdv. – Rhodovulum; Rps. – Rhodopseudomonas; Rsp. – Rhodospirillum; Rvi. – Rubrivivax; TMAO – trimethylaminooxide; UQ – ubiquinone
reducing power, providing the driving force for a linear electron pathway, e.g., from succinate to NAD+ (Knaff, 1978). The ∆µ~H drives an uphill electron transfer from quinol to NAD+ through Complex I (Dupuis et al., 1998). Injection of the succinate electrons into the quinone pool is mediated by Complex II, which, in contrast to Complex I, is not coupled to the ∆µ~H. This light-induced uphill electron transfer from quinol to NAD+ is responsible for the inhibition of the aerobic and anaerobic respiratory processes upon continuous illumination in photosynthetic bacteria (Verméglio et al., 2004). The competition for the reduced periplasmic carrier between the RC and cytochrome oxidase is also consistent with the inhibition of respiratory activity by light. The main steps of the transformation of light into chemical energy can be outlined as follows. After the absorption of a photon by the light-harvesting (LH) complexes, the excitation reaches a RC in less than 100 picoseconds, inducing a charge separation between the excited primary electron donor P, a bacteriochlorophyll dimer, and the primary acceptor QA, a quinone molecule. Stabilization of these separated charges occurs in the microsecond range through subsequent electron transfers from the secondary donor (C) and to the pool of quinones (Q). The reduction of Q to the quinol (QH2) state involves the uptake of two protons from the cytoplasm. The cyclic electron transfer is completed by the reduction of the oxidized carrier C by QH2, catalyzed by the cytochrome bc1 complex. The reactions that take place in the cytochrome bc1 complex are understood according to the modified Q-cycle model (see Table 1) elaborated by Crofts and coworkers (Crofts et al., 1983; for recent reviews see Crofts, 2004 and Chapter 23, Kramer et al.). The outcome of this process is the electrogenic translocation, from cytoplasm to periplasm, of two protons per electron flowing through the cyclic pathway.
Chapter 26
Coupling Between RC and Cytochrome bc1 Complex
511
Fig. 1. Schematic representation of the components involved in the photosynthetic electron chain of purple bacteria. Following the absorption of light by the peripheral (LH2) or core (LH1) light-harvesting complexes, a photochemical charge separation takes place in the reaction center (RC). This initiates an electron flow through, successively, a quinone (Q) from the membrane sequestered pool, the cytochrome bc1 complex (bc1) and a soluble periplasmic carrier (C), which completes the cycle by reducing the donor side of the RC. This cycle is coupled to the translocation of protons across the membrane and to the generation of a membrane potential, eventually used as a driving force by the ATP synthase (ATPase) for producing ATP.
In addition to the electron transfer complexes, the photosynthetic membranes accommodate the pigmented LH (or antenna) complexes, which absorb light and allow resonance energy transfer of the excitation until it hits the RC (Chapter 13, van Grondelle and Novderezhkin; Chapter 15, Şener and Schulten). The LH complexes are oligomeric structures whose building block is a heterodimer of single transmembrane helices (α and β) that binds bacteriochlorophylls and carotenoids. While all photosynthetic bacteria possess the ‘core’ LH1 complexes, only some of them have the peripheral antenna LH2. There is a fixed ratio between LH1 and RC (14 to 16 αβ per RC) whereas the amount of LH2 (up to 40 αβ-LH2 per RC) de-
pends on the illumination conditions during growth. There is also a large excess of Q per RC (~20–30). Ubiquinone molecules, with long isoprenoid chains of 8–10 units, are present in all photosynthetic bacteria, but quinones of lower mid-point potential, like menaquinone and/or rhodoquinone (with similar long tails), are also found in numerous species (Hiraishi et al., 1984; Imhoff and Bias-Imhoff, 1995). On the other hand, in photosynthetically grown bacteria, the levels of the cytochrome bc1 complex and the C carrier are generally lower than that of the RC (e.g., by a factor of two). There has been no systematic study of the rate of the light-induced cyclic process and of its limiting
Table 1. A scheme of the electron transfer reactions in the low potential chain of the cytochrome bc1 complex. The Q-cycle mechanism involves two turnovers, triggered by the transfer of one electron from the high potential chain of the complex (cytochrome c1 and the Iron Sulfur Protein Fe-S) to the mobile carrier C, oxidized by the RC. A quinol (QH2) molecule is then oxidized at the Qo site via a bifurcated reaction in which one electron is injected into the high potential chain and the other into the low potential chain (hemes bL and bH). The electron remains stored on the bH heme (semiquinone formation at site Qi is unfavorable for pH <8.5, see Robertson et al., 1984). A similar series of reactions during the second turnover leads to the formation and release of QH2 from the Qi site and release of a Q from the Qo site. The overall oxidation of the two QH2 in Qo is accompanied by the release of four protons H+p into the periplasmic space; the reduction of one Q at Qi is accompanied by the uptake of two protons Hc+ from the cytoplasm. The overall outcome is thus, per electron transferred to C, the electrogenic translocation of one proton across the membrane in addition to the proton released into the periplasm due to the oxidation of ½ QH2.
512
Jérôme Lavergne, André Verméglio and Pierre Joliot
steps under physiological conditions, but much can be inferred from the knowledge gained under various experimental situations using isolated chromatophores or whole cells. The cyclic transfer implies a conservation of the number of ‘electrons’ looping in the circuit. Therefore, the adjustment of this ‘redox poise’ requires an intervention from outside. The interaction with the respiratory chain is important in this respect. For instance, under anaerobic conditions in the dark, the quinone pool is largely reduced. Under such conditions, the initiation of photosynthetic activity will require a priming process, with the active fraction of RCs generating a ∆µ~H , and extracting electrons from the circuit (through reverse electron transfer in Complex I) to activate a larger number of RCs. The requirement for a substantially oxidized pool is also important for the cytochrome bc1 complex, whose functioning implies both the supply of QH2 at site Qo and of Q at site Qi (see Table 1). This complex will thus function optimally under conditions where the quinone pool is partly oxidized and reduced, i.e., when the ambient redox poise is kept in the region of the midpoint potential of the ubiquinone pool (~100 mV). Under more reducing conditions, the turnover of the system is limited by the transfer of the Q formed at site Qo of the cytochrome bc1 complex to the QB site of the RC. This implies a competition between the two Q binding sites, QB and Qi, that will be discussed in more detail later. Under oxidizing conditions, i.e., when the pool is mainly oxidized, the rate of the cyclic flow could be limited by the local concentration of QH2 in the vicinity of the Qo site of the cytochrome bc1 complex. In these two extreme situations, the confinement of mobile carriers within a supercomplex that associates the RC and the cytochrome bc1 complex can be a way to increase the local concentration of QH2 (oxidizing condition) or Q (reducing) in the vicinity of the Qo or QB site, respectively. Another important parameter is the bc1:RC stoichiometric ratio. This stoichiometry varies significantly depending on the species and growth conditions (light intensity, oxygen tension, etc). For example, it varies between 0.5 for Rhodobacter (Rba.) sphaeroides or Rba. capsulatus to 0.1 for Rhodospirillum (Rsp.) rubrum grown under anaerobic photosynthetic conditions (Crofts et al., 1983; van der Wal and van Grondelle, 1983). For Rba. capsulatus cells grown at very high light intensity it can reach a value of 2 (Garcia et al., 1987). In this chapter, emphasis will be placed on the
different factors that affect the efficiency of lightinduced cyclic electron transfer, with a particular focus on the lateral organization of the membrane complexes, and the degree of mobility of the electron carriers connecting these complexes. II. Structure of the Protein Complexes Structural information on the photosynthetic chain of purple bacteria has been gained at different levels. Three-dimensional (3-D) structures at atomic resolution have been obtained for each of the protein complexes of the photosynthetic chain. The RC of Blastochloris (Blc.) (formerly Rhodopseudomonas (Rps.)) viridis, one of the first membrane proteins to be crystallized, revealed a pseudo-twofold symmetry for the chromophores and the L and M subunits (Deisenhofer et al., 1984; 1985). This quasi-symmetrical arrangement is found in all photosynthetic RCs. The crystallographic structure of the cytochrome bc1 complex was first solved for the mitochondrial complex (Xia et al., 1997). It revealed a homodimeric organization, locked by the swapping of the membrane helix of the iron-sulfur protein subunit. A very similar structure was later obtained for the complex from Rba. capsulatus (Berry et al., 2004; Chapter 22, Berry et al.). The 3-D structures of LH2 complexes isolated from different photosynthetic bacteria (McDermott et al., 1995; Koepke et al., 1996; Chapter 8, Gabrielsen et al.) disclosed their organization as (αβ)n oligomeric rings. At a higher level of organization, the structure of the core complexes formed by the RC and the LH1 was revealed by Miller (1982), using negatively stained intracytoplasmic membranes of Blc. viridis, which naturally form 2-D crystals. This pioneering work established that a ring of LH1 molecules surrounds a single RC. This type of organization has also been observed for native membranes or 2-D crystals of RC–LH1 complexes from several other species (Ectothiorhodospira halochloris (Stark et al., 1984), Rsp. molischianum (Boonstra et al., 1994), Rsp. rubrum (Walz and Ghosh, 1997). In all these cases, the RC is completely surrounded by the LH1 ring (Chapter 9, Bullough et al.). This poses a problem as regards the traffic of quinones that must pass the LH1 barrier when shuttling between the RC and cytochrome bc1 complexes: this issue is discussed in Section IX. The best resolution obtained thus far for a RC-LH1 core complex has been attained in
Chapter 26
Coupling Between RC and Cytochrome bc1 Complex
the case of Rps. palustris (Roszak et al., 2003). The crystal structure of this complex was resolved at 4.8 Å resolution and shows the RC surrounded by an oval ring of 15 αβ LH1 subunits. In this case, the complete closure of the RC by the LH1 is prevented by a single transmembrane helix, denoted W, whose sequence is not known. Open rings of LH1 have also been observed in various Rhodobacter species. In the case of Rba. sphaeroides, negative staining of tubular membranes present in LH2-deleted strains has revealed a dimeric organization of the RC-LH1 complexes (Jungas et al., 1999; Siebert et al., 2004). This S-shaped organization has also been observed for 2D crystals of purified and reconstituted RC-core complexes (Scheuring et al., 2004a; Qian et al., 2005). A dimeric association of RC-LH1 complexes had been previously revealed by freeze-fracture electron microscopy in tubular membranes corresponding to regions of the intracytoplasmic membrane devoid of LH2 complexes (Hunter et al., 1988; Kiley et al., 1988; Golecki et al., 1991; Sabaty et al., 1994). The presence of the S-shaped dimers in native membranes has been confirmed by atomic force microscopy (AFM) in Rba. sphaeroides (Bahatyrova et al., 2004a) and Rba. blasticus (Scheuring et al., 2005a; Chapter 47, Scheuring). The detergent-solubilized dimeric core complexes can be purified (Francia et al., 1999) and retain ‘native-like’ functional properties with regards to excitation transfer and photoreduction of a quinone pool (Francia et al., 2004; Comayras et al., 2005a). The PufX polypeptide is strictly required for the formation of these S-shaped structures in Rba. sphaeroides. PufX is a small protein of 80 amino acids encoded by the pufX gene, localized in the puf (photosynthetic unit formation) operon downstream of the genes encoding the LH1 and the L and M subunits of the RC in several Rhodobacter species (Farchaus et al., 1990; Lilburn et al., 1992; Tsukatani et al., 2004; Chapter 9, Bullough et al.). It is a membrane protein associated with the RC–LH1 complex in a 1:1 ratio (Francia et al., 1999), and it is essential for anaerobic photosynthetic growth in both Rba. capsulatus and Rba. sphaeroides (Farchaus et al., 1990; Lilburn et al., 1992). Other effects of the deletion of PufX are the formation of a complete closed ring of LH1 around the RC, abolishing the dimeric association (Siebert et al., 2004) and, in membranes devoid of LH2, the disappearance of the tubular membrane structure (Verméglio and Joliot, 2002) and of the long-range regular array of (dimeric) core complexes (Frese et
513
al., 2000). The precise location of PufX is still under debate (see Chapter 9, Bullough et al.; Chapter 47, Scheuring). One possibility is that it belongs to the LH1 assembly at the dimer junction (Scheuring et al., 2004a; 2005a); other authors locate PufX between the LH1 ring and the QB site of the RC (Qian et al., 2005). The first model has the advantage of avoiding the arrangement at the dimer junction proposed by Qian et al. (2005) whereby the LH1αβ pair of one half of the dimer is adjacent to an αβ pair of the other ring with an opposite arrangement; the Qian et al. model also locates the β unit of the end pair close to the RC of the other dimer, an approach normally expected of LH1α. Recently, Gubellini et al. (2006) have studied the role of PufX in Rba. veldkampii, a species related to Rba. sphaeroides. Although a PufX subunit is present, the core complex was found to be monomeric. Possibly, the arrangement of the core complex in Rba. veldkampii is similar to the one observed for Rps. palustris where the LH1 ring is discontinuous due to the presence of the polypeptide W but where dimeric association of RCs has not been observed in the crystal structure (Roszak et al., 2003) nor by AFM (Scheuring et al., 2006). In addition to the description of the organization of RC-LH1 core complexes of several purple bacteria (reviewed by Scheuring, 2006; see also Chapter 14, Sturgis and Niederman, Chapter 47, Scheuring), AFM of ‘native’ membranes has provided interesting clues on the supramolecular interactions between RCs and LH complexes. A packed association of LH2 complexes and RC-LH1 core complexes is observed in agreement with the high efficiency of the excitonic energy transfer. On the other hand, LH2 complexes appear sometimes segregated in particular regions of the membrane, adopting a quasi-crystalline arrangement. The LH2-only areas should not be too large, however, otherwise the efficiency of energy transfer to the RC is worsened (see the calculations by Ritz et al., 2001). The cytochrome bc1 complex has not been found in AFM images of membranes from different species studied by this technique thus far (except possibly by Scheuring et al., 2004b). In most cases this does not seem attributable to a resolution problem, but to the actual absence of the complex from the investigated material. This is of course at odds with functional evidence. We suspect (see Section X) that the AFM requirement for flat membrane fragments may be the origin of such anomalies.
514
Jérôme Lavergne, André Verméglio and Pierre Joliot
III. The Electron Donors to the Reaction Center In order to stabilize the light-induced charge separation that takes place in the RC, fast electron transfers occur on both the donor and acceptor sides. The immediate electron donor to the RC is a c-type heme, which, depending on species, can belong to a soluble or membrane-bound monoheme cytochrome c, or to a tri- or tetraheme RC-associated cytochrome c subunit. RC-bound tetraheme cytochromes c are present in the majority of photosynthetic bacteria. Recently, the presence of a triheme RC-bound cytochrome subunit has been discovered for species of the Rhodovulum genus (Masuda et al., 1999; Tsukatani et al., 2004). The periplasmic monoheme cytochrome c2, homologous to the mitochondrial cytochrome c, is the immediate electron donor to the photo-oxidized primary donor for species like Rba. sphaeroides, Rba. capsulatus, Rps. palustris or Rsp. rubrum (Meyer and Donohue, 1995). One exception is found in Rba. capsulatus, which contains, in addition to the cytochrome c2, a cytochrome denoted as cy, anchored to the membrane by an α-helix (Jenney and Daldal, 1993; Jenney et al., 1994). A. Tetraheme Reaction Center Subunits The spatial arrangement of the tetrahemic subunit has been resolved at atomic resolution for the RCs of Bcl. viridis (Deisenhofer et al., 1984) and Thermochromatium tepidum (Nogi et al., 2000). The four hemes, numbered 1 to 4 starting from the heme closer to the RC special pair P, extend beyond the RC into the periplasmic space. They are arranged in a roughly linear manner in a high-low-high-low sequence in terms of redox potential (Nitschke and Rutherford, 1989; Verméglio et al., 1989; Alegria and Dutton, 1991; Nitschke and Dracheva, 1995). The midpoint potentials of the high potential hemes (HP) are found in the range 300 mV to 380 mV and those of the low potential hemes range from –80 mV to 130 mV. This means that, under physiological conditions, the low potential hemes are in the oxidized state (except during transient electron transfer steps, see below). The main conclusions that can presently be derived from a number of investigations on native and modified multiheme donor subunits can be summarized as follows. (i) The low potential hemes are active intermediates in electron transfer. Because of the short interheme distances, the kinetic penalty arising
from the uphill steps remains acceptable (Ortega and Mathis, 1993; Page et al., 1999; Chen et al., 2000; Alric et al., 2006). On the other hand, the distance between two consecutive high potential hemes is too large to allow a significant direct tunneling through or beside a low potential heme. (ii) Significant electrostatic interactions (50–70 meV) take place between neighboring redox centers (P-H1, H1-H2, etc.) (Gunner and Honig, 1991; Alric et al., 2004a, 2006). This implies that the effective height of the uphill steps is lower than inferred from redox titrations. It also implies a somewhat variable energy landscape, depending on the number of electrons present in the complex (Alric et al., 2006). (iii) There is no evidence for (but rather evidence against) a role of the low potential hemes as direct partners of a soluble carrier in a hypothetical low potential cyclic regime. Under reducing conditions in vivo where the low potential hemes are reduced in the dark, their reduction after being photo-oxidized is slow, and rapid cyclic flow requires the photo-oxidation of the high potential hemes (Schoepp et al., 1995; Menin et al., 1997; Verméglio et al., 2002). (iv) Although the insertion of low potential hemes in the chain causes no prohibitive slowing (the overall rate across the tetraheme subunit is in the (10 µs)–1 (i.e., 105 s–1) range, i.e., fast with respect to the few ms rate-limiting reactions of the cyclic flow), there is no obvious reason for the conservation of this ‘roller coaster’ pattern. A drift of a low potential heme towards higher potential would not cause much improvement, but apparently would not do any harm either. Such a drift has not been observed, however, and the energy pattern of the multiheme subunits appears well conserved among rather distantly related species. A possible selective advantage has been put forward by Nitschke and Dracheva (1995) whereby the spacers constituted by the permanently oxidized low potential hemes would prevent too large modifications of the energy landscape caused by neighboring interactions, depending on the redox states of the hemes. The connection between the tetraheme RC subunit and the cytochrome bc1 complex is mediated by two types of electron carriers: a soluble c-type cytochrome or a HiPIP depending upon the species (see Ciurli and Musiani, 2005, for a review on the role of HiPIP in photosynthetic bacteria). In Bcl. viridis, Roseobacter denitrificans and Rhodoblastus acidophilus only c-type cytochromes are found, e.g., either cytochrome c2 or the smaller cytochrome c8. Species such as Rubrivivax (Rvi.) gelatinosus,
Chapter 26
Coupling Between RC and Cytochrome bc1 Complex
Ectothiorhodospira shaposhnikovi, Rhodoferax fermentans, Rhodocyclus tenuis, Allochromatium vinosum, and Marichromatium purpuratum possess both types of electron carriers (Bartsch, 1991). For the species possessing both HiPIP and cytochrome c, the nature of the operative electron donor depends upon parameters like growth conditions in Rvi. gelatinosus (Schoepp et al., 1995; Menin et al., 1999; Nagashima et al., 2002) and Allochromatium vinosum (Verméglio et al., 2002), or ambient redox potential in Rhodocyclus tenuis (Menin et al., 1997). The general trend is that HiPIP synthesis is favored under photosynthetic conditions. The half-time of electron transfer between these soluble carriers and the RC-bound tetraheme cytochrome varies between 110 and 400 µs in various species. This fast electron transfer rate is due, in part, to the formation of a tight complex between the soluble carrier and the tetraheme. This has been clearly demonstrated in the case of Rvi. gelatinosus or Bcl. viridis (Garcia et al., 1993; Lieutaud et al., 2003). The binding sites of the c-type cytochrome or HiPIP to the RC-bound cytochrome have been characterized in the case of Rvi. gelatinosus by the analysis of a series of mutants. These sites are located on the subunit surface close to the most distant low potential heme, implying that the electron transfer from the soluble carrier is an uphill step. The binding sites of the cytochrome c8 and of the HiPIP overlap partially (Knaff et al., 1991) but involve different amino acids (Osyczka et al., 1998, 1999a,b). The interaction between the tetraheme and the HiPIP is dominated by the non-polar interaction through their solvent-exposed hydrophobic regions (Osyczka et al., 1999a,b) while cytochrome c8 interacts via salt bridges (Osyczka et al., 1998; Alric et al., 2004b; see Nogi et al., 2005, for a review). B. Triheme Reaction Center Subunits Although the RC-bound cytochrome of Rhodovulum (Rdv.) sulfidophilum has a high similarity to its homolog from Roseobacter denitrificans (40% identity), it presents an unusual characteristic with respect to the well-characterized tetraheme cytochromes. One of the four conserved heme-binding motifs (Cys-XX-Cys-His), corresponding to the most distal heme in the tetraheme subunit of other species, is not present in Rdv. sulfidophilum. This has been established for the RC-bound cytochrome from all species of the genus Rhodovulum studied thus far (Tsukatani et al.,
515
2004). In the case of Rdv. sulfidophilum, the redox arrangement of the hemes is high-low-ultralow, the ultralow mid-point potential (–160 mV) being due to the replacement of the usual methionine by a cysteine as the sixth axial ligand of the third heme (Alric et al., 2004c). The reduction of the photooxidized primary donor by the high potential heme is essentially monophasic, with 90% of P+ being reduced with a half-time of 1.6 µs. Mutational analyses have shown that the electron carrier which connects the RC and the cytochrome bc1 complex in vivo can be either a cytochrome c2 or a membrane-bound monoheme cytochrome c of 50 kDa denoted as cytochrome c2m (Masuda et al., 2002; Kimura et al., 2006) similar to cytochrome cy mentioned earlier, and described below. The rate of electron transfer between the trihemic cytochrome and these secondary donors has not yet been determined. C. Monoheme Electron Donor Among the species that do not possess a RC-bound cytochrome Rba. sphaeroides and Rba. capsulatus have been the most studied. In the case of Rba. sphaeroides, cytochrome c2 is the only intermediate between the RC and the cytochrome bc1 complex. The donor chain of Rba. capsulatus appears more complex. A series of genetic, biochemical and biophysical works by Daldal and coworkers (Jenney and Daldal, 1993; Jenney et al., 1994; Myllykallio et al., 2000) have clearly demonstrated that two distinct carriers are present. The first pathway, corresponding to about 70% of the RCs, involves cytochrome c2 in the same way as in Rba. sphaeroides. In the remaining fraction (30%) of the RCs, a membrane-bound cytochrome named cy provides the functional connection to the cytochrome bc1 complex (Myllykallio et al., 1997). This cytochrome consists of a 100 amino acid-long membrane anchor domain and a cytochrome c domain. Cytochrome cy is also involved in the respiratory pathway where it shuttles electrons between the cbb3-type cytochrome c oxidase and the cytochrome bc1 complex (Hochkoeppler et al., 1995). IV. Kinetics of P+ Reduction by Mobile Cytochromes There is a wealth of information (see Axelrod and Okamura (2005) and Chapter 17, Axelrod et al. for reviews) on the molecular details of the docking of
516
Jérôme Lavergne, André Verméglio and Pierre Joliot
cytochrome c2 to the RC of Rba. sphaeroides, based on the crystallographic 3-D structure of the complex (Axelrod et al., 2002) together with binding and kinetic studies on a number of mutants (Tetreault et al., 2002; Gong et al., 2003) and Poisson-Boltzmann calculations (Miyashita et al., 2004). According to the analysis developed by this group, the first step of the interaction is the electrostatic capture of the cytochrome by the RC within an ‘encounter’ complex. The location of the cytochrome in the encounter complex spreads over a broad region, only slightly smaller than the periplasmic area of the RC and privileging the M subunit. The electrostatic forces exert an efficient couple that rotates the cytochrome into its docking orientation. The attraction range in the transverse direction (where the binding energy exceeds kBT with kB and T being the Boltzmann’s constant and temperature, respectively) is about 20 Å. A striking result of the calculations is to show that in the encounter complex configurations, the energy minimum does not correspond to direct contact (salt bridges) between the negatively charged residues of the RC and positively charged residues of the cytochrome, but maintains a separation distance of 3–5 Å (Miyashita et al., 2003, 2004). At shorter distances, the energy penalty for the desolvation of the two partners would exceed the ‘electrostatic’ gain; the quotation marks are meant as a reminder that the solvation energy is also electrostatic (Honig and Nicholls, 1995). This situation is not trivial (protein association through salt bridges do exist), but appears to be a rather general feature among electron transfer protein complexes that must easily dissociate (Crowley and Carrondo, 2004). The encounter complex is a ‘hors d’oeuvre’ on the way to the more intimate contact achieved in the ‘bound state’. As shown by the crystallographic structure, this direct contact is established between facing patches of hydrophobic residues but still no salt bridges are formed. The bound state positions the two redox centers at a short distance allowing rapid (~1µs) electron transfer. This is about 100-fold faster than the rate calculated by Miyashita et al. (2005) for the encounter state configuration, i.e., through the ≥3 Å water layer separating the two partners, in agreement with data from mutants where the hydrophobic interactions controlling the bound state were suppressed. The light-induced electron transfer reactions between cytochrome c2 and P+ have been extensively studied with isolated components, chromatophores and intact cells of Rba. sphaeroides (Tiede and Dut-
ton, 1993). In the case of experiments with purified complexes the main facts can be summarized as follows. A complex between the RC and reduced cytochrome c2 is formed, with thermodynamic and kinetic characteristics that depend strongly on the ionic strength of the medium. At low ionic strength (we argue in Section XII that the low ionic strength case is relevant to the situation in vivo), the dissociation constant is in the range of 0.3–1 µM and the second-order rate constant for forming the complex is kon ≈ 109 M–1 s–1 (Tetreault et al., 2002). This implies a dissociation rate koff ≈ 2 × 103 s–1 (i.e., a lifetime of the complex on the order of 500 µs). This pertains, however, to the reduced cytochrome. The oxidized form is somewhat more tightly bound (~four-fold; Moser and Dutton, 1988; Larson and Wraight, 2000), but its release rate is similar (~2 × 103 s–1 according to Graige et al., 1998; see also the work of Gerencser et al., 1999, where mammalian cytochrome c was used). The kon rate is high, approaching the diffusion limit, benefiting from the electrostatic steering effect. For cytochrome concentrations on the order of the dissociation constant, two kinetic phases are observed for P+ reduction. The fast phase has a concentration and viscosity independent rate of about (1 µs)–1 and corresponds to the population of pre-formed RC-c2 complexes. The slow phase corresponds to the second-order encounter of the partners. There has been a debate about what happens at saturating cytochrome concentration (>> 1 µM). In addition to the fast (1 µs) phase, a slower component (100 µs) has been observed. Its rate and relative amplitude does not depend on cytochrome concentration, but it is sensitive to the viscosity of the medium. A model was put forward by Overfield et al. (1979) and elaborated by Moser and Dutton (1988) involving an equilibrium between two conformations (proximal and distal) of the complex. This behavior was not confirmed, however, in the systematic studies carried out by Okamura, Feher and coworkers with wild type and mutant RCs (Rosen et al., 1980; Tetreault et al., 2001). This issue was examined by Tiede and coworkers (Tiede and Dutton, 1993; Tiede et al., 1993) who concluded that the appearance of the slow first-order phase was a variable phenomenon, which they tentatively ascribed to the formation of RC aggregates, in relation to the supercomplex model (see below). An alternative explanation is the occurrence of double photochemical turnovers when using a flash of too long duration, so that the slow phase would reflect the delay for releasing a first (oxidized) cytochrome
Chapter 26
Coupling Between RC and Cytochrome bc1 Complex
and binding a second one (Larson and Wraight, 2000). This explanation cannot apply, however, to experiments where a short laser flash was used (e.g., Overfield et al., 1979; Tiede et al., 1993). A biphasic photooxidation has also been observed in isolated chromatophores or intact cells of Rba. sphaeroides (Overfield et al., 1979) using a short laser flash. An alternative explanation to the ‘proximaldistal’ model has been proposed in the case of intact cells to explain the observed biphasicity (Joliot et al., 1989). In this model, the two phases are caused by the confinement of a single cytochrome c2 within a supercomplex containing two RCs and one cytochrome bc1 complex (see Section VII). The kinetics of cytochrome cy photooxidation have been studied in detail in Rba. capsulatus, in addition to the cytochrome c2 reactions, which are similar to those that occur in Rb. sphaeroides. Cytochrome cy reduces the photo-oxidized RC in two distinct phases with half-times of 5 µs and 40 µs (Myllykallio et al., 1998). The fast phase is attributed to electron donation within a proximal complex between the cytochrome cy and the RC while the 40 µs phase, slowed in the presence of glycerol, is probably binding-limited. The very efficient light-induced cyclic electron transfer in Rba. capsulatus implies a small distance between RC, cytochrome cy and cytochrome bc1 complex, i.e., some supramolecular organization as proposed for Rba. sphaeroides. In Rsp. rubrum, in contrast to Rba. sphaeroides, there is no tight binding of cytochrome c2 on the RC and the electron transfer kinetics are clearly secondorder (van der Wal and van Grondelle, 1983; Joliot et al., 1990). The residues that have been identified in the Rba. sphaeroides studies as playing a key role in the docking of the cytochrome are all present in the RC and cytochrome of Rsp. rubrum, with one exception. This concerns the absence of the Arg 32 of the cytochrome c2 in Rba. sphaeroides, which has no equivalent in Rsp. rubrum. This destroys the cation-π interaction with an aromatic residue (TyrM295 in Rba. sphaeroides—changed to a phenylalanine in the Rsp. rubrum RC), and removes the electrostatic interaction with negatively charged residues on the RC. According to the study of Rba. sphaeroides mutants described by Paddock et al. (2005), the electrostatic effect is predominant. An 80-fold increase of the dissociation constant was observed when replacing the arginine with an alanine, mimicking the behavior of Rsp. rubrum.
517
V. Donor Side Shuttling and Turnover of the Cytochrome bc1 Complex The turnover of the cytochrome bc1 complex can be followed easily by monitoring the light-induced carotenoid band shift. This absorption change is due to an electrochromic effect, and it constitutes a very sensitive endogenous probe of the membrane electric potential. Following a short flash, the signal presents three distinct phases. A very fast first phase corresponds to the initial charge separation events in the RC. A second phase, occurring in the tens of microseconds range, is linked to the electron transfer from the primary (QA) to the secondary (QB) quinone and associated proton uptake and to the reduction of P+ by its secondary donor(s). The slowest phase is indicative of the membrane potential increment due to the electrogenic Q-cycle reactions taking place in the cytochrome bc1 complex. The rise develops over a time ranging from a few milliseconds to some tens of milliseconds, depending on conditions and organism. Its amplitude is similar to that of the combined fast phases. In general, the functioning of the cytochrome bc1 complex is the limiting step of the light-induced cyclic electron transfer in photosynthetic bacteria. This is especially true under physiological conditions and high light intensity, because the Q-cycle reactions are strongly coupled to the generation of the ∆µ~H and their rate is bound to decrease under high load conditions. But even under uncoupled conditions, the cytochrome bc1 complex is generally a bottleneck for the cyclic flux. Two factors are of importance in this respect: the redox state of the quinone pool and the ratio between the cytochrome bc1 complex and the RC. The fastest rate for the flash-induced ‘slow phase’ of the membrane potential rise (1 ms half-time at room temperature) is observed for chromatophores or cells of species containing large amounts of cytochrome bc1 complex (i.e., about 0.5 bc1 monomer per RC), like Rba. sphaeroides or Rba. capsulatus, when the quinone pool is substantially reduced. It is important to note that the functioning of the cytochrome bc1 complex (Table 1) requires an oxidized Q at its Qi site to accept electrons coming from the oxidation of the QH2 at the Qo site through the bL and bH cytochromes. In anaerobic conditions where the quinone pool is fully reduced, the only available oxidized Q is that formed at the site Qo of the cytochrome bc1 complex (Fig. 2). This Q must therefore bind preferentially to the Qi site, rather than to the QB site of
518
Jérôme Lavergne, André Verméglio and Pierre Joliot
Fig. 2. Schematic representation of the light-induced Q-cycle (see Table 1) under anaerobic conditions where all quinones are reduced, according to Joliot et al. (2005). Starting from a dark-adapted state where heme bH is reduced and heme bL oxidized, the Q released from site Qo at the first turnover is transferred to site Qi and reduced by oxidizing the two b hemes. The second turnover restores the initial state bL bH– and the Q released from site Qo migrates to the QB pocket of the RC, allowing reoxidation of QA– and sustaining photochemical activity.
the RC, in order to oxidize the b hemes and allow further turnovers of the cytochrome bc1 complex. In contrast, the Q formed during the second turnover of the cytochrome bc1 complex has to reach the QB site of the RC to permit further charge separations. For species which contain a low bc1:RC ratio, i.e., Rvi. gelatinosus or Bcl. viridis, the completion of the electrogenic reactions in the cytochrome bc1 complex following a saturating flash requires several tens of milliseconds. This rather low rate is due to the multi-turnover process of the cytochrome bc1 complex necessary to rereduce the large amount of photo-oxidized RCs generated by a saturating flash. Presumably, the low bc1:RC ratio reflects an adaptation to weakly illuminated niches, as an alternative (not necessarily exclusive) to increasing the size of the LH antenna. No structural information is yet available for the docking between the cytochrome c2 (or the HiPIP) and the bacterial cytochrome bc1 complex. However these proteins present a high homology with their mitochondrial counterparts, and one expects similar modes of docking. In addition to mutagenesis studies, the analysis of co-crystals of mitochondrial cytochrome c and cytochrome bc1 complex has allowed a detailed characterization of their interactions (Xia et al., 1997; Lange and Hunte, 2002), which are essentially mediated by nonpolar forces. The binding configuration sets a short distance between the two hemes, which are facing each other with their solvent exposed edges, allowing direct electron transfer. As observed for the cytochrome c2-RC interaction, the oxidized form of the cytochrome c has a higher affinity towards the cytochrome bc1 complex (Speck and Margoliash, 1984).
VI. Quinone Reactions The mobile electron (and proton) carrier that shuttles between the RC and the cytochrome bc1 complex is a ‘long tail’ ubiquinone in all photosynthetic anoxygenic bacteria (lower potential quinones are also often present, see Section I). The 8–10 unit long isoprenoid chain implies a severe hydrophobicity of this molecule that must remain confined to the lipid bilayer and to the hydrophobic protein interface. The stable redox states are the fully oxidized Q and the fully reduced QH2, carrying two electrons and two protons. The ‘bucket brigade’ hydrogen transfer (QH2 + Q → Q + QH2) is thermodynamically unfavorable (Rich, 1982) so that the electron transport requires the physical displacement of the quinones between protein binding sites where the two electrons per proton transfer reactions are appropriately catalyzed. The quinones are in marked stoichiometric excess with respect to the electron transfer proteins. For instance, in anaerobically grown Rba. sphaeroides, one has about 20–25 UQ molecules per RC. We use the term quinone pool to acknowledge this fact, leaving for a later discussion the issue of sharing of quinones by the electron transfer complexes. There is reasonably good agreement between the biochemical estimates of the amount of solvent-extractable quinones per RC (Takamiya and Dutton, 1979) and the functional stoichiometry based on the amount of photoreducible quinones per RC (Mezzetti et al., 2003; Comayras et al., 2005a). In other words, approximately all the quinones present in the membrane have an easy access to the QB pocket of one or several RCs. The two electron transfer reactions involved in the Q to QH2 interconversion are handled quite differently at the quinol oxidation site Qo of the cytochrome bc1
Chapter 26
Coupling Between RC and Cytochrome bc1 Complex
complex, and at the quinone reducing sites Qi (bc1) and QB (RC). In the case of Qo, the two electron transfer steps occur in a more or less concerted way involving two distinct acceptor chains. In the Q reducing sites, the reactions occur as a stepwise process, involving a singly reduced intermediate. In the case of the RC, the cycle starts with the binding of an oxidized quinone in the QB pocket. This molecule is reduced by two successive photochemical turnovers, involving a semiquinone radical QB– as the intermediate state, which is stabilized by the RC (this results in a leveling of the redox potentials of the two transitions). The reduced and protonated QH2 is then released from the RC. The overall cycle (in chromatophores with an oxidized pool, saturating illumination and an excess of fast electron donor) takes about 1.6 ms (Comayras et al., 2005a). This includes the two electron transfer reactions (both in the 100 µs range), the associated proton traffic and the delays for binding the Q and releasing the QH2. The time constant for each of the two reduction reactions of P+ by an exogenous donor was also about 100 µs. Thus, roughly, the Q binding-release steps require altogether about 1.2 ms. Milano et al. (2003) analyzed the binding and release rates of the oxidized UQ-10 for RCs in liposomes. The value obtained for the kon was close to the value estimated by these authors for the diffusion-limited case (≈3 × 105 M–1s–1, in terms of the Q concentration in the lipid space). From these results, one may estimate the binding rate in chromatophores with 20 UQs per RC to be ~(40 µs)–1. Furthermore, they estimated the dissociation time constant of the oxidized quinone as ~25 ms. The data indicate a high binding affinity of the Q and predict that the QB site should be 100% occupied in the presence of the oxidized pool. These results suggest that the dominant contribution to the overall turnover time must arise from the release of the Q molecule. If the duration of this process is about 1.2 ms, this implies that the QH2 leaves the QB pocket about 20-fold faster than Q. Assuming that the kon is similar for the two forms, this predicts a ~20-fold lower affinity of the QB pocket for QH2 in reasonable agreement with the factor of 10 estimated from thermodynamic data (Crofts and Wraight, 1983). The transit time of the QH2 released from the RC to its oxidizing site (Qo) on the cytochrome bc1 complex has been estimated from the flash-induced reduction of the high potential heme bH of the cytochrome bc1 complex in the presence of antimycin, which blocks the cytochrome re-oxidation at the Qi site. When the
519
ambient potential is such that the quinone pool is fully oxidized, QH2 can only be produced by the RC every second flash. Then, the time course of heme bH reduction monitors the arrival of the QH2 released from the RC (the experiment is best done at a potential high enough to oxidize the high potential chain of the cytochrome bc1 complex so there is no need to bother about a possible limitation from cytochrome c2). The kinetics start after a ~1 ms lag and the overall reaction has a half-time of about 5–10 ms in Rba. sphaeroides (Crofts et al., 1983). The 1 ms lag is consistent with the delay estimated above for QH2 release from the RC. The subsequent 5–10 ms reaction then represents essentially the time required for the released QH2 to reach the cytochrome bc1 complex, penetrate the intra dimer cavity, and bind at site Qo, competing with the 20-fold excess of oxidized quinones. At low redox potential, the cytochrome bc1 complex can react with QH2 already present: the duration of the lag becomes markedly smaller and the reaction rate is accelerated by up to five-fold, becoming limited by the reactions taking place in the cytochrome bc1 complex (Crofts et al., 1983). VII. Supramolecular Organization in Rhodobacter sphaeroides and Rhodobacter capsulatus There has been a long-standing debate in the Bioenergetics community between the ‘liquid state’ and ‘solid state’ conceptions, following Rich’s terminology (Rich, 1984). The issue concerns the effective size of the functional units, i.e., small aggregates of complexes (solid state view), or, virtually the whole membrane (fluid view). The fluid model puts emphasis on the mobility of the electron transfer shuttles (quinones and cytochrome c), which diffuse rapidly so that the electron transfer is effectively delocalized and mediated by random collisions. In the solid-state model, the various components are arranged in supramolecular assemblies that are able to retain the mobile carriers, and thus ensure local channeling between complexes. Compared with other bioenergetic membranes, photosynthetic membranes have the specific property of containing large amounts of LH complexes. The efficiency of the excitonic energy transfer between these LH complexes and the RC requires some supramolecular arrangement that ensures the closeness of the LH and RC chromophores (Chapter 13, van Grondelle and Novderezhkin;
520
Jérôme Lavergne, André Verméglio and Pierre Joliot
Chapter 14, Sturgis and Niederman; Chapter 15, Şener and Schulten). On the other hand, while surrounding the RC with antenna proteins is advantageous in respect to excitation transfer, it may pose a problem in respect to electron transfer, by increasing the RCbc1 separation and by cramming the intermediary space with obstacles to the diffusion of quinones. In this respect, the fluid membrane picture evoked in the random collision model is difficult to reconcile with the dense packing with the RC-LH1 and LH2 complexes as imaged by electron microscopy or, with much better resolution, by AFM. These images show little space left for the lipids and quinones. This difficulty is reinforced by the fact that one expects the lipids at the boundary of proteins to be immobilized (Li et al., 1989; Kirchhoff et al., 2002). Relatively long range diffusion of the mobile electron shuttles is a clear necessity in membranes where the bc1:RC stoichiometry is low (such as Blc. viridis, Rvi. gelatinosus, Rsp. rubrum). One may then expect a relatively slow diffusion rate of the quinones due to the highly obstructed route. It is noteworthy, however, that the slow diffusion rate can be counteracted by increasing the Q:RC ratio. For instance, a net flux of one electron per ms can be sustained by one Q making a round trip of 1 ms between one RC and one cytochrome bc1 complex; alternatively, the same flux can be sustained for a round trip duration of 20 ms if the Q:RC stoichiometry is 20. Whether long-range diffusion in such species occurs between randomly dispersed protein complexes, or it implies some organization of membrane domains and/or some channeling of the mobile carriers is an open question (Scheuring and Sturgis, 2006). We will turn now to the case of species like Rba. sphaeroides and Rba. capsulatus, which possess a high bc1:RC stoichiometry. A series of thermodynamic and kinetic studies has provided new insight into the organization of the components of the photosynthetic electron chain of these two bacteria. In the case of intact cells of Rba. sphaeroides strain R-26 (a strain devoid of carotenoids and of LH2), a key observation (Fig. 3) was that the ‘apparent equilibrium constant’ between P and its secondary donors (cytochrome c2 and the cofactors of the high potential chain of the cytochrome bc1 complex: heme c1 and the iron sulfur center (FeS)), measured during their photooxidation (with the cytochrome bc1 complex inhibited at the Qo site), was much lower than the true equilibrium constant deduced from the mid-point potentials (Joliot et al., 1989).
‘Low apparent equilibrium constant’ means that the accumulation of oxidized P+ occurred much sooner than would be expected if it could react freely with its pool of secondary donors. The phenomenon was ascribed to a confinement effect, i.e., to the distribution of photochemical hits on small, independent systems — supercomplexes — each containing two RCs, one cytochrome bc1 complex and one cytochrome c2. The two Ps can thus equilibrate with only three carriers (c2, c1 and Fe-S) whose potentials are markedly more reductive. As a first approximation, the amount of P+ present after some illumination time corresponds to supercomplexes that have received more than three photons (1 P+ in supercomplexes which have received just four photons and 2 P+ in those which have received five or more). Because of the small size of the system, the Poisson distribution is broad and P+ accumulates much earlier than would be the case in a larger system where no P+ would be observed until almost all the secondary donors have been oxidized (Lavergne et al., 1989). The equilibration (e.g., through the exchange of cytochrome c2) between these domains occurs on a much slower time scale. Based on chromatophore heterogeneity, an alternative explanation for the low apparent equilibrium constant was proposed by Crofts et al. (1998): we discuss it below in Section X. In addition to the supercomplex [RC]2[c2][bc1], the observations on strain R-26 revealed the presence of a fraction of ‘incomplete’ supercomplexes, lacking the cytochrome bc1 complex. About 30 % of the RCs were present in these [RC]2[c2] structures. A still more dramatic heterogeneity was found when the same analysis was applied to whole cells of ‘wild type’ Rba. sphaeroides (strain Ga) (Verméglio et al., 1993). The supercomplexes [RC]2[c2][bc1] were found to comprise ~80 % of the RCs. The other 20 % of the RCs were interacting with an excess of cytochrome c2 (~7 c2 per RC). This contribution appeared as a marked slow phase in the cytochrome photo-oxidation kinetics. It was observed that (i) oxygenation of the sample preferentially oxidized the cytochromes c2 involved in the slow phase rather than the cytochromes engaged in the supercomplexes, and (ii) the ratio between the fast and the slow phase increased when increasing, through growth conditions, the amount of invaginated membranes. Based on these observations, it was proposed that only the RCs present in the invaginated part of the membrane are organized in supercomplexes while those located in the cytoplasmic part of the membrane share the cytochromes c2
Chapter 26
Coupling Between RC and Cytochrome bc1 Complex
and cytochrome bc1 complexes with the respiratory chains. This sharing of the cytochromes c2 between RCs and cytochrome oxidase in the cytoplasmic part of the membrane is responsible for the direct competition observed between the photosynthetic and respiratory chains (Verméglio and Carrier, 1984). The specific association of cytochrome c2 with the RCs and the cytochrome bc1 complex in the invaginated parts of the membrane limits its interaction with other complexes present in the cytoplasmic membrane or the periplasm, such as the cytochrome oxidase; this favors photosynthetic activity over respiration (Verméglio et al., 2004). An intriguing result in the work with Ga cells is that the confinement of cytochrome c2 responsible for the biphasic photo-oxidation kinetics was dramatically affected when decreasing the pH (the half effect was ~pH 7.5) or in the absence of divalent cations. Under such conditions, the delocalized equilibration of the donor chain occurred over a few tens of milliseconds. The mechanism and possible physiological role of this switch remain unknown. The confinement effects implied by the supercomplex model leads to specific predictions concerning the effect of subsaturating inhibition of the cytochrome bc1 complex. Partially conflicting results have been obtained on this issue. Fernandez-Velasco and Crofts (1991) studied the effect on isolated chromatophores of subsaturating concentrations of stigmatellin, an inhibitor of the cytochrome bc1 complex, and observed a linear dependence of the rate of cytochrome c2 re-reduction on the fraction of inhibited complexes. The results imply that the cytochrome c2 diffuses and interacts with a relatively large number of cytochrome bc1 complexes that could represent the content of the whole vesicle. Using the same type of approach, also with isolated chromatophores, but with myxothiazol as the cytochrome bc1 inhibitor, Joliot et al. (1996) obtained different results. The diffusion of a given cytochrome c2 appeared restricted. On a short time scale (millisecond range) the donor chain equilibrated over a domain with the size of a supercomplex. In the 100 ms time-range, the equilibrium was extended over a two-fold larger domain that could be a dimer of supercomplexes. The discrepancies between the results of both groups remain to be clarified. In intact cells of Rba. sphaeroides strain Ga, Joliot et al. (1996) observed that the addition of a subsaturating concentration of myxothiazol did not affect the rate of electron transfer for the uninhibited complexes. Moreover, the reduction of the cytochrome
521
c2 fraction associated with the inhibited cytochrome bc1 complexes was blocked, showing that it cannot interact with the uninhibited chains. Crofts (2000a,b) also noted a more restricted diffusion of cytochrome c2 (compatible with the size of 1–2 supercomplexes) when using intact cells rather than isolated chromatophores. It is worth noting that the same type of experiment realized with intact cells of Bcl. viridis, a species which contains a large excess of RC over the cytochrome bc1 complex, clearly gave the opposite result, reflecting free diffusion of the mobile cytochrome connecting these two complexes. Following its flash-induced photo-oxidation, the totality of the cytochrome c2 was rereduced irrespective of the myxothiazol concentration, with a rate depending on the fraction of active cytochrome bc1 complexes (Garcia et al., 1993). Another indication of a necessary proximity between the RC, cytochrome c2 and cytochrome bc1 complex is that the complete cyclic photoinduced electron transfer can occur at −20 °C in a frozen medium in intact cells of Rba. sphaeroides (Joliot et al., 1997). Such a behavior is not observed for example in the case of Rsp. rubrum where the light-induced cyclic electron transfer is blocked already at –5 °C (Joliot et al., unpublished results). A supramolecular organization of the photosynthetic chain has also been proposed in the case of Rba. capsulatus (Myllykallio et al., 2000). As already described, this species possesses, in addition to cytochrome c2, a membrane-bound cytochrome cy which efficiently connects the cytochrome bc1 complex to the RC (Jones et al., 1990; Jenney and Daldal, 1993). A series of biochemical and functional approaches has shown that, due to its membrane attachment, the movement of cytochrome cy is restricted to a small number of RCs and cytochrome bc1 complexes, implying, as for Rb. sphaeroides, some clustering of the RC, cytochrome bc1 complex and cytochrome cy (Myllykallio et al., 1998; Verméglio et al., 1998). Two cytochromes cy can only interact with two RCs and one cytochrome bc1 complex. Additional evidence for this clustering has recently been reported by Lee et al. (2006) who engineered a strain of Rba. capsulatus deprived of cytochrome c2 and where the cytochrome domain of cy was fused to the cytochrome bc1 complex: a substantial cyclic transfer allowing photosynthetic growth was restored in this strain. Finally, we wish to briefly discuss the case of Rsp. rubrum. It was proposed that the very biphasic oxidation of cytochrome c2 observed in this bacterium was
522
Jérôme Lavergne, André Verméglio and Pierre Joliot
related to a dimeric association of the RCs (Joliot et al., 1990). However, this interpretation seems now hardly tenable considering the structure found for the core complexes in this bacterium, with a closed LH1 ring forming a roughly circular envelope around the RC (Jamieson et al., 2002). This suggests no plausible attachment site for forming dimers. A product inhibition mechanism similar to the one put forward by Moser and Dutton (1988) may provide a valuable alternative explanation for the results. One has to assume that the oxidized cytochrome is rapidly released from the RC with a reduced P, but somehow rebinds to RCs in the P+ state, hindering their reduction by reduced cytochrome c2. This model was discarded by Joliot et al. (1990) because it would fail to account for the observed dependence on the energy of the flash. On second thoughts, we believe that this statement was wrong, and that this model does predict correctly the quadratic dependence on flash energy. If this is eventually the correct interpretation, it remains to explain why the oxidized cytochrome has a particular affinity for the oxidized RC. The functional aspects of supercomplexes described in this section have implied essentially a confinement of the donor side mobile carrier (cytochrome c2). We will now examine in this respect the other mobile shuttle between the RC and the cytochrome bc1 complex, i.e., the Q acceptor pool. VIII. Quinone Confinement in Rhodobacter sphaeroides The question of the extent of pooling of the quinones can be addressed by asking how many RCs share a common pool. This can be determined by monitoring the kinetics of quinone photoreduction under ‘cul de sac’ conditions, i.e., with inhibited cytochrome bc1 complexes and a non-limiting supply of electron donors to the RC. Then, the effect of a partial inhibition of the RCs will be indicative of the ‘size’ of the quinone pool. If n RCs share a common pool of quinones (for brevity, we denote n as the ‘domain size’) and assuming a fraction f of inhibited centers, the amount of photoreducible quinones will be diminished by a factor f n, i.e., the fraction of domains where all RCs are inactive. Experiments along these lines were carried out in chromatophores of Rba. sphaeroides wild type or PufX– mutant (Comayras et al., 2005a,b). The results were clearly indicative of a small average domain size. For small fractions of
inhibited RCs, the decrease of the size of the photoreducible pool was consistent with a domain size of n ≈ 2 RCs. For larger inhibition levels the data were indicative of larger values of n, implying a significant heterogeneity of the system. The authors retained an overall figure of n ≈ 4 RCs. Evidence for a confinement of the quinones to a supercomplex structure associating the RC and the cytochrome bc1 complex emerged from a kinetic investigation examining how the freshly oxidized quinone released from the Qo site of the cytochrome bc1 complex reaches the RC (Joliot et al., 2005). The experiments were carried out with cells of Rba. sphaeroides under anaerobic conditions. Most of the quinone pool is then reduced, and most RCs lack an oxidized secondary quinone (QB) able to reoxidize rapidly the primary acceptor QA–. For these centers, the decay of QA– following a flash expresses the arrival of an oxidized quinone formed by the cytochrome bc1 complex. About half of the QA– was thus rapidly reoxidized (within ~200 ms at 0 °C), while the remaining fraction was much more slowly reoxidized. The rapid phase was interpreted as the transfer of quinone formed at site Qo to the QB pocket of a RC belonging to the same supercomplex (or possibly to the same dimer of supercomplexes, see Section X). In agreement with this proposal, the fast phase of QA– was specifically inhibited by the cytochrome bc1 inhibitor myxothiazol. The multiphasic slow phase expressed the redistribution of the few oxidized quinones present before flash excitation, involving longer-range diffusion. In agreement with these interpretations, the PufX-deleted mutant (further examined in the next section) displayed only the slow phases of QA– oxidation. This implies that, in the absence of a supramolecular association of membrane complexes, the average distance between site Qo and site QB is larger than in its presence. IX. Quinone Traffic in the PufX– Mutant of Rhodobacter sphaeroides Interestingly, the PufX– mutation caused quite significant modifications as regards the quinone domains (Comayras et al., 2005b). The average number of RCs per domain was much smaller (n ≈ 1.7 RCs) and the dependence of the amount of photoreducible quinones on the fraction of inhibited RCs was indicative of a more homogeneous system. Whereas the domain size was smaller in terms of the number
Chapter 26
Coupling Between RC and Cytochrome bc1 Complex
of RCs sharing a common pool of quinones, the stoichiometric ratio Q:RC was about doubled in the mutant (~50 quinones per RC) compared with a wild type strain (~25–30 quinones per RC in wild type grown under the same semiaerobic conditions; this is slightly larger than the figure of ~20 observed in membranes from photosynthetic cultures). The photosynthetic incompetence of the PufX– strain was related to a drastic slowing of the quinone turnover on the RC. When the effect of deleting the PufX subunit on the structure of the RC-LH1 complex was clarified (see Section II) this seemed to provide an obvious explanation for the inhibition of the quinone shuttling. Whereas the open ‘S’ structure of the RC-LH1 dimers in a wild type strain presents a clear passage for the traffic of quinones, the closed LH1 ring of the RC-LH1 monomers in PufX– appears impassable. There are, however, some difficulties with this explanation. A first point is that, as previously noted, the closed LH1 ring is the native structure in a number of bacteria. The second point concerns the effect of the ambient redox conditions on the quinone turnover in PufX–. Barz et al. (1995a,b) showed that the photoinduced cyclic flow of PufX– cells in vivo was severely inhibited under anaerobic conditions. This was observed from the kinetics of the carotenoid shift (indicating the buildup of the membrane potential) and those of cytochromes c2 and c1 oxidation, under a pulse of strong continuous light. Both signals were markedly diminished in the mutant, consistent with an accumulation of quinol blocking the RCs on the acceptor side. On the other hand, this inhibition was completely relieved when the cell suspension was made aerobic (saturating the medium with oxygen) or when adding TMAO to the medium. The Q pool was largely oxidized by the former treatment and only partially by the latter. The results suggest that Q shuttling is specifically inhibited in the PufX– strain under anaerobic conditions and restored to normal under aerobic, or more oxidizing (TMAO) conditions. It should be realized that models based on a diffusion barrier for quinones (the closed LH1 ring) cannot explain the dramatic contrast between aerobic/anaerobic conditions. Under steady-state shuttling conditions, the kinetic limitation implied by the barrier will limit the flux in the same way, irrespective of the relative abundances of the Q and QH2 forms. The question of the passage of quinones across the LH1 barrier was addressed in the study of Comayras et al. (2005b). According to this work, the overall turnover time for the QB site is increased from ~1.6
523
ms in the wild type strain (see Section VI) to ~3.5 ms in the PufX– mutant. This increment of ~2 ms is the time required to cross the LH1 barrier twice (inward passage of one Q and outward passage of one QH2), so that the individual passage requires ~1 ms. No evidence for a rapidly accessible ‘internal’ pool (between the RC and LH1 ring) was found, suggesting that the LH1 wall is close to the opening of the QB pocket. The quinol diffusion from the RC to the cytochrome bc1 complex was studied in PufX– chromatophores by monitoring the flash-induced reduction of the high potential heme of the cytochrome bc1 complex in the presence of antimycin. As explained in Section VI, when the ambient potential is such that the Q pool is oxidized, the time course of this reaction monitors the arrival of the QH2 released from the RC. This experiment was done by two groups, and conflicting results were obtained. According to Barz et al. (1995b) and Gubellini et al. (2006), a dramatic slowing of this reaction was observed in the PufX– membranes. The initial lag of ~1 ms in the wild type strain was increased to 10 ms and the subsequent rate of the kinetics was slowed more than 10-fold. Much smaller effects were reported by Comayras et al. (2005b) who observed a lag of ~2 ms and a half-time of ~14 ms (versus 7 ms in wild type membranes). The drastic inhibition found by Barz et al. (1995b) is difficult to reconcile with the observations reported in the same paper showing that a normal cyclic flow, implying a high shuttling rate between the two complexes, was observed in aerobic PufX– cells. Keeping this difficulty in mind, we adopt in the following discussion the picture proposed by Comayras et al. (2005b). The unexpectedly modest effects of the PufX– mutation, i.e., a 1 ms delay for Q passage across the LH1 barrier and a 7 ms longer journey to the cytochrome bc1 complex do not suffice to account for the photosynthetic incompetence of this strain. The interpretation proposed by Comayras et al. (2005b) is that the RC functioning is modified in the PufX– RCLH1 complexes, so that it becomes inhibited by QH2 molecules. The mechanism for this effect is not fully clear, but various properties of the QB pocket were indeed found to be modified, such as the affinity for inhibitors and the QA– QB– → QAQBH2 electron transfer rate. These modifications may be due to inappropriate interactions between the modified LH1 ring and the RC. This model accounts for the specific inhibition of the cyclic flow in anaerobic cells and its restoration
524
Jérôme Lavergne, André Verméglio and Pierre Joliot
in aerobic conditions. The rapid (1 ms) passage of quinones across the LH1 ring in a PufX– mutant is consistent with the fact that many strains natively possess a closed LH1 ring. The question is, however, how does it work? The best-resolved crystallographic information on the oligomeric association of the αβ subunits was obtained for the core complex of Rps. palustris (Roszak et al., 2003). Leaving aside the open region of the ‘horseshoe’ arrangement of this particular core complex, this does give an idea of the tightness of the LH1 wall — which presents no obvious interstice through which quinones might thread their way. However, the real world is not a crystal, and local deformations as well as Brownian shaking of the assembly might provide sufficient openings. The AFM images are more realistic in this respect. Notably, these images show that the RC-LH1 complex is elliptical in bacteria where the LH1 surrounds the RC, whereas the LH1 ring is cylindrical in the absence of the RC (Scheuring, 2006). Indication of a particular flexibility of the core complex has been reported by Bahatyrova et al. (2004b) who have pinpointed a difference with LH2 as regards the hydrogen-bonding mode of the subunits that could account for a greater lability of the LH1 subunits. In recent molecular dynamics study (Aird et al., 2007) a relatively rapid (≤1 µs) passage time of the quinones across the LHI ring of Rsp. rubrum was estimated. A last remark on this subject is that the LH1 structure (and presumably, that of LH2 as well) does not provide the fence that would contain the quinones and account for the confinement described in Section VIII: we will come back to this in Section XIII. X. The Supercomplex Model: Difficulties and Alternative Possibilities In the supercomplex model, the low apparent equilibrium constant observed during the photooxidation of the electron donor chain (P, c2, c1 and FeS) was ascribed to the distribution of photochemical hits on a small system containing 5 redox centers, as explained in Section VII. Another situation that can produce a low apparent equilibrium constant is that of a distribution affecting the stoichiometric composition of the system (Lavergne and Joliot, 1991). Crofts et al. (1998) proposed a reinterpretation of the apparent equilibrium data based on such a possibility. In this model, cytochrome c2 diffuses freely
over the whole internal surface of the chromatophore. Due to the small size of the vesicle, the components of the photosynthetic chain are present in small numbers and one expects fluctuations resulting in a spread of the stoichiometric distribution. Some chromatophores will have a higher than average ratio of P over secondary donors, and will consequently display an earlier photooxidation of P. As pointed out by Crofts, this effect can account for the observed low apparent equilibrium constant, provided the ‘size of the chromatophore’ is small enough. This is illustrated in Fig. 3, showing the data reproduced from the paper by Joliot et al. (1989), the simulation based on the supercomplex model (curve 1) and on the Crofts’ model (curves 2 and 3). Curve 2 uses the mean values recommended by Crofts, i.e., 16 RCs, eight cytochromes c2 and four cytochrome bc1 dimers per chromatophore. These values are smaller than previous estimates, i.e., 50–60 (Saphon et al., 1975) or 30 RCs per chromatophore (Crofts et al., 1983). The mean diameter of chromatophores has been measured as 60 nm (Saphon et al., 1975; Crofts et al., 1983; Drews and Golecki, 1995), which can accommodate ~30 RCs taking into account the sizes of the RC-LH1 and LH2 complexes. Despite the twofold smaller figures adopted for computing curve 2, it still fails to mimic the observed low equilibrium constant (but is nevertheless far from the true equilibrium corresponding to curve 4). It is of interest to see how small the system size should actually be in order to match the data. Such is about the case for curve 3, which corresponds to half the size used by Crofts, (i.e., eight RCs, four cytochromes c2 and two cytochrome bc1 complex dimers). This is clearly much smaller than the chromatophore content. The supercomplex model for Rba. sphaeroides was formulated as a structural prediction based on functional data. The arrangement of the LH1 around the RC was not clear at that time, and the dimeric structure of the cytochrome bc1 complex was not established either. An updated view of the supercomplex must somehow cope with these new facts. One feature of the model’s structural predictions has been confirmed, concerning the dimeric association of the RCs, tied up by the S-shaped chain of LH1-PufX. On the other hand, things are not as clear in respect of the cytochrome bc1 complex. A first problem is that this latter complex does not seem to be present in the AFM images obtained thus far. Tentatively, one may surmise that the flattening of the intracytoplasmic membrane attached to a mica
Chapter 26
Coupling Between RC and Cytochrome bc1 Complex
Fig. 3. The problem of the low ‘apparent equilibrium constant’ in the donor chain of Rba. sphaeroides. The fraction of reduced P is plotted vs. the fraction of reduced cytochromes (summing signals from the mobile carrier c2 and heme c1 of the cytochrome bc1 complex) during the photo-oxidation kinetics monitored under a continuous illumination of relatively weak intensity. The cytochrome bc1 complex is inhibited by myxothiazol, blocking QH2 oxidation at the Qo site. The datapoints (taken from Joliot et al., 1989) were obtained at various times during the kinetics, starting from the top right corner (dark-adapted state with the fully reduced donor chain) and ending at the bottom left corner (complete photo-oxidation). Curve 1 results from the simulation based on the supercomplex model, assuming that 70% of the RCs belong to complete supercomplexes (one dimer of RCs and one cytochrome bc1 complex interacting with a single cytochrome c2) and 30% to incomplete supercomplexes lacking the cytochrome bc1 complex. Curve 2 was computed from the assumptions put forward by Crofts et al. (1998), whereby cytochrome c2 equilibrates with all the RCs and the cytochrome bc1 complexes present in individual chromatophores. The stoichiometries of the components are distributed over the population of chromatophores, assuming Poisson distributions around mean values of 16 RCs, 8 cytochromes c2 and 4 cytochrome bc1 complex dimers (99% of the chromatophore population was taken into account in the simulation). Curve 3 was obtained following the same procedure, but assuming two-fold smaller mean values. Curve 4 is the equilibrium relationship that would be obtained if the mobile cytochrome c2 were able to interact with a large number of partners (with a 2:1:1 stoichiometry for RC, c2 and cytochrome bc1 monomer, respectively). The calculations were carried out according to Lavergne et al. (1989), but without the assumption of an infinite equilibrium constant between P and its partners. The midpoint potentials used for all simulations were 450, 340, 270, and 300 mV for P, c2, c1 and FeS, respectively.
plate for AFM exploration causes a disruption of the supercomplexes, expelling the cytochrome bc1 complex out of the flat region (or alternatively that the procedure selects flat regions which are lacking the cytochrome bc1 complex). It is clear that the ar-
525
rangement of proteins and lipids, which is responsible for the very small radius (~30 nm) of curvature of the chromatophores can hardly be compatible with flattening, and it is not a surprise that large rearrangements should occur. The membrane protein content plays a crucial role for determining the shape of the intracytoplasmic membrane. For instance, deletion of the LH2 in Rba. sphaeroides causes the appearance of tubular regions with regular arrays of dimeric core complexes, excluding the cytochrome bc1 complex (Siebert et al., 2004). When the dimers do not form because of the deletion of PufX, the tubular structure is not formed either. The intrinsic curvature of the dimeric core complexes appeared to generate a corrugated surface with alternating protein orientations in reconstituted membranes examined by AFM (Scheuring et al., 2004a): this is a clear example that the constraints for obtaining a flat surface impose a certain type of arrangement of the protein complexes. In other words, AFM can provide non-invasive images of native supramolecular arrangements only insofar as it can respect the native membrane structure. A second and more serious problem is to confront the supercomplex model with the dimeric structure of the cytochrome bc1 complex. On the one hand, this results in a doubling of the overall size of the supramolecular arrangement, which becomes [RC]4[bc1]2[c2]2. Crofts et al. (1998) pointed out that such a big motif should be obvious in electron micrographs. This may not be unquestionable, though, if the structural arrangement has some flexibility. The essential requirement of the model is a confinement of cytochrome c2 to a few partners and, depending on the physical mechanism responsible for this confinement (see Section XII), the underlying supramolecular arrangement may not be visually obvious in electron micrographs. On the other hand, the ‘supercomplex dimer’ raises precisely a problem concerning the confinement of cytochrome c2. Indeed, the model’s requirement that each of the cytochromes c2 be confined to only one half (i.e., one supercomplex in the original definition) of the dimeric overall structure is problematic. In fact, the inhibitor titrations with isolated chromatophores reported by Joliot et al. (1996) indicated a confinement domain for cytochrome c2 that included two cytochrome bc1 complexes, consistent with the supercomplex dimer hypothesis. The Q domain size of ~4 RCs found by Comayras et al. (2005a) is also consistent with a dimer of supercomplexes. One can actually envisage the possibility of a model where there would be no strict stoichiometry for the
526
Jérôme Lavergne, André Verméglio and Pierre Joliot
supramolecular associations, but a distributed one, similar to the variations observed for mitochondrial supercomplexes (see next section). One might for example have some RC-LH1 dimers associated with 0 or one cytochrome bc1 complex dimer or some cytochrome bc1 complex dimers associated with two RC-LH1 dimers. Such heterogeneity was actually present in the interpretation proposed by Joliot et al. (1989), which involved, besides the supercomplex fraction (~70%), a 30% contribution from incomplete supercomplexes devoid of the cytochrome bc1 complex. This flexibility with respect to the original model may be what is needed to solve the difficulties raised by the inclusion of a dimeric cytochrome bc1 complex, and explain differences between observations in vivo and in vitro or the effect of varying the components’ stoichiometry (Crofts et al., 1998). Irrespective of the precise nature of the supramolecular arrangements, the results obtained for both of the mobile electron transfer carriers in Rba. sphaeroides, quinones and cytochrome c2, are indicative of a restricted mobility, whereby these carriers bind to a very small cluster of protein complexes but react nevertheless quite rapidly with all members of this cluster. The physical mechanisms by which this sort of confinement can be explained are by no means obvious, and below we discuss some possibilities. Before doing so, however, we would like to cast a glance at recent results supporting a supercomplex organization of mitochondrial complexes, emphasizing that the issues under discussion may thus have a broader scope. XI. Mitochondrial Supercomplexes As for the photosynthetic chain of purple bacteria, the organization of the mitochondrial respiratory chain is still a matter of debate. Until recently, the random collision model was generally accepted (reviewed by Hackenbrock et al., 1986). This view was based on the (sufficiently) rapid diffusion properties found for quinones and for cytochrome c in model systems or mitochondrial membranes (Gupte et al., 1984; Gupte and Hackenbrock, 1988a; 1988b; Rajarathnam et al., 1989). Kinetic evidence supporting the random collision model was obtained in experiments showing the ‘pool behavior’ of both the quinones and of cytochrome c. The most straightforward demonstration of a homogeneous pool of quinones molecules was based on the observation of the non-linear inhibition
of the respiratory activity by antimycin A for a variety of mitochondrial systems (Kröger et al., 1973). The random collision model has been challenged, however, by a series of recent reports, addressing both structural and functional aspects. Thanks to the improvement of techniques for isolating fragile assemblies of integral membrane protein complexes, substantial evidence has been gained, showing that the respiratory complexes are associated, forming supercomplexes of defined stoichiometries. The use of gentle solubilization of membranes by non-ionic detergents followed by separation by ‘blue native polyacrylamide gel electrophoresis’ has thus allowed characterization of different specific associations between the various complexes (complex I, complex III or cytochrome bc1, complex IV or cytochrome c oxidase). Supercomplexes, such as [I][III]2, [III]2[IV]1–2, [I][III]2[IV]1–4, have been found in mitochondria isolated from plants, yeast, or mammals (Schägger and Pfeiffer, 2000; Dudkina et al., 2006). The term ‘respirasome’ has been coined for supercomplex [I][III]2[IV]1–4 (Schägger and Pfeiffer, 2000). Single particle electron microscopy studies have enabled a structural characterization of the [I][III]2 supercomplex (Dudkina et al., 2005). A further indication of supramolecular associations came from the observation that the deletion of subunits of individual complexes affects the functioning or stability of other complexes of the respiratory chain (Acin-Perez et al., 2004; Grad and Lemire, 2004). The structural organization in supercomplexes is suggestive of — but does not warrant — a functional role. In the case of yeast mitochondria, inhibitor titration of respiratory activity with antimycin A shows that neither Q nor cytochrome c exhibit a pool behavior but behave as confined carriers (Boumans et al., 1998). Interestingly, however, pool behavior for both carriers could be observed after addition of chaotropic agents to the mitochondria. Functional evidence for ‘solid state’ behavior was also obtained in mammalian mitochondria for the supercomplex [I][III]2 (Bianchi et al., 2004). Of particular interest is the case of a yeast mutant unable to synthesize cardiolipin, a phospholipid exclusively found in the bacterial and inner mitochondrial membranes. Contrary to the linear relationship between respiratory activity and the concentration of antimycin A observed in the case of the wild type, the cardiolipin-lacking mitochondria present a hyperbolic relationship indicative of a pool behavior for the cytochrome c (Zhang et al., 2005b). This demonstrates that cardiolipin is essential for the
Chapter 26
Coupling Between RC and Cytochrome bc1 Complex
supramolecular association between complexes III, IV and cytochrome c in intact yeast mitochondria. A deficiency affecting cardiolipin was also shown to destabilize the supercomplexes in human mitochondria from patients suffering from Barth syndrome, and this was proposed to be the basis of this genetically inherited pathology (McKenzie et al., 2006). It is not obvious, especially for outsiders, to get a clear appreciation of the contradiction between the ‘traditional’ results supporting the liquid state model and the more recent results in favor of functional supercomplexes (see Lenaz, 2001). It seems that the latter results have been obtained with whole mitochondria, while most of the liquid state results were obtained using submitochondrial particles. This suggests that the containment of mobile carriers implied by the supercomplex mode of functioning is not a very robust property, and requires particularly intact membranes. This may also concern the supramolecular associations, which are probably not tightly bound, stationary entities but may be subjected to more dynamic dissociation equilibria. In this context it is worth noting that the relative ratio and the nature of supercomplexes were found to depend highly upon the growth conditions in the case of yeast (Schägger and Pfeiffer, 2000). To conclude this incursion into mitochondrial affairs, we note that, at least in some cases, the respiratory supercomplexes achieve the confinement of both the mobile carriers, as observed for the photosynthetic supercomplexes of some Rhodobacter species, and that, possibly, similar mechanisms are involved in both systems. XII. Diffusion and Confinement of Cytochrome c2: Possible Mechanisms The supercomplex model implies that in vivo one— and only one — cytochrome remains confined, over durations extending to at least 100 ms, to a domain including two RCs and one cytochrome bc1 complex. Within this domain the cytochrome exchanges rapidly between the three binding sites. We would like to discuss the mechanisms that may be envisaged to account for these rather stringent constraints. There are two possibilities: attachment by a tether or electrostatic confinement. Tethering could be achieved by attachment to a lipid or to a protein loop. There is an abundant record for the association of cytochrome c to anionic lipids. In mitochondrial membranes, specific attachment to
527
the non bilayer forming cardiolipin (diphosphatidylglycerol) has been documented (Vik et al., 1981; Froud and Ragan, 1984; Tuominen et al., 2002). The hypothesis of the attachment through ‘extended lipid anchorage’ (Rytömaa and Kinnunen, 1995; Tuominen et al., 2002) where one lipid chain is inserted in the membrane and the other in the cytochrome resembles the membrane-anchored cytochrome cy of Rba. capsulatus (see Section III). The current view is that the lipid-attached cytochrome c is the respiratory-active avatar of the protein—its release from the membrane transforms it into a messenger of death, triggering apoptosis (Iverson and Orrenius, 2004). Cardiolipin is found attached to all complexes of the respiratory chain. It has been shown to play a key role in the formation of mitochondrial supercomplexes (Zhang et al., 2005a), and its presence is essential to the function of the mitochondrial cytochrome bc1 complex (Gomez and Robinson, 1999; Lange et al., 2002) and of the cytochrome c oxidase (Sedlak and Robinson, 1999). Cardiolipin is present in the membrane of Rba. sphaeroides and it has been found associated with the RC of Rba. sphaeroides in crystal structures (see Jones, 2007, for a review). Considering the key role of this lipid in mitochondrial supercomplexes, acting both as a glue between the protein complexes and as an anchor for cytochrome c, it is very tempting to speculate that it has similar functions in Rba. sphaeroides. Interestingly, Rba. capsulatus has no cardiolipin (Aygun-Sunar et al., 2006), which may be related to the presence in this bacterium of cytochrome cy endowed with an α-helix membrane anchor. On the other hand, this work reveals a crucial role in Rba. capsulatus for ornithine lipid as regards the steady-state amount of membrane cytochrome complexes, which suggests that this lipid may play a similar role to cardiolipin for stabilizing supramolecular assemblies. The alternative to tethering is electrostatic confinement. In essence, this would mean that the attractive region forming the encounter complex with the RC (see Section IV) extends in the supercomplex over the three partners—and is surrounded by a non-attractive belt (possibly LH1 or LH2) preventing the escape of the cytochrome. The spacing maintained by solvating water would avoid too intense local attractors, ensuring ‘lubricated’ (Axelrod et al., 2002) diffusion of the cytochrome. Such an electrostatic gluing of the cytochrome is certainly possible at low ionic strength, but is this the case under physiological conditions? The answer is clearly positive if one takes
528
Jérôme Lavergne, André Verméglio and Pierre Joliot
into account the very high effective concentration of the cytochrome. Assuming a typical chromatophore diameter of 45 nm, accommodating ~25 RCs, one would have ~12 cytochromes c2 for an internal volume of 14 × 10–26 L. This corresponds (if the cytochromes were all dissociated from the membrane) to a concentration of 140 M (which exceeds the Kd for the attachment to a RC by 8 orders of magnitude). Conversely, if there is a 100 mM concentration of a salt in the periplasmic space, the probability to find one of its ions in a given chromatophore is ~8×10–3. Thus, most chromatophores have no salt at all, and a small fraction has just one ion that can screen the electrostatic interaction of only one of the 12 cytochromes present. A further constraint of the supercomplex model is that it requires only one cytochrome c2 to be present on a supercomplex. This is easier to imagine in the case of the tether mechanism, where a single tether can be present per supercomplex. In the electrostatic confinement hypothesis, it may still be plausible that the occupation by one cytochrome of the attraction region causes sufficient hindrance to decrease significantly the probability of trapping a second cytochrome. In this respect, it is interesting to consider the result obtained for the co-crystallization of cytochrome c with the dimeric cytochrome bc1 complex (Lange and Hunte, 2002): only one cytochrome binds to the dimeric complex, which could not be explained by any obvious hindrance or significant dissymmetry of the complex. This result could mean that, for as yet unclear reasons, the binding of one cytochrome exerts a long-range repulsion disfavoring the binding at the second site. XIII. Diffusion and Confinement of Quinones: Possible Mechanisms We would like to discuss now the case of confined Q diffusion as observed in Rba. sphaeroides. The finding of a small domain size (n ≈ 4 RCs) is reminiscent of similar results obtained with chloroplasts (Joliot et al., 1992; Lavergne et al., 1992). The interpretation put forward in this work was based on the crowding of the membrane by protein complexes, above the percolation threshold. This is expected to result in the formation of closed lipid/quinone cells of irregular shape and size surrounded by proteins. However, as noted by several authors, (Drepper et al., 1993; Tremmel et al., 2003; Scheuring and Sturgis, 2006),
this effect should be significantly alleviated by the Brownian jiggling of the protein complexes. A theoretical treatment for diffusion of small objects among a crowd of diffusing obstacles has been developed by Saxton (1987) and applied to the case of thylakoid grana by Kirchhoff et al. (2002). A Monte Carlo simulation was also carried out by Scheuring and Sturgis (2006) using the AFM information concerning the dynamics of the complexes in Rsp. photometricum. Considering the purple bacterial membrane, the fact that quinones manage to permeate the LH1 ring (see Section IX) is clearly at odds with the view that the protein complexes constitute fences containing the quinones within small and more or less permanent puddles. This motivated an alternative model accounting for Q confinement, where membrane protein complexes act as attractors rather than obstacles (Comayras et al., 2005a,b). We discuss below various arguments supporting this hypothesis. The AFM studies on membranes of purple bacteria have revealed that the protein complexes (at least the core complexes and LH2) adopt a dense packing arrangement, often excluding a large fraction of the lipids that form regions devoid of proteins (Scheuring et al., 2005b). Thus, when shuttling between the RC and cytochrome bc1 complex, quinones are perhaps interacting more with proteins and with other quinones than with lipids. It is also likely that they partition more favorably within the regions of protein clustering than in the pure bilayer areas. Indeed, when isolated under mild conditions, some membrane proteins or complexes retain a large amount of quinones, i.e., many more than could be predicted from the known quinone binding sites. This was found for mitochondrial complexes (Lass and Sohal, 1999). This paper also shows that Q retention is selective compared with lipids, and that it is not due to a particular affinity for detergent. Q retention was also reported for the cytochrome bc1 complex from a purple bacterium (Montoya et al., 1999), and for the RC-LH1 complexes from Rba. sphaeroides (Francia et al., 2004; Comayras et al., 2005a), for which figures of 6–15 UQ per RC were reported (implying that typically ~30% of the UQ pool present in the chromatophores can be retained in the detergent-solubilized complexes). Interestingly, in the work by Francia and coworkers, the LH2 complexes purified in the same way were devoid of quinones, showing that the affinity for quinones is not a trivial property of all membrane proteins. As previously noted (Fig 2 and Section V), the efficient
Chapter 26
Coupling Between RC and Cytochrome bc1 Complex
operation of the cyclic machinery under reducing conditions requires that the Q formed at the Qo site of the cytochrome bc1 complex is somehow passed to the Qi site before it is released from the complex, as experimentally verified (Joliot et al., 2005. This channeling is more easily understood if the protein hydrophobic surface provides a preferential diffusion space for quinones. On longer timescales, Q confinement in Rba. sphaeroides is likely to concern the dimer of supercomplexes. There are various indications suggesting that besides the interaction with proteins, UQ-UQ interactions may also be important. A cooperative distribution of the quinones among isolated RC-LH1 complexes was reported by Comayras et al (2005a). Various nuclear magnetic resonance studies suggest that long chain quinones can form a specific phase in the membrane, with mobility and isotropy properties differing from those of the lipids (Kingsley and Feigenson, 1981; Ulrich et al., 1985; Ondarroa and Quinn, 1986; Cornell et al., 1987). From studies of model bilayers, a preferential location of the quinones in the midplane of the membrane has been proposed (Ulrich et al., 1985; Marchal et al., 1998; Hauss et al., 2005; see also the molecular dynamics study described by Söderhäll and Laaksonen, 2001). The protein/lipid interface may provide another region where quinones tend to partition. Evidence against the free diffusion of quinones in mitochondria was reported by Jorgensen et al. (1985) and Lass and Sohal (1998), showing that distinct pools were reduced when adding succinate or NADH. These pools amounted to 80–90% of the total Q content. This does not support the view that the mitochondrial supercomplexes are sequestering one Q molecule shuttling between the complexes active sites, whilst the bulk of the quinones would be an inactive pool diffusing in the surrounding membrane. On the other hand, it is consistent with the model proposed by Comayras et al. (2005a,b) where quinones form aggregates around ‘friendly’ protein complexes. The quinones would be highly mobile in these patches, allowing rapid electron transfer between neighboring complexes connected by the same patch. This hypothesis of Q aggregates around particular proteins, if correct, raises a number of intriguing questions. What are the determinants (amino acids/ structure) of the Q-protein association? What is the partition ratio for the protein associated patches and escape into the bilayer? In membranes where the scarcity of the cytochrome bc1 complex implies the
529
diffusion of quinones over relatively long distances, are the quinones still aggregated around protein complexes with migration occurring through exchanges during collisions between these complexes? XIV. Conclusions Photosynthetic bacteria possess the fascinating property of accommodating in the same membrane the respiratory and photosynthetic machineries, which have several essential components in common. This sharing allows an efficient regulation between the different bioenergetic pathways for an optimal use of the available energy, especially for responding to rapid changes in the environmental conditions (light, oxygen concentration, etc). In many species the use of several types of periplasmic carriers appears to be one way of controlling the fluxes through the various pathways. Membrane differentiation and specialization is also an essential arrangement in this respect. The photosynthetic apparatus is by far the most demanding in terms of membrane space and organization. Its need for light is manifested as a proliferating membrane (the invaginated intracytoplasmic membrane) paved with LH complexes. The respiratory complexes are mostly excluded from the intracytoplasmic membrane and appear confined to the non-invaginated regions, possibly mixed with a minor fraction of the photosynthetic complexes. In the intracytoplasmic membrane, the electron transfer components must cope with the cluttering by antenna complexes, find some room to settle, or—in the case of quinones—manage to carry out their role as mobile carriers in a highly inconvenient landscape. This complex set of constraints requires appropriate arrangements concerning the membrane shape and the supramolecular organization of the proteins. These mesoscopic aspects are perhaps presently the less well understood in the study of photosynthesis, and offer an exciting field for future research. This is especially true considering the diversity of solutions adopted by the bacteria as regards the structure of the intracytoplasmic membrane, the presence and arrangement of the peripheral antenna, the stoichiometric ratios between electron transfer components, the diffusion range of the mobile carriers, and the nature of these carriers. For example, there are probably diverse arrangements for the ‘photosynthetic units’ (location of LH2 and core complexes; Scheuring et al., 2005b; Chapter 14, Sturgis and Niederman), but all must
530
Jérôme Lavergne, André Verméglio and Pierre Joliot
avoid the formation of overly large or disconnected LH2 regions. Mechanisms that could regulate such arrangements are not known (see however Scheuring and Sturgis, 2005). In this review, we have given a particular focus on the question of supercomplexes and confinement of mobile carriers, although this type of organization is clearly particular to certain bacteria. Nevertheless the questions posed by this structure (what are the interactions that stabilize the assembly of core and cytochrome bc1 complexes? what are the interactions responsible for the confinement of the mobile carriers?) are obviously of a broad interest. Besides the questions of how supercomplexes are held together, we wish to address the question of the physiological advantage(s) that are likely to account for their emergence. There are obvious kinetic and thermodynamic advantages in forcing the mobile carriers to react rapidly with the correct partner rather than roaming a hazardous diffusion pathway. This strategy of ‘metabolic channeling’ within supramolecular associations has been described for soluble enzymes involved in cellular metabolism (Welch and Easterby, 1994). Supramolecular associations have also been observed for membrane complexes involved in the respiratory chain of a few bacteria and, as outlined above, in mitochondria. Such channeling can be of particular importance in the case of membrane of photosynthetic bacteria, as a means for controlling the multiple bioenergetic pathways. In addition to the general benefits offered by the local channeling of intermediate carriers, we believe that an essential role of the association of the cytochrome bc1 complex with the RC in some bacterial species may be for priming the photosynthetic activity under reducing, anaerobic conditions. The proximity of these two complexes in the supercomplex, and the confinement over this structure of both mobile carriers, allows an efficient light-induced cyclic electron transfer even at very low concentration of oxidized quinones by channeling the oxidized Q formed at the cytochrome bc1 complex level to the acceptor site of the RC. The above proposal appears to agree with a correlation that seems to take place between the presence of low potential quinones (mena or rhodoquinones), in addition to ubiquinones (Hiraishi et al., 1984; Imhoff and Bias-Imhoff, 1995), in bacteria devoid of polypeptide PufX or its equivalent such as W in Rps. palustris (Hiraishi et al., 1984). Tentatively, we surmise that the PufX-containing bacteria are generally capable of making a specific association between
core complexes and the cytochrome bc1 complex. On the other hand, bacteria containing menaquinone or rhodoquinone will not encounter the problem of inactivation under anaerobic conditions, because part of their acceptor pool consists of these low potential quinones that remain substantially oxidized. If such bacteria are unable to form supercomplexes because they lack PufX and have closed LH1 rings, this may be unimportant for efficient cyclic electron transfer under reducing conditions. Further work is clearly needed to test this conjecture. The spectacular development of AFM investigations in the recent years is extremely promising, because it is presently the only method that can provide information at the mesoscopic scale relevant to an integrated vision of the photosynthetic machinery. We are confident that the progress of both functional and structural investigations will result in important advances in this field. References Acin-Perez R, Bayona-Bafaluy MP, Fernandez-Silva P, MorenoLoshuertos R, Perez-Martos A, Bruno C, Moraes CT and Enriquez JA (2004) Respiratory complex III is required to maintain complex I in mammalian mitochondria. Mol Cell 13: 805–815 Aird A, Wrachtrup J, Schulten K and Tietz C (2007) Possible pathways for ubiquinone shuttling in Rhodospirillum rubrum revealed by molecular dynamics simulation. Biophys J 92: 23–33 Alegria G and Dutton PL (1991) Langmuir-Blodgett monolayer films of the Rhodopseudomonas viridis reaction center: Determination of the order of the hemes in the cytochrome c subunit. Biochim Biophys Acta 1057: 258–1972 Alric J, Cuni A, Maki H, Nagashima KVP, Verméglio A and Rappaport F (2004a) Electrostatic interaction between redox cofactors in photosynthetic reaction centers. J Biol Chem 279: 47849–47855 Alric J, Yoshida M, Nagashima KVP, Hienerwadel R, Parot P, Verméglio A, Chen SW and Pellequer JL (2004b) Two distinct binding sites for high potential iron-sulfur protein and cytochrome c on the reaction center-bound cytochrome of Rubrivivax gelatinosus. J Biol Chem 279: 32545–32553 Alric J, Tsukatani Y, Yoshida M, Matsuura K, Shimada K, Hienerwadel R, Schoepp-Cothenet B, Nitschke W, Nagashima KVP and Verméglio A (2004c) Structural and functional characterization of the unusual triheme cytochrome bound to the reaction center of Rhodovulum sulfidophilum. J Biol Chem 279: 26090–26097 Alric J, Lavergne J, Rappaport F, Verméglio A, Matsuura K, Shimada K and Nagashima KVP (2006) Kinetic performance and energy profile in a roller coaster electron transfer chain: A study of modified tetraheme-reaction center constructs. J Amer Chem Soc 128: 4136–4145
Chapter 26
Coupling Between RC and Cytochrome bc1 Complex
Axelrod HL and Okamura MY (2005) The structure and function of the cytochrome c2:reaction center electron transfer complex from Rhodobacter sphaeroides. Photosynth Res 85: 101–114 Axelrod HL, Abresch EC, Okamura MY, Yeh AP, Rees DC and Feher G (2002) X-ray structure determination of the cytochrome c2:reaction center electron transfer complex from Rhodobacter sphaeroides. J Mol Biol 319: 501–515 Aygun-Sunar S, Mandaci S, Koch H-G, Murray IVJ, Goldfine H and Daldal F (2006) Ornithine lipid is required for optimal steady-state amounts of c-type cytochromes in Rhodobacter capsulatus. Mol Microbiol 61: 418–435 Bahatyrova S, Frese RN, Siebert CA, Olsen JD, Van Der Werf KO, van Grondelle R, Niederman RA, Bullough PA, Otto C and Hunter CN (2004a) The native architecture of a photosynthetic membrane. Nature 430: 1058–1062 Bahatyrova S, Frese RN, Van Der Werf KO, Otto C, Hunter CN and Olsen JD (2004b) Flexibility and size heterogeneity of the LH1 light harvesting complex revealed by atomic force microscopy: Functional significance for bacterial photosynthesis. J Biol Chem 279: 21327–21333 Bartsch RG (1991) The distribution of soluble metallo-redox proteins in purple phototrophic bacteria. Biochim Biophys Acta 1058: 28–30 Barz WP, Francia F, Venturoli G, Melandri BA, Verméglio A and Oesterhelt D (1995a) Role of PufX protein in photosynthetic growth of Rhodobacter sphaeroides. 1. PufX is required for efficient light-driven electron transfer and photophosphorylation under anaerobic conditions. Biochemistry 34: 15235–15247 Barz WP, Verméglio A, Francia F, Venturoli G, Melandri BA and Oesterhelt D (1995b) Role of the PufX protein in photosynthetic growth of Rhodobacter sphaeroides. 2. PufX is required for efficient ubiquinone/ubiquinol exchange between the reaction center QB site and the cytochrome bc1 complex. Biochemistry 34: 15248–15258 Berry EA, Huang L-S, Saechao LK, Pon NG, Valkova-Valchanova M and Daldal F (2004) X-ray structure of Rhodobacter capsulatus cytochrome bc1: Comparison with its mitochondrial and chloroplast counterparts. Photosynth Res 81: 251–275 Bianchi C, Genova ML, Parenti CG and Lenaz G (2004) The mitochondrial respiratory chain is partially organized in a supercomplex assembly: Kinetic evidence using flux control analysis. J Biol Chem 279: 36562–36569 Boonstra AF, Germeroth L and Boekema EJ (1994) Structure of the light harvesting antenna from Rhodospirillum molischianum studied by electron microscopy. Biochim Biophys Acta 1184: 227–234 Boumans H, Grivell LA and Berden JA (1998) The respiratory chain in yeast behaves as a single functional unit. J Biol Chem 273: 4872–4877 Chen IP, Mathis P, Koepke J and Michel H (2000) Uphill electron transfer in the tetraheme cytochrome subunit of the Rhodopseudomonas viridis photosynthetic reaction center: Evidence from site-directed mutagenesis. Biochemistry 39: 3592–3602 Ciurli S and Musiani F (2005) High potential iron–sulfur proteins and their role as soluble electron carriers in bacterial photosynthesis: Tale of a discovery. Photosynth Res 85: 115–131 Comayras F, Jungas C and Lavergne J (2005a) Functional consequences of the organization of the photosynthetic apparatus in Rhodobacter sphaeroides: I. Quinone domains and excitation transfer in chromatophores and reaction center antenna complexes. J Biol Chem 280: 11203–11213
531
Comayras F, Jungas C and Lavergne J (2005b) Functional consequences of the organization of the photosynthetic apparatus in Rhodobacter sphaeroides: II. A study of PufX– membranes. J Biol Chem 280: 11214–11223 Cornell BA, Keniry MA, Post A, Robertson RN, Weir LE and Westerman PW (1987) Location and activity of ubiquinone 10 and ubiquinone analogues in model and biological membranes. Biochemistry 26: 7702–7707 Crofts AR (2000a) Photosynthesis in Rhodobacter sphaeroides. Trends Microbiol 8: 105–106 Crofts AR (2000b) Response from Crofts. Trends Microbiol 8: 107–108 Crofts AR (2004) The cytochrome bc1 complex: Function in the context of structure. Annu Rev Physiol 66: 689–733 Crofts AR and Wraight CA (1983) The electrochemical domain of photosynthesis. Biochim Biophys Acta 726: 149–185 Crofts AR, Meinhardt SW, Jones KR and Snozzi M (1983) The role of the quinone pool in the cyclic electron-transfer chain of Rhodopseudomonas spaeroides. A modified Q-cycle mechanism. Biochim Biophys Acta 723: 202–218 Crofts AR, Guergova-Kuras M and Hong S (1998) Chromatophore heterogeneity explains phenomena seen in Rhodobacter sphaeroides previously attributed to supercomplexes. Photosynth Res 55: 357–362 Crowley PB and Carrondo MA (2004) The architecture of the binding site in redox protein complexes: Implications for fast dissociation. Proteins 55: 603–612 Deisenhofer J, Epp O, Miki K, Huber R and Michel H (1984) Xray structure analysis of a membrane protein complex: Electron density map at 3Å resolution and a model of the chromophores of the photosynthetic reaction center from Rhodopseudomonas viridis. J Mol Biol 180: 385–398 Deisenhofer J, Epp O, Miki K, Huber R and Michel H (1985) Structure of the protein subunits in the photosynthetic reaction centre of Rhodopseudomonas viridis at 3Å resolution. Nature 318: 618–624 Drepper F, Carlberg I, Andersson B and Haehnel W (1993) Lateral diffusion of an integral membrane protein: Monte Carlo analysis of the migration of phosphorylated light-harvesting complex II in the thylakoid membrane. Biochemistry 32: 11915–11922 Drews G and Golecki JR (1995) Structure, Molecular Organization, and Biosynthesis of Membranes of Purple Bacteria. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 231–257. Kluwer Academic Publishers, Dordrecht Dudkina NV, Eubel H, Keegstra W, Boekema EJ and Braun H-P (2005) Structure of a mitochondrial supercomplex formed by respiratory-chain complexes I and III. Proc Natl Acad Sci USA 102: 3225–3229 Dudkina NV, Heinemeyer J, Sunderhaus S, Boekema EJ and Braun H-P (2006) Respiratory chain supercomplexes in the plant mitochondrial membrane. Trends Plant Sci 11: 232–240 Dupuis A, Chevallet M, Darrouzet E, Duborjal H, Lunardi J and Issartel JP (1998) The Complex I from Rhodobacter capsulatus. Biochim Biophys Acta 1364: 147–165 Farchaus JW, Gruenberg H, Gray KA, Wachtveitl J, DeHoff B, Kaplan S, Oesterhelt D, Drews G and Dawes EA (1990) The pufB, A, L, M genes are not sufficient to restore the photosynthetic plus phenotype to a pufL, M, X deletion strain. In: Drews G and Dawes EA (eds) Molecular Biology of Membrane-Bound
532
Jérôme Lavergne, André Verméglio and Pierre Joliot
Complexes in Phototrophic Bacteria, pp 65–76. Plenum Press, New York, London Fernandez-Velasco J and Crofts AR (1991) Complexes or super complexes: Inhibitor titrations show that electron transfer in chromatophores from Rhodobacter sphaeroides involves a dimeric UQH2:cytochrome c2 oxidoreductase, and is delocalized. Biochem Soc Trans 19: 588–593 Francia F, Wang J, Venturoli G, Melandri BA, Barz WP and Oesterhelt D (1999) The reaction center-LH1 antenna complex of Rhodobacter sphaeroides contains one PufX molecule which is involved in dimerization of this complex. Biochemistry 38: 6834–6845 Francia F, Dezi M, Rebecchi A, Mallardi A, Palazzo G, Melandri BA and Venturoli G (2004) Light-harvesting complex 1 stabilizes P+QB– charge separation in reaction centers of Rhodobacter sphaeroides. Biochemistry 43: 14199–14210 Frese RN, Olsen JD, Branvall R, Westerhuis WH, Hunter CN and van Grondelle R (2000) The long-range supraorganization of the bacterial photosynthetic unit: A key role for PufX. Proc Natl Acad Sci USA 97: 5197–5202 Froud RJ and Ragan CI (1984) Cytochrome c mediates electron transfer between ubiquinol-cytochrome c reductase and cytochrome c oxidase by free diffusion along the surface of the membrane. Biochem J 217: 561–571 Garcia AF, Venturoli G, Gad’on N, Fernandez-Velasco JG, Melandri BA and Drews G (1987) The adaptation of the electron transfer chain of Rhodopseudomonas capsulata to different light intensities. Biochim Biophys Acta 890: 335–345 Garcia D, Richaud P and Verméglio A (1993) The photoinduced cyclic electron transfer in whole cells of Rhodopseudomonas viridis Biochim Biophys Acta 1144: 295–301 Gerencser L, Laczko G and ti P (1999) Unbinding of oxidized cytochrome c from photosynthetic reaction center of Rhodobacter sphaeroides is the bottleneck of fast turnover. Biochemistry 38: 16866–16875 Golecki JR, Ventura S and Oelze J (1991) The architecture of unusual membrane tubes in the B800-850 light-harvesting bacteriochlorophyll-deficient mutant 19 of Rhodobacter sphaeroides FEMS Microbiol Lett 77: 335–340 Gomez B, Jr. and Robinson NC (1999) Phospholipase digestion of bound cardiolipin reversibly inactivates bovine cytochrome bc1. Biochemistry 38: 9031–9038 Gong XM, Paddock ML and Okamura MY (2003) Interactions between cytochrome c2 and photosynthetic reaction center from Rhodobacter sphaeroides: Changes in binding affinity and electron transfer rate due to mutation of interfacial hydrophobic residues are strongly correlated. Biochemistry 42: 14492–14500 Grad LI and Lemire BD (2004) Mitochondrial complex I mutations in Caenorhabditis elegans produce cytochrome c oxidase deficiency, oxidative stress and vitamin-responsive lactic acidosis. Hum Mol Genet 13: 303–314 Graige MS, Feher G and Okamura MY (1998) Conformational gating of the electron transfer reaction QA–•QB → QAQB–• in bacterial reaction centers of Rhodobacter sphaeroides determined by a driving force assay. Proc Natl Acad Sci USA 95: 11679–11684 Gubellini F, Francia F, Busselez J, Venturoli G and Lévy D (2006) Functional and structural analysis of the photosynthetic apparatus of Rhodobacter veldkampii. Biochemistry 45: 10512–10520
Gunner MR and Honig B (1991) Electrostatic control of midpoint potentials in the cytochrome subunit of the Rhodopseudomonas viridis reaction center. Proc Natl Acad Sci USA 88: 9151–9155 Gupte SS and Hackenbrock CR (1988a) Multidimensional diffusion modes and collision frequencies of cytochrome c with its redox partners. J Biol Chem 263: 5241–5247 Gupte SS and Hackenbrock CR (1988b) The role of cytochrome c diffusion in mitochondrial electron transport. J Biol Chem 263: 5248–5253 Gupte S, Wu E-S, Hoechli L, Hoechli M, Jacobson K, Sowers AE and Hackenbrock CR (1984) Relationship between lateral diffusion, collision frequency, and electron transfer of mitochondrial inner membrane oxidation-reduction components. Proc Natl Acad Sci USA 81: 2606–2610 Hackenbrock CR, Chazotte B and Gupte SS (1986) The random collision model and a critical assessment of diffusion and collision in mitochondrial electron transport. J Bioener Biomem 18: 331–368 Hauss T, Dante S, Haines TH and Dencher NA (2005) Localization of coenzyme Q10 in the center of a deuterated lipid membrane by neutron diffraction. Biochim Biophys Acta 1710: 57–62 Hiraishi A, Hoshino Y and Kitamura H (1984) Isoprenoid quinone composition in the classification of Rhodospirillaceae. J Gen Appl Microbiol 30: 197–210 Hochkoeppler A, Jenney FE, Jr., Lang SE, Zannoni D and Daldal F (1995) Membrane-associated cytochrome cy of Rhodobacter capsulatus is an electron carrier from the cytochrome bc1 complex to the cytochrome c oxidase during respiration. J Bacteriol 177: 608–613 Honig B and Nicholls A (1995) Classical electrostatics in biology and chemistry. Science 268: 1144–1149 Hunter CN, Pennoyer JD, Sturgis JN, Farrelly D and Niederman RA (1988) Oligomerization states and associations of lightharvesting pigment-protein complexes of Rhodobacter sphaeroides as analyzed by lithium dodecyl sulfate-polyacrylamide gel electrophoresis. Biochemistry 27: 3459–3467 Imhoff JM and Bias-Imhoff U (1995) Lipids, quinones and fatty acids of anoxygenic phototrophic bacteria. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 179–205. Kluwer Academic Publishers, Dordrecht Iverson SL and Orrenius S (2004) The cardiolipin-cytochrome c interaction and the mitochondrial regulation of apoptosis. Arch Biochem Biophys 423: 37—46 Jamieson SJ, Wang P, Qian P, Kirkland JY, Conroy MJ, Hunter CN and Bullough PA (2002) Projection structure of the photosynthetic reaction centre-antenna complex of Rhodospirillum rubrum at 8.5 Å resolution. EMBO J 21: 3927–3935 Jenney FE and Daldal F (1993) A novel membrane-associated ctype cytochrome, cyt cy, can mediate the photosynthetic growth of Rhodobacter capsulatus and Rhodobacter sphaeroides. EMBO J 12: 1283–1292 Jenney FE, Prince RC and Daldal F (1994) Roles of the soluble cytochrome c2 and membrane-associated cytochrome cy of Rhodobacter capsulatus in photosynthetic electron transfer. Biochemistry 33: 2496–2502 Joliot P, Verméglio A and Joliot A (1989) Evidence for supercomplexes between reaction centers, cytochrome c2 and cytochrome bc1 complex in Rhodobacter sphaeroides whole cells. Biochim Biophys Acta 975: 336–345
Chapter 26
Coupling Between RC and Cytochrome bc1 Complex
Joliot P, Verméglio A and Joliot A (1990) Electron transfer between primary and secondary donors in Rhodospirillum rubrum: Evidence for a dimeric association of reaction centers. Biochemistry 29: 4355–4361 Joliot P, Lavergne J and Béal D (1992) Plastoquinone compartmentation in chloroplasts. 1. Evidence for domains with different rates of photo-reduction. Biochim Biophys Acta 1101: 1–12 Joliot P, Verméglio A and Joliot A (1996) Supramolecular organization of the photosynthetic chain in chromatophores and cells of Rhodobacter sphaeroides. Photosynth Res 48: 291–299 Joliot P, Joliot A and Verméglio A (1997) Photo-induced cyclic electron transfer operates in frozen cells of Rhodobacter sphaeroides. Biochim Biophys Acta 1318: 374–384 Joliot P, Joliot A and Verméglio A (2005) Fast oxidation of the primary electron acceptor under anaerobic conditions requires the organization of the photosynthetic chain of Rhodobacter sphaeroides in supercomplexes. Biochim Biophys Acta 1706: 204–214 Jones MR (2007) Lipids in photosynthetic reaction centres: Structural roles and functional holes. Progress in Lipid Research 46: 56–87 Jones MR, McEwan AG and Jackson JB (1990) The role of c-type cytochromes in the photosynthetic electron transport pathway of Rhodobacter capsulatus. Biochim Biophys Acta 1019: 59–66 Jorgensen BM, Rasmussen HN and Rasmussen UF (1985) Ubiquinone reduction pattern in pigeon heart mitochondria. Identification of three distinct ubiquinone pools. Biochem J 229: 621–629 Jungas C, Ranck JL, Rigaud JL, Joliot P and Verméglio A (1999) Supramolecular organization of the photosynthetic apparatus of Rhodobacter sphaeroides. EMBO J 18: 534–542 Kiley PJ, Varga AR and Kaplan S (1988) Physiological and structural analysis of light-harvesting mutants of Rhodobacter sphaeroides. J Bacteriol 170: 1103–1115 Kimura Y, Alric J, Verméglio A, Masuda S, Hagiwara Y, Matsuura K, Shimada K and Nagashima KVP (2006) A new membranebound cytochrome c works as an electron donor to the photosynthetic reaction center complex in the purple bacterium, Rhodovulum sulfidophilum. J Biol Chem 282: 6463–6472 Kingsley PB and Feigenson GW (1981) 1H-NMR study of the location and motion of ubiquinones in perdeuterated phosphatidylcholine bilayers. Biochim Biophys Acta 635: 602–618 Kirchhoff H, Mukherjee U and Galla HJ (2002) Molecular architecture of the thylakoid membrane: Lipid diffusion space for plastoquinone. Biochemistry 41: 4872–4882 Knaff D (1978) Reducing potentials and the pathway of NAD+ reduction. In: Clayton RK and Sistrom WR (ed) The Photosynthetic Bacteria, pp 629–640. Plenum Press, New York Knaff DB, Willie A, Long JE, Kriauciunas A, Durham B and Millett F (1991) Reaction of cytochrome c2 with photosynthetic reaction centers from Rhodopseudomonas viridis Biochemistry 30: 1303–1310 Koepke J, Hu X, Muenke C, Schulten K and Michel H (1996) The crystal structure of the light-harvesting complex II (B800-850) from Rhodospirillum molischianum. Structure 4: 581–597 Kröger A, Klingenberg M and Schweidler S (1973) Further evidence for the pool function of ubiquinone as derived from the inhibition of the electron transport by antimycin. Eur J Biochem 39: 313–323 Lange C and Hunte C (2002) Crystal structure of the yeast cyto-
533
chrome bc1 complex with its bound substrate cytochrome c. Proc Natl Acad Sci USA 99: 2800–2805 Lange C, Nett JH, Trumpower BL and Hunte C (2002) Specific roles of protein-phospholipid interactions in the yeast cytochrome bc1 complex structure. EMBO J 20: 6591–6600 Larson JW and Wraight CA (2000) Preferential binding of equine ferricytochrome c to the bacterial photosynthetic reaction center from Rhodobacter sphaeroides. Biochemistry 39: 14822–14830 Lass A and Sohal RS (1998) Electron transport-linked ubiquinonedependent recycling of α-tocopherol inhibits autooxidation of mitochondrial membranes. Arch Biochem Biophys 352: 229–236 Lass A and Sohal RS (1999) Comparisons of coenzyme Q bound to mitochondrial membrane proteins among different mammalian species. Free Radic Biol Med 27: 220–226 Lavergne J and Joliot P (1991) Restricted diffusion in photosynthetic membranes. Trends Biochem Sci 16: 129–134 Lavergne J, Joliot P and Verméglio A (1989) Partial equilibration of photosynthetic electron carriers under weak illumination: A theoretical and experimental study. Biochim. Biophys. Acta 975: 346–354 Lavergne J, Bouchaud JP and Joliot P (1992) Plastoquinone compartmentation in chloroplasts. 2. Theoretical aspects. Biochim Biophys Acta 1101: 13–22 Lee DW, Ozturk Y, Mamedova A, Osyczka A, Cooley JW and Daldal F (2006) A functional hybrid between the cytochrome bc1 complex and its physiological membrane-anchored electron acceptor cytochrome cy in Rhodobacter capsulatus. Biochim Biophys Acta 1757: 346–352 Lenaz G (2001) A critical appraisal of the mitochondrial coenzyme Q pool. FEBS Lett 509: 151–155 Li G, Knowles PF, Murphy DJ, Nishida I and Marsh D (1989) Spin-label ESR studies of lipid-protein interactions in thylakoid membranes. Biochemistry 28: 7446–7452 Lieutaud C, Nitschke W, Verméglio A, Parot P and SchoeppCothenet B (2003) HiPIP in Rubrivivax gelatinosus is firmly associated to the membrane in a conformation efficient for electron transfer. Biochim Biophys Acta 1557: 83–90 Lilburn TG, Haith CE, Prince RC and Beatty JT (1992) Pleiotropic effects of pufX gene deletion on the structure and function of the photosynthetic apparatus of Rhodobacter capsulatus. Biochim Biophys Acta 1100: 160–170 Marchal D, Boireau W, Laval JM, Moiroux J and Bourdillon C (1998) Electrochemical measurement of lateral diffusion coefficients of ubiquinones and plastoquinones of various isoprenoid chain lengths incorporated in model bilayers. Biophys J 74: 1937–1948 Masuda S, Yoshida M, Nagashima KVP, Shimada K and Matsuura K (1999) A new cytochrome subunit bound to the photosynthetic reaction center in the purple bacterium, Rhodovulum sulfidophilum. J Biol Chem 274: 10795–10801 Masuda S, Tsukatani Y, Kimura Y, Nagashima KVP, Shimada K and Matsuura K (2002) Mutational analyses of the photosynthetic reaction center-bound triheme cytochrome subunit and cytochrome c2 in the purple bacterium Rhodovulum sulfidophilum. Biochemistry 41: 11211–11217 McDermott G, Prince SM, Freer AA, Hawthornthwaite AM, Papiz MZ, Cogdell RJ and Isaacs NW (1995) Crystal structure of an integral membrane light-harvesting complex from photosynthetic bacteria. Nature 374: 517–521
534
Jérôme Lavergne, André Verméglio and Pierre Joliot
McKenzie M, Lazarou M, Thorburn DR and Ryan MT (2006) Mitochondrial respiratory chain supercomplexes are destabilized in Barth syndrome patients. J Mol Biol 361: 462–469 Menin L, Schoepp B, Parot P and Verméglio A (1997) Photoinduced cyclic electron transfer in Rhodocyclus tenuis cells: Participation of HiPIP or cyt c8 depending on the ambient redox potential. Biochemistry 36: 12183–12188 Menin L, Yoshida M, Jaquinod M, Nagashima KV, Matsuura K, Parot P and Verméglio A (1999) Dark aerobic growth conditions induce the synthesis of a high midpoint potential cytochrome c8 in the photosynthetic bacterium Rubrivivax gelatinosus. Biochemistry 38: 15238–15244 Meyer TE and Donohue TJ (1995) Cytochromes, iron-sulfur and copper proteins mediating electron transfer from the cyt bc1 complex to photosynthetic reaction center complexes. In: Blankenship RE, Madigan MT and Bauer CE (ed) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 725–745. Kluwer Academic Publishers, Dordrecht Mezzetti A, Leibl W, Breton J and Nabedryk E (2003) Photoreduction of the quinone pool in the bacterial photosynthetic membrane: Identification of infrared marker bands for quinol formation. FEBS Lett 537: 161–165 Milano F, Agostiano A, Mavelli F and Trotta M (2003) Kinetics of the quinone binding reaction at the QB site of reaction centers from the purple bacteria Rhodobacter sphaeroides reconstituted in liposomes. Eur J Biochem 270: 4595–4605 Miller KR (1982) Three-dimensional structure of a photosynthetic membrane. Nature 300: 53–55 Miyashita O, Onuchic JN and Okamura MY (2003) Continuum electrostatic model for the binding of cytochrome c2 to the photosynthetic reaction center from Rhodobacter sphaeroides. Biochemistry 42: 11651–11660 Miyashita O, Onuchic JN and Okamura MY (2004) Transition state and encounter complex for fast association of cytochrome c2 with bacterial reaction center. Proc Natl Acad Sci USA 101: 16174–16179 Miyashita O, Okamura MY and Onuchic JN (2005) Interprotein electron transfer from cytochrome c2 to photosynthetic reaction center: Tunneling across an aqueous interface. Proc Natl Acad Sci USA 102: 3558–3563 Montoya G, Kaat K, Rodgers S, Nitschke W and Sinning I (1999) The cytochrome bc1 complex from Rhodovulum sulfidophilum is a dimer with six quinones per monomer and an additional 6-kDa component. Eur J Biochem 259: 709–718 Moser CC and Dutton PL (1988) Cytochrome c and c2 binding dynamics and electron transfer with photosynthetic reaction center protein and other integral membrane redox proteins. Biochemistry 27: 2450–1961 Myllykallio H, Jenney FE, Jr., Moomaw CR, Slaughter CA and Daldal F (1997) Cytochrome cy of Rhodobacter capsulatus is attached to the cytoplasmic membrane by an uncleaved signal sequence-like anchor. J Bacteriol 179: 2623–2631 Myllykallio H, Drepper F, Mathis P and Daldal F (1998) Membrane-anchored cytochrome cy mediated microsecond time range electron transfer from the cytochrome bc1 complex to the reaction center in Rhodobacter capsulatus. Biochemistry 37: 5501–5510 Myllykallio H, Drepper F, Mathis P and Daldal F (2000) Electron-transfer supercomplexes in photosynthesis and respiration. Trends Microbiol 8: 493–494
Nagashima KVP, Matsuura K, Shimada K and Verméglio A (2002) High-potential iron-sulfur protein (HiPIP) is the major electron donor to the reaction center complex in photosynthetically growing cells of the purple bacterium Rubrivivax gelatinosus. Biochemistry 41: 14028–14032 Nitschke W and Dracheva SM (1995) Reaction centers associated cytochromes. In: Blankenship RE, Madigan MT and Bauer CE (ed) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 775–805. Kluwer Academic Publishers, Dordrecht, The Netherlands Nitschke W and Rutherford AW (1989) Tetraheme cytochrome c subunit of Rhodopseudomonas viridis characterized by EPR. Biochemistry 28: 3161–3168 Nogi T, Fathir I, Kobayashi M, Nozawa T and Miki K (2000) Crystal structures of photosynthetic reaction center and highpotential iron-sulfur protein from Thermochromatium tepidum: Thermostability and electron transfer. Proc Natl Acad Sci USA 97: 13561–13566 Nogi T, Hirano Y and Miki K (2005) Structural and functional studies on the tetraheme cytochrome subunit and its electron donor proteins: The possible docking mechanisms during the electron transfer reaction. Photosynth Res 85: 87–99 Ondarroa M and Quinn PJ (1986) Proton magnetic resonance spectroscopic studies of the interaction of ubiquinone-10 with phospholipid model membranes. Eur J Biochem 155: 353–361 Ortega JM and Mathis P (1993) Electron transfer from the tetraheme cytochrome to the special pair in isolated reaction centers of Rhodopseudomonas viridis Biochemistry 32: 1141–1151 Osyczka A, Nagashima KVP, Sogabe S, Miki K, Yoshida M, Shimada K and Matsuura K (1998) Interaction site for soluble cytochromes on the tetraheme cytochrome subunit bound to the bacterial photosynthetic reaction center mapped by site-directed mutagenesis. Biochemistry 37: 11732–11744 Osyczka A, Nagashima KVP, Shimada K and Matsuura K (1999a) Interaction site for high-potential iron-sulfur protein on the tetraheme cytochrome subunit bound to the photosynthetic reaction center of Rubrivivax gelatinosus. Biochemistry 38: 2861–2865 Osyczka A, Nagashima KVP, Sogabe S, Miki K, Shimada K and Matsuura K (1999b) Comparison of the binding sites for highpotential iron-sulfur protein and cytochrome c on the tetraheme cytochrome subunit bound to the bacterial photosynthetic reaction center. Biochemistry 38: 15779–15790 Overfield RE, Wraight CA and Devault D (1979) Microsecond photooxidation kinetics of cytochrome c2 from Rhodopseudomonas sphaeroides: In vivo and solution studies. FEBS Lett 105: 137–142 Paddock ML, Weber KH, Chang C and Okamura MY (2005) Interactions between cytochrome c2 and the photosynthetic reaction center from Rhodobacter sphaeroides: The cation-π interaction. Biochemistry 44: 9619–9625 Page CC, Moser CC, Chen X and Dutton PL (1999) Natural engineering principles of electron tunnelling in biological oxidation-reduction. Nature 402: 47–52 Qian P, Hunter CN and Bullough PA (2005) The 8.5 Å projection structure of the core RC-LH1-PufX dimer of Rhodobacter sphaeroides. J Mol Biol 349: 948–960 Rajarathnam K, Hochman J, Schindler M and Ferguson-Miller S (1989) Synthesis, location, and lateral mobility of fluorescently labeled ubiquinone 10 in mitochondrial and artificial mem-
Chapter 26
Coupling Between RC and Cytochrome bc1 Complex
branes. Biochemistry 28: 3168–3176 Rich PR (1982) Electron and proton transfers in chemical and biological quinone systems. Faraday Discuss Chem Soc 74: 349–364 Rich PR (1984) Electron and proton transfers through quinones and cytochrome bc complexes. Biochim Biophys Acta 768: 53–79 Ritz T, Park S and Schulten K (2001) Kinetics of excitation migration and trapping in the photosynthetic unit of purple bacteria. J Phys Chem B 105: 8259–8267 Robertson DE, Prince RC, Bowyer JR, Matsuura K, Dutton PL and Ohnishi T (1984) Thermodynamic properties of the semiquinone and its binding site in the ubiquinol-cytochrome c (c2) oxidoreductase of respiratory and photosynthetic systems. J Biol Chem 259: 1758–1763 Rosen D, Okamura MY and Feher G (1980) Interaction of cytochrome c with reaction centers of Rhodopseudomonas sphaeroides R-26: Determination of number of binding sites and dissociation constants by equilibrium dialysis. Biochemistry 19: 5687–5692 Roszak AW, Howard TD, Southall J, Gardiner AT, Law CJ, Isaacs NW and Cogdell RJ (2003) Crystal structure of the RC-LH1 core complex from Rhodopseudomonas palustris. Science 302: 1969–1972 Rytömaa M and Kinnunen PKJ (1995) Reversibility of the binding of cytochrome c to liposomes. Implications for lipid-protein interactions. J Biol Chem 270: 3197–3202 Sabaty M, Jappe J, Olive J and Verméglio A (1994) Organization of electron transfer components in Rhodobacter sphaeroides forma sp. denitrificans whole cells. Biochim Biophys Acta 1187: 313–323 Saphon S, Jackson JB, Lerbs V and Witt HT (1975) The functional unit of electrical events and phosphorylation in chromatophores from Rhodopseudomonas sphaeroides. Biochim Biophys Acta 408: 58–66 Saxton MJ (1987) Lateral diffusion in an archipelago. The effect of mobile obstacles. Biophys J 52: 989–997 Schägger H and Pfeiffer K (2000) Supercomplexes in the respiratory chains of yeast and mammalian mitochondria. EMBO J 19: 1777–1783 Scheuring S (2006) AFM studies of the supramolecular assembly of bacterial photosynthetic core-complexes. Curr Opin Chem Biol 10: 387–393 Scheuring S and Sturgis JN (2005) Chromatic adaptation of photosynthetic membranes. Science 309: 484–487 Scheuring S and Sturgis JN (2006) Dynamics and diffusion in photosynthetic membranes from Rhodospirillum photometricum. Biophys J 91: 3707–3717 Scheuring S, Francia F, Busselez J, Melandri BA, Rigaud JL and Lévy D (2004a) Structural role of PufX in the dimerization of the photosynthetic core complex of Rhodobacter sphaeroides. J Biol Chem 279: 3620–3626 Scheuring S, Sturgis JN, Prima V, Bernadac A, Lévy D and Rigaud JL (2004b) Watching the photosynthetic apparatus in native membranes. Proc Natl Acad Sci USA 101: 11293–11297 Scheuring S, Busselez J and Lévy D (2005a) Structure of the dimeric PufX-containing core complex of Rhodobacter blasticus by in situ atomic force microscopy. J Biol Chem 280: 1426–1431 Scheuring S, Lévy D and Rigaud JL (2005b) Watching the com-
535
ponents of photosynthetic bacterial membranes and their in situ organisation by atomic force microscopy. Biochim Biophys Acta 1712: 109–127 Scheuring S, Gonçalves RP, Prima V and Sturgis JN (2006) The photosynthetic apparatus of Rhodopseudomonas palustris: Structures and organization. J Mol Biol 358: 83–96 Schoepp B, Parot P, Menin L, Gaillard J, Richaud P and Verméglio A (1995) In vivo participation of a high potential iron-sulfur protein as electron donor to the photochemical reaction center of Rubrivivax gelatinosus. Biochemistry 34: 11736–11742 Sedlak E and Robinson NC (1999) Phospholipase A2 digestion of cardiolipin bound to bovine cytochrome c oxidase alters both activity and quaternary structure. Biochemistry 38: 14966–14972 Siebert CA, Qian P, Fotiadis D, Engel A, Hunter CN and Bullough PA (2004) Molecular architecture of photosynthetic membranes in Rhodobacter sphaeroides: The role of PufX. EMBO J 23: 690–700 Söderhäll JA and Laaksonen A (2001) Molecular dynamics simulations of ubiquinone inside a lipid bilayer. J Phys Chem B 105: 9308–9315 Speck SH and Margoliash E (1984) Characterization of the interaction of cytochrome c and mitochondrial ubiquinol-cytochrome c reductase. J Biol Chem 259: 1064–1072 Stark W, Kühlbrandt W, Wildhaber I, Wehrli E and Mühlethaler K (1984) The structure of the photoreceptor unit of Rhodopseudomonas viridis. EMBO J 3: 777–783 Takamiya KI and Dutton PL (1979) Ubiquinone in Rhodopseudomonas sphaeroides. Some thermodynamic properties. Biochim Biophys Acta 546: 1–16 Tetreault M, Rongey SH, Feher G and Okamura MY (2001) Interaction between cytochrome c2 and the photosynthetic reaction center from Rhodobacter sphaeroides: Effects of charge-modifying mutations on binding and electron transfer. Biochemistry 40: 8452–8462 Tetreault M, Cusanovich M, Meyer T, Axelrod H and Okamura MY (2002) Double mutant studies identify electrostatic interactions that are important for docking cytochrome c2 onto the bacterial reaction center. Biochemistry 41: 5807–5815 Tiede DM and Dutton PL (1993) Electron transfer between bacterial reaction centers and mobile c-type cytochromes. In: Deisenhofer J and Norris JR (ed) The Photosynthetic Reaction Center. pp 257–288. Academic Press, San Diego Tiede DM, Vashishta AC and Gunner MR (1993) Electron-transfer kinetics and electrostatic properties of the Rhodobacter sphaeroides reaction center and soluble c-cytochromes. Biochemistry 32: 4515–4531 Tremmel IG, Kirchhoff H, Weis E and Farquhar GD (2003) Dependence of plastoquinol diffusion on the shape, size, and density of integral thylakoid proteins. Biochim Biophys Acta 1607: 97–109 Tsukatani Y, Matsuura K, Masuda S, Shimada K, Hiraishi A and Nagashima KVP (2004) Phylogenetic distribution of unusual triheme to tetraheme cytochrome subunit in the reaction center complex of purple photosynthetic bacteria. Photosynth Res 79: 83–91 Tuominen EKJ, Wallace CJA and Kinnunen PKJ (2002) Phospholipid-cytochrome c interaction. Evidence for the extended lipid anchorage. J Biol Chem 277: 8822–8826 Ulrich EL, Girvin ME, Cramer WA and Markley JL (1985) Location and mobility of ubiquinones of different chain lengths in
536
Jérôme Lavergne, André Verméglio and Pierre Joliot
artificial membrane vesicles. Biochemistry 24: 2501–2508 van der Wal HN and van Grondelle R (1983) Flash-induced electron transport in b- and c-type cytochromes in Rhodospirillum rubrum. Evidence for a Q-cycle. Biochim Biophys Acta 725: 94–103 Verméglio A and Carrier JM (1984) Photoinhibition by flash and continuous light of oxygen uptake by intact photosynthetic bacteria. Biochim Biophys Acta 764: 233–238 Verméglio A and Joliot P (2002) Supramolecular organisation of the photosynthetic chain in anoxygenic bacteria. Biochim Biophys Acta 1555: 60–64 Verméglio A, Richaud P and Breton J (1989) Orientation and assignment of the four cytochrome hemes in Rhodopseudomonas viridis reaction centers. FEBS Lett 243: 259–263 Verméglio A, Joliot P and Joliot A (1993) The rate of cytochrome c2 photooxidation reflects the subcellular distribution of reaction centers in Rhodobacter sphaeroides Ga cells. Biochim Biophys Acta 1183: 352–360 Verméglio A, Joliot A and Joliot P (1998) Supramolecular organization of the photosynthetic chain in mutants of Rhodobacter capsulatus deleted in cytochrome c2. Photosynth Res 56: 329–337 Verméglio A, Li J, Schoepp-Cothenet B, Pratt N and Knaff DB (2002) The role of high-potential iron protein and cytochrome c8 as alternative electron donors to the reaction center of Chromatium vinosum. Biochemistry 41: 8868–8875 Verméglio A, Borghese R and Zannoni D (2004) Interactions
between photosynthesis and respiration in facultative anoxygenic phototrophs. In: Zannoni D (ed) Respiration in Archaea and bacteria: diversity of prokaryotic respiratory systems (Advances in Photosynthesis and Respiration, Vol 16), pp 279–295. Springer, Dordrecht Vik SB, Georgevich G and Capaldi RA (1981) Diphosphatidylglycerol is required for optimal activity of beef heart cytochrome c oxidase. Proc Natl Acad Sci USA 78: 1456–1460 Walz T and Ghosh R (1997) Two-dimensional crystallization of the light-harvesting I-reaction centre photounit from Rhodospirillum rubrum. J Mol Biol 265: 107–111 Welch GR and Easterby JS (1994) Metabolic channeling versus free diffusion: Transition-time analysis Trends Biochem. Sci. 19: 193 –197 Xia D, Yu CA, Kim H, Jia-Shi Xia JZ, Kachurin AM, Zhang L, Yu L and Deisenhofer J (1997) Crystal structure of the cytochrome bc1 complex from bovine heart mitochondria. Science 277: 60–66 Zhang M, Mileykovskaya E and Dowhan W (2005a) Cardiolipin is essential for organization of complexes III and IV into a supercomplex in intact yeast mitochondria. J Biol Chem 280: 29403–29408 Zhang M, Mileykovskaya E and Dowhan W (2005b) Cardiolipin is essential for organization of complexes III and IV into a supercomplex in intact yeast mitochondria. J Biol Chem 280: 29403–29408
Chapter 27 Respiration and Respiratory Complexes Davide Zannoni* Department of Biology, University of Bologna, I-40126 Bologna, Italy
Barbara Schoepp-Cothenet Laboratoire de Bioénergétique et Ingénierie des Protéines, Institut de Biologie Structurale et Microbiologie (IFR), F-13402 Marseille Cedex 20, France
Jonathan Hosler Department of Biochemistry, University of Mississippi Medical Center, Jackson, MS 39216 U.S.A.
Summary ............................................................................................................................................................... 538 I. Aerobic Respiration ........................................................................................................................................ 538 A. Respiratory Complexes and Electron Transport Carriers................................................................. 538 1. NADH Dehydrogenase (NADH:Q Oxidoreductase) ................................................................ 538 2. Cytochrome bd-type Terminal Oxidase ................................................................................. 539 3. Cytochrome cbb3-type Terminal Oxidase ............................................................................... 540 4. Cytochrome aa3-type Terminal Oxidase ................................................................................. 541 5. A Putative caa3-type Cytochrome c Oxidase and a Putative Quinol Oxidase ........................ 542 6. Cytochromes c and HiPIPs (High-Potential Iron-sulfur Proteins)............................................ 542 B. Possible Reasons for Complex Respiratory Systems ...................................................................... 543 C. Structure-function Studies of the aa3-type Cytochrome c Oxidase ................................................. 543 1. Proton Pathways and Proton Pumping. ................................................................................. 544 2. Protection from Oxidative Damage ........................................................................................ 545 3. Cytochrome c Oxidase Assembly .......................................................................................... 545 D. Structure-function Studies of the cbb3-type Oxidase ...................................................................... 545 II. Respiration Utilizing Substrates other than Oxygen ....................................................................................... 546 A. Dimethylsulfoxide and Trimethylamine-N-oxide Respiration ............................................................ 546 1. Organization of Dimethylsulfoxide - Trimethylamine-N-oxide Respiratory Chains.................. 548 B. Arsenics as Bioenergetic Substrates................................................................................................ 548 1. Arsenite Oxidation ................................................................................................................... 548 2. Arsenate Respiration............................................................................................................... 549 C. Selenate Respiration ........................................................................................................................ 550 D. The Selenate Reductase-type Class of Enzymes ............................................................................ 550 E. Non-conventional Substrates but Conventional Bioenergetics ........................................................ 550 F. The Enzymes: Variations on a Theme ............................................................................................. 552 III. Respiration vs. Photosynthesis: Which One Came First? .............................................................................. 553 IV. Respiration and Photosynthesis are Intermingled .......................................................................................... 553 Acknowledgments ................................................................................................................................................. 555 References ............................................................................................................................................................ 555 *Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 537–561. © 2009 Springer Science + Business Media B.V.
538
Davide Zannoni, Barbara Schoepp-Cothenet and Jonathan Hosler
Summary Respiration in facultative phototrophs is a flexible metabolic process that involves various electron donors and acceptors. A good example of such respiratory flexibility can be found in Rhodobacter species, most of them being equipped with genes that encode five distinct oxidases having different oxygen affinities. One of these, the cytochrome cbb3 oxidase is prevalent at low oxygen tensions, and terminates a highly coupled electron transfer pathway which is formed by a ‘core’ of redox components, e.g., quinones, the cytochrome bc1 complex and cytochrome c, in common with the photosynthetic apparatus. Thus, by modulating expression of different terminal oxido-reductases that lock onto a core electron transfer pathway, Rhodobacter species can survive in a range of oxic, micro-oxic, and anoxic environments either in the dark or in the light. This chapter covers first the types and basic characteristics of the terminal oxidases in a few Rhodobacter species; then, respiratory substrates other than oxygen are examined. These substrates include orthodox anaerobic electron acceptors such as DMSO or TMAO but also arsenics as unconventional bioenergetics substrates. Finally, a synopsis of the data examining the functional interactions between photosynthetic and respiratory ETP is given along with a phylogenetic scenario suggesting that respiration is more ancient than both anoxygenic and oxygenic photosynthesis. I. Aerobic Respiration The term ‘aerobic respiration’ refers to the process of transferring high-energy electrons, derived from reduced organic or inorganic substrates, through a series of electron carriers to O2. Organic and transition metal electron carriers are organized into multisubunit integral membrane protein complexes in the bacterial cytoplasmic membrane, while NADH, water soluble or lipid anchored cytochromes c, and lipid soluble quinone transfer electrons between these complexes. Three of these multi-subunit enzymes, the proton pumping NADH dehydrogenase, the cytochrome bc1 complex and the terminal oxidases are electrogenic, in that each of them has evolved one or more mechanisms to convert the energy released by electron transfer reactions into a voltage gradient across the cytoplasmic membrane. The voltage gradient is used by the F1Fo ATP synthase to synthesize ATP from ADP and inorganic phosphate. In the longest known respiration chain, NADH, produced by NAD+ reduction in cytosolic reactions, is oxidized by NADH dehydrogenase (Fig. 1). This enzyme transfers electrons to membrane-soluble ubiquinone, which diffuses into the bilayer. Reduced quinone is then oxidized by the cytochrome bc1 complex that Abbreviations: C. – Chloroflexus; CcO – cytochrome c oxidase; DMS – dimethylsulfide; DMSO – dimethylsulfoxide; E. – Escherichia; HiPIP – high-potential iron-sulfur protein; Rba. – Rhodobacter; RC – photochemical reaction center; Rps. – Rhodopseudomonas; Rsb. – Roseobacter; Rvu. – Rhodovulum; T, – Thermus; TMAO – trimethylamine-N-oxide; TMPD – N,N,N´,N´-tetramethyl-p-phenylenediamine; UQ – ubiquinone
conveys electrons to cytochrome c. Finally, reduced cytochrome c is oxidized by a cytochrome c oxidase, which uses these electrons to reduce O2 to H2O. Reduced quinone (i.e., quinol) may also be oxidized by a quinol terminal oxidase, in a shorter chain that bypasses the cytochrome c pathway. Electrons may enter the respiratory system at various points, as NADH from cytosolic reactions, from reactions that reduce UQ (such as succinate dehydrogenase) or from periplasmic oxidation reactions that reduce various cytochromes c. Rhodobacter (Rba.) sphaeroides and Rba. capsulatus are two of the most intensively studied prokaryotes in terms of their structural, functional and genetic features of aerobic respiration; as such they provide useful focal points for this discussion. As studies of the cytochrome bc1 complex and the ATP synthase are presented elsewhere in this volume (Chapter 22, Berry et al.; Chapter 23, Kramer et al.; Chapter 24, Feniouk and Junge), they will not be discussed here. A. Respiratory Complexes and Electron Transport Carriers 1. NADH Dehydrogenase (NADH:Q Oxidoreductase) The aerobic respiration system of Rba. sphaeroides begins with the presence of two versions of the proton-pumping NADH dehydrogenase (termed NDH-1), as predicted by the genome (Joint Genome Institute Microbial Genomics, 2006, http://genome. jgi-psf.org/mic_home.html). Both predicted com-
Chapter 27
Respiratory Complexes
539
Fig. 1. The aerobic respiratory system of Rhodobacter sphaeroides. Note that expression of the genes for the two terminal oxidases containing fusions of subunits I and III is yet to be demonstrated.
plexes have high similarity to the core of the more elaborate mitochondrial enzyme (Other groups of bacteria also synthesize a single-subunit, non-proton pumping NADH dehydrogenase (NDH-2), and/or a sodium-translocating NADH dehydrogenase similar to NDH-1 but, thus far, these enzymes have not been found in the purple photosynthetic bacteria). NADH dehydrogenase is the largest of the bacterial electron transfer complexes, consisting of 14 protein subunits that contain a flavin cofactor and nine iron sulfur (Fe-S) centers (Friedrich and Scheide, 2000; Yagi et al., 2001). The characteristic L-shaped structure of the NADH dehydrogenase complex consists of a membrane-embedded arm plus an extramembrane domain that extends into the cytoplasm (Yagi and MatsunoYagi, 2003). A recent structure of the eight-subunit hydrophilic domain of Thermus (T.) thermophilus NADH dehydrogenase confirms that all of the redox centers are bound within the extramembrane domain (Sazanov and Hinchliffe, 2006). NADH is oxidized by flavin mononucleotide near the top of the extramembrane domain and the electrons are transferred to the site of quinone reduction, located at the interface of the extramembrane and membrane domains, near the cytoplasmic surface of the membrane. The structure of T. thermophilus NADH dehydrogenase emphasizes the well-recognized question of how electron transfer in the extramembrane domain drives proton pumping through the membrane domain of the complex. The Rba. sphaeroides 2.4.1 genome [as well as those of strains 2.4.3 (ATCC 17025) and 2.4.9 (ATCC 17029)] includes two nuo operons encoding this large complex, which are differentially regulated (Pappas et al., 2004). Under low oxygen conditions, the expression of one nuo operon is down-regulated while the other is
enhanced. Interestingly, other members of the purple photosynthetic bacteria, including Rba. capsulatus, Rhodospirillum rubrum and Rhodopseudomonas (Rps.) palustris contain only one nuo operon, corresponding to the one that is upregulated in the presence of O2 in Rba. sphaeroides. Electrons from NADH dehydrogenase are transferred into the ubiquinone (UQ) pool in the membrane, from which they may enter the cytochrome bc1 complex to flow via the ‘cytochrome c’ pathway to O2, or the electrons may be shunted directly to a quinoloxidizing terminal oxidase that reduces O2. Here, the respiratory system is elaborated in two ways. First, the genome of Rba. sphaeroides encodes no fewer than five different terminal oxidases, and second, a variety of c-type cytochromes are present and capable of donating electrons to the cytochrome c oxidases. Rba. sphaeroides is more economical with regards to the cytochrome bc1 complex; one operon encodes this complex, and the same enzyme participates in both the respiratory and photosynthetic electron transfer pathways. We will first discuss the types and basic characteristics of the terminal oxidases. Because the literature on terminal oxidases is vast, many of the citations refer the reader to reviews or recent articles where references to the original research may be found. 2. Cytochrome bd-type Terminal Oxidase Two super-families of bacterial terminal oxidases are the cytochrome bd-type quinol oxidases, in which the site of O2 reduction contains two heme groups that are buried within the largest subunit, and the heme-Cu oxidases that contain one five-coordinate
540
Davide Zannoni, Barbara Schoepp-Cothenet and Jonathan Hosler
heme and a closely-associated copper atom at the similarly buried site of O2 reduction (Garcia-Horsman et al., 1994). Cytochrome bd-type oxidases are common throughout aerobic eubacteria, but the most studied is that of Escherichia (E.) coli (Junemann, 1997). Of the two integral membrane subunits that form this complex, one, CydA, contains all of the redox centers as well as the proposed quinol binding site near the periplasmic surface (Dueweke and Gennis, 1991; Junemann, 1997; Mogi et al., 2006). A six-coordinate b-type heme mediates electron flow from quinol to the di-heme active site, composed of a b-type heme and an O2-binding d-type heme. The bd-type oxidase is not a proton pump, although it is electrogenic because the protons required for O2 reduction are taken up from the n (negative) side of the cytoplasmic membrane, through an undefined pathway, while the electrons from quinol are delivered from the p (positive) side of the membrane (Zhang et al., 2004) (In fact, all of the terminal oxidases are electrogenic in this way, whether or not they are also proton pumps). The architecture of the di-heme active site is argued to be responsible for the fact that the bd-type oxidase of E. coli has a high affinity for O2, with a Km of 3–8 nM (D’Mello et al., 1996; Borisov et al., 1994). A general characteristic of these enzymes is that a significantly higher concentration of cyanide is required to inhibit activity, as compared to hemeCu oxidases (Junemann, 1997). Compared to the bd-type oxidase of E. coli, relatively little is known about the bd-type oxidase of Rba. sphaeroides. Its transcript abundance is unaffected by ambient O2 concentration (Pappas et al., 2004). It is present in the membranes of aerobically grown Rba. sphaeroides cells, but it accounts for much less of the total O2 reduction activity than the cbb3- or the aa3-type oxidases (Hosler, unpublished). However, when the ‘cytochrome c pathway’ is inactivated in Rba. sphaeroides, thereby preventing electron flow to terminal oxidases that require reduced cytochrome c, the bd-type oxidase is capable of supporting aerobic growth (Yun et al., 1990; Cox et al., 2001). Rba. sphaeroides makes no heme d, thus its bd-type oxidase contains a heme bb active site, as is the case in several other bacteria. The heme bb-subtype has not been purified and characterized, therefore it is not known to what extent, if any, the heme substitution affects O2 affinity or oxidase activity. Apart from genomic information, the existence of a bd-type terminal oxidase in other members of the non-sulfur purple bacteria has been deduced from cyanide
resistant O2 reduction activity that is simultaneously resistant to inhibitors of the cytochrome bc1 complex (Bonora et al., 1998). 3. Cytochrome cbb3-type Terminal Oxidase The cbb3-type cytochrome c oxidases, members of the heme-Cu superfamily, are also widely distributed throughout the eubacteria. The catalytic core consists of three subunits, where the largest, CcoN, contains the heme b3-CuB O2 reduction site along with a low spin heme b that transfers electrons to the active site (Pitcher and Watmough, 2004). Electrons are transferred into CcoN by two other subunits, CcoO and CcoP, which consist of membrane anchoring helices and periplasmic domains containing c-type hemes. CcoO contains a single c-type heme, while CcoP contains two c-type hemes. Whether these two subunits function independently or in series to transfer electrons from cytochrome c to CcoN is not clear, although a homology model of the cbb3-type cytochrome c oxidase (CcO) of Rba. sphaeroides suggests that the latter is more likely (Sharma et al., 2006). Based on the homology of CcoN and CcoO to the two subunits of NO reductase, CcoO is thought to be the immediate electron donor to the low spin heme of CcoN. Whether di-heme CcoP is absolutely required for the activity of all cbb3-type CcOs is yet unclear. Inactivation of the gene for CcoP in Bradyrhizobium japonicum eliminates much but not all of the activity of the complex (Zufferey et al., 1996), while inactivation of its homolog in Rba. capsulatus eliminates all activity (Koch et al., 1998). In addition, the site(s) at which soluble cytochrome c or membrane-bound cytochrome cy bind to donate electrons remains to be identified. A fourth subunit, CcoQ, is a small peptide with a single transmembrane helix that appears to stabilize the CcoNOP complex, possibly by minimizing proteolysis (Oh and Kaplan, 2002). The ccoGHIS operon encoding proteins required for the assembly of the cbb3-type CcO is located adjacent to the ccoNOQP operon encoding this enzyme. The function of the ccoGHIS gene products is discussed in Chapter 21 (Sanders et al.). Transcript abundance for cbb3-type CcO of Rba. sphaeroides increases as ambient O2 decreases (Pappas et al., 2004), although the cytoplasmic membranes of aerobically-grown Rba. sphaeroides cells still contain large amounts of the enzyme. In fact, the cbb3type CcO of Rba. sphaeroides appears to be present under all laboratory growth habits. The affinity of the
Chapter 27
Respiratory Complexes
cbb3-type CcO for O2 has only been measured for the Bradyrhizobium japonicum enzyme, where it is only expressed under microaerobic or anaerobic conditions. Even though this oxidase contains a heme-Cu active site, a Km of 7 nM O2 was obtained (Preisig et al., 1996), essentially the same as for the bd-type oxidases (see above). Unlike the bd-type oxidases, the cbb3-type CcOs pump protons from the n side to the p side of the cytoplasmic membrane, in addition to generating a transmembrane voltage gradient by consuming substrate protons from the n side. This proton pumping activity has been demonstrated in several species, including Rba. sphaeroides (ToledoCuevas et al., 1998). Particularly intriguing is the role of the cbb3-type CcO in gene regulation. Synthesis of the photosynthetic apparatus of Rba. sphaeroides normally requires anaerobic or microaerophilic growth conditions. However, when the genes for the cbb3-type CcO are deleted from the cell, photosynthesis proteins accumulate during aerobic growth (Oh and Kaplan, 1999). Preventing electron flow to the cbb3-type CcO by deleting genes directing the synthesis of the cytochrome bc1 complex or cytochromes c2 and cy has the same effect (Daldal et al., 2001; Rios-Velazquez et al., 2003). The connection between the cbb3-type CcO and the photosynthesis genes is thought to occur through the PrrA/PrrB (also called RegA/RegB in Rba. capsulatus) response regulator/histidine kinase system. Normally, the level of electron flow through the cbb3-type CcO modulates the kinase/phosphatase activities of PrrB such that greater electron flow leads to greater repression of photosynthetic gene expression (Oh et al., 2004). The biochemistry that links electron flux through the cbb3-type CcO to PrrB activity remains to be elucidated. The redox state of the quinone pool, which will vary with the rate of electron flow to O2, affects the autophosphorylation activity of RegB of Rba. capsulatus, the homolog of PrrA in Rba. sphaeroides (Swem et al., 2006). However, recent experiments show that mutation of a single conserved histidine residue of in Rba. sphaeroides CcoN, not a heme ligand, breaks the sensor linkage in that electron flux through the mutant cbb3-type CcO is the same as through the wild-type enzyme but the photosynthetic genes are not repressed under aerobic conditions (Oh, 2006). This argues against the involvement of the quinone, since the high electron transfer activity of the mutant oxidase should maintain the same redox poise of the quinone pool as wild-type cbb3-type CcO.
541 4. Cytochrome aa3-type Terminal Oxidase The aa3-type CcO of Rba. sphaeroides is the best characterized terminal oxidase of the purple photosynthetic bacteria, with three published crystal structures (Svensson-Ek et al., 2002; Qin et al., 2006) and numerous structure/function studies via site-directed mutagenesis (Richter and Ludwig, 2003; Branden et al., 2006a; Hosler et al., 2006). The twelve transmembrane helices of subunit I follow the architecture of other members of the heme-Cu oxidase family, binding the six-coordinate heme a center plus five-coordinate heme a3 with a nearby copper (CuB), which function together as the O2 reduction center. Subunit II contains two transmembrane helices that anchor a periplasmic domain containing the di-copper CuA center. Subunit III consists of seven transmembrane helices, bound to the face of subunit I opposite that of the helices of subunit II. Its function will be discussed below. A small subunit IV, of unknown function, contains a single transmembrane helix that lies against subunits I and III. Subunits I, II and III are closely related to the catalytic core of mitochondrial CcO, making this Rba. sphaeroides enzyme an appropriate experimental model for human CcO. In the catalytic mechanism of the aa3-type CcO, cytochrome c binds to subunit II and donates its electron to CuA (Wang et al., 1999). The electrons then flow to heme a in subunit I, and then to the heme a3-CuB center (Hill, 1994). One or two substrate protons (those destined for water) are taken up from the n side of the membrane during the reduction of the active site (reviewed in Hosler et al., 2006). Once both heme a3 and CuB are reduced, O2 binds transiently to CuB and then to heme a3 (Lemon et al., 1993). Current evidence indicates that O2 is rapidly reduced by four electrons (Proshlyakov et al., 1998; Babcock, 1999; Morgan et al., 2001), two from the reduced heme a3, one from reduced CuB, plus one from a covalently cross-linked histidinetyrosine group in the active site (Yoshikawa et al., 1998) that forms a radical (Proshlyakov et al., 2000; Proshlyakov, 2004). The histidine of the cross-linked His-Tyr pair is one of the three ligands of CuB. With this concerted four-electron reduction of O2, the O-O bond is broken and the first water is formed, which binds to CuB as hydroxide, while the other oxygen atom forms an oxoferryl of heme a3 (a34+=O2-). The oxoferryl and the histidine-tyrosine radical persist in the active site until two more electrons are delivered by the cytochrome c-CuA-heme a pathway, and the
542
Davide Zannoni, Barbara Schoepp-Cothenet and Jonathan Hosler
remaining substrate protons are delivered (Adelroth and Brzezinski, 2004). These electrons and protons 1) protonate the hydroxide on CuB, releasing water; 2) reduce the oxoferryl to heme a33+ plus a hydroxide, which binds to CuB; and 3) reduce the tyrosine radical. The aa3-type CcO is an efficient proton pump, with one proton being transported through the protein, across the membrane, for every proton (or electron) that is delivered to O2 (Wikström, 2004; Hosler et al., 2006). The aa3-type CcO proteins have a lower affinity for O2 than the cbb3- or bd-type oxidases, with reported Km values of 0.4 to 5 µM (Poole et al., 1979; García-Horsman et al., 1991; Riistama et al., 2000; Pils and Schmetterer, 2001). Transcription of the genes for the aa3-type CcO of Rba. sphaeroides requires high ambient O2 (Pappas et al., 2004), although the mechanism of O2 regulation is not yet clear. As an example of the requirement for high O2, the aa3type CcO cannot be detected spectroscopically in cells taken from aerobic plate cultures whereas it is present once cells are grown in liquid culture with vigorous aeration (Hosler, unpublished). Since the biochemical functions of the cbb3- and aa3-type CcOs are the same (both are cytochrome c oxidases and both are proton pumps), but the cbb3-type CcO is synthesized at both low and high ambient O2, it becomes a question as to why the aa3-type CcO is retained. Indeed, Rba. capsulatus synthesizes only the cbb3-type enzyme. Increased transcription of the genes for the aa3-type CcO of Rba. sphaeroides occurs in response to the addition of H2O2 (Zeller et al., 2005), raising the possibility that the aa3-type CcO helps relieve oxidative stress. 5. A Putative caa3-type Cytochrome c Oxidase and a Putative Quinol Oxidase Genes for two additional terminal oxidases that have yet to be expressed under laboratory growth conditions are found in the genome of Rba. sphaeroides. The first of these is a putative caa3-type CcO since the gene for its subunit II predicts that it should contain both CuA and a cytochrome c (Mackenzie et al., 2001). The presence of a caa3-type enzyme in Rba. sphaeroides is remarkable, since these enzymes are generally found in Gram-positive bacteria that lack soluble, periplasmic, c-type cytochromes. Even more remarkable is the subunit I for this enzyme, predicted to be a fusion of a typical subunit I domain with an integral membrane protein that contains seven trans-
membrane helices. The predicted organization of this latter domain is similar to subunit III of the aa3-type CcOs, although the amino acid sequence of this region has little homology to known subunit III sequences. Such gene fusions (both subunit II-cytochrome c and subunit I-subunit ‘III’) are found in Archaea that grow in extreme environments. The genes for a fifth potential oxidase of Rba. sphaeroides appear to encode a heme-Cu oxidase with the ability to oxidize quinol (Mouncey et al., 2000). The largest subunit of this enzyme also appears to be a fusion of usual genes for subunit I plus a subunit III-like integral membrane protein. The notion that this terminal oxidase accepts electrons from quinol is derived from the observation that its predicted subunit II is similar to those of other quinol oxidases in the heme-Cu family. The genes for these two oxidases do not appear to be expressed under conditions tested thus far (Mouncey et al., 2000), and attempts to express them from other promoters have not met with success. Both sets of genes are found in the genomes of the three strains of Rba. sphaeroides that have been sequenced to date, and the genes for the caa3 form with fused subunits I and III are present in the genomes of Rps. palustris strains as well (Joint Genome Institute Microbial Genomics, 2006, http://genome.jgi-psf.org/mic_home.html). This suggests that members of the purple photosynthetic bacteria can perform aerobic respiration under some specialized, perhaps extreme, condition that has yet to be identified. 6. Cytochromes c and HiPIPs (High-Potential Iron-sulfur Proteins) Rba. sphaeroides synthesizes several c-type cytochromes that mediate electron transfer between the cytochrome bc1 complex and the various CcOs, including the soluble cytochrome c2 (Meyer and Cusanovich, 1985), isocytochrome c2 (Rott et al., 1993) and membrane-bound cytochrome cy (Jenney and Daldal, 1993). All three of these proteins are efficient donors to both the cbb3- and the aa3-type CcOs (Hochkoeppler et al., 1995b; Daldal et al., 2001). The relative affinities of all three cytochromes for each CcO and the maximum rates of O2 reduction that they can support are all similar (Drosou et al., 2002; Donohue and Hosler, unpublished). Therefore, if there is a preferred cytochrome c substrate for either oxidase in vivo it has yet to be identified. All three cytochromes c are present in aerobically grown cells. Transcript abundance for Rba. sphaeroides
Chapter 27
Respiratory Complexes
cytochrome c2 is increased in the absence of O2, consistent with its role in photosynthetic electron transfer, but transcription of the genes for isocytochrome c2 and cytochrome cy are higher in the presence of O2 (Pappas et al., 2004). In the middle of the nineties, after almost three decades of studies on the fundamental role of the soluble cytochrome c2 in both photosynthesis and respiration of purple phototrophs, it became clear that highpotential iron sulfur proteins (HiPIPs) are soluble electron carriers alternative to cytochrome(s) c, in most phototrophic species of the genera Rhodoferax, Rhodocyclus, Rhodospirillum, Chromatium and Ectothiorhodospira (Meyer and Donohue, 1995; Ciurli and Musiani, 2005). In Rhodoferax fermentans, a facultative phototroph lacking soluble cytochrome c2 but expressing high amounts of HiPIP, lightinduced respiration (also called LIOU, for LightInduced Oxygen Uptake; Zannoni et al., 1998) is dependent on the concentration of the HiPIP (E0´ = + 351 mV) (Hochkoeppler et al., 1995a; Hochkoeppler et al., 1995c). Cells of the halophilic facultative phototroph Rhodospirillum salinarum (reclassified as Rhodovibrio salexigens by Imhoff et al., 1998) grown aerobically in the dark contain small amounts of cytochrome c´ along with two isoforms of HiPIPs (Moschettini et al., 1999; Hochkoeppler et al., 1999). Since in isolated membrane fragments the HiPIPiso1 was shown to greatly accelerate the aa3-type CcO activity, it has been suggested that this isoform is indeed a functional component of Rsp. salinarum respiratory chain (Hochkoeppler et al., 1999). B. Possible Reasons for Complex Respiratory Systems While mitochondria contain a single route for electron transfer from NADH to O2, purple photosynthetic bacteria such as Rba. sphaeroides contain numerous pathways. Several reasons for such complexity can be proposed, even though definitive evidence for these functions is not strong, and may be difficult to obtain by standard methods. The most obvious reason for multiple terminal oxidases is to allow the bacterium to adapt to differing ambient O2 concentrations. Transcript abundance data from gene chip analyses are consistent with this notion, however the cbb3-type CcO, with its higher affinity for O2, is also present in large amounts along with the aa3-type CcO in aerobically-grown cells of Rba. sphaeroides. In addition, the bd-type quinol oxidase, which may have
543 the highest affinity for O2, is expressed at both high and low ambient O2. Determination of the amount of each oxidase present, and the relative flux of electrons through each, at differing O2 concentrations, is required to fully understand how various respiratory systems respond to O2. The cbb3- and bd-type oxidases almost certainly protect the photosynthetic apparatus from damage by O2, by removing O2 at the cytoplasmic membrane and keeping the environment of the intracytoplasmic membrane anaerobic. A possible related use of these oxidases is to poise the redox levels of the cytochrome c and the quinone pools for optimum cyclic electron transfer during photosynthesis. Since over-reduction of either pool may slow cyclic electron transfer, rendering photosynthesis less efficient for ATP production, either the bd- or the cbb3-type oxidase may be used to shunt excess electrons to trace amounts of O2 during photosynthesis. Another obvious, but unproven, use for multiple respiratory pathways is to allow the cell to modulate its efficiency of ATP synthesis. For example, the transfer of two electrons from NADH to O2 via the bd-type terminal oxidase should result in a total of six charge separation events (equivalent to six pumped protons), while the transfer of the same two electrons to O2 via the cytochrome bc1 complex and the CcOs should result in a total of ten charge separation events (ten pumped protons). Thus, the choice of pathway may lead to a 167% increase in the efficiency of ATP synthesis. Whether or not the cell actually utilizes more efficient pathways in response to energetic demand, or simply increases the rate and amount of less efficient pathways, requires more investigation. Shorter respiratory pathways may also help the cell regenerate NAD+ under conditions where the ATP/ ADP pool is already highly phosphorylated. Other reasons for the occurrence of specific oxidases have already been discussed (e.g., the regulatory function of the cbb3-type CcO) C. Structure-function Studies of the aa3-type Cytochrome c Oxidase The close similarity of the aa3-type CcO of Rba. sphaeroides to the catalytic core of mitochondrial CcO makes it a useful experimental model. Structurefunction analyses of Rba. sphaeroides CcO became possible with two key advances. The first was the development of a mutagenesis system (Shapleigh et al., 1992; Hosler et al., 1993), while the second
544
Davide Zannoni, Barbara Schoepp-Cothenet and Jonathan Hosler
was the release of a high resolution crystal structure containing all four subunits, the redox centers, the Mg center between subunits I and II and six tightlybound lipids (Svensson-Ek et al., 2002). A higher resolution structure has recently been obtained for subunits I and II (Qin et al., 2006). The following is a short discussion of a few of the salient findings from studies of the Rba. sphaeroides enzyme, although studies of other systems, particularly the aa3-type CcO of Paracoccus denitrificans, have also contributed greatly to the current understanding of CcO function (Richter and Ludwig, 2003). 1. Proton Pathways and Proton Pumping. Two proton pathways lead from the inner surface of CcO (the cytoplasmic or negative surface) toward the buried heme a3-CuB center. These were identified in CcO structures with the help of knowledge obtained in previous mutagenesis studies (Hosler et al., 1993; Fetter et al., 1995; Mitchell et al., 1996). For the D pathway, protons enter via the surface residue Asp 132 of subunit I and are transferred ~26 Å through a series of hydrogen-bonded water molecules to Glu 286, located midway between hemes a and a3 (Svensson-Ek et al., 2002). Substrate protons (those used to reduce O2) flow from Glu 286 to the active site through a shorter series of waters. Because Glu 286 is buried, it has an unusually high pKa (9.4), and it remains protonated due to rapid proton delivery from Asp 132 (Namslauer and Brzezinski, 2004). The K pathway for protons begins at Glu 101 of subunit II, on the inner surface but far from Asp 132 (Tomson et al., 2003). K pathway protons move ~ 26 Å via amino acid side chains and bound waters to Tyr 288, which is covalently linked to His 284, a CuB ligand. The D and K pathways operate at different times during the catalytic cycle. The K pathway imports one (possibly two) proton as heme a3 and CuB are being reduced prior to the binding of O2 (Adelroth et al., 1998; Ruitenberg et al., 2000). This proton combines with a hydroxyl group on CuB, releasing the second water of the overall cycle, as discussed above in section I.D. Once O2 binds, the D pathway brings in the remaining substrate protons (Adelroth and Brzezinski, 2004). The mechanism of how the active site accepts protons from one pathway or the other remains unsolved. The D pathway provides all four of the protons that are pumped through CcO, across the membrane, as one O2 is reduced to two waters (Branden et al.,
2006a; Hosler et al., 2006). The pump is thought to be closely linked to the active site in subunit I, although the actual site and mechanism of pumping are not yet clear. The cleavage of the O-O bond by the concerted four-electron reduction of O2 initiates a series of active site intermediates, all with the capability to draw protons into the protein (Branden et al., 2006a). Glu 286 of the D pathway appears to be a branch point from which protons are directed to O2 reduction intermediates or to the proton pump. At four sequential conversions between active site intermediates, two protons are taken up from the negative surface of the membrane and one proton is ejected to the positive surface of the membrane. Of the two protons that are taken up from the negative surface, one is pumped while the other combines with an electron from the positive surface in the production of electrically neutral water. Solving the pumping mechanism requires ingenuity. Mutagenesis of residues in or near the active site often inactivates the enzyme, but mutagenesis of more distant residues can provide important insights. For example, placing a carboxylate residue within the D pathway eliminates proton pumping while maintaining O2 reduction (Pawate et al., 2002). It appears that electrostatic interactions raise the pKa of Glu 286 to the point where it fails to protonate the pump, but Glu 286 can still be deprotonated by the O2 reduction intermediates which have very high affinity for protons (Branden et al., 2006b). A key component of CcO is water, not all of which appears in the crystal structures. As the positions of waters become more defined, it is possible to deduce the sites of proton binding within the protein, e.g., those in water clusters, in hydrogen bonds and on amino acid side chains such as the ligands of CuB in the active site. This detail allows the formulation of more chemically precise pumping mechanisms such as one recently derived by first describing the positions of ‘conserved’ waters in subunit I by phylogenetic and structural analyses (Sharpe et al., 2005). Higher resolution structures increase the confidence of structural water assignments (Qin et al., 2006). In addition to mutagenesis, testing of proton transfer mechanisms is being pursued by a wide variety of computational methods (reviewed in Hosler et al., 2006). Another mystery to be solved is the path that transfers protons from the site of pumping to the outer surface of CcO. A single proton exit pathway is not apparent in the CcO structures. An important finding in this search is that of reverse proton flow,
Chapter 27
Respiratory Complexes
presumably through the normal exit pathway (Mills et al., 2002). Under conditions (such as a high transmembrane voltage gradient) that inhibit proton uptake from the inner surface, CcO maintains a slower flow of protons from the outer surface to the active site. An increasing collection of mutants that affect proton backflow should help map this pathway (Mills et al., 2005). Proton backflow may have an important protective effect for CcO (see below). 2. Protection from Oxidative Damage Enzymes that reduce O2 are prone to oxidative damage. As described above, once O2 binds to reduced heme a3, a four electron reduction rapidly converts diatomic O2 to a bound oxoferryl (a34+=O2–) and a hydroxyl anion on CuB. The radical of the Tyr 288–His 284 group is also generated. The rapid reduction mechanism has the benefit that the one and two electron reduction intermediates, superoxide and H2O2, cannot be released into the cell. A disadvantage is that two potent oxidizing agents, the oxoferryl and the tyrosine radical, persist within the protein until further electrons and protons are delivered to the buried active site. Interaction of these oxidizing agents with the surrounding protein, if allowed to occur, is likely to be detrimental. For the aa3-type CcO, a major part of the solution to this problem appears to be subunit III. Subunit III is as conserved as subunit I, but it contains no metal centers, and thus plays no direct role in electron transfer or O2 reduction. Subunit III does bind two structural lipids within a deep cleft; one of the lipids lies against subunit I and comes within 15 Å of the active site (Svensson-Ek et al., 2002). Subunit III can be removed from CcO using strong non-ionic detergent. The resulting sub-complex is highly stable, has near normal O2 reduction activity and pumps protons. However, with continued catalytic turnover subunit III-deficient CcO rapidly and irreversibly loses activity, or suicide inactivates (Bratton et al., 1999; Hosler, 2004). Suicide inactivation has been traced to events at the heme a3-CuB center, and it is much more probable under conditions that extend the lifetime of the oxoferryl during the catalytic cycle, pointing toward this intermediate as the causative agent (Mills and Hosler, 2005), although the tyrosine radical might also be involved. Thus far, studies indicate that subunit III normally functions to prevent suicide inactivation in three ways. First, subunit III enhances proton uptake into the D pathway, which shortens the lifetime of the oxoferryl during
545 the catalytic cycle (Gilderson et al., 2003; Adelroth and Hosler, 2006). Second, subunit III promotes the activity of the proton backflow pathway in subunit I (the apparent reverse of the exit pathway for pumped protons) (Mills et al., 2003). This helps to limit the lifetime of the oxoferryl under conditions where proton uptake from the inner surface of CcO is inhibited. Finally, normal binding of the structural lipids in the cleft of subunit III is required to prevent suicide inactivation (Varanasi et al., 2006). The lipids appear to function as structural modifiers, perhaps decreasing protein movements associated with proton pumping that would bring the oxoferryl close to an internal target. 3. Cytochrome c Oxidase Assembly The post-translational assembly of mitochondrial CcO requires a large number of proteins specific for this process. Several of these proteins have homologs in Rba. sphaeroides. A key advantage of studying CcO assembly in the bacterium, as opposed to mitochondria, is that partially assembled CcO forms accumulate in the membrane, from which they can be isolated and characterized (Bratton et al., 2000). For example, the isolation of a CcO form containing all of the redox centers except for CuB provides the direct physical evidence necessary to conclude that the copper-binding protein Cox11p is absolutely required for the assembly of CuB of the aa3-type CcO (and not for the assembly of CuA) in both Rba. sphaeroides and in mitochondria (Hiser et al., 2000). Another assembly protein, Surf1, is not absolutely required for CcO assembly but early events in the assembly of the complex in mitochondria are markedly affected by its absence, leading to a deficiency of CcO (Stiburek et al., 2005). Recent work in Rba. sphaeroides reveals that Surf1 is required for the efficient assembly of heme a3 of the active site in subunit I, but not for the assembly of heme a in subunit I or CuA in subunit II (Smith et al., 2005). Further details of CcO assembly are discussed elsewhere in this volume (Chapter 21, Sanders et al.) D. Structure-function Studies of the cbb3-type Oxidase Analysis of the cbb3-type CcO has lagged behind that of the aa3-type enzymes. Emerging studies are of high interest since the comparison of proton pumping CcOs with considerable variations in structure
546
Davide Zannoni, Barbara Schoepp-Cothenet and Jonathan Hosler
reveals common features necessary for O2 reduction and proton pumping. A crystal structure of a cbb3-type CcO is not yet available, but homology models of CcoN (subunit I) of the Rba. sphaeroides and Vibrio cholerae enzymes have been constructed (Hemp et al., 2005; Sharma et al., 2006). A proton channel analogous to that of the K pathway of the aa3-type CcOs leads from the inner surface of CcoN to a His-Tyr group at the heme b3-CuB active site (Sharma et al., 2006). The His-Tyr group at the active site appears analogous to that present in the active site of the aa3type CcOs but the tyrosine is located on a different transmembrane helix (Hemp et al., 2005; Sharma et al., 2006). Alteration of the tyrosine to a phenylalanine eliminates activity and the existence of the His-Tyr cross-link has recently been confirmed in both the Rba. sphaeroides and Vibrio cholerae cbb3-type CcOs (Hemp et al., 2006; Rauhamaki et al., 2006). Thus, a His-Tyr group at the active site must be an essential feature in the heme-Cu terminal oxidases since it is found throughout this wide group regardless of other structural differences. A major difference between the aa3- and cbb3-type oxidases is the lack of an analog of the D pathway in the latter group. Further studies of the cbb3-type CcO may help reveal the reason for the existence of two proton uptake pathways (D and K) in the aa3-type enzymes. II. Respiration Utilizing Substrates other than Oxygen Many members of the α-, β- and γ-subgroups, in particular the facultative phototrophs, are commonly considered to be the most versatile of all prokaryotes with respect to energy metabolism. Species such as Rba. capsulatus and Rba. sphaeroides may grow aerobically and photosynthetically using either organic or inorganic substrates but also anerobically (in darkness) with trimethylamine-N-oxide (TMAO) or dimethylsulfoxide (DMSO) as electron acceptors. Oddly, the question of whether anaerobic reduction of TMAO or DMSO is a true respiratory process, or a peculiar example of fermentative process requiring accessory oxidants to proceed, is still open to debate (McEwan et al., 1985; Zannoni, 1995). Some strains of the species Rps. palustris, Roseobacter (Rsb.) denitrificans and Rba. sphaeroides can reduce nitrate (NO3–) into dinitrogen (N2) via nitrite (NO2–), and in some cases also nitric oxide (NO) and nitrous oxide (N2O) (see Chapter 31, Shapleigh). In
addition to widespread substrates such as O2, NO3–, H2 or organic acids and alcohols, a plethora of less well-known redox chemicals present in the environments are recruited for bioenergetic purposes. These redox chemicals comprise ‘nasty’ substances such as arsenics, chlorates and perchlorates or halogenated aromates (see also Chapter 29, Harwood). Admittedly, not all of these substrates have been shown to be used by photosynthetic members of the proteobacteria (i.e., the purple bacteria). We consider, however, that this is at least partly due to the lack of experimental data rather than the actual absence of phototrophic proteobacteria performing these reactions. For decades, studies of purple bacterial energy conservation have focused on the photosynthetic pathways, or at best on the oxygen-respiratory mechanisms operative under aerobic conditions (i.e., at relatively high oxygen tensions). Consequently, the respiratory systems working under low oxygen tensions in the dark (i.e., the microaerophilic lifestyles), are only about to be elucidated in several purple bacteria. At the same time, genomic data on the phototrophs are scarce, and the purple bacteria are certainly under-represented in genome sequencing projects, most probably due to their low priority under a biomedical perspective of the microbial world. These considerations led us to include in this chapter not only those ‘exotic’ or unconventional types of energy conservation that have actually been observed in purple bacteria, but also those which we think are very likely to exist in various members of this class of proteobacteria. The respiratory reactions involving inorganic nitrogen are considered to be ‘common’, and the reader is referred to recent reviews covering these topics (Richardson, 2000; Ferguson and Richardson, 2004) as well as to Chapter 31 (Shapleigh). On the other hand, although microbial DMSO and TMAO respiration has been reviewed recently (McEwan et al., 2004; McCrindle et al., 2005), for the sake of completeness a brief description of this important anaerobic process is included here. A. Dimethylsulfoxide and Trimethylamine-Noxide Respiration Organic sulfur compounds originating from dimethylsulfonioproprionate are largely present in marine habitats. Dimethylsulfonioproprionate is converted to dimethylsulfide (DMS, volatile) and acrylic acid (Taylor and Kiene, 1989; Visscher et al., 1995), and then DMS is used as electron donor in bacterial
Chapter 27
Respiratory Complexes
Fig. 2. Schematic representation of the electrochemical properties of electron donors and acceptors to the enzymes DMS dehydrogenase, ethylbenzene dehydrogenase, arsenite oxidase, arsenate reductase, chlorate and selenate reductase and the respective redox reactions. Em values are given with respect to the standard hydrogen electrode (NHE) and the extent of the grey bars corresponds to the range of 10% to 90% reduction of the respective redox chemicals. The Em value of the substrate to nitrate reductase is indicated for comparison.
photosynthesis (Visscher and Taylor, 1993). As a consequence, DMSO has been detected in seawater at relatively high amounts (Kelly and Smith, 1990), as it also returns in rainfall from the atmosphere where DMS is photochemically converted to DMSO. DMSO can be used in a variety of nutritional modes by prokaryotes mainly under anaerobic conditions as exemplified by Rhodobacter species (Zannoni, 1995). In this latter bacterial group, DMSO reductase is a periplasmic enzyme and possesses a twin-arginine (Tat) secretory N-terminal signal peptide to direct its export (Shaw et al., 1996). In Rba. capsulatus, DMSO reductase is encoded by the dorA gene located in the same operon as the dorC gene, encoding a pentaheme c-type cytochrome of the NirT family (Shaw et al.,
547 1999b). In Rhodobacter species the Km of this enzyme for DMSO is in the micromolar range while its Km for TMAO, another molecule acting as electron acceptor in bacterial respiration, is in the millimolar range (Johnson and Rajagopalan, 2001). The latter high Km value for TMAO would suggest that this electron acceptor is physiologically non-relevant to most purple phototrophs. However, the marine aerobic photosynthetic species Rsb. denitrificans has been shown to reduce TMAO (Arata et al., 1992). On the other hand, the TMAO-reductase of Rsb. denitrificans is slightly different from that present in Rhodobacter species suggesting that its primary role might be in the TMAO respiration. Studies in E. coli show that the key feature that distinguishes TMAO reductase from DMSO reductase is that the former enzyme has a very low activity towards S-oxides (Iobbi-Nivol et al., 1996). The E0´ of the DMSO/DMS couple is +160 mV while that of the TMAO/Trimethylamine (TMA) couple is +130 mV (Wood, 1981; Gon et al., 2001), indicating that both DMS and TMA are potential electron donors (and acceptors) in bacterial respiration and in photosynthetic metabolism. Indeed, the purple phototroph Rhodovulum (Rvu.) sulfidophilum strain SH1 was shown to be capable of photolithotrophic growth when DMS was used as the sole electron donor and DMSO accumulated as a product (Hanlon et al., 1994). DMS dehydrogenase has been purified from Rvu. sulfidophilum, and it has been shown to contain bis(molybdopterin guanine dinucleotide)Mo (McDevitt et al., 2002a). DMS dehydrogenase is coded by the ddh gene cluster, which comprises the ddhA, B, C and D genes. The ddhA gene encodes a polypeptide with highest sequence similarity to the molybdopterin-containing subunits of selenate reductase and ethylbenzene dehydrogenase, while ddhB contains cysteine-rich sequence motifs suggesting that it contains multiple iron-sulfur clusters (McDevitt et al., 2002b). DMS dehydrogenase is part of a larger family of enzymes which will be treated below. The ddhD gene product is deduced to be cytoplasmic and water soluble, and may act as a molecular chaperone specific for the assembly of DdhAB. The observation that DdhD is not a component of purified DMS dehydrogenase is consistent with this view. It is interesting that DdhD is related to SerD (selenate reductase) and NarJ (nitrate reductase). This suggests a common origin in line with the evolutionary relationship between these type II enzymes, according to an earlier classification of enzymes of the DMSO
548
Davide Zannoni, Barbara Schoepp-Cothenet and Jonathan Hosler
reductase family (Trieber et al., 1996; Blasco et al., 2001). In contrast to the DMSO reductase mentioned above, the ddhC product appears to be secreted via the Sec-mediated pathway of secretion, as an apoprotein into the periplasm, where it folds and incorporates its heme b prosthetic group. There are very few examples of periplasmic b-type cytochromes; the majority of cytochromes in that compartment are of c-type, with covalently attached heme (Thony-Meyer, 1997). How the DdhAB complex and DdhC are assembled together in the periplasm to form the mature DMS dehydrogenase is unknown. 1. Organization of Dimethylsulfoxide - Trimethylamine-N-oxide Respiratory Chains Analysis of the organization of a variety of anaerobic respiratory electron transport chains has revealed that they use QH2 dehydrogenases bypassing the cytochrome bc1 complex containing pathway (Richardson, 2000). The NirT family of QH2 nitrite reductases (Jungst et al., 1991) is mainly formed by tetraheme c-type cytochromes that are membrane bound but facing the periplasm. Thus, they are ideally located to transfer electrons from QH2 to periplasmic terminal reductases. The c-type multiheme cytochrome involved in DMSO reduction by Rba. capsulatus is DorC encoded by the dorC gene (Shaw et al., 1999a). It is a pentaheme protein composed of a tetraheme domain similar to other proteins of NirT type, and a C-terminal domain that contains an additional c-type cytochrome. The redox potentials of the heme centers of the purified DorC have been shown to range from –34 to –276 mV (Shaw et al., 1999a), with the highest potential being assumed to correspond to the monoheme that acts as electron donor to DMSO reductase. TorC is a pentaheme protein, which is composed of a tetraheme domain with strong sequence similarity to DorC (Gon et al., 2001). The redox potentials of the hemes in the tetraheme domain are below 0 mV while that of the monomeric c-type heme is about +120 mV. These thermodynamic features seem to fit with a model in which the tetraheme is involved in menaquinol (MQH2) oxidation while the monoheme transfers electrons to the TMAO reductase. Cytochrome c2 is essential for electron transfer from DMS dehydrogenase to the reaction center (RC) during phototrophic growth of Rvu. sulfidophilum (McDevitt et al., 2002). Although DMS dehydrogenase does not appear to be essential for growth under
aerobic conditions with DMS as the sole energy and carbon source, accumulation of DMSO indicates that DMS is oxidized during this growth mode (Hanlon et al., 1994). Under these conditions, DMS dehydrogenase is expected to pass electrons to an energy conserving CcO via a soluble cytochrome c2. Thus, DMS respiration would contribute to the energy metabolism of the cell even during oxidation of carbon substrates by the cell (myxotrophy). Recent analyses have shown that at least three sub-families exist within the DMSO reductase family (McEwan et al., 2002; for a phylogenetic grouping, see Fig. 4). Type I enzymes such as periplasmic nitrate reductase (Nap), formate dehydrogenases and assimilatory nitrate reductase (Nas) contain a [4Fe4S] cluster at the N-terminus of the Mo-containing α-subunit, and have a cysteine or selenocysteine side chain as a ligand to the Mo atom. An exception to this rule is the arsenite oxidase (see section E.1. of this chapter) where the N-terminal iron-sulfur cluster is also of the cubane type but there appears to be no Mo-ligand amino acid side chain. Type II enzymes such as respiratory nitrate reductases (Nar), selenate reductases (Ser) and DMS dehydrogenases (Ddh) possess an N-terminal His/Cys motif that is involved in iron-sulfur cluster formation (Bertero et al., 2003; Jormakka et al., 2004). Type III enzymes exemplified by Rhodobacter DMSO reductase and TMAO reductase lack a cysteine rich motif at their N-termini, and are not associated with a β-subunit that contains multiple [Fe-S] clusters. The ligand to the Mo atom in enzymes of this sub-family is a serine residue. Additional structural and molecular details on DMSO reductases in Rhodobacter sp. can be found in McEwan et al. (2004). B. Arsenics as Bioenergetic Substrates 1. Arsenite Oxidation Although considered the mother of poisons, arsenics are rather widespread components of our environment. Both aquatic habitats and soil contain significant amounts of arsenics (0.1 to more than 1000 ppm in soil, 50 to 400 ppm in atmospheric dust, 2.6 ppb in sea water; Mukhopadhyay et al., 2002). In aquatic environments, which are not at extreme pH values, the biorelevant forms of arsenics are their oxyanions H3AsO3/H2AsO3– (reduced form: arsenite) and H2AsO4–/HAsO42– (oxidized form: arsenate). Genome surveys provided evidence for the presence of the
Chapter 27
Respiratory Complexes
enzyme arsenite oxidase in numerous bacterial and archaeal species (Lebrun et al., 2003; Lebrun et al., 2006). Although arsenite oxidation often appears as a simple detoxification process, its implication in bioenergetics has been demonstrated in a range of non-photosynthetic α- and γ-proteobacteria (Oremland and Stolz, 2005), as well as in the phototroph Chloroflexus aurantiacus (Duval and Schoepp, in preparation). Among the photosynthetic organisms possessing arsenite oxidase genes are, in addition to Chloroflexus, the green sulfur bacteria Chlorobium phaeobacteroides and Chlorobium limicola (see also Table 3 of Chapter 33, Borsetti et al.). It has been speculated that in Chloroflexus electrons originating from the oxidation of arsenite might enter the photosynthetic electron chains (Zannoni and Ingledew, 1985; Lebrun et al., 2003), and subsequent experiments have shown that expression of the enzyme is not upregulated from its constitutive basal level when the bacterium is grown anaerobically in the light and in the presence of arsenite. However, Chloroflexus appears to be able to use arsenite as an electron donor under microaerophilic photosynthetic conditions (Duval and Schoepp, unpublished) like Rhizobium sp. str. NT-26 or Hydrogenophaga sp. str. NT-14 (Santini et al., 2000; van den Hoven and Santini, 2004). The identification of the oxidases and electron carriers involved in these electron transfer chains are under study. In the case of the strains NT-26 and NT-14, a ctype cytochrome seems to be the electron acceptor of arsenite oxidases (van den Hoven and Santini, 2004; Santini et al., 2007). Implication of a cytochrome c oxidase has therefore been suggested. A unique case of a bacterium using arsenite as donor, and nitrate (instead of O2) as terminal electron acceptor has been reported for Ectothiorhodospira strain MLHE-1 (Oremland et al., 2002). Since green sulfur bacteria are, in contrast to the facultatively photosynthetic Chloroflexus, considered to be obligate phototrophs, arsenite oxidase may in these species indeed donate electrons to the (RCI-type) photosynthetic reaction centers. A more in-depth understanding of the role of arsenite oxidation in green sulfur bacteria and possibly in purple bacteria has to await further studies. The actual cellular localization of the enzyme arsenite oxidase is still under debate. Three previously purified enzymes (i.e., those of Alcaligenes faecalis, Rhizobium sp. strain NT-26 and strain NT-14; Anderson et al., 1992; Santini and van den Hoven, 2004; van den Hoven and Santini, 2004) are soluble. By contrast, nearly all arsenite oxidase activity is found
549 in the spheroplast fractions obtained from Alcaligenes faecalis and Herminiimonas arsenicoxydans (Anderson et al., 1992; Muller et al., 2003), and in the membrane fragments (resulting from French press treatment) of C. aurantiacus (Lebrun et al., 2003; Duval and Schoepp, unpublished). The basic building blocks of arsenite oxidases are a small subunit, which is a member of the Rieske [2Fe-2S] protein family (Lebrun et al., 2003; Lebrun et al., 2006) (rendering membrane-attachment likely), and a molybdopterincontaining catalytic subunit belonging to the DMSOreductase superfamily. A more detailed discussion of the subunit composition will be given below. The structure of the soluble form of the enzyme has been determined (Ellis et al., 2001). 2. Arsenate Respiration In more oxidizing environments, i.e., where the oxidized form of arsenic, arsenate, is predominant, it can be used as terminal electron acceptor with acetate, pyruvate or lactate as an electron donor. This mechanism has been shown to operate in several γ−, δ- and ε-proteobacteria, as well as in a few members of the low GC Gram-positive bacteria, of the Deinococci and of Crenarchaeota (Oremland and Stolz, 2003). Therefore arsenate respiration currently appears much less widespread than arsenite oxidation. The basic structure of the arsenate reductase consists of a molybdopterin containing catalytic subunit belonging to the DMSO-reductase superfamily, and a small Fe-S subunit containing four cubane centers, products of the arrA and arrB genes, respectively. As suggested by the results obtained on one of the two enzymes purified (Krafft and Macy, 1998), and by the majority of the identified sequences (Stolz et al., 2006; Duval and Schoepp, unpublished), such a two-subunit respiratory arsenate reductase would be soluble. The process by which these Arr enzymes receive electrons from the quinol pool therefore remains to be established. The Arr enzyme from Chrysiogenes arsenatis is proposed to be periplasmic, thus how it contributes to the formation of a proton gradient is enigmatic. The only presently known example of a membrane-bound enzyme is that from Bacillus selenitireducens (Afkar et al., 2003). However, in several arr operons detected in genome surveys a third gene named arrC, homologous to the psrC gene of polysulfide reductase, is present immediately upstream of arrA (see below Fig. 4) (Stolz et al., 2006). Since PsrC is the membrane anchor subunit of polysulfide
550
Davide Zannoni, Barbara Schoepp-Cothenet and Jonathan Hosler
reductase (Krafft et al., 1992), the presence of this gene in arr operons is in favor of a membrane association of respiratory arsenate reductases. C. Selenate Respiration The microbial reduction of selenate mediated by bioenergetic electron transport in α-, β-, γ-, ε- proteobacteria and Gram-positive bacteria (Oremland et al., 1994; Switzer Blum et al., 1998), and in particular in their photosynthetic representatives (such as Rba. sphaeroides) was reported 15 years ago (Moore and Kaplan, 1992; Chapter 33, Borsetti et al.). In some of these species, selenate reduction was subsequently shown (Sabaty et al., 2001) to be a side-reaction of a more common anaerobic energy conserving pathway, i.e., denitrification, and more specifically the respiration of nitrate performed by enzymes of the nitrate reductase family (Nap and Nar) (Avazéri et al., 1997; Sabaty et al., 2001; Watts et al., 2005). More recently, the existence of an enzyme complex truly dedicated to selenate reduction was demonstrated in Thauera selenatis (Schröder et al., 1997), Sulfurospirillum barnesii (Stolz and Oremland, 1999) and Enterobacter cloacae SLD1a-1 (Watts et al., 2003). Selenate reductase, just as arsenite oxidase and respiratory arsenate reductase, belongs to the large superfamily of molybdopterin containing enzymes, and consists of, in addition to the large catalytic subunit harboring the molybdopterin moiety, a tetracubane Fe-S protein and a peculiar b-type cytochrome. A more comprehensive description of the structural features of this enzyme will be given below. Cellular localization of this enzyme seems to vary, since that from Thauera selenatis was reported to be soluble and in the periplasm, whereas that from Enterobacter cloacae has been demonstrated as being membranebound and facing the periplasm. Thus, as evoked for arsenate reductase above, how these enzymes are connected to the quinol pool, and how they generate proton gradient remain open questions. In recent years, selenate reductase has turned out to be a representative of a well defined and quite conserved superfamily of enzymes, the members of which handle extremely divergent substrates that all function in energy conserving photosynthetic and respiratory chains.
D. The Selenate Reductase-type Class of Enzymes At the time being, five distinct enzymes belonging to this class are known. These are selenate reductase (Schröder et al., 1997), chlorate reductase (Danielsson Thorell et al., 2003), DMS dehydrogenase (McDevitt et al., 2002), ethylbenzene dehydrogenase (Kniemeyer and Heider, 2001) and perchlorate reductase (Bender et al., 2005). All these enzymes share the same subunit architecture (perchlorate reductase being slightly divergent, see Fig. 3) with a catalytic molybdopterin subunit, a tetracubane iron sulfur protein and a b-type cytochrome, and are located in the periplasm of the organisms producing them. DMS-dehydrogenase oxidizes dimethylsulfide in the purple non-sulfur bacterium Rvu. sulfidophilum (McDevitt et al., 2002), and injects the resulting reducing equivalents into a soluble cytochrome c which in turn serves as a reductant to the photooxidized special pair of the RCII-type reaction centre. Chlorate and perchlorate reductase, as well as ethylbenzene dehydrogenase, have so far only been detected in non-photosynthetic proteobacteria. The three dimensional structure of ethylbenzene dehydrogenase from Aromatoleum aromaticum strain EbN1 has recently been solved (Kloer et al., 2006), providing an excellent structural model for the whole class of enzymes, considering their conserved subunit composition and the sequence similarity of the individual subunits. E. Non-conventional Substrates but Conventional Bioenergetics Although the above detailed substrates may appear strange, as they are mostly toxic, the energetics involved in the respective electron transport chains are nothing abnormal. Selenate and in particular chlorate are, due to their very high redox potentials at pH 7 (Fig. 2), ideal electron acceptors providing enormous throughput driving forces which, in the case of chlorate, comes close to those attainable by oxygen reduction. Arsenite is a reasonably good electron donor to aerobic respiration provided that it is present in sufficiently high concentrations in the environment, which is the case in most hydrothermal habitats. Consequently, it is not surprising that C. aurantiacus should be able to use this arsenic oxyanion for bioenergetic means. Finally, DMS is present in many marine habitats and thus lends itself as an
Chapter 27
Respiratory Complexes
551
Fig. 3. Construction kit arrangement of the molybdoenzymes arsenite oxidase, arsenate reductase, polysulfide reductase, selenate reductase superfamily (including DMS dehydrogenase, ethylbenzene dehydrogenase and chlorate reductase) as well as formate dehydrogenase and respiratory nitrate reductase. The top part of the figure gives a schematic representation of the operon structure of the various complexes, whereas the bottom part shows the overall structures of the redox subunits involved using known 3-D structures of subunits from the set of enzymes. The structure of the membrane subunit of polysulfide reductase and respiratory arsenate reductase is not known so far and only a schematic version is represented.
alternative source of reducing equivalents in marine phototrophs. The only bioenergetic pathway that may be difficult to rationalize in terms of its energetics is arsenate respiration. Only a small ∆G0´ can be
harvested from the electron transfer from pyruvate to arsenate (Fig. 2), so it is likely that the details of arsenate respiration are not yet well understood.
552
Davide Zannoni, Barbara Schoepp-Cothenet and Jonathan Hosler
F. The Enzymes: Variations on a Theme The enzyme complexes catalyzing the oxidoreduction reactions of these exotic substrates all use a molybdopterin protein belonging to the DMSO reductase superfamily as a catalytic subunit. In addition to this ‘large’ catalytic subunit, a small subunit, frequently a cytochrome, serves as entry or exit point for reducing equivalents. In between these two, a tetracubane iron sulfur protein plays the role of an electron wire, or an electron storage device, in most cases. The enzymes described above, together with the more common formate dehydrogenases, nitrate reductases and DMSO/TMAO reductases, provide examples of the modular composition observed in many bioenergetic enzymes (Baymann et al., 2003). Fig. 3 schematically summarizes the subunit composition of the respective complexes. In this context, it may seem surprising that the subunit composition of the arsenite oxidizing and the arsenate reducing enzymes differ so substantially. Rather than mutating one enzyme, e.g., the oxidase, in specific places to modify electron transfer properties and thus reverse catalytic directionality, Nature apparently has chosen to invent a new enzyme. Instead of a Rieske protein, a tetracubane iron sulfur protein and most likely a large transmembrane subunit are present in arsenate reductase. Even the catalytic molybdopterin subunits of both enzymes show rather low sequence similarities. The Mo subunits of arsenate reductases are, in fact, closer to true DMSO reductases than to those of the arsenite oxidases. The phylogenetic tree built from the sequences of diverse classes of enzymes using the molybdopterin subunit suggests a rationalization for the dissimilarity of arsenite oxidase and arsenate reductase. The tree in Fig. 4 shows all arsenite oxidases clustering together on a tight clade within the tree. The arsenite oxidase subtree itself is split into an archaeal and a bacterial cluster. This suggests that the evolutionary origin of arsenite oxidase pre-dates the divergence of Bacteria and Archaea more than 3 billion years ago (Lebrun et al., 2003). This early origin of arsenite oxidase is corroborated by the position of its Rieske subunit on a tree encompassing arsenite oxidases and Rieske/Cytochrome b complexes (Lebrun et al., 2006). The respective phylogenies of the Mo and the Rieske subunits quite clearly demonstrate that the enzyme arose at a time when the oxidation state of the environment of the Earth was relatively low. Under these conditions virtually no arsenate was
Fig. 4. Schematic representation of the phylogeny of the molybdopterin-containing catalytic subunits observed in arsenite oxidase (Aox), respiratory arsenate reductase (Arr), polysulfide reductase (Psr), membrane-bound nitrate reductase (Nar), periplasmic nitrate reductase (Nap) and DMSO reductase (Dor). Archaeal radiations are shown in grey shading. Phylogenetic trees were reconstructed based on the neighbor joining algorithm of Saitou and Nei (1987) using Clustal X and allowing for multiple substitutions.
present, hence there was no driving force to develop arsenate respiration. The branch of the arsenate reductases, by contrast, shows all properties characteristic of an enzyme derived from another one at a later stage in evolution. Rather than from the stem of the entire tree, the reductase clade diverges from the subtree of the polysulfide reductases and does not feature an Archaea/Bacteria cleavage. Further, the branches of the arsenate reductases are quite mixed up as compared to the phylogenetic tree of their parent species, arguing for a distribution via lateral gene transfer. The comparable subunit composition of arsenate and polysulfide reductases (apart from a reshuffling of the operon structure, see Fig. 3) corroborates the scenario suggested by the phylogeny of the Mo subunits. The gradual advent of an oxidized environment brought about by oxygenic photosynthesis, and the concommitant accumulation of oxidized arsenic oxyanions therefore allowed for arsenate respiration to become energetically profitable and induced the evolution of a polysulfide reductase into an arsenate reductase (Duval and Schoepp, in preparation). The energy conserving processes based on arsenic oxyanions thus provide a textbook example of the intertwined
Chapter 27
Respiratory Complexes
evolutionary history of bioenergetics and of the geochemical environment of the ancient Earth. III. Respiration versus Photosynthesis: Which One Came First? The concept that the early Earth’s atmosphere was basically devoid of oxygen, with values <<1% of present atmospheric levels, is relatively widely accepted (Holland, 1994). However, reinterpretation of geological data, including the analysis of paleosols as well as iron buried in archaean sediments, has led to the proposal that significant amounts of oxygen would have been present in the primordial atmosphere (Towe, 1990, 1994, 1996; Ohmoto 1996, 1997). According to these studies, the amount of oxygen in the early atmosphere would have been at least 1 to 2% of the present values (Towe 1996; Ohmoto 1997). Accepting this latter proposal, it is evident that geological events, such as submarine hydrothermal vents, have contributed to the production of a wide range of reduced inorganic molecules such as H2S, H2, CO and NH4+, able to sustain a variety of aerobic and anaerobic respiratory processes (Madigan et al., 1997; Reysenbach and Shock, 2002). However, ATP-producing mechanisms using electron transfer to generate a proton gradient require both a well-formed membrane and the presence of an ATP synthase. Due to a parallel need for these two complex biological systems, respiration probably originated later than fermentation. Conversely, the order of appearance of chlorophyll-based photosynthesis and respiration seems to favor the latter one. Until recently, it was commonly assumed that, after fermentation, some type of primitive photosynthesis evolved in the most ancient cells (Cavalier-Smith, 2001). This assumption was based on 3500 Mya fossils taken as tracers of cyanobacteria (Schopf et al., 2002). However, these fossils have recently been reinterpreted as artifacts that arose from amorphous graphite (Brasier et al., 2002). Additionally, molecular phylogenesis and comparative analysis of key enzymes involved in photosynthesis and respiration support the view of a late emergence of photosynthesis relative to respiratory pathways (Castresana, 2001). In this context, there are five lineages within Bacteria with some type of bacteriochlorophyll-based photosystem. Purple bacteria (e.g., Rhodobacter, Rhodopseudomonas) and green non-sulfur bacteria (Chloroflexus) use a non-oxygen evolving type-II photosystem. Green sulfur bacteria
553 (Chlorobium) and heliobacteria (Heliobacterium) use a type-I Photosystem. Finally, cyanobacteria use both Photosystems I and II for the more complex oxygenic photosynthesis. Since these five groups are not phylogenetically related, it is likely that the evolutionary pathways that gave rise to the different photosynthetic organisms are not as straightforward as it appears from the literature (Blankenship and Hartman, 1998; Baymann et al., 2001; Xiong and Bauer, 2002). Indeed, the close proximity of some non-oxygenic phototrophs to other non-phototrophs indicates that non-oxygenic photosynthesis has been lost at different stages in some evolutionary lines. In contrast, all known cyanobacteria perform oxygenic photosynthesis, and no other prokaryote outside this group has a similar metabolic option. This continuous maintenance of the same bioenergetic mechanism in such a large prokaryotic group, and for such a long time-scale period (Mya), is unique in the prokaryotic world. Though most textbooks still assume that oxygenic photosynthesis arose very early during the evolution of life, it is obvious from phylogenetic studies and biochemical considerations that oxygenic photosynthesis arose quite late, after the divergence of several other phototrophic bacterial groups (Brochier and Philippe, 2002). Thus photosynthesis, and particularly oxygenic photosynthesis, is not an ancient metabolic pathway. The presence of respiration before cyanobacterial type photosynthesis is consistent with the high affinity of the some terminal oxidases for oxygen, e.g., the cbb3-type cytochrome c oxidases (see Section I.A.3) are functional at O2 concentrations as low as 10–9 M (Preisig et al., 1996). Additionally, other bacterial populations living in niches totally devoid of oxygen would have used electron acceptors alternative to oxygen to sustain anaerobic respiration. These considerations, taken together, are sufficient to conclude that oxygenic photosynthesis appeared after respiration, and that it arose in Bacteria after their divergence from Archaea (Nitschke et al., 1997; Olson, 1999). IV. Respiration and Photosynthesis are Intermingled Continuous illumination of Rhodobacter cells partially inhibits both respiratory and denitrification activities, this latter process being present only under anaerobic conditions (Sabaty et al., 1993). Two nonmutually exclusive mechanisms were proposed to
554
Davide Zannoni, Barbara Schoepp-Cothenet and Jonathan Hosler
explain the way respiration and photosynthesis might interact. First, the membrane potential generated by photosynthesis may affect the activity of electrogenic respiratory complexes in the membrane (Rugolo and Zannoni, 1983; Cotton et al., 1983). Second, there may be direct interaction between redox carriers common to both photosynthetic and respiratory chains (Zannoni et al., 1978). What are the results that support the above proposals? The carotenoid-band shift, a specific indicator of the membrane potential present across the plasma membrane, is identical when induced either by photosynthesis or respiration (Wraight et al., 1978). This indicates that the ∆µH+ is delocalized over the entire cell membrane. This finding is supported by the fact that uncouplers prevent the inhibition of the respiratory activity by light (Rugolo and Zannoni, 1983; Richaud et al., 1986). The inhibition of respiration is essentially at the level of complex I, which is linked to light-dependent reversal of the electron flow from the Q pool (Cotton et al., 1983). Indeed, no back pressure of photosynthesis on respiration is seen at the level of the cytochrome bc1 complex or CcOs, because continuous illumination has no effect on tetramethyl-p-phenylenediamine (TMPD)-dependent oxygen reduction (Sabaty et al., 1993). A direct interaction between the diverse redox complexes has been shown by several approaches. Induction of a quinol oxidase activity by light with exogenous electron donors such as reduced horse heart cytochrome c or TMPD is observed with both membrane vesicles and hybrid-membrane fragments generated from electron transport mutants of Rba. capsulatus (Zannoni et al., 1978, 1986). This phenomenon might explain why under anaerobic conditions, and in the presence of reduced carbon sources such as succinate or butyrate, photosynthesis is partially inhibited by the reduction of the Q pool (McEwan et al., 1985). Under these strong reducing conditions continuous illumination can reactivate most of the RCs. This phenomenon might result from a fraction of functional RCs that are able to generate a membrane potential sufficient to reoxidize the Q pool by reversing the electron flow from quinol to NAD+ (Herter et al., 1998). Reoxidation of the Q pool is also caused by oxygenation of the bacterial suspension, or by the addition of NO3– or DMSO (McEwan et al., 1985). This type of interaction is particularly important for aerobic photosynthetic bacteria such as Roseobacter, since under anaerobic conditions the primary quinone acceptor, QA, is totally reduced and
no photochemistry occurs (Takamiya et al., 1987; Candela et al., 2001). Surprisingly, it has been shown that membranes from the facultative phototroph Rba. capsulatus, when subjected to continuous illumination, can generate ATP only under oxic conditions. Alternatively, light-induced phosphorylation can be measured under anoxic conditions, and at suitable redox poise, which is close to the Em7.0 of the Q pool (~ +90 mV) (Candela et al., 2001). Whatever the molecular mechanism, it is apparent that the photosynthetic and respiratory chains must interact to favor optimal phosphorylation conditions. Another demonstration of this direct interaction has been provided by Richaud and coworkers (1986) using membrane fragments of Rba. sphaeroides. After a short flash of light, one electron is diverted from the respiratory chain to the photooxidized RC at the level of cytochrome c2. Since two electrons are required to generate a QH2 molecule at the secondary electron acceptor level, QB (gating mechanism; Verméglio, 1977), stimulation of respiration is seen after an even number of flashes of light. Photosynthetic activity inhibits every step of the reduction of nitrate into dinitrogen (Sabaty et al., 1993). This inhibitory effect is not prevented by uncouplers, and therefore does not need the formation of a ∆µH+. It is postulated that this inhibition is due to the diversion of electrons from the denitrification chain to the photosynthetic cycle at the level of the cytochrome bc1 complex and of soluble cytochrome c2, components shared by the two pathways (Richardson et al., 1991; Sabaty et al., 1994). The redox state of the Q pool modulates not only photosynthesis but also the activity of the terminal oxidases of aerobic respiration (Zannoni and Moore, 1990). While electron transport through the cytochrome bc1-CcO pathways is linearly related to the redox state of the Q pool (between 4% and 30% of Q reduction), electron flux through the quinol oxidase pathway is strongly limited until the Q pool is ~25 % reduced. Based on a Q pool formed by approximately 60 Q per RC, it has been estimated that the Km of QH2 at the Qo site of the cytochrome bc1 complex (Cramer and Knaff, 1990) is equivalent to 2.4 to 3 QH2 per RC, while at the Qo site of the QH2-bb3 pathway (quinol oxidase), the Km of QH2 is close to 15 QH2 per RC (Zannoni and Moore, 1990). Thus, the quinol oxidase pathway has a lower affinity for QH2 than the cytochrome bc1-CcO segment of the chain. In conclusion, the organization of the bioenergetic systems in facultative phototrophs satisfies two op-
Chapter 27
Respiratory Complexes
posing needs, namely 1) the photosynthetic apparatus has to interact with the aerobic and/or anaerobic respiratory chains for a better use of the available energy, and 2) this latter interaction must prevent the over-reduction of the photocyclic chain at the level of quinone to avoid inhibition of photosynthesis. In this respect, it is worth noting the recent findings that the membrane associated thiol:disulfide oxidoreductase (DsbB) of Rba. capsulatus may allow oxidation of the membrane embedded quinols in the presence of the water soluble metalloid tellurite (TeO32–) (Borsetti et al., 2007). This finding suggests that metalloids might also support cell growth under unfavorable reducing conditions often experienced by facultative phototrophs (see Chapter 33, Borsetti et al.). Acknowledgments This work was supported by grants from NIH GM 56824 (J.H) and MIUR (PRIN 2005) (D.Z). References Ädelroth P and Brzezinski P (2004) Surface-mediated protontransfer reactions in membrane-bound proteins. Biochim Biophys Acta 1655: 102–115 Ädelroth P and Hosler JP (2006) Surface proton donors for the D-pathway of cytochrome c oxidase in the absence of subunit III. Biochemistry 45: 8308–8318 Ädelroth P, Gennis RB and Brzezinski P (1998) Role of the pathway through K(I-362) in proton transfer in cytochrome c oxidase from R. sphaeroides. Biochemistry 37: 2470–2476 Afkar E, Lisak J, Saltikov C, Basu P, Oremland RS and Stolz JF (2003) The respiratory arsenate reductase from Bacillus selenitireducens strain MLS10. FEMS Microbiol Lett 226: 107–112 Anderson G, Williams J and Hille R (1992) The purification and characterization of arsenite oxidase from Alcaligenes faecalis, a molybdenum-containing hydroxylase. J Biol Chem 267: 23674–23682 Arata H, Shimizu M and Takamiya K (1992) Purification and properties of trimethylamine-N-oxide reductase from the aerobic photosynthetic bacterium Roseobacter denitrificans. J Biochem 112: 470–475 Babcock GT (1999) How oxygen is activated and reduced in respiration. Proc Natl Acad Sci USA 96: 12971–12973 Baymann F, Brugna M, Muhlenhoff U and Nitschke W (2001) Daddy, where did (PS)I come from? Biochim Biophys Acta 1507: 291–310 Baymann F, Lebrun E, Brugna M, Schoepp-Cothenet B, Giudici-Orticoni M-T and Nitschke W (2003) The redox protein construction kit: Pre-last universal common ancestor evolution of energy-conserving enzymes. Phil Trans Roy Soc Lond B 358: 267–274
555 Bender KS, Shang C, Chakraborty R, Belchik SM, Coates JD and Achenbach LA (2005) Identification, characterization, and classification of genes encoding perchlorate reductase. J Bacteriol 187: 5090–5096 Bertero MG, Rothery RA, Palak M, Hou C, Lim D, Blasco F, Weiner JH and Strynadka NCJ (2003) Insights into the respiratory electron transfer pathway from the structure of nitrate reductase A. Nature Struct Biol 10: 681–687 Blankenship RE and Hartman H (1998) The origin and evolution of oxygenic photosynthesis. Trends Biochem Sci 23: 94–97 Blasco G, Guigliarelli B, Magalon A, Asso M, Giordano G and Rothery RA (2001) The coordination and function of the redox centers of the membrane bound nitrate reductases. Cell Molec Life Sci 58: 179–193 Bonora P, Principi I, Hochkoeppler A, Borghese R and Zannoni D (1998) The respiratory chain of the halophilic anoxygenic purple bacterium Rhodospirillum sodomense. Arch Microbiol 170: 435–441 Borisov VB, Liebl U, Rappaport F, Martin JL, Zhang J, Gennis RB, Konstantinov AA and Vos MH (2002) Interactions between heme d and heme b595 in quinol oxidase bd from Escherichia coli: A photoselection study using femtosecond spectroscopy. Biochemistry 41: 1654–1662 Borsetti F, Francia F, Turner RJ and Zannoni D (2007) The thiol: disulfide oxidoreductase DsbB mediates the oxidizing effects of the toxic metalloid tellurite on the plasma membrane redox system of the facultative phototroph Rhodobacter capsulatus. J Bacteriol 189: 851–859 Branden G, Gennis RB and Brzezinski P (2006a) Transmembrane proton translocation by cytochrome c oxidase. Biochim Biophys Acta 1757: 1052–1063 Branden G, Pawate AS, Gennis RB and Brzezinski P (2006b) Controlled uncoupling and recoupling of proton pumping in cytochrome c oxidase. Proc Natl Acad Sci USA 103: 317–322 Brasier MD, Green OR, Jephcoat AP, Kleppe AK, Van Kranendonk MJ, Lindsay JF, Steele A and Grassineau NV (2002) Questioning the evidence for Earth’s oldest fossils. Nature 616: 76–81 Bratton MR, Pressler MA and Hosler JP (1999) Suicide inactivation of cytochrome c oxidase: Catalytic turnover in the absence of subunit III alters the active site. Biochemistry 38: 16236–16245 Bratton MR, Hiser L, Antholine WE, Hoganson C and Hosler JP (2000) Identification of the structural subunits required for formation of the metal centers in subunit I of cytochrome c oxidase of Rhodobacter sphaeroides. Biochemistry 39: 12989–12995 Brochier C and Philippe H (2002) Phylogeny: A non hyperthermophilic ancestor for bacteria. Nature 417:244 Candela M, Zaccherini E and Zannoni D (2001) Respiratory electron transport and light induced energy transduction in membranes from the photosynthetic bacterium Roseobacter denitrificans. Arch Microbiol 175: 169–177 Castresana J (2001) Comparative genomics and bioenergetics. Biochim Biophys Acta 1506: 147–162 Cavalier-Smith T (2001) Obcells as proto-organisms: Membrane heredity, lithophosphorylation, and the origins of the genetic code, the first cell, and photosynthesis. J Mol Evol 53: 555–595 Ciurli S and Musiani F (2005) High potential iron-sulfur proteins and their role as soluble electron carriers in bacterial photosynthesis: Tale of a discovery. Photosynth Res 85: 115–131
556
Davide Zannoni, Barbara Schoepp-Cothenet and Jonathan Hosler
Cotton NJP, Clark AJ and Jackson JB (1983) Interaction between the respiratory and photosynthetic electron transport chains of intact cells of Rhodopseudomonas capsulata mediated by membrane potential. Eur J Biochem 130: 581–587 Cox RL, Patterson C and Donohue TJ (2001) Roles for the Rhodobacter sphaeroides CcmA and CcmG proteins. J Bacteriol 183: 4643–4647 Cramer WA and Knaff DB (1990) Energy Transduction in Biological Membranes. A Textbook of Bioenergetics. SpringerVerlag, New York Daldal F, Mandaci S, Winterstein C, Myllykallio H, Duyck K and Zannoni D (2001) Mobile cytochrome c2 and membraneanchored cytochrome cy are both efficient electron donors to the cbb3- and aa3-type cytochrome c oxidases during respiratory growth of Rhodobacter sphaeroides. J Bacteriol 183: 2013–2024 Danielsson Thorell H, Beyer NH, Heegaard NH, Ohman M and Nilsson T (2004) Comparison of native and recombinant chlorite dismutase from Ideonella dechloratans. Eur J Biochem 271: 3539–3346 D’Mello R, Hill S and Poole RK (1994) Determination of the oxygen affinities of terminal oxidases in Azotobacter vinelandii using the deoxygenation of oxyleghaemoglobin and oxymyoglobin: Cytochrome bd is a low affinity oxidase. Microbiology 140: 1395–1402 D’Mello R, Hill S and Poole RK (1996) The cytochrome bd quinol oxidase in Escherichia coli has an extremely high oxygen affinity and two oxygen-binding haems: Implications for regulation of activity in vivo by oxygen inhibition. Microbiology 142: 755–763 Drosou V, Reincke B, Schneider M and Ludwig B (2002) Specificity of the interaction between the Paracoccus denitrificans oxidase and its substrate cytochrome c: Comparing the mitochondrial to the homologous bacterial cytochrome c552, and its truncated and site-directed mutants. Biochemistry 41: 10629–10634 Dueweke TJ and Gennis RB (1991) Proteolysis of the cytochrome d complex with trypsin and chymotrypsin localizes a quinol oxidase domain. Biochemistry 30: 3401–3406 Dupuis A, Prieur I and Lunardi J (2001) Toward a characterization of the connecting module of complex I. J Bioenerg Biomembr 33: 159–168 Ellis PT, Conrads T, Hille R and Kuhn P (2001) Crystal structure of the 100 kDa arsenite oxidase from Alcaligenes faecalis in two crystal forms at 1.6 Å and 2.03 Å. Structure 9: 125–132 Ferguson SJ and Richardson DJ (2004) The enzymes and bioenergetics of bacterial nitrate, nitrite, nitric oxide and nitrous oxide respiration. In: Zannoni D (ed) Respiration in Archaea and Bacteria. Diversity of Prokaryotic Respiratory Systems (Advances in Photosynthesis and Respiration, Vol 16), pp 169–206 Springer, Dordrecht Fetter JR, Qian J, Shapleigh J, Thomas JW, Garcia-Horsman A, Schmidt E, Hosler J, Babcock GT, Gennis RB and FergusonMiller S (1995) Possible proton relay pathways in cytochrome c oxidase. Proc Natl Acad Sci USA 92: 1604–1608 Finel M (1996) Genetic inactivation of the H+-translocating NADH:ubiquinone oxidoreductase of Paracoccus denitrificans is facilitated by insertion of the ndh gene from Escherichia coli. FEBS Lett 393: 81–85 Friedrich T and Scheide D (2000) The respiratory complex I of bacteria, archaea and eukarya and its module common with
membrane-bound multisubunit hydrogenases. FEBS Lett 479: 1–5 García-Horsman JA, Barquera B and Escamilla JE (1991) Two different aa3-type cytochromes can be purified from the bacterium Bacillus cereus. Eur J Biochem 199: 761–768 García-Horsman JA, Barquera B, Rumbley J, Ma J and Gennis R (1994) The superfamily of heme-copper respiratory oxidases. J Bacteriol. 176: 5587–5600 Gilderson G, Salomonsson L, Aagaard A, Gray J, Brzezinski P and Hosler J (2003) Subunit III of cytochrome c oxidase of Rhodobacter sphaeroides is required to maintain rapid proton uptake through the D pathway at physiologic pH. Biochemistry 42: 7400–7409 Gon S, Giudici-Orticoni MT, Mejean V and Iobbi-Nivol C (2001) Electron transfer and binding of the c-type cytochrome TorC to the trimethylamine-N-oxide reductase of Escherichia coli. J Bacteriol 182: 5779–5786 Hanlon SP, Holt RA, Moore GR and McEwan AG (1994) Isolation and characterization of a strain of Rhodobacter sulfidophilus: A bacterium which grows autotrophically with dimethylsulfide as electron donor. Microbiology 140: 1953–1958 Hemp J, Christian C, Barquera B, Gennis RB and Martinez TJ (2005) Helix switching of a key active-site residue in the cytochrome cbb3 oxidases. Biochemistry 44: 10766–10775 Hemp J, Robinson DE, Ganesan KB, Martinez TJ, Kelleher NL and Gennis RB (2006) Evolutionary migration of a post-translationally modified active-site residue in the proton-pumping hemecopper oxygen reductases. Biochemistry 45: 15405–15410 Herter SM, Kortlüle CM and Drews G (1998) Complex I of Rhodobacter capsulatus and its role in reverted electron transport. Arch Microbiol 169: 98–105 Hill BC (1994) Modeling the sequence of electron transfer reactions in the single turnover of reduced, mammalian cytochrome c oxidase with oxygen. J Biol Chem 269: 2419–2425 Hiser L, Di Valentin M, Hamer AG and Hosler JP (2000) Cox11p is required for stable formation of the CuB and magnesium centers of cytochrome c oxidase. J Biol Chem 275: 619–623 Hochkoeppler A, Ciurli S, Venturoli G and Zannoni D (1995a) The high potential iron-sulphur protein (HiPIP) from Rhodoferax fermentans is competent in photosynthetic electron transfer. FEBS Lett 357:70–74 Hochkoeppler A, Jenney FE, Jr., Lang SE, Zannoni D and Daldal F (1995b) Membrane-associated cytochrome cy of Rhodobacter capsulatus is an electron carrier from the cytochrome bc1 complex to the cytochrome c oxidase during respiration. J Bacteriol 177: 608–613 Hochkoeppler A, Moschettini G and Zannoni D (1995c) The electron transport system of the facultative phototrophic bacterium Rhodoferax fermentans. I. A functional, thermodynamic and spectroscopic study of the membrane bound respiratory chain of dark- and light-grown cells. Biochim Biophys Acta 1229:73–80 Hochkoeppler A, Principi I, Bonora P, Ciurli S and Zannoni D (1999) On the role of soluble redox carriers alternative to cytochrome c2 as donors to tetraheme-type reaction centers and cytochrome oxidases. In: Peschek GA, Loffelhardt W and Schmetterer G (eds) The Phototrophic Prokaryotes, pp 293–302. Kluwer Academic / Plenum Publishers, New York Holland HD (1994) Early proterozoic atmospheric change. In: Bengtson S (ed) Early Life on Earth, pp 237–244. Columbia University Press, New York
Chapter 27
Respiratory Complexes
Hosler JP (2004) The influence of subunit III of cytochrome c oxidase on the D pathway, the proton exit pathway and mechanism-based inactivation in subunit I. Biochim Biophys Acta 1655: 332–339 Hosler JP, Ferguson-Miller S, Calhoun MW, Thomas JW, Hill J, Lemieux L, Ma J, Georgiou C, Fetter J, Shapleigh J, Teklenburg MMJ, Babcock GT and Gennis RB (1993) Insight into the active-site structure and function of cytochrome oxidase by analysis of site-directed mutants of bacterial cytochrome aa3 and cytochrome bo. J Bioenerg Biomembr 25: 121–136 Hosler JP, Ferguson-Miller S and Mills DA (2006) Energy transduction: Proton transfer through the respiratory complexes. Annu Rev Biochem 75: 165–187 Imhoff JF, Petri R and Suling J (1998) Reclassification of species of the spiral-shaped phototrophic purple non-sufur bacteria of the alpha-Proteobacteria: Description of the new genera Paeospirillum gen. nov., Rhodovibrio gen. nov., Rhodothalassium gen. nov. and Roseospira gen. nov. as well as transfer of Rhodospirillum fulvum to Phaeospirillum fulvum comb. nov., of Rhodospirillum molischianum to Paeospirillum molischianum com. nov., of Rhodospirillum salinarum to Rhodovibrio salexigens. Int J Syst Bacteriol 48: 793–798 Iobbi-Nivol C, Pommier J, Simala-Grant J, Mejean V and Giordano G (1996) High substrate specificity and induction characteristics of trimethylamine-N-oxide reductase of Escherichia coli. Biochim Biophys Acta 1294: 157–162 Jenney Jr. FE and Daldal F (1993) A novel membrane-associated c-type cytochrome Cyt cy, can mediate the photosynthetic growth of Rhodobacter capsuatus and Rhodobacter sphaeroides. EMBO J 12: 1283–1292 Johnson KE and Rajagopalan KV (2001) An active site tyrosine influences the ability of the dimethylsulfoxide reductase family of molybdopterin enzymes to reduce S-oxides. J Biol Chem 276: 13178–13185 Jormakka M, Richardson D, Byrne B and Iwata S (2004) Architecture of NarGH reveals a structural classification of Mo-bisMGD enzymes. Structure 12: 95–104 Junemann S (1997) Cytochrome bd terminal oxidase. Biochim Biophys Acta 1321: 107–127 Jungst A, Wakabayashi S, Matsubara H and Zunft WG (1991) The nirSTBM region coding for cytochrome cd1-dependent nitrite respiration in Pseudomonas stutzeri consists of a cluster of mono-, di-, and tetraheme proteins. FEBS Lett 279: 205–209 Kelly DP and Smith NA (1990) Organic sulphur compounds in the environment-biogeochemistry, microbiology, and ecological aspects. Adv Microb Ecol 11: 345–385 Kloer DP, Hagel C, Heider J and Schulz GE (2006) Crystal structure of ethylbenzene dehydrogenase from Aromatoleum aromaticum. Structure 14: 1377–1388 Kniemeyer O and Heider J (2001) Ethylbenzene dehydrogenase, a novel hydrocarbon-oxidizing molybdenum/iron-sulfur/heme enzyme. J Biol Chem 276: 21381–21386 Koch HG, Hwang O and Daldal F (1998) Isolation and characterization of Rhodobacter capsulatus mutants affected in cytochrome cbb3 oxidase activity. J Bacteriol 180: 969–978 Krafft T and Macy JM (1998) Purification and characterization of the respiratory arsenate reductase of Chrysiogenes arsenatis. Eur J Biochem 255: 647–653 Krafft T, Bokranz M, Klimmer O, Schröder I, Fahrenholz F, Kojro E and Kröger A (1992) Cloning and nucleodide sequence of the
557 psrA gene of Wolinella succinogenes polysulphide reductase. Eur J Biochem 206: 503–510 Lebrun E, Brugna M, Baymann F, Muller D, Lievremont D, Lett M-C and Nitschke W (2003) Arsenite oxidase, an ancient bioenergetic enzyme. Mol Biol Evol 20: 686–693 Lebrun E, Santini JM, Brugna M, Ducluzeau AL, Ouchane S, Schoepp-Cothenet B, Baymann F and Nitschke W (2006) The Rieske-protein: A case study on the pitfalls of multiple sequence alignments and phylogenetic reconstruction. Mol Biol Evol 23: 1180–1191 Lemon DD, Calhoun MW, Gennis RB and Woodruff WH (1993) The gateway to the active site of heme-copper oxidases. Biochemistry 32: 11953–11956 Madigan MT, Martinko JM and Parker J (1997) Brock. Biology of Microorganisms. 10th Edition. Printice Hall, Upper Saddle River Mackenzie C, Choudhary M, Larimer FW, Predki PF, Stilwagen S, Armitage JP, Barber RD, Donohue TJ, Hosler JP, Newman JE, Shapleigh JP, Sockett RE, Zeilstra-Ryalls J and Kaplan S (2001) The home stretch, a first analysis of the nearly completed genome of Rhodobacter sphaeroides 2.4.1. Photosynth Res 70: 19–41 McDevitt CA, Hugenholtz P, Hanson GR and McEwan AG (2002a) Molecular analysis of dimethylsulfide dehydrogenase from Rhodovulum sulfidophilum; its place in the DMSO reductase family of microbial molybdenum-containing enzymes. Mol Microbiol 44: 1576–1587 McDevitt CA, Hanson GR, Noble CJ, Cheesman MR and McEwan AG (2002b) Characterization of the redox centers in dimethyl sulfide dehydrogenase from Rhodovulum sulfidophilum. Biochemistry 41: 15234–15244 McEwan AG, Cotton NPJ, Ferguson SJ and Jackson JB (1985) The role of auxillary oxidants in the maintenance of balanced redox poise for photosynthesis in bacteria. Biochim Biophys Acta 810: 140–147 McEwan AG, Ridge JP, McDevitt CA and Hugenholtz P (2002) The DMSO reductase family of microbial molybdenum enzymes; molecular properties and role in the dissimilatory reduction of toxic elements. Geomicrobiol J 19: 3–21 McEwan AG, Kappler U and McDevitt CA (2004) Microbial molybdenum-containing enzymes in respiration: structural and functional aspects. In: Zannoni D (ed) Respiration in Archaea and Bacteria. Diversity of Prokaryotic Electron Transport Carriers (Advances in Photosynthesis and Respiration, Vol 15), pp 175–202. Kluwer Academic Publishers, Dordrecht Meyer TE and Cusanovich MA (1985) Soluble cytochrome composition of the purple phototrophic bacterium Rhodopseudomas sphaeroides ATCC 17023. Biochim Biophys Acta 807: 308–391 Meyer TE and Donohue TJ (1995) Cytochromes, iron-sulfur, and copper proteins mediating electron transfer from Cyt bc1 complex to photosynthetic reaction center complexes. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 725–745. Kluwer Academic Publishers, Dordrecht Mills DA, Schmidt B, Hiser C, Westley E and Ferguson-Miller S (2002) Membrane potential-controlled inhibition of cytochrome c oxidase by zinc. J Biol Chem. 277: 14894–14901 Mills DA and Hosler JP (2005) Slow proton transfer through
558
Davide Zannoni, Barbara Schoepp-Cothenet and Jonathan Hosler
the pathways for pumped protons in cytochrome c oxidase induces suicide inactivation of the enzyme. Biochemistry 44: 4656–4666 Mills DA, Tan Z, Ferguson-Miller S and Hosler J (2003) A role for subunit III in proton uptake into the D pathway and a possible proton exit pathway in Rhodobacter sphaeroides cytochrome c oxidase. Biochemistry 42: 7410–7417 Mills DA, Geren L, Hiser C, Schmidt B, Durham B, Millett F and Ferguson-Miller S (2005) An arginine to lysine mutation in the vicinity of the heme propionates affects the redox potentials of the hemes and associated electron and proton transfer in cytochrome c oxidase. Biochemistry 44: 10457–10465 Mogi T, Akimoto S, Endou S, Watanabe-Nakayama T, MizuochiAsai E and Miyoshi H (2006) Probing the ubiquinol-binding site in cytochrome bd by site-directed mutagenesis. Biochemistry 45: 7924–7930 Moore MD and Kaplan S (1992) Identification of intrinsic highlevel resistance to rare-earth oxides and oxyanions in members of the class Proteobacteria: Characterization of tellurite, selenite, and rhodium sesquioxide reduction in Rhodobacter sphaeroides. J Bacteriol 174: 1505–1514 Morgan JE, Verkhovsky MI, Palmer G and Wikström M (2001) Role of the Pr intermediate in the reaction of cytochrome c oxidase with O2. Biochemistry 40: 6882–6892 Moschettini G, Bonora P, Zaccherini E, Hochkoeppler A, Principi I and Zannoni D (1999) The primary quinone acceptor and the membrane-bound c-type cytochromes of the halophilic purple nonsulfur bacterium Rhodospirillum salinarum: a spectroscopic and thermodynamic study. Photosynth Research 62:43–53 Mouncey NJ, Gak E, Choudhary M, Oh J and Kaplan S (2000) Respiratory pathways of Rhodobacter sphaeroides 2.4.1T: Identification and characterization of genes encoding quinol oxidases. FEMS Microbiol Lett 192: 205–210 Mukhopadhyay R, Rosen BP, Phung LT and Silver S (2002) Microbial arsenic: From geocycles to genes and enzymes. FEMS Microbiol Rev 26: 311–325 Muller D, Lièvremont D, Dancheva Simeonova D, Hubert J-C and Lett M-C (2003) Arsenite oxidase aox genes from a metalresistant !-proteobacterium. J Bacteriol 185: 135–141 Namslauer A and Brzezinski P (2004) Structural elements involved in electron-coupled proton transfer in cytochrome c oxidase. FEBS Lett. 567: 103–110 Nitschke W, Muhlenhoff U and Liebl U (1997) Evolution. In: Raghavendra A (ed) Photosynthesis: A Comprehensive Treatise, pp 285–304. Cambridge University Press, Cambridge Oh JI (2006) Effect of mutations of five conserved histidine residues in the catalytic subunit of the cbb3 cytochrome c oxidase on its function. J Microbiol 44: 284–292 Oh JI and Kaplan S (1999) The cbb3 terminal oxidase of Rhodobacter sphaeroides 2.4.1: Structural and functional implications for the regulation of spectral complex formation. Biochemistry 38: 2688–2696 Oh JI and Kaplan S (2002) Oxygen adaptation. The role of the CcoQ subunit of the cbb3 cytochrome c oxidase of Rhodobacter sphaeroides 2.4.1. J Biol Chem 277: 16220–16228 Oh JI, Ko IJ and Kaplan S (2004) Reconstitution of the Rhodobacter sphaeroides cbb3-PrrBA signal transduction pathway in vitro. Biochemistry 43: 7915–7923 Ohmoto H (1996) Evidence in pre-2.2 Ga paleosols for the early evolution of atmospheric oxygen and terrestrial biota. Geology 24: 1135–1138 Ohmoto H (1997) When did the Earth’s atmosphere become oxic?
Geochemical News 93: 12–29 Olson JM (1999) Early evolution in chlorophyll-based photosynthesis. Chemtracts 12: 468–482 Oremland RS and Stolz JF (2003) The ecology of Arsenic. Science 300: 939–944 Oremland RS and Stolz JF (2005) Arsenic, microbes and contaminated aquifers. Trends in Microbiology 13: 45–49 Oremland RS, Blum JS, Culbertson CW, Visscher PT, Miller LG, Dowdle P and Strohmaier FE (1994) Isolation, growth, and metabolism of an obligately anaerobic, selenate-respiring bacterium, strain SES-3. Appl Environ Microbiol 60: 3011–3019 Oremland RS, Hoeft SE, Santini JM, Bano N, Hollibaugh RA and Hollibaugh RT (2002) Anaerobic oxidation of arsenite in Mono Lake water and by a facultative, arsenite-oxidizing chemoautotroph, strain MLHE-1. Appl Environ Microbiol 68: 4795–4802 Pappas CT, Sram J, Moskvin OV, Ivanov PS, Mackenzie RC, Choudhary M, Land ML, Larimer FW, Kaplan S and Gomelsky M (2004) Construction and validation of the Rhodobacter sphaeroides 2.4.1 DNA microarray: Transcriptome flexibility at diverse growth modes. J Bacteriol 186: 4748–4758 Pawate AS, Morgan J, Namslauer A, Mills D, Brzezinski P, Ferguson-Miller S and Gennis RB (2002) A mutation in subunit I of cytochrome oxidase from Rhodobacter sphaeroides results in an increase in steady-state activity but completely eliminates proton pumping. Biochemistry 41: 13417–13423 Pils D and Schmetterer G (2001) Characterization of three bioenergetically active respiratory terminal oxidases in the cyanobacterium Synechocystis sp. Strain PCC 6803. FEMS Microbiol Lett 203: 217–222 Pitcher RS and Watmough NJ (2004) The bacterial cytochrome cbb3 oxidases. Biochim Biophys Acta 1655: 388–399 Poole RK, Lloyd D and Chance B (1979) The reaction of cytochrome oxidase with oxygen in the fission yeast Schizosaccharomyces pombe 972H-. Studies at subzero temperatures and measurement of apparent oxygen affinity. Biochem J 184: 555–563 Preisig O, Zufferey R, Thoeny-Meyer L, Appleby CA and Hennecke H (1996) A high-affinity cbb3-type cytochrome oxidase terminates the symbiosis-specific respiratory chain of Bradyrhizobium japonicum. J Bacteriol 178: 1532–1538 Proshlyakov DA (2004) UV optical absorption by protein radicals in cytochrome c oxidase. Biochim Biophys Acta 1655: 282–289 Proshlyakov DA, Pressler MA and Babcock GT (1998) Dioxygen activation and bond cleavage by mixed-valence cytochrome c oxidase. Proc Natl Acad Sci USA 95: 8020–8025 Proshlyakov DA, Pressler MA, DeMaso C, Leykam JF, DeWitt DL and Babcock GT (2000) Oxygen activation and reduction in respiration: Involvement of redox-active tyrosine 244. Science 290: 1588–1591 Qin L, Hiser C, Mulichak A, Garavito RM and FergusonMiller S (2006) Identification of conserved lipid/detergentbinding sites in a high-resolution structure of the membrane protein cytochrome c oxidase. Proc Natl Acad Sci USA 103: 16117–16122 Rauhamaki V, Baumann M, Soliymani R, Puustinen A and Wikström M (2006) Identification of a histidine-tyrosine cross-link in the active site of the cbb3-type cytochrome c oxidase from Rhodobacter sphaeroides. Proc Natl Acad Sci USA 103: 16135–16140 Reysenbach AL and Shock E (2002) Merging genomes with
Chapter 27
Respiratory Complexes
geochemistry in hydrothermal ecosystems. Science 296: 1077–1082 Richardson DJ (2000) Bacterial respiration: a flexible process for a changing environment. Microbiology 146: 551–571 Richardson DJ, Bell LC, McEwan AG, Jackson JB and Ferguson SJ (1991) Cytochrome c2 is essential for electron transfer to nitrous oxide reductase from physiological substrates in Rhodobacter capsulatus and acts as electron donor to the reductase in vivo. Correlation with photoinhibition studies. Eur J Biochem 199: 677–683 Richaud P, Marrs BL and Verméglio A (1986) Two modes of interaction between photosynthetic and respiratory electron chains in whole cells of Rhodopseudomonas capsulata. Biochim Biophys Acta 850: 256–263 Richter OM and Ludwig B (2003) Cytochrome c oxidase–structure, function, and physiology of a redox-driven molecular machine. Rev Physiol Biochem Pharmacol 147: 47–74 Riistama S, Puustinen A, Verkhovsky MI, Morgan JE and Wikström M (2000) Binding of O2 and its reduction are both retarded by replacement of valine 279 by isoleucine in cytochrome c oxidase from Paracoccus denitrificans. Biochemistry 39: 6365–6372 Rott MA, Witthuhn VC, Schike BA, Soranno M, Ali A and Donohue TJ (1993) Genetic evidence for the role of isocytochrome c2 in photosynthetic growth of Rhodobacter sphaeroides spd mutants. J Bacteriol 175: 358–366 Rugolo M and Zannoni D (1983) Oxygen induced inhibition of light dependent uptake of tetraphenylphosphonium ions as a probe of a direct interaction between photosynthetic and respiratory components in cells of Rhodopseudomonas capsulata. Biochem Biophys Res Commun 113:155–162 Ruitenberg M, Kannt A, Bamberg E, Ludwig B, Michel H and Fendler K (2000) Single-electron reduction of the oxidized state is coupled to proton uptake via the K pathway in Paracoccus denitrificans cytochrome c oxidase. Proc Natl Acad Sci USA 97: 4632–4636 Sabaty M, Gans P and Verméglio A (1993) Inhibition of nitrate reduction by light and oxygen in Rhodobacter sphaeroides forma sp. denitrificans. Arch Microbiol 159: 153–159 Sabaty M, Jappé J, Olive J and Verméglio A (1994) Organization of electron transfer components in Rhodobacter sphaeroides forma sp. denitrificans. Biochim Biophys Acta 1187: 313–323 Sabaty M, Avazéri C, Pignol D and Verméglio A (2001) Characterization of the reduction of selenate and tellurite by nitrate reductases. Appl Environ Microbiol 67: 5122–5126 Saitou N and Nei M (1987) The neighbor-joining method: A new method for reconstructing phylogenetic trees. Mol Biol Evol 4: 406–425 Santini JM and vanden Hoven RN (2004) Molybdenum-containing arsenite oxidase of the chemolithoautotrophic arsenite oxidizer NT-26. J Bacteriol 186: 1614–1619 Santini JM, Sly LI, Schnagl RD and Macy JM (2000) A new chemolithoautotropic arsenite-oxidizing bacterium isolated from a gold mine: Phylogenetic, physiological, and preliminary biochemical studies. App Env Microbiol 66: 92–97 Santini JM, Kappler U, Ward SA, Honeychurch MJ, van den Hoven RN and Bernhard PV (2007) The NT-26 cytochrome c552 and its role in arsenite oxidation. Biochim Biophys Acta 1767: 189–196 Sazanov LA and Hinchliffe P (2006) Structure of the hydrophilic domain of respiratory complex I from Thermus thermophilus.
559 Science 311: 1430–1436 Shapleigh J, Hosler JP, Tecklenburg MJ, Ferguson-Miller S, Babcock GT and Gennis RB (1992) Identification of the heme axial ligands for the cytochrome a component of cytochrome c oxidase. Proc Natl Acad Sci USA 89: 4786–4790 Sharma V, Puustinen A, Wikström M and Laakkonen L (2006) Sequence analysis of the cbb3 oxidases and an atomic model for the Rhodobacter sphaeroides enzyme. Biochemistry 45: 5754–5765 Sharpe M, Qin L and Ferguson-Miller S (2005) Proton entry, exit and pathways in cytochrome oxidase: Insight from ‘conserved’ water. In: Wikström M (ed) Biophysical and Structural Aspects of Bioenergetics, pp 26–54. RSC Publishing, Cambridge Shaw AL, Hanson GR and McEwan AG (1996) Cloning and sequence analysis of the dimethylsulfoxide reductase structural gene from Rhodobacter capsulatus. Biochim Biophys Acta 1276: 176–180 Shaw AL, Hochkoeppler A, Bonora P, Zannoni D, Hanson GR and McEwan AG (1999a) Characterization of DorC from Rhodobacter capsulatus, a c-type cytochrome involved in electron transfer to dimethylsulfoxide reductase. J Biol Chem 274: 9911–9914 Shaw AL, Leimkuehler S, Klipp W, Hanson GR and McEwan AG (1999b) Mutational analysis of the dimethylsulfoxide respiratory (dor) operon of Rhodobacter capsulatus. Microbiol 145: 1409–1420 Schopf JW, Kudryavtsev AB, Agresti DG, Wdowiak TJ and Czaja AD (2002) Laser-Raman imagery of Earth’s earliest fossils. Nature 416: 73–76 Schröder I, Rech S, Krafft T and Macy JM (1997) Purification and characterization of the selenate reductase from Thauera selenatis. J Biol Chem 272: 23765–23768 Smith D, Gray J, Mitchell L, Antholine WE and Hosler JP (2005) Assembly of cytochrome c oxidase in the absence of assembly protein Surf1p leads to loss of the active site heme. J Biol Chem 280: 17652–17656 Stiburek L, Vesela K, Hansikova H, Pecina P, Tesarova M, Cerna L, Houstek J and Zeman J (2005) Tissue-specific cytochrome c oxidase assembly defects due to mutations in Sco2 and Surf1. Biochem J 392: 625–632 Stolz JF and Oremland RS (1999) Bacterial respiration of arsenic and selenium. FEMS Microbiol Rev 23: 615–627 Stolz JF, Basu P, Santini JM and Oremland RS (2006) Arsenic and selenium in microbial metabolism. Annu Rev Microbiol 60: 107–130 Svensson-Ek M, Abramson J, Larsson G, Tornroth S, Brzezinski P and Iwata S (2002) The X-ray crystal structures of wild-type and EQ(I-286) mutant cytochrome c oxidases from Rhodobacter sphaeroides. J Mol Biol 321: 329–339 Swem LR, Gong X, Yu CA and Bauer CE (2006) Identification of a ubiquinone-binding site that affects autophosphorylation of the sensor kinase RegB. J Biol Chem 281: 6768–6775 Switzer Blum J, Burns Bindi A, Buzzeli J, Stolz JF and Oremland RS (1998) Bacillus arsenicoselenatis, sp. nov., and Bacillus selenitireducens, sp. nov.: Two haloalkaliphiles from Mono Lake, California that respire oxyanions of selenium and arsenic. Arch Microbiol 171: 19–30 Takamiya K, Iba K and Okamura K (1987) Reaction center complex from an aerobic photosynthetic bacterium Erythrobacter species OCh114. Biochim Biophys Acta 890: 127—133 Taylor BF and Kiene RP (1989) Microbial-metabolism of dimeth-
560
Davide Zannoni, Barbara Schoepp-Cothenet and Jonathan Hosler
ylsulfide. ACS Symposium Series 393: 202–221 Thöny-Mayer L (1997) Biogenesis of respiratory cytochromes in bacteria. Microbiol Mol Biol Rev 61: 337–376 Toledo-Cuevas M, Barquera B, Gennis RB, Wikström M and García-Horsman JA (1998) The cbb3-type cytochrome c oxidase from Rhodobacter sphaeroides, a proton-pumping heme-copper oxidase. Biochim Biophys Acta 1365: 421–434 Tomson FL, Morgan JE, Gu G, Barquera B, Vygodina TV and Gennis RB (2003) Substitutions for glutamate 101 in subunit II of cytochrome c oxidase from Rhodobacter sphaeroides result in blocking the proton-conducting K-channel. Biochemistry 42: 1711–1717 Towe KM (1990) Aerobic respiration in the Archaean? Nature 348: 54–56 Towe KM (1994) Earth’s early atmosphere: Constraints and opportunities for early evolution. In: Bengston S (ed) Early Life on Earth, pp. 36–47. Columbia University Press, New York Towe KM (1996) Environmental oxygen conditions during the origina and early evolution of life. Adv Space Res 18: 7–15 Trieber CA, Rothery RA and Weiner JH (1996) Engineering a novel iron-sulfur cluster into the catalytic subunit of Escherichia coli dimethyl-sulfoxide reductase. J Biol Chem 271: 4620–4626 Unemoto T and Hayashi M (1993) Na+-translocating NADH-quinone reductase of marine and halophilic bacteria. J Bioenerg Biomembr 25: 385–391 vanden Hoven RN and Santini JM (2004) Arsenite oxidation by the heterotroph Hydrogenophaga sp. str. NT-14: the arsenite oxidase and its physiological electron acceptor. Biochim Biophys Acta 1656: 148–155 Varanasi L, Mills D, Murphree A, Gray J, Purser C, Baker R and Hosler J (2006) Altering conserved lipid binding sites in cytochrome c oxidase of Rhodobacter sphaeroides perturbs the interaction between subunits I and III and promotes suicide inactivation of the enzyme. Biochemistry 45: 14896–14907 Verméglio A (1977) Secondary electron transfer in reaction centers of Rhodopseudomonas sphaeroides: out-of-phase periodicity of two for the formation of ubisemiquinone and fully reduced ubiquinone. Biochim Biophys Acta 459: 516–524 Visscher PT, Taylor BF and Kiene RP (1995) Microbial consumption of dimethyl sulfide and methanediol in coastal marinesediments. FEMS Microbiol Ecol 18: 145–153 Visscher PT and Taylor BF (1993) Organic thiols as organolithotrophic substrates for growth of phototrophic bacteria. Appl Environ Microbiol 59: 93–96 Wang K, Zhen Y, Sadoski R, Grinnell S, Geren L, Ferguson-Miller S, Durham B and Millett F (1999) Definition of the interaction domain for cytochrome c on cytochrome c oxidase. I. Rapid kinetic analysis of electron transfer from cytochrome c to Rhodobacter sphaeroides cytochrome oxidase surface mutants. J Biol Chem 274: 38042–38050 Watts CA, Ridley H, Condie KL, Leaver JT, Richardson DJ and Butler CS (2003) Selenate reduction by Enterobacter cloacae SDL1a-1 is catalysed by a molybdenum-dependent membranebound enzyme that is distinct from the membrane-bound nitrate reductase. FEMS Microbiol Lett 228: 273–279 Watts CA, Ridley H, Dridge EJ, Leaver JT, Reilly AJ, Richardson DJ and Butler CS (2005) Microbial reduction of selenate and nitrate: common themes and variations. Biochem Soc Trans 33: 173–175 Wikström M (2004) Cytochrome c oxidase: 25 years of the elusive
proton pump. Biochim Biophys Acta 1655: 241–247 Wood PM (1981) The redox potential of dimethylsulfoxide reduction to dimethylsulfide-evaluation and biochemical implications. FEBS Lett 124: 11–14 Wraight CA, Cogdell RJ and Chance B (1978) Ion transport and electrochemical gradients in photosynthetic bacteria. In: Clayton RK and Sistrom RW (eds) The Photosynthetic Bacteria, pp 471–502. Plenum Press, New York Xiong J and Bauer CE (2002) Complex evolution of photosynthesis. Annu Rev Plant Biol 53: 503—521 Xu X and Yagi T (2001) Identification of the NADH binding subunit of energy transducing NADH-quinone oxidoreductase (NDH-1) of Thermus thermophilus HB-8. Biochem Biophys Res Commun 174: 667–672 Xu X, Matsuno-Yagi A and Yagi T (1991a) Characterization of the 25-kilodalton subunit of the energy transducing NADH-ubiquinone oxidoreductase of Paracoccus denitrificans: Sequence similarity to the 24-kilodalton subunit of the flavoprotein fraction of mammalian complex I. Biochemistry 30: 8678–8684 Xu X, Matsuno-Yagi A and Yagi T (1991b) The NADH-binding subunit of the energy transducing NADH-ubiquinone oxidoreductase of Paracoccus denitrificans: Gene cloning and deduced primary structure. Biochemistry 30: 6422–6428 Xu X, Matsuno-Yagi A and Yagi T (1992a) Gene cluster of the energy-transducing NADH-quinone oxidoreductase of Paracoccus denitrificans: Characterization of four structural gene products. Biochemistry 31: 6925–6932 Xu X, Matsuno-Yagi A and Yagi T (1992b) Structural features of the 66 kilodalton subunit of the energy transducing NADH-ubiquinone oxidoreductase (NDH-1) of Paracoccus denitrificans. Arch Biochem Biophys 296: 40–48 Xu X, Matsuno-Yagi A and Yagi T (1993) DNA sequencing of the seven remaining structural genes of the gene cluster encoding the energy-transducing NADH-quinone oxidoreductase of Paracoccus denitrificans. Arch Biochem Biophys 250: 302–311 Yagi T (1993) The bacterial energy-transducing NADH-quinone oxidoreductases. Biochim Biophys Acta 1141: 1–17 Yagi T and Matsuno-Yagi A (2003) The proton-translocating NADH-quinone oxidoreductase in the respiratory chain: The secret unlocked. Biochemistry 42: 2266–2274 Yagi T, Seo BB, Di Bernardo S, Nakamaru-Ogiso E, Kao MC and Matsuno-Yagi A (2001) NADH dehydrogenases: From basic science to biomedicine. J Bioenerg Biomembr 33: 233–242 Yoshikawa S, Shinzawa-Itoh K, Nakashima R, Yaono R, Yamashita E, Inoue N, Yao M, Fei MJ, Libeu CP, Mizushima T, Yamaguchi H, Tomizaki T and Tsukihara T (1998) Redox-coupled crystal structural changes in bovine heart cytochrome c oxidase. Science 280: 1723–1729 Yun CH, Beci R, Crofts AR, Kaplan S and Gennis RB (1990) Cloning and DNA sequencing of the fbc operon encoding the cytochrome bc1 complex from Rhodobacter sphaeroides. Characterization of fbc deletion mutants and complementation by a site-specific mutational variant. Eur J Biochem 194: 399–411 Zannoni D (1995) Aerobic and anaerobic electron transport chains in anoxygenic phototrophic bacteria. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 449–971. Kluwer Academic Publishers, Dordrecht Zannoni D and Ingledew JW (1985) A thermodynamic analysis of the plasma membrane electron transport components in
Chapter 27
Respiratory Complexes
phototrophically grown cells of Chloroflexus aurantiacus: An optical and electron paramagnetic resonance study. FEBS Letters 193: 93–98 Zannoni D and Moore AL (1990) Measurement of the redox state of the ubiquinone pool in Rhodobacter capsulatus membrane fragments. FEBS Letters 271:123–127 Zannoni D, Jasper P and Marrs BL (1978) Light induced oxygen uptake as a probe of electron transport between respiratory and photosynthetic components in membranes of Rhodopseudomonas capsulata Arch Biochem Biophys 191: 625–631 Zannoni D, Peterson S and Marrs BL (1986) Recovery of the alternative oxidase dependent electron flow by fusion of membrane vesicles from Rhodobacter capsulatus mutant strains.
561 Arch Microbiol 144: 375–380 Zeller T, Moskvin OV, Li K, Klug G and Gomelsky M (2005) Transcriptome and physiological responses to hydrogen peroxide of the facultatively phototrophic bacterium Rhodobacter sphaeroides. J Bacteriol 187: 7232–7242 Zhang J, Barquera B and Gennis RB (2004) Gene fusions with beta-lactamase show that subunit I of the cytochrome bd quinol oxidase from E. coli has nine transmembrane helices with the O2 reactive site near the periplasmic surface. FEBS Lett 561: 58–62 Zufferey R, Preisig O, Hennecke H and Thöny-Meyer L (1996) Assembly and function of the cytochrome cbb3 oxidase subunits in Bradyrhizobium japonicum. J Biol Chem 271: 9114–9119
Chapter 28 Carbon Dioxide Metabolism and its Regulation in Nonsulfur Purple Photosynthetic Bacteria Simona Romagnoli† and F. Robert Tabita* Department of Microbiology, Ohio State University, 484 West 12th Avenue, Columbus, OH 43210-1292 U.S.A.
Summary .............................................................................................................................................................. 563 I. Introduction..................................................................................................................................................... 564 II. Regulation of cbb Gene Expression .............................................................................................................. 564 A. Rhodobacter sphaeroides .............................................................................................................. 566 B. Rhodobacter capsulatus ................................................................................................................. 569 1. The CbbR Proteins and their Co-inducers ............................................................................. 570 C. Rhodopseudomonas palustris ........................................................................................................ 572 Acknowledgments ................................................................................................................................................. 575 References ............................................................................................................................................................ 575
Summary Nonsulfur purple (NSP) photosynthetic bacteria are able to photosynthetically metabolize and grow at the expense of a wide range of reduced and oxidized organic compounds as well as using inorganic carbon dioxide under anaerobic conditions. In addition, these organisms, for the most part, are able to oxidize these compounds under aerobic conditions as well. Thus, NSP bacteria are widely dispersed and survive in diverse environments. Carbon dioxide reduction is important for both photoheterotrophic and photoautotrophic metabolism, and CO2 serves as an essential electron acceptor for the maintenance of cellular redox homeostasis when highly reduced organic carbon is used as the electron donor for growth. Since earlier reviews, much progress towards elucidating molecular mechanisms governing carbon dioxide assimilation has been made, primarily in two representative species, Rhodobacter (Rba.) sphaeroides and Rba. capsulatus. These studies established the importance of the main transcriptional regulator, CbbR, and its interaction with specific promoter sequences. Furthermore, the redox-sensitive two-component global regulator system, RegAB/PrrAB, was shown to be important for controlling cbb operon gene expression and appeared to be linked to the regulation of redox balancing mechanisms. A further level of complexity is found in Rhodopseudomonas (Rps.) palustris, where a unique three-protein two-component system, in addition to CbbR, was recently shown to contribute to the regulation of CO2 fixation. This chapter will focus on recent progress made in understanding the mechanism of regulation of CO2 metabolism in NSP bacteria.
*Author for correspondence, email: [email protected] † Present address: Department of Biology, University of Bologna, Via Selmi 3, 40126 Bologna, Italy
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 563–576. © 2009 Springer Science + Business Media B.V.
564 I. Introduction A prominent characteristic of the nonsulfur purple (NSP) bacteria is their metabolic versatility, as these organisms are unique in their ability to employ all the known modes of metabolism; i.e., they are able to grow photolithoautotrophically, photoheterotrophically, chemiolithoautotrophically and chemoheterotrophically. NSP bacteria are also able to fix nitrogen, evolve hydrogen, and utilize oxygen or diverse alternative electron acceptors for respiratory purposes, and these organisms are able to use a large variety of organic and inorganic electron donors. In addition to obvious functions of supplying needed carbon, CO2 assimilation and organic carbon oxidation play a pivotal role in maintaining the redox balance of the organism. With respect to molecular mechanisms governing CO2 assimilation, studies have primarily focused on Rba. sphaeroides and Rba. capsulatus. Although these species are closely related, they differ in the organization and regulation of the cbb operons: Rba. sphaeroides contains two copies of most of the important structural genes (Chen et al., 1991; Gibson et al., 1991), whereas Rba. capsulatus possesses two structurally distinct ribulose 1,5-bisphosphate carboxylase/oxygenase (Rubisco) genes but single copies of genes that encode other enzymes of the Calvin-Benson-Bassham (CBB) pathway (Paoli et al., 1995). A single cbbR gene, divergently transcribed from the cbbI operon, regulates transcription of both cbb operons in Rba. sphaeroides, although to a different extent (Gibson and Tabita, 1993; Dubbs and Tabita, 1998). On the other hand, in Rba. capsulatus there are two cbbR genes, each divergently transcribed from and regulating its cognate cbb operon (Paoli et al., 1998b; Vichivanives et al., 2000; Dubbs and Tabita, 2004) (Fig. 1). CbbR is a positive transcriptional regulator of the LysR-type family (Schell, 1993) and typically binds to a T-N11-A consensus region (Goethals et al., 1992), actively regulating the downstream gene(s). Another trait in common to Rba. sphaeroides and Rba. capsulatus is the role of the global RegAB/PrrAB regulator system (Joshi and Tabita, 1996) in Abbreviations: bp – base pair(s); CBB – Calvin-BensonBassham; CbbRRS – Calvin Benson Bassham response regulators/sensor kinase; LTTR – LysR-type transcriptional regulator; NSP – nonsulfur purple; PAS – Per-Arnt-Sim sensing module; Rba. – Rhodobacter; Rps. – Rhodopseudomonas; Rubisco – ribulose 1,5-bisphospate carboxylase/oxygenase; RuBP – ribulose 1,5-bisphosphate
Simona Romagnoli and F. Robert Tabita modulating the expression of the cbb operons (Dubbs et al., 2000; Vichivanives et al., 2000; Dubbs and Tabita, 2003, 2004). Surprisingly, the Reg system is not involved in controlling cbb operon expression in Rps. palustris (Romagnoli and Tabita, unpublished). As discussed later, in Rps. palustris a unique dedicated system appears to be responsible for redox regulation of the cbbI operon (Romagnoli and Tabita, 2006). The metabolism of reduced and oxidized organic compounds and the subsequent physiological consequences on the control of CO2 assimilation were extensively reviewed in the previous edition of this book (Tabita, 1995), as were the main biochemical properties of the individual enzymes of the CBB cycle. This chapter will focus on mechanisms governing cbb operon expression and regulation, and the necessity for maintaining cellular redox poise. A detailed update on the structure-function relationships of Rubisco and the interesting and newly discovered Rubisco-like proteins will be presented elsewhere (Tabita et al., 2007). II. Regulation of cbb Gene Expression In the NSP bacteria, the major metabolic route for carbon dioxide assimilation is via the Calvin-Benson-Bassham (CBB) reductive pentose phosphate cycle, with Rubisco catalyzing the actual CO2 fixation reaction (Tabita, 1995). Through the action of Rubisco and other enzymes of the pathway, necessary organic compounds may be synthesized to sustain all the major metabolic requirements of the cell. In NSP bacteria, the genes that encode enzymes responsible for this process are for the most part organized in two separate loci, the cbbI and cbbII operons (Fig. 1). Each operon contains distinct structural genes encoding different forms of Rubisco; the cbbLS genes of the cbbI operon encode, respectively, the large and small subunits of form I Rubisco, while the cbbM gene of the cbbII operon encodes the single subunit of form II Rubisco. These loci also contain structural genes in single or double copies that encode other CBB cycle enzymes, in addition to short open reading frames of unknown function (reviewed in Gibson, 1995). Environmental growth conditions play a major role in controlling the expression of the cbb genes. Expression of both the cbbI and cbbII operons is quite low under dark aerobic chemoheterotrophic conditions. However, under photosynthetic growth conditions, expression of the genes from both operons
Chapter 28
Regulation of Carbon Dioxide Metabolism
565
Fig. 1. Organization of the cbb operons in three representative species of NSP photosynthetic bacteria, Rba. sphaeroides, Rba. capsulatus, and Rps. palustris. Note the divergently transcribed transcriptional regulator cbbR gene and the unique cbbRRS gene cluster in Rps. palustris. Abbreviations: R= CbbR transcriptional regulator; RR1= response regulator 1; RR2= response regulator 2; SR= hybrid sensor kinase/response regulator; F= fructose 1,6/sedoheptulose 1,7-bisphosphatase; P= phosphoribulokinase; A= fructose 1,6/sedoheptulose 1,7-bisphosphate aldolase; G= glyceraldehyde-3-phosphate dehydrogenase; T= transketolase; L= large-subunit of form I (L8S8) ribulose-1,5-bisphosphate carboxylase/oxygenase; S= small-subunit of form I (L8S8) ribulose-1,5-bisphosphate carboxylase/oxygenase; M= large-type subunit of form II ribulose-1,5-bisphosphate carboxylase/oxygenase; E, O, Q, X, Y, Z= genes with no assigned function.
is derepressed, with each operon responding independently to a number of environmental parameters such as the level of CO2 and the reduction state of the organic carbon compounds supplied for growth (Tabita, 1995). In general, growth under photoheterotrophic conditions, with a fixed (organic) carbon source, results in an excess of cbbII expression over cbbI. Maximal expression from both operons is observed under photoautotrophic conditions; i.e., when CO2 is used as the sole source of carbon, cbbI operon expression exceeds that of the cbbII operon (Gibson et al., 1991; Dubbs and Tabita, 1998). Thus, under photoautotrophic growth conditions, where CO2 functions as the sole carbon source, the CBB reductive pentose phosphate cycle provides nearly all cellular carbon (Tabita, 1988; Shively et al., 1998). Photosynthetic growth in the presence of fixed carbon sources (i.e., photoheterotrophic growth) changes the primary role of the CBB cycle. In this case, the CBB cycle facilitates the use of CO2 as an electron sink and terminal electron acceptor for reducing equivalents generated by carbon oxidation and photosynthesis (Tabita, 1995). Thus, a close relationship exists between the oxidation-reduction state of the
organic carbon supplied and subsequent regulation of the cbb operons. For many years, our laboratory has focused its efforts on two closely related organisms, Rba. sphaeroides and Rba. capsulatus, where cbb operon organization and mechanisms of regulation are distinct despite the close phylogenic relationship of these organisms. Moreover, in all cases described thus far, cbb transcription is regulated by a specific positive regulator protein, CbbR, encoded by the divergently transcribed cbbR gene. CbbR is a member of the large family of LysR-type transcriptional regulators (LTTRs) (Schell, 1993) and functions by binding conserved consensus sites within cbb promoter regions (Dubbs and Tabita, 1998, 2003). Recently a new regulatory feature emerged from genomic sequencing of Rps. palustris strain CGA009 (Larimer et al., 2004), the first photosynthetic bacterium to be completely sequenced and an organism with much ecological and physiological importance and applied potential. In Rps. palustris, a unique three-protein two-component system was found to be juxtaposed between the cbbLS genes of the cbbI operon and the cbbR transcriptional regulator gene, adding a further
566 level of complexity and diversity to the mechanism by which the cbb genes and CO2 assimilation are regulated in NSP bacteria. A. Rhodobacter sphaeroides Rba. sphaeroides has for a long time served as a model system for the regulation of cbb gene expression due to its accessible genetics and straightforward control of gene expression. The CBB cycle genes in this species are organized in two major loci, the cbbI and cbbII operons, located on two separate genetic elements (Suwanto and Kaplan, 1989; 1992; Gibson, 1995). The cbbI operon consists of the cbbFPALS structural genes that encode, respectively, the CBB cycle enzymes fructose 1,6/sedoheptulose 1,7-bisphosphatase (cbbFI), phosphoribulokinase (cbbPI), and fructose 1,6/sedoheptulose 1,7-bisphosphate aldolase (cbbAI), as well as the large- and small-subunit genes of form I (L8S8) Rubisco (cbbLcbbS) (Fig. 1). The organization of the cbbII operon is comparable to the cbbI operon, with similar, but not identical, copies of the cbbF, cbbP, and cbbA genes, in addition to genes encoding transketolase (cbbTII) and glyceraldehyde3-phosphate dehydrogenase (cbbGII) along with the gene encoding the single large-type subunit of the distinct form II-type Rubisco (cbbM). A single copy of the gene coding for the transcriptional regulator, cbbR, is present upstream of cbbFI and divergently transcribed (Gibson and Tabita, 1993). CbbR functionally coordinates the expression of both operons after binding consensus sequences at the 5’ promoter regions of the cbbI and cbbII gene clusters. Initial studies at the protein level had shown a complex pattern of expression of the cbb structural genes in Rba. sphaeroides (Jouanneau and Tabita, 1986; Hallenbeck et al., 1990). During aerobic chemoheterotrophic growth, the expression of both cbb operons is repressed. Under photosynthetic growth conditions, the expression of both the cbbI and cbbII operons is independently affected by the CO2 concentration and the redox state of the carbon source supplied for growth, with more-reduced carbon sources resulting in higher levels of cbb gene expression (Tabita, 1988, 1995). During photoheterotrophic growth with oxidized carbon sources, both operons are expressed, with the cbbII gene products generally predominating. With more reduced carbon sources (electron donors), such as butyrate, exogenous CO2 must be added to the growth medium, and cbb gene expression is considerably up-regulated, with cbbI
Simona Romagnoli and F. Robert Tabita operon expression greatly exceeding cbbII gene expression. Maximal expression of both operons occurs under photoautotrophic conditions (that is, in a 1.5% CO2, 98.5% H2 atmosphere), with the expression of the cbbI operon even more predominant compared to cbbII operon expression (Joanneau and Tabita, 1986). This differential control of cbbI and cbbII operon expression led to the proposal that form II Rubisco functions primarily to allow the CBB cycle to facilitate the use of CO2 as a terminal electron acceptor, thus maintaining the redox balance of the cell in the presence of oxidized organic carbon sources (electron donors). On the other hand, the function of the form I enzyme, encoded by the cbbI operon, is more attuned to allowing the CBB pathway to provide the cell with fixed carbon under growth conditions where more reduced organic electron donors are available, or under conditions where CO2 is the sole carbon source in the presence of hydrogen as reductant (Wang et al., 1993; Gibson and Tabita, 1996). Although cbbI and cbbII operon expression is clearly independently controlled, mutagenesis studies revealed that there is communication between the two operons. In fact, insertional inactivation of genes in either operon gives rise to a compensatory increase in the expression of the unaffected operon, resulting in enzyme levels that are equal to or higher than those of wild-type cells. This compensatory effect is mediated by the cbbR gene (Gibson et al., 1991; Gibson and Tabita, 1993). A cbbR Rba. sphaeroides knockout mutant strain is capable of growth under chemoheterotrophic and photoheterotrophic conditions but not under photoautotrophic conditions (Gibson and Tabita, 1993); moreover this strain does not synthesize form I Rubisco but expresses cbbM (form II Rubisco) at low levels compared to that of the wild-type strain. The results of these experiments clearly indicate that cbbR is obligatorily required for regulated expression of the cbbI operon, and substantiallyrequired for cbbII expression. CbbR is therefore a positive regulator of both operons, controlling cbb gene expression by binding specifically to AT-rich sites within the cbbR-cbbL intergenic regions of cbbI and to consensus regions at the 5’ of the cbbII promoter (Dubbs et al, 2000; Dubbs and Tabita, 2003). In a number of aerobic chemoautotrophic bacteria such as Ralstonia eutropha (Kusian and Bowein, 1995) and Xanthobacter flavus (van Keulen et al., 1998), Thiobacillus ferrooxidans (Kusano and Sugawara, 1993) and Hydrogenophilus thermoluteolus (Terazono et al., 2001), as well as the purple bacterium Chromatium
Chapter 28
Regulation of Carbon Dioxide Metabolism
vinosum (Viale et al., 1991), CbbR binds to similar promoter regions. The binding of CbbR and the organization of the promoters of both operons in Rba. sphaeroides have been thoroughly investigated (Dubbs and Tabita, 1998, 2003; Dubbs et al., 2000). The region between cbbR and cbbFI was dissected and subjected to in vitro binding assays with the purified recombinant CbbR protein. Gel mobility shift binding assays showed that CbbR binds within 103 bp of sequence 5´ to the Rba. sphaeroides cbbFI transcription start. This region contains two imperfect inverted repeat elements, each of which contains the CbbR/LysRtype consensus binding motif T-N11-A. In addition cbbI::lacZ translational fusions were constructed and their expression monitored under a variety of growth conditions (Dubbs and Tabita, 1998). These lacZ translational fusion plasmids were constructed such that they spanned from 103 to 1242 bp of DNA upstream of the cbbFI transcription start. When these plasmids were introduced into Rba. sphaeroides and the resulting strains assayed for β-galactosidase activity under chemoheterotrophic, photoheterotrophic and photoautotrophic growth conditions, it was clear that the highest β-galactosidase activity was detected in cells grown under photosynthetic conditions with CO2 as the sole source of carbon, while photosynthetic growth in the presence of malate resulted in significantly lower activity. Aerobic growth in the malate medium yielded the lowest levels of lacZ expression with all reporter plasmids. These translational fusion studies verified the general pattern of control obtained from the protein studies (Jouanneau and Tabita, 1986), however additional important information was revealed. Indeed, after analyzing the reporter gene results, along with the in vitro binding studies, it was concluded that the cbbI promoter consists of two domains. One domain was found within 103 bp 5´ to the cbbFI transcription start; this region confers the proper pattern of CbbR-dependent regulation (Dubbs and Tabita, 1998). However, most significantly, a more distal domain, that is necessary for maximal levels of cbbI expression, but does not interact with CbbR, was discovered. Since it was impossible to detect CbbR binding to this region in gel mobility shift assays, the presence of a possible second regulator intervening on the more distal promoter region of the cbbI operon was suggested. The binding sites for CbbR within the cbbI promoter-proximal regulatory domain were subsequently mapped by using DNase I footprinting experiments (Dubbs et al., 2000), as was
567
the cbbII region (Dubbs and Tabita, 2003), which also illustrated the importance of upstream regulatory regions that do not bind CbbR. In 1996, it was discovered that the RegAB/PrrAB two-component system was important for cbb control (Qian and Tabita, 1996) and this work, along with additional studies, for the first time indicated that the Reg/Prr system is an important global regulator for both cbb and nitrogen fixation (nif) control (Joshi and Tabita, 1996), as well as functioning in its more traditional role of controlling photosystem biosynthesis (Mosley et al., 1994; Eraso and Kaplan, 1995; Inoue et al., 1995). In fact, studies of the role of the Reg/Prr system provided a strong impetus to resolving the importance of the distal regulatory regions of the cbb operons of Rba. sphaeroides. Although originally identified as a regulator of photosystem biosynthesis genes in both Rba. capsulatus and Rba. sphaeroides, the regA-regB (prrA-prrB) genes have subsequently been found to control many important physiological processes, acting as global regulators of many energy-related cellular events (Dubbs and Tabita, 2004; Elsen et al., 2004). With respect to cbb control, Qian and Tabita (1996) showed that a Rba. sphaeroides regB insertion mutant exhibited reduced cbbI and cbbII expression during photoautotrophic growth (i.e., in a 1.5% CO2/98.5% H2 atmosphere). The isolation of a Rba. capsulatus regA mutant protein, RegA* (Du et al., 1998), allowed a detailed in vitro analysis of the molecular interaction between the Reg system and the cbb promoter-operator regions. RegA* is a constitutively active RegA mutant protein that functions independently of the cognate sensor kinase, RegB. It also possesses enhanced DNA binding activity relative to wild-type RegA, as first demonstrated by investigating the direct interaction of RegA* with discrete sites within the puc and puf photosynthesis gene promoters (Du et al., 1998). Once the functional complementation of an Rba. sphaeroides regA (prrA) insertion mutant with Rba. capsulatus regA was demonstrated, in vitro footprinting studies with purified RegA* proved the presence of multiple RegA* binding sites within the cbbI promoter-operator region that extend over the cbbI promoter proximal regulatory region and its more distal upstream activating region. In particular, RegA* interacts with the cbbI promoter at both the promoter proximal region (partially overlapping the CbbR binding site) and the previously described upstream activating regions (Dubbs and Tabita, 1998). RegA* binds to the Rba. sphaeroides cbbI promoter at four
568 distinct sites, delimiting a promoter region of cbbI much larger than originally thought, as demonstrated by using radiolabeled probes covering a region of the Rba. sphaeroides cbbI promoter from 161 to 2600 bp. Indeed, a strong correlation was made with respect to the strength and relative affinity of RegA* binding to upstream sequences (Dubbs et al. 2000; Dubbs and Tabita, 2004) that yielded maximum lacZ reporter activity (Dubbs and Tabita, 1998). These studies suggested a model in which RegA, when bound to distinct sites, loops over and interacts with CbbR and RNA polymerase at the transcription start site (Dubbs at al., 2000) (Fig. 2). More recently, complex formation between RegA and CbbR has been verified (Dangel and Tabita, unpublished). Similar to the cbbI promoter, the regulation of cbbII expression was analyzed in vivo by using different-length promoter regions translationally fused to lacZ. In addition, in vitro studies identified the binding sites for both CbbR and RegA via DNase I footprinting (Dubbs and Tabita, 2003). A region of the cbbII promoter from –38 to –227 bp contained a CbbR binding site and conferred low level regulated cbbII expression. The region from –227 to –1025 bp contained six RegA binding sites and conferred enhanced cbbII expression under all growth conditions. Unlike the cbbI operon, in which the region proximal to the transcriptional start has a prominent regulatory role, the region between –227 and –545 bp that contains one RegA binding site was responsible for the majority of the observed enhancement of cbbII gene expression. Therefore, both RegA and CbbR were required for maximal cbbII expression. The cbbII::lacZ fusion studies suggested that other regulatory factors in addition to CbbR and RegA might be involved in modulating the expression of the cbbII operon under aerobic chemoautotrophic growth conditions. This was particularly evident because a cbbII::lacZ fusion plasmid containing 227 bp of sequence upstream of the cbbII transcription start showed a higher level of lacZ expression under chemoautotrophic growth conditions in a regA mutant background, compared to the parental strain. In addition, the involvement of potential alternative regulators was consistent with the observation that cbbII expression during aerobic chemoautotrophic growth was significantly lower than during anaerobic photoautotrophic growth. To determine whether additional cbbII regulatory proteins were present in Rba. sphaeroides, a crude extract of chemoautotrophically grown Rba. sphaeroides CAC::regA was subjected
Simona Romagnoli and F. Robert Tabita
Fig. 2. Potential DNA-looping mechanism in Rba. sphaeroides to explain the ability of distal RegA-bound promoter sites to interact with promoter-proximal sites bound to CbbR and RegA at the transcription start site. From Dubbs et al. (2000), with permission.
to ammonium sulfate fractionation. The 70% saturation precipitate fraction was further resolved by anionic exchange chromatography after elution with a KCl gradient. The resulting fractions were tested for cbbII promoter binding activity in a gel mobility shift assay. The results indicated that at least two DNA binding activities with different affinities for the cbbII promoter were present in crude extracts of Rba. sphaeroides (Dubbs and Tabita, 2003). Moreover, these two potentially novel and specific cbbII promoter-binding proteins did not interact with the cbbI promoter region. These results, combined with the observation that chemoautotrophic expression of the cbbI operon is RegA independent, indicated that the mechanisms controlling cbbI and cbbII operon expression during chemoautotrophic growth are quite different. In conclusion, the finding that maximal aerobic chemoautotrophic expression of cbbII required regA was unexpected, given that chemoautotrophic expression of cbbI is regA-independent (Gibson et al., 2002). This indicates that the molecular mechanisms involved with regulating the two operons are quite distinct under chemoautotrophic versus photoautotrophic conditions. This specific regulation of the cbbII operon would allow the CBB pathway to play a somewhat more specialized role such that CO2 may be employed as a terminal electron acceptor, suggesting that the Reg/Prr control over cbbII gene expression during chemoautotrophic growth may give Rba. sphaeroides an enhanced ability to regulate the cellular redox poise when growing under these conditions. Finally, CbbR mutants were selected that allowed constitutive cbb gene expression under all growth conditions (Dangel et al., 2005), indicating
Chapter 28
Regulation of Carbon Dioxide Metabolism
that changes in this protein is the most important factor with respect to affecting cbb gene expression. B. Rhodobacter capsulatus The organization and regulation of the genes required for CO2 assimilation in Rba. capsulatus are quite different from those of other autotrophic organisms (Paoli et al., 1995; 1998a,b). Unlike most CO2-fixing organisms, Rba. capsulatus possesses two CbbR proteins, namely CbbRI and CbbRII, each required for the regulation of its own cognate operon (Paoli et al., 1998b), although some cross regulation has been observed (Vichivanives et al., 2000). From comparisons of deduced amino acid sequences, it is apparent that cbbRI, plus the genes of the cbbI operon and adjacent downstream genes, were acquired by some form of horizontal transfer from a phylogenetically distinct chemoautotrophic bacterium (Paoli et al., 1998a). The characterization of Rba. capsulatus mutant strains proved to be essential for understanding the role of the two cbb operons. As observed in Rba. sphaeroides (Gibson et al., 1991), inactivation of either of the two Rubisco structural genes in Rba. capsulatus induced enhanced and compensatory synthesis of the other enzyme in the mutant strain (Paoli et al., 1998b), providing an indication that expression of the separate Rubisco genes is somehow coordinated even though the cbb operons belong to independent CbbR regulons. Although form I Rubisco is not normally detectable in photoheterotrophically-grown wild-type Rba. capsulatus cells, either form I or form II Rubisco can support photoheterotrophic, photoautotrophic and chemoautotrophic growth; however, the doubling times for the mutant strains are slightly longer than for the wild-type under photoheterotrophic and photoautotrophic conditions (Paoli et al., 1998b). Unlike Rba. sphaeroides, Rba. capsulatus appears to have only a single copy of cbbP (Paoli et al., 1995). Phosphoribulokinase is the only enzyme, other than Rubisco, that is unique to the CBB pathway; therefore, disruption of the Rba. capsulatus cbbP gene should abolish the CBB pathway. Indeed, in the absence of a functional Calvin cycle, Rba. capsulatus grows photoheterotrophically only when dimethyl sulfoxide is supplied as an exogenous electron acceptor (Paoli et al., 1998b). With respect to the control of gene expression, knockout mutant strains for either of the two cbbR genes provided important insights. Amino acid sequence comparisons and phylogeneticanalyses (Paoli
569
et al., 1998a) of the Rba. capsulatus CbbR proteins with other CbbRs showed that Rba. capsulatus CbbRII is most similar to Rba. sphaeroides CbbR (55.2% identity) and less similar to Rba. capsulatus CbbRI (42.5% identity). Thus, in order to determine the role of the two cbbR genes, their function was studied in specific cbbR mutant strains (Paoli et al., 1998b). When the cbbRI gene was inactivated, no phenotype was noted under photoheterotrophic conditions, and Rubisco activity in this strain was not significantly different than the level in the wild-type strain. Interestingly, about one-half of the wild-type Rubisco activity was detected in the cbbRI strain under photoautotrophic growth conditions. Since form I Rubisco is not synthesized in photoheterotrophically grown wild-type Rba. capsulatus, these results (wild-type Rubisco activity under photoheterotrophic conditions and reduced Rubisco activity under photoautotrophic conditions) were consistent with lack of synthesis of form I Rubisco in the absence of a functional CbbRI. Western immunoblot analysis confirmed that Rubisco activity in this strain was due solely to form II Rubisco, with about the same levels of form II protein in the wild-type and mutant strains under both photoheterotrophic and photoautotrophic conditions. No synthesis of form I Rubisco occurs under photoautotrophic conditions in the cbbRI mutant strain. On the other hand, a strain in which cbbRII was insertionally inactivated was unable to grow photo- or chemoautotrophically, but grew photoheterotrophically on malate although at a reduced rate (Paoli et al., 1998b). In Rba. capsulatus, the cbbI operon, namely cbbL and cbbS, is much more tightly controlled than the cbbII operon (Fig.1), since neither cbbLS transcripts nor form I Rubisco protein were detected under photoheterotrophic growth conditions when malate, for instance, was used as the major electron donor and carbon source. Similarly, no form I Rubisco was synthesized under chemoautotrophic growth conditions (i.e., 45% H2, 5% CO2, 50% air). Under photoautotrophic conditions, with CO2 present in excess of 5–10% of the gas atmosphere, there was no form I Rubisco detected. However, at lower levels of CO2 (1.5% CO2, 98.5% H2), the accumulation of form I Rubisco protein and the level of cbbI gene expression were maximal (Paoli et al., 1995). In contrast to the selective control of cbbI gene expression, the cbbII operon, comprised of the cbbFPTGAM genes, is expressed under all growth conditions tested (Paoli et al., 1998b) but expression is maximal under
570 photoautotrophic growth conditions in a 1.5% CO2, 98.5% H2 atmosphere. The redox response regulator, RegA, was required for maximum cbbI and cbbII gene expression when cultures were switched from photoheterotrophic to photoautotrophic conditions at 1.5% CO2. Under these conditions, only 14% and 10% of cbbI and cbbII promoter activity, respectively, was observed in a photoautotrophically-grown regA strain. Further detailed analysis of the structure of the promoter regions of the Rba. capsulatus cbb operon was carried out by Vichivanives et al. (2000). In vitro DNase I footprinting studies revealed the presence of at least two RegA* binding sites within the cbbI promoter regions. The upstream RegA* binding site spans a 20 bp region between –102 to –121 that possesses a very high affinity for RegA*, and the position of this site within the promoter is consistent with the role of RegA as an activator for cbbI expression. The downstream RegA* binding site appears to be weaker; it covers a region of 16 bp between position –4 to –19 and is partially overlapped with the CbbR binding site, suggesting a potential negative role for RegA at this site, with RegA acting as a possible repressor of transcription of the cbbI operon. Two strong RegA* binding sites, very close to each other, are present within the cbbII promoter region at nucleotides –101 to –116 and –124 to –169. The distance of these binding sites compared to the transcriptional start site is similar to that of the strong RegA*- binding site found within the cbbI promoter region, suggesting a possible positive role of RegA in the transcription activation of cbbII. Since all the regulator proteins involved have different capacities to affect transcription under specialized growth conditions, a working model involving both specific and global regulators, namely CbbRI, CbbRII and RegA, in the control of cbbI, cbbII, as well as cbbRI, and cbbRII expression was proposed by Vichivanives et al. (2000) (Fig. 3). Under photoautotrophic growth conditions, cbbI expression is activated by RegA and CbbRI; likewise the level of cbbII expression is maximized by RegA and CbbRII. Under chemoautotrophic, photoheterotrophic and chemoheterotrophic growth conditions, there does not appear to be any role of RegA and there is no cbbI gene expression. Obviously the possibility remains that as yet unknown factors may alter the pattern of expression of cbb operons under specific growth conditions. In conclusion, for Rba capsulatus, it is clear that CbbRII is not required for full induction of cbbII, while cbbI expression is absolutely dependent on the presence of CbbRI (Vichivanives et al., 2000).
Simona Romagnoli and F. Robert Tabita Unlike CbbRII, CbbRI also negatively regulates its own expression under photoautotrophic and chemoautotrophic conditions. Negative autoregulation has also been shown for other LysR-type regulatory proteins (Schell, 1993). 1. The CbbR Proteins and their Co-inducers Quite early in investigations of the role of CbbR, characterization of the phenotypes of mutant strains clearly suggested that the balance of various intermediates of the CBB pathway might regulate gene expression (Paoli et al., 1998b), encouraging further investigation in this area. In particular, it was recognized that key metabolic intermediates might act as co-inducers or effectors to influence CbbR-mediated transcription, much like other LTTR regulators (Schell, 1993). This idea gained credence as there was no indication that the level of CbbR itself became altered, as recently shown by microarray studies (Pappas et al., 2004), upon growth under conditions where cbb transcription was vigorously affected. Thus, if CbbR levels do not change, and CbbR is obligately required for cbb transcription, then some posttranslational modification must be involved with altering the ability of CbbR to modulate cbb transcription under different growth conditions. Furthermore, genetic and physiological studies had indicated that the intracellular pool of RuBP might effect CbbR-mediated transcription (Smith and Tabita, 2002; Tichi and Tabita, 2002). Recently, Dubbs et al. (2004) conducted a comprehensive in vitro study of the effect of potential co-inducer molecules on the DNA binding ability of Rba. capsulatus CbbRI and CbbRII. These were mainly intermediates of the CBB cycle, since the binding of an effector, usually an intermediate or end product of the pathway regulated, is a common requirement of the members of the LTTR family (Schell, 1993). The presence of the co-inducer often increases the binding activity of the transcriptional regulator, an effect that can be easily detected by gel mobility shift and DNAse I footprinting assays. It was shown that the formation of the complex between CbbRI and the cbbI promoter was facilitated by adding 2-phosphoglycerate, 2-phosphoglycolate, 3-phosphoglycerate, phosphoenolpyruvate, KH2PO4 and also ATP to some extent, all at 1 mM concentration. The addition of 1 mM or more of RuBP, a specific substrate for Rubisco, to the binding reaction mix appeared to inhibit CbbRI-DNA binding. However, lower concentrations of RuBP stimulated the binding of CbbRI to the DNA (Dubbs et al., 2004).
Chapter 28
Regulation of Carbon Dioxide Metabolism
571
Fig. 3. Transcriptional regulation model of the Rba. capsulatus cbbI and cbbII operons involving CbbRI, CbbRII, and RegA under different growth conditions. Bent arrows in front of each gene indicate directions of transcription. The relative thickness of the arrows represents the relative abundance of the transcript under the different growth conditions. Solid curved arrows represent direct effects of regulatory proteins on transcriptional regulation, with a plus sign indicating an activator role and a minus sign indicating a repressor role. The thin curved arrows depict regulatory proteins that may directly or indirectly affect transcriptional regulation. Dashed curved arrows represent regulatory proteins that are absolutely necessary for gene expression. P symbolizes phosphate. (a) Left panel; regulation under photoautotrophic conditions (PA) indicating that the level of cbbI expression is activated by RegA and CbbRI, while cbbII expression is maximized by RegA and CbbRII. CbbRI negatively regulates its own expression. (a) Right panel; regulation under chemoautotrophic conditions (CA), indicating that cbbI is not expressed, CbbRI downregulates its own expression, and CbbRII is required for cbbII expression. (b) Left panel; regulation under photoheterotrophic conditions (PH), indicating that there is no cbbI expression and there is a considerable reduction in cbbII expression compared to photoautotrophic growth conditions. (b) Right panel; regulation under chemoheterotrophic conditions (CH), indicating that cbbI is not expressed while cbbII expression is at its lowest level. From Vichivanives et al. (2000), with permission.
The binding activity of CbbRII was also influenced by the presence of RuBP, 2-phosphoglycolate, 3-phosphoglycerate, phosphoenolpyruvate, and fructose1,6-bisphosphate, and KH2PO4. These results, along with the formation of multiple complexes, possibly indicated either a change in the oligomerization state
of CbbRII or a further bending of the DNA molecule. Neither CbbRI nor CbbRII were influenced by the presence of 2-phosphoglycerate, NADPH, NADH, fructose-6-phosphate and ribose-5-phosphate, whereas ATP and 2-phosphoglycerate were found to affect only CbbRI binding, while fructose-1,6-
572 bisphosphate altered the binding properties of only CbbRII, confirming that different metabolic signals may converge on either CbbR protein. As for the effect of RuBP, it was shown that the pattern of binding was modified depending upon the concentration of the effector added to the binding reaction, with low concentrations of RuBP (1 and 10 µM) stimulating the formation of three CbbRII-DNA complexes; in contrast, RuBP concentrations above 10 µM resulted in the disappearance of the highest molecular weight complex along with an increase in the intensity of complexes 1 and 2. The effects produced by positive inducers were also manifest in DNAse I protection assays, with RuBP in particular causing a concentration-dependent reduction of the protected sequence, and alterations of the hypersensitive sites on both cbbI and cbbII promoters (Dubbs et al., 2004). The general positive role of RuBP on the binding activity of CbbR proteins was also recently confirmed in Rba. sphaeroides (Dangel et al., 2005). By selecting in vivo for constitutively active CbbR mutant proteins, that allow constant, growth condition-independent cbb transcription, it was possible to identify several amino acid residues critical for the DNA binding activity of CbbR, many of which were found to be within the predicted co-inducer binding pocket of CbbR. The effect of RuBP on wild-type and mutant CbbR proteins was thoroughly confirmed by in vitro binding assays. To reiterate, the effect of RuBP is particularly important, since it reinforces earlier physiological and genetic studies of Rba. capsulatus and Rba. sphaeroides that indicate that RuBP, and possibly another CBB cycle intermediate(s), acts as a positive inducer of CbbR-mediated cbb gene expression (Smith and Tabita, 2002; Tichi and Tabita, 2002). Negative effectors, and other positive effectors, might also be important, as recently shown in chemoautotrophic bacteria (van Keulen et al., 1998; Grzeszik et al., 2000). C. Rhodopseudomonas palustris Amongst the NSP bacteria, Rps. palustris provides the most noticeable example of metabolic versatility and adaptability to diverse environments, as it can assimilate carbon dioxide and grow anaerobically in the light and aerobically by oxidizing organic compounds as well as thiosulfate, hydrogen and other inorganic electron donors. Rps. palustris also shows the unique ability to degrade aromatic compounds
Simona Romagnoli and F. Robert Tabita and to use them as a source of carbon, as well as energy, in addition to performing nitrogen fixation and producing large amounts of hydrogen (Larimer et al., 2004). With respect to carbon dioxide metabolism, Rps. palustris represents an interesting exception, because of the unique gene arrangement of the cbbI region. Genome sequencing of this species revealed the presence of an unprecedented three-protein and putative two-component system juxtaposed between the typically divergently transcribed cbbR gene and the structural genes of form I Rubisco (cbbLS). Because of its location within the cbbI region, this cluster of genes was named the CbbRRS system, i.e., the Calvin Benson Bassham response regulators/sensor kinase system. The role of the CbbRRS proteins as potential regulatory partners for modulating the expression of the cbb operons, was recently investigated (Romagnoli and Tabita, 2006). The CbbRRS system consists of a transmembrane hybrid sensor kinase and two distinct response regulators (Fig. 4). An evident protein sequence homology to modular domains of well-characterized kinase and response regulator proteins qualify the CbbRRS proteins as members of the two-component family of signal transduction proteins. However, these proteins lack a discernible DNA binding domain (or a typical effector domain of response regulators), and the presence of multiple phosphoacceptor modules sets this unusual system apart from more traditional two-component systems. In addition, two PAS motifs, i.e., sensory motifs distinctive of proteins generally involved in redox sensing and signal transduction activities (Taylor and Zhulin, 1999), are predicted in the N-terminal region of the sensor kinase. This raises the possibility that a dedicated system for the regulation of form I Rubisco (CbbLS) synthesis might be present in this organism, that might function by integrating multiple metabolic and redox signals through a phosphorylation cascade. Thus, two approaches were taken in order to gain insights on the role of the CbbRRS system: (i) strains containing translationally in-frame (non-polar) knockouts of key regulatory and structural genes were constructed; (ii) all the potential phosphotransfer reactions catalyzed by each of the distinct functional domains within the three CbbRRS proteins were examined in vitro in order to determine all the possible signal transduction routes. From characterization of mutant strains, some interesting features emerged. Firstly, the transcriptional regulator CbbR selectively controls only the
Chapter 28
Regulation of Carbon Dioxide Metabolism
573
Fig. 4. Domain organization within proteins of the CbbRRS system, as predicted by PFAM analysis (http://www.ncbi.nlm.nih.gov/BLAST). The hybrid sensor kinase protein CbbSR possesses two canonical transmembrane regions (TM), two PAS motifs, a transmitter domain (HisKA – histidine-containing phosphoacceptor domain; HATPase – ATPase domain with an autophosphorylatable histidine residue) at the N-terminal region, and a C-terminal Che Y-homologous receiver domain (REC). Response regulator 1, CbbRR1, presents an HPt (histidine-containing phosphotransfer) domain and a receiver domain (REC) while response regulator 2, CbbRR2, presents two receiver domains (REC). The conserved residues involved in the autophosphorylation and phosphotransfer reactions are highlighted.
expression of the cbbLS genes (cbbI operon, form I Rubisco) in Rps. palustris, whereas the cbbII operon genes, and therefore form II Rubisco, appear to be constitutively expressed. This scenario is in sharp contrast to what transpires in Rba. sphaeroides and Rba. capsulatus. Secondly, both Rubisco proteins efficiently supported growth under photoautotrophic growth conditions, with no over-expression compensatory effect or impairment in growth observed in either cbbLS or cbbM mutant strains (Romagnoli and Tabita, 2006). Although the CbbRRS system appears to be dispensable for photoautotrophic growth, deletions of single genes within the CbbRRS system resulted in the accumulation of less form I Rubisco and significantly lower total Rubisco activity; this effect was further amplified in the absence of CbbM (form II Rubisco). In fact, strains deleted for components of the CbbRRS system in a cbbM background showed between 10 and 25% of the wild-type Rubisco activity and displayed a significant growth phenotype, with a prolonged lag phase and a doubling time three times longer than the wild-type strain. In addition, investigations in progress in our laboratory suggest that
benzoate-grown cells specifically require only form I Rubisco, enabling CO2 to function as an electron sink in the presence of a reduced organic carbon source. Therefore, bearing in mind the seemingly constitutive expression of the cbbII operon in Rps. palustris, these observations suggest that there might be some mechanism of redox regulation of form I Rubisco, requiring the CbbRRS system. This regulation would allow Rps. palustris to overcome redox stress at low levels of carbon under photoautotrophic growth conditions, or when the overall intracellular redox balance is changed by culturing cells with reduced organic carbon. Detailed analyses of phosphotransfer reactions were performed using recombinant wild-type and sitedirected mutant constructs of the component proteins of the CbbRRS system. These studies provided clear indications of the phosphotransfer pathways and allowed a coherent model of signal transduction to be offered (Fig. 5) that took into account the phenotype of various mutant strains (Romagnoli and Tabita, 2007). Specifically, it was shown that the full length sensor kinase was able to phosphorylate both response
574
Simona Romagnoli and F. Robert Tabita
Fig. 5. Summary of a hypothetical model of regulation to explain signal transduction via the CbbRRS system of Rps. palustris, as proposed by Romagnoli and Tabita (2007). (a) The phosphorylation cascade is activated by a putative redox signal ultimately controlling the expression of the cbbI operon. (b) Multiple phosphotransfer reactions take place within the CbbRRS system as investigated in vitro by utilizing different length truncated constructs of the sensor kinase in combination with wild-type and site-directed mutant of all three proteins. Following autophosphorylation of residue His-409 the phosphate group may be transferred to Asp-696 within the kinase receiver domain and then to the His-171 residue of the CbbRR1 HPt domain; the phosphate group may also be transferred from His-409 of CbbSR to residue Asp-54 of the receiver domain of CbbRR1; the phosphate may be transferred from His-409 of CbbSR to either Asp-54 or Asp-195 of the two receiver domains of CbbRR2, with a preference for Asp-54. (c) Once phosphorylated, the sensor kinase, CbbSR, may assume at least three separate signaling states: (i) an inactive mode, whereby phosphorylation of the kinase is ultimately terminated by phosphotransfer from His-409 to first its own receiver domain and then to the HPt domain of CbbRR1; (ii) an active mode whereby CbbSR can catalyze phosphorylation to CbbRR1; and (iii) a second active mode in which CbbSR can catalyze phosphorylation of CbbRR2. (d) The phosphorylated response regulators may interact with the transcriptional regulator, CbbR, modulating its activity.
regulators, ruling out the necessity of a multi-step phosphorelay (Fig. 5b). Secondly, by utilizing two different-length truncated kinase constructs, it was demonstrated that the integrity of the N-terminal region of the sensor kinase, containing the first PAS motif, is important for discriminating which of the two response regulators might receive the phosphate from the sensor kinase. Finally, the receiver domain of the sensor kinase, along with the N-terminal region, contributes to the ATP affinity and stability of the sensor kinase, and consequently the final signaling activity of the system. To integrate the physiological data with in vitro characterization of the phosphotransfer reactions, it was envisaged that multiple signaling states of the sensor kinase control downstream reactions in response to activating signals (Fig. 5c). The net effect of these concerted interactions at distinct regions of the sensor kinase would allow this protein to function
as an internal molecular switch to coordinate a unique branched phosphorelay system, with an imbalance in the phosphorylation cascade subsequently resulting in changes in the levels of form I Rubisco. Since neither response regulator seems to bear a noticeable DNA binding motif, and thus lacks the capacity to directly affect the expression of the cbbLS genes, it was proposed that the phosphorylated response regulators might modify the interaction of the main transcriptional regulator protein, CbbR, with the cbbLS promoter (Fig. 5d). Indeed, it is well established in other species that CbbR binding and/or its ability to activate transcription can be enhanced or reduced by co-inducers, or post-translational modifications (Smith and Tabita, 2002; Tichi and Tabita, 2002; Dubbs et al. 2004; Dangel et al., 2005). In conclusion, the added feature of the CbbRRS signal transduction system in Rps. palustris provides
Chapter 28
Regulation of Carbon Dioxide Metabolism
an interesting twist and renewed interest for studying the regulation of cbb gene expression in NSP bacteria. Acknowledgments These studies were supported by grants from the National Institute of Health (GM 45404) and the Department of Energy Genomics:GTL Program (DE-FG02-01ER63241). References Chen JH, Gibson JL, McCue LA and Tabita FR (1991) Identification, expression, and deduced primary structure of transketolase and other enzymes encoded within the form II CO2 fixation operon of Rhodobacter sphaeroides. J Biol Chem 266: 20447–20452 Dangel AW, Gibson JL, Janssen AP and Tabita FR (2005) Residues that influence in vivo and in vitro CbbR function in Rhodobacter sphaeroides and identification of a specific region critical for co-inducer recognition. Mol Microb 57: 1397–1414 Du S, Bird TH and Bauer CE (1998) DNA binding characteristics of RegA. A constitutively active anaerobic activator of photosynthesis gene expression in Rhodobacter capsulatus. J Biol Chem 273: 18509–18513 Dubbs JM and Tabita FR (1998) Two functionally distinct regions upstream the cbbI operon of Rhodobacter sphaeroides regulate gene expression. J Bacteriol 180: 4903–4911 Dubbs JM and Tabita FR (2003) Interactions of the cbbII promoter-operator region with CbbR and RegA (PrrA) regulators indicate distinct mechanisms to control expression of the two cbb operons of Rhodobacter sphaeroides. J Biol Chem 278: 16443–16450 DubbS JM and Tabita FR (2004) Regulators of nonsulfur purple phototrophic bacteria and the interactive control of CO2 assimilation, nitrogen fixation, hydrogen metabolism and energy generation. FEMS Microbiology Reviews 28: 353–376 Dubbs JM, Bird TH, Bauer CE and Tabita FR (2000) Interaction of CbbR and RegA* transcription regulators with the Rhodobacter sphaeroides cbbI promoter-operator region. J Biol Chem 275: 19224–19230 Dubbs P, Dubbs JM and Tabita FR (2004) Effector-mediated interaction of CbbRI and CbbRII regulators with target sequences in Rhodobacter capsulatus. J Bacteriol 186: 8026–8035 Elsen S, Swem LR, Swem DL and Bauer, CE (2004) RegB/RegA, a highly conserved redox-responding global two-component regulatory system. Microbiol Mol Biol Rev 68: 263–279 Eraso JM and Kaplan S (1995) Oxygen-insensitive synthesis of the photosynthetic membranes of Rhodobacter sphaeroides: A mutant histidine kinase. J Bacteriol 177: 2695–2706 Gibson JL (1995) Genetic analysis of CO2 fixation genes. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 1107–1124. Kluwer Academic Publishers, Dordrecht
575
Gibson JL and Tabita FR (1993) Nucleotide sequence and functional analysis of CbbR, a positive regulator of the Calvin cycle operons of Rhodobacter sphaeroides. J Bacteriol 175: 5778–5784 Gibson JL and Tabita FR (1996) The molecular regulation of the reductive pentose phosphate pathway in proteobacteria and cyanobacteria. Arch Microbiol 166: 141–150 Gibson JL, Falcone DL and Tabita FR (1991) Nucleotide sequence, transcriptional analysis and expression of genes encoded within the form I CO2 fixation operon of Rhodobacter sphaeroides. J Biol Chem 266: 14646–14653. Gibson JL, Dubbs JM and Tabita FR (2002) Differential expression of the CO2 fixation operons of Rhodobacter sphaeroides by the Prr/Reg two-component system during chemoautotrophic growth. J Bacteriol 184: 6654–6664 Goethals K, Van Montague M and Holsters M (1992) Conserved motifs in a divergent nod box of Azorhizobium caulinodans ORS571 reveal a common structure in promoters regulated by LysR-type proteins. Proc Natl Acad Sci USA 89: 1646–1650 Grzeszik C, Jeffke T, Schaferjohann, Kusian B, and Bowien B (2000) Phosphoenolpyruvate is a signal metabolite in transcriptional control of the cbb CO2 fixation operons in Ralstonia eutropha. J Mol Microbiol Biotechnol 2: 311–320 Hallenbeck PL, Lerchen R, Hessler P and Kaplan S (1990) Phosphoribulokinase activity and regulation of CO2 fixation critical for photosynthetic growth of Rhodobacter sphaeroides. J Bacteriol 172: 1749–1761 Inoue K, Kouadio JL, Mosley CS and Bauer CE (1995) Isolation and in vitro phosphorylation of sensory transduction components controlling anaerobic induction of light harvesting and reaction center gene expression in Rhodobacter capsulatus. Biochemistry 34: 391–396 Joshi H and Tabita FR (1996) A global two component signal transduction system that integrates the control of photosynthesis, carbon dioxide assimilation, and nitrogen fixation. Proc Natl Acad Sci USA 93: 14515–14520 Jouanneau Y and Tabita FR (1986) Independent regulation of synthesis of form I and form II ribulose bisphosphate carboxylase-oxygenase in Rhodopseudomonas sphaeroides. J Bacteriol 165: 620–624 Kusian B and Bowein B (1995) Operator binding of the CbbR protein, which activates the duplicate cbb CO2 assimilation operons of Alcaligenes eutrophus. J Bacteriol 177: 6568–6574 Kusano T and Sugawara K (1993) Specific binding of Thiobacillus ferrooxidans RbcR to the intergenic sequence between the rbc operon and the rbcR gene. J Bacteriol 175: 1019–1025 Larimer FW, Chain P, Hauser L, Lamerdin J, Malfatti S, Do L, Land ML, Pelletier DA, Beatty JT, Lang AS, Tabita FR, Gibson JL, Hanson TE, Bobst C, Torres JL, Peres C, Harrison FH, Gibson J and Harwood CS (2004) Complete genome sequence of the metabolically versatile photosynthetic bacterium Rhodopseudomonas palustris. Nat Biotechnol 22: 55–61 Mosley CS, Suzuki JY and Bauer CE (1994) Identification and molecular genetic characterization of a sensor kinase responsible for coordinately regulating light harvesting and reaction center gene expression in response to anaerobiosis. J Bacteriol 176: 7566–7573 Paoli GC, Strom-Morgan N, Shivelyand JM and Tabita FR (1995) Expression of the cbbLcbbS and cbbM genes and distinct organization of the cbb Calvin cycle structural genes of Rhodobacter capsulatus. Arch Microbiol 164: 396–405
576 Paoli GC, Soyer F, Shively JM and Tabita FR (1998a) Rhodobacter capsulatus genes encoding form I ribulose-1,5-bisphosphate carboxylase/oxygenase (cbbLS) and neighboring genes were acquired by a horizontal gene transfer. Microbiology 144: 219–227 Paoli GC, Vichivanives P and Tabita FR (1998b) Physiological control and regulation of the Rhodobacter capsulatus cbb operons. J Bacteriol 180: 4258–4269 Pappas CT, Sram J, Moskvin OV, Ivanov PS, Mackenzie RC, Choudhary M, Land ML, Larimer FW, Kaplan S, and Gomelsky M (2004) Construction and validation of the Rhodobacter sphaeroides 2.4.1 DNA microarray: Transcriptome flexibility at diverse growth modes. J Bacteriol 186: 4748–4758 Qian Y and Tabita FR (1996) A global signal transduction system regulates aerobic and anaerobic CO2 fixation in Rhodobacter sphaeroides. J Bacteriol 178: 12–18 Romagnoli S and Tabita FR (2006) A novel three-protein twocomponent system provides a regulatory twist on an established circuit to modulate expression of the cbbI region of Rhodopseudomonas palustris CGA010. J Bacteriol 188: 2780–2791 Romagnoli S and Tabita FR (2007) Phosphotransfer reactions of the three-protein CbbRRS two-component system from Rhodopseudomonas palustris CGA010 appear to be controlled by an internal molecular switch on the sensor kinase. J Bacteriol 189: 325–335 Schell MA (1993) Molecular biology of the LysR family of transcriptional regulators. Ann Rev Microbiol 47: 597–626 Shively JM, van Keulen G and Meijer WG (1998) Something from almost nothing: Carbon dioxide fixation in chemoautotrophs Ann Rev Microbiol 52: 191–230 Smith SA and Tabita FR (2002) Up-regulated expression of the cbbI and cbbII operons during photoheterotrophic growth of a Rubisco deletion mutant of Rhodobacter sphaeroides. J Bacteriol 184: 6721–6724 Suwanto A and Kaplan S (1989) Physical and genetic mapping of the Rhodobacter sphaeroides genome: Presence of two unique circular chromosomes. J Bacteriol 171: 5850–5859 Suwanto A and Kaplan S (1992) Chromosome transfer in Rhodobacter sphaeroides: Hfr formation and genetic evidence for two unique circular chromosomes. J Bacteriol 174: 1135–1145 Tabita FR (1988) Molecular and cellular regulation of autotrophic carbon dioxide fixation in microorganisms. Microbiol Rev 52: 155–189
Simona Romagnoli and F. Robert Tabita Tabita FR (1995) The biochemistry and metabolic regulation of carbon metabolism and CO2 fixation in purple bacteria. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria, pp 885–914. Kluwer Academic Publishers, The Netherlands Tabita FR, Hanson TE, Li H, Satagopan, S, Singh J and Chan S (2007) Function, structure, and evolution of the Rubisco like proteins and their Rubisco homologs. Microbiol Mol Biol Rev 71: 576–599 Terazono K, Hayashi NR and Igarashi Y (2001) CbbR, a LysRtype transcriptional regulator from Hydrogenophilus thermoluteolus, binds two cbb promoter regions. FEMS Microbiol Lett 198: 151–157 Tichi MA and Tabita FR (2002) Metabolic signals that lead to control of cbb gene expression in Rhodobacter capsulatus. J Bacteriol 184: 1905–1915 van den Berg ERE, Dijkhuizen L and Meijer WG (1993) CbbR, a LysR-type transcriptional activator, is required for expression of the autotrophic CO2 fixation enzymes of Xanthobacter flavus. J Bacteriol 175: 6097–6104 van Keulen G, Girbal L, van den Berg ER, Dijkhuizen L and Meijer WG (1998) The LysR-type transcriptional regulator CbbR controlling autotrophic CO2 fixation by Xanthobacter flavus is an NADPH sensor. J Bacteriol 180: 1411–1417 Viale AM, Kobayashi H, Akazawa T and Henikoff S (1991) rbcR, a gene coding for a member of the LysR family of transcriptional regulators, is located upstream of the expressed set of ribulose-1,5-bisphosphate carboxylase/oxygenase genes in the photosynthetic bacterium Chromatium vinosum. J Bacteriol 173: 5224–5229 Vichivanives P, Bird TH, Bauer CE and Tabita FR (2000) Multiple regulators and their interaction in vivo and in vitro with the cbb regulons of Rhodobacter capsulatus. J Mol Biol 300: 1079–1099 Wang X., Falcone DL and Tabita FR (1993) Reductive pentose phosphate-independent CO2 fixation in Rhodobacter sphaeroides and evidence that ribulose bisphosphate carboxylase/oxygenase activity serves to maintain the redox balance of the cell. J Bacteriol 175: 3372–3379 Windhovel U and Bowein B (1991) Identification of cfxR, an activator gene of autotrophic CO2 fixation in Alcaligenes eutrophus. Mol Microbiol 5: 2695–2705
Chapter 29 Degradation of Aromatic Compounds by Purple Nonsulfur Bacteria Caroline S. Harwood* Department of Microbiology, The University of Washington, Seattle, WA 98195-7242, U.S.A.
Summary ............................................................................................................................................................... 577 I. Introduction..................................................................................................................................................... 578 II. Biochemical Themes ..................................................................................................................................... 578 III. Species that Degrade Aromatic Compounds ................................................................................................. 579 IV. Aerobic Degradation of Aromatic Compounds by Rhodopseudomonas palustris ........................................ 579 V. Anaerobic Benzoate Degradation ................................................................................................................. 580 A. A Historical Overview of Anaerobic Benzoate Degradation ............................................................ 580 B. Rhodopseudomonas palustris as a Model Organism ...................................................................... 580 C. Organization of the Anaerobic Benzoate Degradation Gene Cluster in Rhodopseudomonas palustris ........................................................................................................................................... 581 D. The Enzymes of Anaerobic Benzoate Degradation ........................................................................ 581 1. Properties and Mechanism of Benzoyl-CoA Reductase ......................................................... 581 a. The Natural Source of Reductant for Benzoyl-CoA Reduction...................................... 584 2. The β-oxidation-like Enzymes ................................................................................................ 584 3. Benzoate-CoA Ligases .......................................................................................................... 585 4. The Lower Benzoate Pathway: Pimelate Degradation ........................................................... 586 E. Peripheral Pathways ........................................................................................................................ 586 1. 4-Hydroxybenzoate Degradation ........................................................................................... 586 2. Cyclohexanecarboxylate Degradation ................................................................................... 587 3. Phenylalkanecarboxylate Degradation: Cinnamate and 4-Hydroxycinnamate ....................... 588 4. Anaerobic Phenylacetate Degradation.................................................................................... 588 5. Chlorobenzoate Degradation ................................................................................................. 589 VI. The Molecular Regulation of Anaerobic Aromatic Compound Degradation .................................................. 589 VII. Comparative Aspects .................................................................................................................................... 590 Acknowledgments ................................................................................................................................................ 591 References ............................................................................................................................................................ 591
Summary Aromatic compounds are among the more difficult groups of naturally occurring organic compounds to degrade because of the high resonance stability of benzene rings. Some purple nonsulfur bacteria have a well-developed ability to degrade green plant-derived aromatic compounds including a variety of lignin monomers as well as some man-made compounds, including chlorobenzoates and toluene. Peripheral pathways modify these compounds to form a small number of common intermediates that enter pathways leading to ring cleavage. Depending on the compound and the species involved, degradation can occur aerobically under chemohet*Email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 577–594. © 2009 Springer Science + Business Media B.V.
Caroline S. Harwood
578
erotrophic conditions or anaerobically under photoheterotrophic conditions. Two different biochemical strategies, one involving oxygenases and the other involving aromatic ring reduction, take place depending on the availability of oxygen. The best-studied aromatic compound-degrading species, Rhodopseudomonas (Rps.) palustris, has served as a model organism to elucidate a central reductive pathway of benzoate degradation that is used to process most compounds anaerobically. The main features of this pathway appear to apply to other metabolic groups of bacteria that degrade aromatic compounds under anaerobic conditions. These include a novel enzyme, benzoyl-CoA reductase, that relieves the resonance stability of the aromatic ring, and a sequence of β-oxidation-like reactions leading to ring cleavage by a new type of ring cleavage enzyme. The genes for anaerobic benzoate and 4-hydroxybenzoate degradation have been located on the sequenced genome of Rps. palustris strain CGA009. A similar gene cluster is present in three other recently sequenced strains of Rhodopseudomonas. The genome sequence of one of the strains, BisB5, revealed that this strain can degrade an expanded set of aromatic compounds converging on and including phenylacetate under photoheterotrophic conditions. I. Introduction About 25% of the terrestrial biomass on earth is comprised of benzene rings joined by ether linkages to form lignin, a major component of the woody tissues of green plants. White rot fungi produce powerful oxidative extracellular enzymes that degrade lignin randomly, a process that results in the release of aromatic lignin monomers, primarily hydroxycinnamates, into soils and waters where they serve as carbon sources for many different kinds of bacteria. Some species of purple nonsulfur bacteria have a welldeveloped ability to utilize plant-derived aromatic compounds under photoheterotrophic conditions. One of these, Rps. palustris, has served as a model organism to elucidate a central metabolic pathway of anaerobic benzene ring reduction and cleavage that starts with benzoyl-coenzyme A (benzoyl-CoA). Many different kinds of human activities also result in the release of compounds with benzene rings into the environment. Purple nonsulfur bacteria have a more limited ability to degrade these generally more toxic compounds, but there are exceptions. Anaerobic toluene, phenol, and 3-chlorobenzoate (3-CBA) degradation by various species of purple nonsulfur bacteria are well documented. This chapter will focus on studies of the physiology, biochemistry and molecular basis for aromatic compound degradation by purple nonsulfur bacteria that have been carried out since this subject was reviewed in the Blankenship et al. (1995) volume on anoxygenic photosynthetic bacteria (Gibson and Abbreviations: 3-CBA – 3-Chlorobenzoate; 4-HBA – 4-Hydroxybenzoate; Benzoyl-CoA – Benzoyl-coenzyme A; CoA – Coenzyme A; Rps. – Rhodopseudomonas; Rsp. – Rhodospirillum; Rvi. – Rubrivivax; T. – Thauera
Harwood, 1995). Genome sequence data that have become available in the last few years have revealed that some strains of Rps. palustris have aromatic degradation capabilities even beyond those that had been elucidated by physiological studies. This information will also be summarized. II. Biochemical Themes The biodegradation of aromatic compounds presents special biochemical challenges for microorganisms because about 125 kJ per mole of resonance energy stabilizes the benzene ring itself. Very different biodegradation strategies are used depending on whether or not molecular oxygen is available. Under aerobic conditions, ring resonance is relieved by introducing hydroxyl groups followed by ring cleavage reactions either between or adjacent to vicinal hydroxyls. Depending on the specific aromatic compound under consideration, one or the other or both of these steps is accomplished by enzymes that use molecular oxygen as a substrate. Over 10 different aerobic ring cleavage pathways have been described in bacteria (Dagley, 1986; Zaar et al., 2004; Nogales et al., 2007). Although a convincing report of the degradation of benzoate in the absence of air was published in 1934 (Tavin and Buswell, 1934), the prevailing view throughout the first half of the twentieth century was that aromatic compounds were recalcitrant to degradation in anaerobic environments. This view gradually changed as radioactively labeled compounds became available, allowing unequivocal demonstration that methanogenic enrichments completely mineralized benzoate to carbon dioxide and methane (Clarke and Fina, 1952; Fina and Fiskin, 1960). Since that time benzoate has emerged as the aromatic compound that
Chapter 29
Degradation of Aromatic Compounds
is most commonly degraded by anaerobes. The pathway used starts with benzoyl-CoA and the benzene ring is destabilized by reduction of the resonating double bonds. Diverse peripheral pathways degrade aromatic hydrocarbons, chlorinated aromatic compounds, phenolics and phenylalkanecarboxylates by diverse pathways to form benzoyl-CoA as a common intermediate (Heider and Fuchs, 1997a,b; Harwood et al., 1999; Gibson and Harwood, 2002). III. Species that Degrade Aromatic Compounds Rps. palustris is the species that most often predominates in enrichments for light-dependent growth in anoxic minimal medium with benzoate as a sole carbon source. In addition, Rhodospirillum fulvum (Pfennig et al., 1965; renamed Phaeospirillum fulvum in Imhoff et al., 1998), Rhodocyclus purpureus (Pfenning, 1978), Rps. acidophila (Yamanaka et al., 1983) (renamed Rhodoblastus acidophilus in Imhoff, 2001), Rhodomicrobium (Rmi.) vannieli (Wright and Madigan, 1991), and Blastochloris (Blc.) sulfoviridis strain ToP1 (Zengler et al., 1999) grow photoheterotrophically with benzoate. More recently a new species, Rubrivivax (Rvi.) benzoatilyticus, isolated from rice paddy soil, has been described that degrades aromatic compounds (Ramana Ch et al., 2006). Only a few studies have surveyed growth on a range of different kinds of aromatic compounds. Depending on the strain, Rps. palustris uses, in addition to benzoate, 4-hydroxybenzoate (4-HBA), 4-hydroxycinnamate, cinnamate, chorobenzoates, phenylacetate, phenol, methoxylated compounds, as well as aromatic aldehydes and alcohols for growth (Harwood and Gibson, 1988; Madigan and Gest, 1988; van der Woude et al., 1994; Noh et al., 2002). Rmi. vannieli grows well with the green-plant derived methoxylated compounds syringate and vannilate (Wright and Madigan, 1991), and Rvi. benzoatilyticus is able to utilize phenylalanine anaerobically (Ramana Ch et al., 2006). To date, just one strain of anoxygenic phototroph has been described that can utilize an aromatic hydrocarbon under anaerobic conditions. Blc. sulfoviridis strain ToP1 was enriched and isolated from activated sludge with toluene supplied as a carbon source. The isolated strain grew on toluene with a doubling time of 30–35 h; about the same rate of growth as was seen on benzoate (Zengler et al., 1999).
579 IV. Aerobic Degradation of Aromatic Compounds by Rhodopseudomonas palustris The ability of purple nonsulfur bacteria to degrade aromatic compounds aerobically has been addressed in only a few studies, mainly with Rps. palustris. An early publication by Hegeman suggested that Rps. palustris degrades 4-HBA via the dihydoxylated compound protocatechuate by a meta-cleavage pathway (Hegeman, 1967). The Rps. palustris CGA009 genome sequence (Larimer et al., 2004) verified that genes for this degradation sequence are present (RPA4695-4703). The Rps. palustris CGA009 genome (Fig. 1) also encodes oxygenase-dependent ring cleavage pathways for the aerobic degradation of the aromatic compounds homoprotocatechuate (RPA3755-3762), homogentisate (RPA4670-4675) and phenylacetate (RPA1723-24 and RPA37653768), and we have verified that strain CGA009 does in fact grow on each of these compounds (Harwood, unpublished). Three of the four other strains of Rhodopseudomonas that have been sequenced at the time of this writing, HaA2, BisB5, and BisA53, each encode the four aerobic ring cleavage pathways shown in Fig 1. The fourth strain, BisB18, lacks genes for aerobic protocatechuate and homoprotecatechuate degradation. The strain CGA009 sequence also encodes many mono- and dioxygenases that likely participate in peripheral pathways for the degradation of diverse aro-
Fig. 1. Ring cleavage pathways for aromatic compound degradation in Rps. palustris strain CGA009. The approximate location of the gene cluster for each degradation pathway (deg) is indicated on this map of the circular chromosome.
580 matic compounds that terminate at the ring cleavage substrates shown in Fig. 1. Strain CGA009 has genes for seven P450 enzymes, four of which have been purified. The mono-oxygenase activity of one these (CYP199A2; RPA1871) was reconstituted with an associated 2Fe-2S protein named palustrisredoxin A (RPA1872) and a Pseudomonas putida ferredoxin reductase and found to catalyze the O-demethylation of 4-methoxybenzoic acid to form 4-HBA (Bell et al., 2006). V. Anaerobic Benzoate Degradation A. A Historical Overview of Anaerobic Benzoate Degradation The first progress in elucidating the mechanism of anaerobic benzoate degradation came from work with pure cultures of Rps. palustris by W. C. Evans and co-workers in Wales and by G. Hegeman and co-workers in the United States in the 1960s. These investigators postulated that the initial steps in benzoate degradation would be reductive rather than oxidative. Dutton and Evans demonstrated that hypothesized unlabeled metabolic intermediates of benzoate degradation became radioactive when added exogenously to cell suspensions that were incubated with 14C-labeled benzoate (Dutton and Evans, 1969).
Caroline S. Harwood At the same time Guyer and Hegeman (1969) isolated Rps. palustris mutants that were unable to grow on two hypothesized intermediates of anaerobic benzoate degradation. Such mutants failed to grow anaerobically with benzoate. These early studies in isotope trapping and mutant analysis supported the concept that benzoate is reduced to form an alicyclic acid that is further metabolized by a series of reactions reminiscent of fatty acid β-oxidation, culminating in ring cleavage. An early pathway proposed for anaerobic benzoate degradation by Rps. palustris is shown in Fig. 2. Once the outlines of the pathway of anaerobic benzoate degradation were established from work with whole cells, investigators experienced difficulties moving to the next step of measuring expected enzymatic activities in cell extracts. Eventually, it became clear that several proposed reactions in the benzoate degradation pathway could be measured provided that coenzyme A (CoA) and ATP were added to cell extracts. This supported the concept that the intermediates of anaerobic benzoate degradation are CoA thioesters (Whittle et al., 1976). B. Rhodopseudomonas palustris as a Model Organism Since the work of Dutton and Evans, Rps. palustris has served as a model organism for studies of
Fig. 2. The view of anaerobic benzoate photometabolism by Rps. palustris as it existed in 1978. From Dutton and Evans (1978).
Chapter 29
Degradation of Aromatic Compounds
anaerobic benzoate and 4-HBA degradation, and has been joined by Thauera (T.) aromatica K172, a denitrifying species that, in addition, has served as a model for studies of anaerobic toluene, phenylacetate, and phenol degradation. Strain K172, originally isolated from enrichments with phenol, is a beta proteobacterium that clusters with other Thauera species and with Azoarcus species, many of which degrade aromatic hydrocarbons and aromatic acids anaerobically (Anders et al., 1995). Rps. palustris is an alpha proteobacterium, closely related to Bradyrhizobium japonicum. The strain of Rps. palustris (strain CGA009) that has been studied most grows on diverse aromatic acids of the types derived from lignin degradation (Harwood and Gibson, 1988) and its genome has been sequenced (Larimer et al., 2004). The pathways for anaerobic benzoate degradation that are used by T. aromatica and Rps. palustris differ slightly and are shown in Fig. 3. From biochemical work and molecular analyses of genes involved in anaerobic benzoate degradation a few themes are evident: 1. A novel enzyme, benzoyl-CoA reductase, relieves the resonance stability of the benzene ring. 2. The ring reduction substrate and subsequent intermediates are CoA thioesters. 3. The biochemical theme of β-oxidation has been modified by evolution to prepare the alicyclic ring for cleavage. 4. The ring cleavage enzymes are new. C. Organization of the Anaerobic Benzoate Degradation Gene Cluster in Rhodopseudomonas palustris The benzoate degradation gene cluster from Rps. palustris strain CGA009 includes genes for cyclohexanecarboxylate (ali), benzoate (bad) and 4-HBA (hba) degradation (Fig. 4). Some of the Rps. palustris CGA009 genes were assigned functions based on matches between the deduced N-terminal amino acid sequences of gene products and the experimentally determined N-terminal amino acid sequences of purified enzymes that carried out known reactions (Pelletier and Harwood, 1998), but in most cases,
581 functions were assigned based on the phenotypes of mutants that were constructed by directed molecular methods. For example, strains with disruptions in badD or badE were unable to convert benzoyl-CoA to reduced products and thus had the phenotypes of benzoyl-CoA reductase mutants (Egland et al., 1997). Similar sets of genes are found in three of the four recently sequenced strains of Rhodopseudomonas. One of the strains (HaA2; GenBank Accession number CP000250), lacks the cluster and is unable to grow anaerobically with aromatic compounds. The benzoate degradation gene cluster in strain BisB18 (CP000301) is identical to that of strain CGA009 with the exception that a glutaryl-CoA dehydrogenase gene (RPC1037) is present adjacent to badK. The BisA53 (CP000463) and BisB5 (CP000283) ali-ben-hba regions match that of CGA009, but with four extra genes inserted. Three of the genes likely encode a 2-oxoglutarate:ferredoxin oxidoreductase and associated ferredoxin for transfer of electrons to benzoyl-CoA reductase (see below). The fourth gene (RPD1527 in BisB5 and RPE0597 in BisA53) is a predicted N-acetyltransferase with an unknown role in anaerobic aromatic compound degradation. Strain BisB5 is missing the hbaHGFE genes encoding a predicted ABC transport system for 4-HBA. BisB5, as noted below, has genes for anaerobic phenylacetate degradation next to its 4-HBA degradation genes. D. The Enzymes of Anaerobic Benzoate Degradation 1. Properties and Mechanism of Benzoyl-CoA Reductase A redox potential of –3.1 V is theoretically required for the initial one-electron reduction of benzene, but this requirement may be changed to about –1.9 V due to the coenzyme A group on benzoyl-CoA (Boll et al., 2000c). In T. aromatica, a Chromatium vinosumtype ferredoxin with a midpoint potential possibly as low as –0.59 V serves as a natural electron donor for benzoyl-CoA reductase (Boll and Fuchs, 1998; Boll et al., 2000b). Although this is a low midpoint potential, it still falls far short of that needed for substrate reduction. From this it is clear that a major function of benzoyl-CoA reductase is to generate a low potential electron donor for ring reduction. When Boll and Fuchs first purified benzoyl-CoA reductase from T. aromatica in 1995 (Boll and
582
Caroline S. Harwood
Fig. 3. Anaerobic benzoate degradation by Rps. palustris and T. aromatica. Reactions involved in funneling cyclohexane-1-carboxylate into the benzoate degradation pathway in Rps. palustris are also shown. Solid arrows indicate enzymes that have been purified from either Rps. palustris or T. aromatica. Dashed arrows indicate expected enzymatic reactions. Assignment of gene products from Rps. palustris and T. aromatica to specific steps is indicated. RedBadB, reduced BadB; redFdx, reduced ferredoxin. From Pelletier and Harwood (2000).
Fuchs, 1995), they found its activity to be strictly dependent on MgATP as a co-substrate, in addition to the provision of a strong reductant. This was an exciting finding because the only other enzyme
known to stochiometrically couple ATP hydrolysis to a reduction reaction is nitrogenase (Fisher and Newton, 2002). Benzoyl-CoA reductase is a 170 kD heterotetramer (Fig. 5). Subsequently, genes encoding
Chapter 29
Degradation of Aromatic Compounds
583
Fig. 4. The benzoate degradation gene cluster from Rps. palustris CGA009 (genome accession no.BX571963). The T. aromatica (accession no. AJ224959) (Breese et al., 1998) benzoate degradation gene cluster is shown for comparison. Hba genes associated with 4-HBA degradation are contiguous with benzoate degradation genes on the Rps. palustris chromosome and are also shown. The genes from Rps. palustris encode enzymes of benzoate and 4-HBA degradation as indicated in Figs 3 and 7. The korAB genes are predicted to encode 2-oxoglutarate: ferredoxin oxidoreductase (Dörner and Boll, 2002; Ebenau-Jehle et al., 2003). The functions in anaerobic benzoate degradation of the badC, badL and unnamed genes are not known. The hbaEFGH genes are predicted to encode an ABC transport system for 4-HBA, but this has not been demonstrated experimentally. Arrows indicate the direction of transcription. Genes in different clusters that have the same pattern of shading are homologous. Arrows above genes in Rps. palustris indicate predicted operons. The number above a given gene indicates the percentage amino acid identify between that gene product (protein) and the homologous protein from T. aromatica.
its four subunits were identified from T. aromatica and from Rps. palustris and the encoded proteins were found to be about 70% identical between the species (Egland et al., 1997; Breese et al., 1998). Benzoyl-CoA reductase does not, however, share any sequence similarity with nitrogenase and the two enzymes have substantially different mechanisms (Unciuleac and Boll, 2001; Möbitz and Boll, 2002). It seems that nitrogenase and benzoyl-CoA reductase have solved a similar difficult biochemical problem by independent evolutionary paths. The reduction of benzoyl-CoA has been likened to a biological Birch reduction, a chemical process whereby aromatic compounds are reduced by exposure to metallic sodium in liquid ammonia (Birch and Smith, 1958; Boll, 2005). A Birch reduction proceeds in alternating electron transfer and protonation steps. By analogy, a critical step in the reaction cycle of benzoyl-CoA reductase was proposed to be a transient one-electron reduction of the thioester to generate a radical anion intermediate (Buckel and Keese, 1995). Recent kinetic studies using various substrate analogs support a modified Birch reduction mechanism involving a proton-assisted electron transfer to give a cyclohexadieneyl radical as the first step in the reaction catalyzed by benzoyl-CoA reductase (Möbitz and Boll, 2002; Boll, 2005). Biochemical and biophysical studies suggest that the T. aromatica benzoyl-CoA reductase overcomes the high barrier to substrate reduction by converting the chemical energy of ATP to redox potential through a step involving ATP binding (ATP binding switch)
Fig. 5. A proposed arrangement of benzoyl-CoA reductase subunits and associated [4Fe-4S] centers. The arrow depicts a predicted path of electron flow based on an observed magnetic interaction between the clusters. ATP binding by the predicted activase subunits is indicated. From Möbitz et al. (2004).
and a second step involving ATP hydrolysis and the formation of an enzyme-phosphate intermediate (ATP hydrolysis switch). Conformational changes that accompany ATP binding and hydrolysis are thought to influence the spin alignments of the three [4S-4Fe] centers present in benzoyl-CoA reductase, one of which changes from a low-spin (S=1/2) to a high-spin state (S=7/2), a transition that has been calculated to lower the potential of the cluster by as much as 500 mV. This low-spin/ high-spin switch is thought to be critical to the delivery of an appropriately reduced
584 electron to the substrate, benzoyl-CoA (Boll et al., 2000a , 2001; Unciuleac and Boll, 2001; Möbitz et al., 2004). The Rps. palustris benzoyl-CoA reductase has proved much more recalcitrant to investigation than the same enzyme from T. aromatica. Despite intensive effort, investigators have failed to consistently detect Rps. palustris benzoyl-CoA reductase activity in cell extracts. Thus, whereas two-dimensional nuclear magnetic resonance spectroscopy has shown that cyclohexa-1,5-diene-1-carboxyl-CoA is the sole product formed by T. aromatica benzoyl-CoA reductase (Boll et al., 2000c), the product formed by the Rps. palustris enzyme is not known with certainty. Cyclohexa-2,5-diene-1-carboxylate and cyclohexa-1,4-diene-1-carboxylate were extracted from Rps. palustris cells grown on benzoate following treatment with alkali to cleave the CoA thioester bond (Gibson and Gibson, 1992). Although more work will be required to determine if one or both of these compounds is the product of benzoyl-CoA reduction by Rps. palustris, it is worth noting that cyclohexa-2,5-diene-1-carboxylate is the compound formed from benzoate in a chemical Birch reduction (Birch and Smith, 1958). Rps. palustris grows well on cyclohexa-2,5-diene-1-carboxylate, cyclohexa1,4-diene-1-carboxylate, and cyclohexa-1,5-diene1-carboxylate (Egland et al., 1997). a. The Natural Source of Reductant for Benzoyl-CoA Reduction The Rps. palustris benzoate degradation gene cluster encodes a ferredoxin (BadB) that probably functions in vivo to deliver reduced electrons to benzoyl-CoA reductase. The Km of the reduced T. aromatica ferredoxin for benzoyl-CoA reductase is about 10 µM and the ring reduction rate with the ferredoxin is about three times higher than with artificial electron donors (Boll and Fuchs, 1998). The T. aromatica ferredoxin has 2 [4Fe-4S] centers and a midpoint redox potential for one of the iron-sulfur clusters that is unusually low (–587 mV) (Boll et al., 2000b). Sequence comparisons suggest that the Rps. palustris BadB ferredoxin is likely to have similar properties. A 2-oxoglutarate:ferredoxin oxidoreductase has been identified as the enzyme that reduces the T. aromatica ferredoxin (Dörner and Boll, 2002). It catalyses the conversion of 2-oxoglutarate + oxidized ferredoxin → succinyl-CoA + CO2 + reduced ferredoxin. The korAB genes, which lie adjacent to the
Caroline S. Harwood T. aromatica benzoate degradation cluster, encode 2-oxoglutarate: ferredoxin oxidoreductase (Fig. 4). The enzyme is induced about 50-fold by growth on benzoate. Rps. palustris strains CGA009 encodes a threesubunit putative 2-oxoglutarate: ferredoxin oxidoreductase (RPA1226-1228), which could play a similar role in benzoate degradation. However, these genes are not in close proximity to the benzoate degradation gene cluster and their synthesis is not induced by growth on benzoate (Harwood, unpublished; Larimer et al., 2004). By contrast Rhodopseudomonas strains BisA53 and BisB5 each have 2-oxoglutarate: ferredoxin oxidoreductase genes (RPE0600-0602 and RPD1530-1532) located within their benzoate degradation gene clusters. The predicted encoded proteins are about 60% identical those of the T. aromatica 2-oxoglutarate:ferredoxin oxidoreductase. 2. The β-oxidation-like Enzymes Once benzoyl-CoA is reduced to a cyclohexadienecarboxyl-CoA intermediate, the Rps. palustris and T. aromatica pathways diverge, although each employs a series of β-oxidation–like enzymes that results in the cleavage of an alicyclic ring, to form pimelyl-CoA in the case of Rps. palustris and 3-hydroxypimelyl-CoA in the case of T. aromatica (Fig. 3). Radiolabeled cyclohex-1-ene-1-carboxylate was used with Rps. palustris to demonstrate the formation of cyclohex-1-ene-1-carboxyl-CoA and its conversion to 2-ketocyclohexanecarboxyl-CoA and pimelyl-CoA. Enzymatic activities for the conversion of cyclohex-1-ene-1-carboxyl-CoA to pimelyl-CoA were demonstrated in cell extracts of benzoate grown cells using substrates that were synthesized chemically (Perrotta and Harwood, 1994). Demonstration of the T. aromatica variant of the benzoate degradation pathway was much more laborious, because the pathway intermediates could not be prepared synthetically and instead had to be generated enzymatically. The identities of the T. aromatica pathway intermediates were determined directly by two-dimensional nuclear magnetic resonance techniques (Koch et al., 1993). The three enzymes required for the conversion of cyclohex-1-ene-1-carboxyl-CoA to pimelyl-CoA by Rps. palustris have been purified (Pelletier and Harwood, 1998, 2000). The properties of the enzymes are given in Table 1. A Rps. palustris enzyme for conversion of the cyclohexadienecarboxyl-CoA product of benzoyl-CoA reduction to cyclohex-1-ene-
Chapter 29
Degradation of Aromatic Compounds
585
Table 1. Properties of the Rps. palustris β-oxidation-like enzymes of anaerobic benzoate degradation Enzyme
Gene name
Enzyme family
Native MW
Km
Specific activity (µmol– 1 min–1mg protein)
Reference
Cylohex-1-ene-1-carboxyl- CoA hydratase
badK
enoyl-CoA hydratase
N.D.
N.D.
1.5
(Pelletier and Harwood, 2000)
2-Hydroxycyclohexane- carboxyl-CoA dehydrogenase
badH
short chain alcohol dehydrogenase
115,000 (α4)
70 µM
6.5
(Pelletier and Harwood, 2000)
N.D. 9.0 (Pelletier and Har2-Ketocyclohexane- carboxylbadI napthoate syn134,000 CoA hydrolase thase (α4) wood, 1998) N.D.= not determined. Rps. palustris does not have a detectable cyclohexa-1,5-diene–1-carboxyl-CoA hydratase activity (Laempe et al., 1998).
1-carboxyl-CoA has yet to be identified. We cannot exclude that benozyl-CoA reductase catalyzes two successive two-electron reductions to generate cyclohex-1-ene-1-carboxylate from benzoyl-CoA. The ring cleavage enzyme is second novel feature, in addition to the benzoyl-CoA reductase, of the Rps. palustris anaerobic benzoate degradation pathway. It catalyzes ring cleavage by a hydrolytic rather than thiolytic mechanism, as would be expected for a traditional fatty acid β-oxidation pathway. 2Ketocyclohexanecarboxyl-CoA hydrolase (BadI) is homologous to bacterial dihydroxynaphthoate synthases (about 50% amino acid identity with BadI), enzymes that catalyze a ring closure reaction in the biosynthesis of menaquinone, a low potential electron carrier (Sharma et al., 1992). The ring cleavage hydrolase is a member of the crotonase superfamily, a large family of mechanistically diverse enzymes that includes dehalogenases, isomerases, and hydratases (Holden et al., 2001). All form an enolate-thioester intermediate during catalysis. The sterochemical course of the BadI enzyme-catalyzed reaction has been determined (Eberhard and Gerlt, 2004). 3. Benzoate-CoA Ligases Benzoate degradation is initiated by the formation of benzoyl-CoA (Fig. 3). This reaction is easily assayed and has been examined in Rps. palustris (Geissler et al., 1988). During growth on benzoate Rps. palustris cells synthesize three CoA ligases that are active with benzoate. These include a benzoate-CoA ligase, encoded by badA, that has a high affinity for benzoate, a cyclohexanecarboxylate-CoA ligase, encoded by aliA (Küver et al., 1995), and a 4-HBA-CoA ligase
encoded by hbaA (Gibson, 1994). The properties of these enzymes are given in Table 2. Each of the three CoA ligases that are present in benzoate-grown cells of Rps. palustris can function to support growth on 100 µM and, possibly lower, concentrations of benzoate (Egland et al., 1995). An Rps. palustris badA mutant grew at almost the same rate as the wild type in media supplemented with 100 µM benzoate. Cell extracts of this mutant had low levels of benzoate-CoA ligase activity when assayed with 10 µM benzoate but wild type levels of ligase activity when assayed with 1 mM benzoate. This suggested that a ligase with a relatively low affinity for benzoate was responsible for benzoate-CoA formation by the badA mutant. A badA hbaA double mutant grew on benzoate at a slightly slower rate than wild type cells (doubling time of 25 h for the badA hbaA mutant vs. 14 h for the wild type). This indicates that 4-HBA-CoA ligase contributes to the ability of an Rps. palustris badA mutant to activate benzoate with CoA, but also suggests that a least one additional ligase that is active with benzoate must be present in cell extracts. Addition of cyclohexanecarboxylate to cell extracts to a level sufficient to saturate purified cyclohexanecarboxylate-CoA ligase dramatically reduced the amount of benzoate-CoA ligase activity exhibited by the badA hbaA double mutant, implicating cyclohexanecarboxylate-CoA ligase as a third ligase that can play a role in activating benzoate with CoA during growth of Rps. palustris cells on benzoate. This remarkable redundancy of enzyme function may reflect the fact that CoA ligases often tend to be active with structurally related substrates. It may also reflect that Rps. palustris genes for benzoate, 4-HBA and cyclohexanecarboxylate degradation are
Caroline S. Harwood
586 Table 2. Properties of purified Rps. palustris CoA ligases active with benzoate Enzyme
Gene name
Benzoate-CoA ligase
badA
Native Mol Mass (kDa) 60
4-HydroxybenzoateCoA ligase
hbaA
117
Km
Relative activityb
ATP
Specific activitya benzoate
Ben
4-HBA
Chc
∆−1
100
3
27.0
100%
1%
1%
13%
400
90
36.0
100%
55%
ND
<1%
Ben
CoASH
1 400
Cyclohexanecarboxyl- aliA 59 144 119 41 2.8 100% <1% 265% N.D. ate-CoA ligase a Specific activities are given as micromoles per minute per mg of protein. b Ben, benzoate; 4-HBA, 4-hydroxybenzoate; Chc, cyclohexanecarboxylate; ∆-1, cyclohex-1-ene-1-carboxylate. Km values are given as micromolar concentrations. The Rps. palustris benzoate-CoA and 4-hydroxybenzoate CoA ligases each have good activity with cyclohex-1,4-diene-1-carboxylate and cyclohex-2,5-diene-1-carboxylate. Data are from (Geissler et al., 1988; Gibson, 1994; Küver et al., 1995; Samanta and Harwood, 2005).
physically linked and share a degree of coordinate regulation so that expression of all three ligases is induced by growth on benzoate (Fig. 4). 4. The Lower Benzoate Pathway: Pimelate Degradation Rps. palustris forms pimelyl-CoA as the product of ring cleavage during benzoate degradation (Fig. 3). The degradation of pimelyl-CoA to three molecules of acetyl-CoA plus one CO2 can be considered the ‘lower’ benzoate pathway. A pathway for the degradation of the seven carbon dicarboxylic acid pimelate has been proposed in which pimelyl-CoA is formed and then degraded by β-oxidation to glutaryl-CoA and acetyl-CoA (Gallus and Schink, 1994; Harrison and Harwood, 2005) (Fig. 6). Rps. palustris grows well on glutarate as a carbon source and encodes a glutaryl-CoA dehydrogenase (RPA1094) that is 65% identical to the well-studied enzyme from humans (Fu et al., 2004). This enzyme is bifunctional and also has a glutaconyl-CoA decarboxylase activity. GlutarylCoA dehydrogenase and glutaconyl decarboxylase activities have been detected in Rps. palustris and are three to six fold higher in benzoate-grown cells than in acetate-grown cells (Härtel et al., 1993). A pimelyl-CoA dehydrogenase has been purified from benzoate-grown Rps. palustris cells. This enzyme is unusual in that it is a flavin-containing dehydrogenase composed of two heterologous subunits to give an active enzyme with a α2β2 configuration rather than the typical structure of α4 (Emig and Harwood, unpublished). The N-terminal amino acid sequences from the two pimelyl-CoA dehydrogenase subunits were used to identify the corresponding genes in the Rps. palustris genome. The two genes, now named
pimC and pimD are located adjacent to each other in a cluster of genes (RPA3713-3717) that has been named pim for pimelate degradation (Harrison and Harwood, 2005). This cluster of genes also encodes a predicted acyl-CoA ligase, enoyl-CoA hydratase and acyl-CoA transferase in addition to having two acyl-CoA dehydrogenase genes (Harrison and Harwood, 2005). The acyl-CoA ligase (PimA) has been purified and found to be active with fatty acids and di-carboxylic acids ranging in length from six carbons to 16 carbons. A pim gene cluster deletion mutant grew slowly with several different dicarboxylic acids, including pimelate, indicating that the pim genes have a general role in the degradation of medium chain length dicarboxylic acids, in addition to a role in benzoate degradation (Harrison and Harwood, 2005). E. Peripheral Pathways 1. 4-Hydroxybenzoate Degradation Rps. palustris degrades 4-HBA by a short pathway initiated by activation of the substrate with CoA followed by a reductive dehydroxylation of 4-hydroxybenzoyl-CoA to benzoyl-CoA (Fig. 7). The Rps. palustris 4-HBA-CoA ligase is a homodimer that has good activity with benzoate and cyclohexadienecarboxylates (Table 2) (Gibson, 1994). A Rps. palustris 4-HBA-CoA ligase (hbaA) mutant failed to grow on 4-HBA, indicating that 4-hydroxybenzoylCoA formation is an obligatory step in the degradation pathway and that the HbaA enzyme in the only ligase encoded by Rps. palustris that mediates 4-HBA degradation (Gibson, 1994). A 4-hydroxybenzoyl-CoA reductase (dehydroxyl-
Chapter 29
Degradation of Aromatic Compounds
587
Fig. 6. The β-oxidation pathway of pimelate degradation. Pimelyl-CoA is the ring cleavage intermediate formed in the pathway of benzoate degradation by Rps. palustris. The Rps. palustris Pim proteins indicated are proposed to catalyze reactions leading to the formation of glutaryl-CoA and acetyl-CoA. From Gallus and Schink (1994) and Harrison and Harwood (2005).
ating) of (αβγ)2 configuration was purified from T. aromatica (Brackmann and Fuchs, 1993) and its crystal structure has been reported (Unciuleac et al., 2004). The three genes encoding this enzyme have been identified in both T. aromatica and Rps. palustris (Breese and Fuchs, 1998; Gibson et al., 1997). The predicted α (75 kDa), β (35 kDa) and γ (17 kDa) polypeptides, encoded by hbaC, hbaD and hbaB in Rps. palustris, and hcrA, hcrB and hcrC in T. aromatica, are similar in size and sequence to molybdenum cofactor-containing enzymes of the xanthine oxidase family. Xanthine oxidase catalyzes
the hydroxylation from water of xanthine to uric acid. 4-Hydroxybenzoyl-CoA reductase is unusual in that it is sensitive to oxygen and catalyzes a reaction, the reductive dehydroxylation of a substrate, that is the reverse of the reaction carried out by most xanthine oxidase family members. 2. Cyclohexanecarboxylate Degradation Alicyclic acids occur in nature and can be derived from plant secondary products and from crude oil (Trudgill, 1986). One of the most structurally simple
Caroline S. Harwood
588
Fig. 7. The pathway of anaerobic 4-HBA degradation by Rps. palustris. From Gibson et al. (1997).
of these, cyclohexanecarboxylate, is easily degraded by many bacteria, including Rps. palustris, which converts this compound to cyclohex-1-ene-1-carboxyl-CoA, an intermediate of benzoate degradation (Fig. 3). Two enzymes accomplish this, a cyclohexanecarboxylate CoA ligase, encoded by aliA and a cyclohexanecarboxyl-CoA dehydrogenase encoded by aliB (formerly named badJ). The N-terminus of the predicted AliA protein matches the experimentally determined N-terminal sequence of purified cyclohexanecarboxylate-CoA ligase (Küver et al., 1995). The main properties of this enzyme are given in Table 2. Cyclohexanecarboxyl-CoA dehydrogenase has been purified from E. coli cells expressing a cloned aliB gene from a pT7-7 expression vector (Emig and Harwood, unpublished). The purified enzyme has a specific activity of 0.7 µmol/min/mg protein, and a Km value of 48 µM with cyclohexanecarboxyl-CoA. It is not active with cyclohex-1-ene-1-carboxyl-CoA. The AliB enzyme is a flavin-containing homotetramer as is typical of acyl-CoA dehydrogenases. A mutant with an antibiotic cassette insertion in the aliB gene does not utilize cyclohexanecarboxylate as a growth substrate but is unimpaired in growth on benzoate (Emig and Harwood, unpublished). 3. Phenylalkanecarboxylate Degradation: Cinnamate and 4-Hydroxycinnamate Phenyalkanecarboxylates with side chains ranging in length from three (3-phenylpropionate) to eight (8-phenyloctanoate) are excellent growth substrates for virtually all Rps. palustris strains. (Harwood and Gibson, 1988; Madigan and Gest, 1988; Elder et al., 1992), as are several lignin monomers in this structural group including cinnamate and 4-hydroxycinnamate (4-coumarate). The precise enzymes involved in the degradation of these compounds have not yet been identified, but the pathway involves the successive removal of acetyl-CoA units from the side chain to yield benzoate (when there are odd numbers of car-
bons in the side chain) or phenylacetate (when there are even numbers of carbons in the side chain) (Elder et al., 1992). Many Rps. palustris strains also remove the side chains from the lignin monomers caffeate and ferulate to yield protocatechuate and vanillate as products. Although most strains of Rps. palustris can degrade benzoate, anaerobic degradation of phenylacetate, protocatechuate (3,4,-hydroxybenzoate) and vanillate (3-methoxy, 4-HBA) is a strain-dependent trait (Harwood and Gibson, 1988). The available evidence suggests that each of these compounds is converted to benzoate or benzoyl-CoA prior to further degradation by the anaerobic benzoate degradation pathway. However, only the pathway for anaerobic phenylacetate degradation has been determined. 4. Anaerobic Phenylacetate Degradation Phenylacetate is formed from phenylalkanecarboxylates that have an even number of side-chain carbons (e.g., 4-phenylbutyrate, 6-phenylhexanoate), as well as from phenylalanine and flavanoids (Elder et al., 1992; Schneider et al., 1997; Winter et al., 1991). We have recently determined that Rhodopseudomonas strain BisB5 can grow anaerobically with phenylacetate and it is likely that Rvi. benzoatilyticus also has this ability (Ramana Ch et al., 2006). The anaerobic pathway for phenylacetate degradation has been determined in T. aromatica and includes activation by a phenylacetate CoA ligase, oxidation to phenylglyoxylate by a three-subunit oxidoreductase and oxidative decarboxylation to benzoyl-CoA by a second oxidoreductase comprised of five subunits (Hirsch et al., 1998; Rhee and Fuchs, 1999). The genes for phenylacetate degradation have been described from the denitrifying Azoarcus sp. strain EbN1(Rabus et al., 2005). A set of genes (RPD1507-RPD1515) that is identical in organization and encodes proteins with high amino sequence identities to those for EbN1 phenylacetate degradation is present in Rhodopseudomonas strain BisB5 next to its anaerobic
Chapter 29
Degradation of Aromatic Compounds
benzoate and 4-HBA degradation genes. A set of five genes (RPD1516–1520) adjacent to the strain BisB5 phenylacetate degradation genes is predicted to encode an ABC transport system that may be specific for phenylacetate. 5. Chlorobenzoate Degradation Several strains of Rps. palustris have been isolated that grow photoheterotrophically with 3-CBA as a sole carbon source (Kamal and Wyndham, 1990; van der Woude et al., 1994; Egland et al., 2001). Many more strains of Rps. palustris can degrade 3-CBA when it is supplied together with benzoate as a co-substrate (Oda et al., 2004). This suggests that benzoate may need to be present initially as an effector to induce enzymes for 3-CBA degradation. Some strains that were initially isolated in benzoate enrichments were able to grow with 3-CBA as a sole carbon source after long periods of incubation in media that contained only 3-CBA (Oda et al., 2001). This suggests that only one or a few mutations are required for many Rps. palustris strains to acquire the ability to grow with 3-CBA as a sole substrate. Although 3-CBA is the most easily utilized of the chlorinated benzoates, some Rps. palustris strains can also degrade 2-, 4- or 3,5 chlorobenzoates once they are adapted to growth on 3-CBA (van der Woude et al., 1994). The pathway for 3-CBA degradation by Rps. palustris strain RCB100 involves conversion to 3-chlorobenzoyl-CoA, followed by a reductive dehalogenation of 3-chlorobenzoyl-CoA to benzoyl-CoA (Egland et al., 2001). 3-CBA-CoA ligase was purified from this strain and found to have an amino acid sequence that is identical to that of the cyclohexanecarboxylate CoA ligase (AliA) from Rps. palustris CGA009 except for a single amino acid change of a serine to a threonine at position 208 of the enzyme (Samanta and Harwood, 2005). This amino acid change rendered the AliA protein from strain RCB100 10-fold more active with 3-CBA than the corresponding enzyme from CGA009, a strain that cannot grow with 3-CBA (Samanta and Harwood, 2005). The CGA009 enzyme was not sufficiently active with 3-CBA to complement an RCB100 aliA mutant for growth with this compound. Conversely, the introduction of the aliA gene from RCB100 into an aliA mutant of strain CGA009, did not enable it to grow on 3-CBA. This implies that CGA009 is lacking in more than one trait necessary for it to degrade 3-CBA.
589 VI. The Molecular Regulation of Anaerobic Aromatic Compound Degradation Whereas work with T. aromatica has emphasized the enzymatic basis for anaerobic benzoate and 4-HBA degradation, Rps. palustris has served as the organism of choice for molecular studies. This organism is easily manipulated genetically and mutants and lacZ transcriptional gene reporter fusions have been constructed that have enabled investigators to analyze the expression of genes involved in benzoate and 4-HBA degradation. More recently, Rps. palustris cells grown photoheterotrophically with benzoate and 4-coumarate have been subjected to both proteomic and transcriptomic analyses (VerBerkmoes et al., 2006; Pan et al., unpublished). All of these types of studies have revealed that the benzoate degradation gene cluster (Fig. 4) is induced during anaerobic growth with aromatic acids. Molecular methods of reverse transcription-polymerase chain reaction amplification and primer extension of total RNA isolated from benzoate-grown Rps. palustris cells have been used to define the operon structure of portions the aromatic acid degradation gene cluster in Rps. palustris (Fig 4). The badHIaliABbadK genes are organized as an operon that is referred to as the cyclohexanecarboxylate degradation operon (Pelletier and Harwood, 2000), and the badDEFGAB genes comprise the benzoylCoA reductase operon (Egland et al., 1997; Peres and Harwood, 2006). The transcriptional start sites of these operons have been determined, as have the transcriptional start sites of the hbaB and hbaA genes (Egland, 1997; Egland and Harwood, 2000); hbaB is the first gene in what is likely a hbaBCD operon. The transcriptional regulators, AadR, BadR and BadM participate in regulating the expression of bad genes in response to anaerobiosis and either benzoate or benzoyl-CoA (Dispensa et al., 1992; Egland and Harwood, 1999; Peres and Harwood, 2006). A badR mutant grows slowly on benzoate but is unimpaired in growth on cyclohex-1-ene-1-carboxylate. The BadR protein is responsible for about a five-fold induction of expression of a badE-lacZ chromosomal transcriptional fusion in cells grown with benzoate (Egland and Harwood, 1999). The predicted badR product is a member of the MarR family of transcriptional regulators, but it does not have strong sequence similarity to any well-described family member. The best-studied MarR family members, including MarR itself, regulate the expression of antibiotic resistance
590 genes (Alekshun and Levy, 1999; Alekshun et al., 2001). MarR homologs have been described, however, that regulate expression of degradation genes involved in the catabolism of homoprotocatechuate (HpcR) (Roper et al., 1993), the hydrolysis of cinnamoyl esters (CinR) (Dalrymple and Swadling, 1997), and aerobic hydroxycinnamate metabolism (HcaR) (Parke and Ornston, 2003). Most MarR regulators repress gene transcription and repression is relieved when the regulator binds an effector molecule and is released from binding to a promoter region. BadR, however, behaves as an activator as its presence is needed for expression of the badDEFG genes in response to benzoate. It is likely that BadR binds to the badDEFG promoter to activate gene expression after it binds either benzoate or benzoyl-CoA as an effector. Benzoyl-CoA is the more likely effector because several compounds, including 4-HBA, that also induce expression of the benzoyl-CoA reductase operon, are metabolized by Rps. palustris to form benzoyl-CoA, but not benzoate. AadR responds to anaerobiosis to regulate expression of benzoyl-CoA reductase independently of BadR (Egland and Harwood, 1999). AadR is a member of the Crp-Fnr superfamily of transcriptional regulators (Korner et al., 2003). It has been proposed to sense oxygen because it has conserved cysteine residues that have been shown to be essential for oxygen sensing by E. coli Fnr (Kiley and Beinert, 1998). The aadR gene is not located near the aromatic acid degradation gene cluster in the Rps. palustris chromosome and was discovered in a screen for Rps. palustris mutants defective in photoheterotrophic growth with benzoate or 4-HBA, but proficient for anaerobic growth with succinate (Dispensa et al., 1992). The promoter region of the badDEFGAB operon has an Fnr consensus binding sequence (Spiro et al., 1990) centered at –39.5 relative to its transcription start site (Egland and Harwood, 1999). A badR aadR double mutant is completely defective in growth on benzoate whereas aadR and badR single mutants can grow on benzoate, albeit slowly (Dispensa et al., 1992; Egland and Harwood, 1999). Recently, BadM was identified as a regulator that represses transcription of the benzoyl-CoA reductase operon (Peres and Harwood, 2006). BadM is a small protein that belongs to the Rrf2 family of transcriptional regulators. This was a surprising discovery because AadR and BadR had appeared to account for the full range of induced expression of the benzoyl-CoA reductase genes. The mechanism by which BadM
Caroline S. Harwood acts to control transcription remains to be elucidated. However, the finding that three transcriptional regulators act independently and apparently in concert, to allow benzoyl-CoA reductase expression only in situations when benzoate is present and oxygen is absent underscores the importance of having this ATP-requiring enzyme under tight control. HbaR controls expression of the hbaA and hbaBCD genes in response to 4-HBA (Egland and Harwood, 2000). An hbaR mutant does not grow on 4-HBA under anaerobic conditions, but it grows well with benzoate and it grows at wild-type rates aerobically on 4-HBA. HbaR is a member of the Fnr-Crp family of transcriptional regulators (Egland and Harwood, 2000), but more distantly related to E. coli Fnr than AadR (Korner et al., 2003). It does not have the conserved cysteine residues that are known in Fnr to be involved in oxygen sensing (Kiley and Beinert, 1998). An AadR mutant is unable to grow photoheterotrophically with 4-HBA (Dispensa et al., 1992). This is likely due to the fact that AadR controls anaerobic 4-HBA degradation in response to oxygen deprivation by modulating expression of HbaR. The AadR homolog Fnr activated a hbaR-lacZ fusion in E. coli (Egland and Harwood, 2000), and there is an Fnr-like binding site upstream of the hbaR regulatory gene (Egland and Harwood, 2000). These findings indicate that AadR and HbaR operate in a regulatory cascade, as do other members of the Fnr-Crp superfamily. For example, regulation of denitrification in P. aeruginosa requires the Fnr homolog Anr, which senses anaerobiosis and activates expression of the dnr gene, whose product senses nitrous oxide and activates expression of denitrification genes (Arai et al., 1997). Neither BadR or HbaR regulates the expression of the cyclohexanecarboxylate degradation operon, which is induced at least 10-fold by growth on benzoate. The Rps. palustris genome must therefore encode additional regulators, probably outside the aromatic degradation gene cluster, that control expression of these other genes. VII. Comparative Aspects The breadth and depth of the studies described in this chapter show that the scientific community’s understanding of how aromatic compounds are degraded in the absence of oxygen has advanced substantially in the last 15 years. Studies with Rps. palustris and with
Chapter 29
Degradation of Aromatic Compounds
the denitrifier T. aromatica are currently being used to guide investigations of anaerobic benzoate degradation by obligate anaerobes, including iron-reducers, sulfate-reducers and fermentative bacteria (Elshahed and McInerney, 2001; Peters et al., 2004; Wischgoll et al., 2005). The available evidence indicates that obligate anaerobes use the T. aromatica variant of the pathway for benzoyl-CoA metabolism (Fig. 3) (Peters et al., 2007). At this point it is unclear why a slightly different version of the pathway evolved in Rps. palustris. It will be of interest to determine whether all benzoate-degrading purple nonsulfur bacteria use a Rps. palustris-type pathway. If they do, then this may indicate a link between the type of pathway used and some physiological or ecological feature that is specific to purple nonsulfur phototrophs. Acknowledgments I thank Georg Fuchs for years of lively competition that stimulated the pace of research on anaerobic benzoate degradation in the author’s laboratory. I am also grateful to Jane Gibson for introducing me to Rps. palustris and especially for agreeing to enter into a collaboration that has endured even after Dr. Gibson’s official retirement from academia. Work from the laboratory of C.S. Harwood was supported by grants from the U.S. Army Research Office and the Division of Energy Biosciences, U.S. Department of Energy. References Alekshun MN and Levy SB (1999) The mar regulon: Multiple resistance to antibiotics and other toxic chemicals. Trends Microbiol 7: 410–413 Alekshun MN, Levy SB, Mealy TR, Seaton BA and Head JF (2001) The crystal structure of MarR, a regulator of multiple antibiotic resistance, at 2.3 Å resolution. Nat Struct Biol 8: 710–714 Anders H-J, Kaetzke A, Kämpfer P, Ludwig W and Fuchs G (1995) Taxonomic position of aromatic-degrading denitrifying pseudomonad strains K 172 and KB 740 and their description as new members of the genera Thaurea, as Thaurea aromatica sp. nov., and Azoarcus, as Azoarcus evansii sp. nov., respectively, members of the beta subclass of the Proteobacteria. Int J Syst Bacteriol 45: 327–333 Arai H, Kodama T and Igarashi Y (1997) Cascade regulation of the two CRP/FNR related transcriptional regulators (ANR and DNR) and the denitrification enzymes in Pseudomonas aeruginosa. Mol Microbiol 25: 1141–1148 Bell SG, Hoskins N, Xu F, Caprotti D, Rao Z and Wong LL (2006) Cytochrome P450 enzymes from the metabolically diverse
591 bacterium Rhodopseudomonas palustris. Biochem Biophys Res Commun 342: 191–196 Birch AJ and Smith H (1958) Reduction by metal-amine solutions: Application in the synthesis and determination of structure. Q Rev Chem Soc Lond 12: 17–33 Blankenship RE, Madigan MT and Bauer CE (1995) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2). Kluwer Academic Publishers, Dordrecht Boll M (2005) Key enzymes in the anaerobic aromatic metabolism catalysing Birch-like reductions. Biochim Biophys Acta 1707: 34–50 Boll M and Fuchs G (1998) Identification and characterization of the natural electron donor ferredoxin and of FAD as a possible prosthetic group of benzoyl-CoA reductase (dearomatizing), a key enzyme of anaerobic aromatic metabolism. Eur J Biochem 251: 946–954 Boll M and Fuchs G (1995) Benzoyl-coenzyme A reductase (dearomatizing), a key enzyme of anaerobic aromatic metabolism. ATP dependence of the reaction, purification and some properties of the enzyme from Thauera aromatica strain K172. Eur J Biochem 234: 921–933 Boll M, Fuchs G, Meier C, Trautwein A and Lowe DJ (2000a) EPR and Mossbauer studies of benzoyl-CoA reductase. J Biol Chem 275: 31857–31868 Boll M, Fuchs G, Tilley G, Armstrong FA and Lowe DJ (2000b) Unusual spectroscopic and electrochemical properties of the 2[4Fe-4S] ferredoxin of Thauera aromatica. Biochemistry 39: 4929–4938 Boll M, Laempe D, Eisenreich W, Bacher A, Mittelberger T, Heinze J and Fuchs G (2000c) Non-aromatic products from anoxic conversion of benzoyl-CoA with benzoyl-CoA reductase and cyclohexa-1,5-diene-1-carbonyl-CoA hydratase. J Biol Chem 275:21889–21895 Boll M, Fuchs G and Lowe DJ (2001) Single turnover EPR studies of benzoyl-CoA reductase. Biochemistry 40: 7612–7620 Brackmann R and Fuchs G (1993) Enzymes of anaerobic metabolism of phenolic compounds. 4-Hydroxybenzoyl-CoA reductase (dehydroxylating) from a denitrifying Pseudomonas species. Eur J Biochem 213: 563–571 Breese K and Fuchs G (1998) 4-Hydroxybenzoyl-CoA reductase (dehydroxylating) from the denitrifying bacterium Thauera aromatica — prosthetic groups, electron donor, and genes of a member of the molybdenum-flavin-iron-sulfur proteins. Eur J Biochem 251: 916–923 Breese K, Boll M, Alt-Morbe J, Schagger H and Fuchs G (1998) Genes coding for the benzoyl-CoA pathway of anaerobic aromatic metabolism in the bacterium Thauera aromatica. Eur J Biochem 256: 148–154 Buckel W and Keese R (1995) One electron redox reactions of CoASH esters in anaerobic bacteria — a mechanistic proposal. Angew Chem (Int. Ed.) 34: 1502–1506 Clarke F M and Fina LR (1952) The anaerobic decomposition of benzoic acids during methane fermentation. Arch Biochem Biophys 36: 26–32 Dagley S (1986) Biochemistry of aromatic hydrocarbon degradation in Pseudomonads. In: Sokatch JR and Ornston N (eds), The Bacteria (The Biology of Pseudomonas, Vol. 10), pp. 527–556. Academic Press, Orlando Dalrymple, B.P., and Swadling, Y. (1997) Expression of a Butyrivibrio fibrisolvens E14 gene (cinB) encoding an enzyme with cinnamoyl ester hydrolase activity is negatively regulated
592 by the product of an adjacent gene (cinR). Microbiology 143: 1203–1210 Dispensa M, Thomas CT, Kim M-K, Perrotta, JA, Gibson J and Harwood CS (1992) Anaerobic growth of Rhodopseudomonas palustris on 4-hydroxybenzoate is dependent on AadR, a member of the cyclic AMP receptor protein family of transcriptional regulators. J Bacteriol 174: 5803–5813 Dutton PL and Evans WC (1969) The metabolism of aromatic compounds by Rhodopseudomonas palustris. Biochem J 113: 525–536 Dutton PL and Evans WC (1978) The metabolism of aromatic compounds by the Rhodospirillaceae. In: Clayton RK and Sistrom WR (eds)The Photosynthetic Bacteria, pp. 719–726. Plenum, New York Dörner E and Boll M (2002) Properties of 2-oxoglutarate:ferredoxin oxidoreductase from Thauera aromatica and its role in enzymatic reduction of the aromatic ring. J Bacteriol 184: 3975–3983 Ebenau-Jehle C, Boll M and Fuchs G (2003) 2-Oxoglutarate: NADP+ oxidoreductase in Azoarcus evansii: Properties and function in electron transfer reactions in aromatic ring reduction. J Bacteriol 185: 6119–6129 Eberhard ED and Gerlt JA (2004) Evolution of function in the crotonase superfamily: The stereochemical course of the reaction catalyzed by 2-ketocyclohexanecarboxyl-CoA hydrolase. J Am Chem Soc 126: 7188–7189 Egland PG (1997) The molecular basis of anaerobic benzoate degradation. PhD Dissertation. University of Iowa, Iowa City Egland PG and Harwood CS (1999) BadR, a new MarR family member, regulates anaerobic benzoate degradation by Rhodopseudomonas palustris in concert with AadR, an Fnr family member. J Bacteriol 181: 2102–2109 Egland PG and Harwood CS (2000) HbaR, a 4-hydroxybenzoate sensor and FNR-CRP superfamily member, regulates anaerobic 4-hydroxybenzoate degradation by Rhodopseudomonas palustris. J Bacteriol 182: 100–106 Egland PG, Gibson J and Harwood CS (1995) Benzoate-coenzyme A ligase, encoded by badA, is one of three ligases able to catalyze benzoyl-coenzyme A formation during anaerobic growth of Rhodopseudomonas palustris on benzoate. J Bacteriol 177: 6545–6551 Egland PG, Pelletier DA, Dispensa M, Gibson J and Harwood CS (1997) A cluster of bacterial genes for anaerobic benzene ring biodegradation. Proc Natl Acad Sci USA 94: 6484–6489 Egland PG, Gibson J and Harwood CS (2001) Reductive, coenzyme A-mediated pathway for 3-chlorobenzoate degradation in the phototrophic bacterium Rhodopseudomonas palustris. Appl Environ Microbiol 67: 1396–1399 Elder DJ, Morgan P and Kelly DJ (1992) Anaerobic degradation of trans-cinnamate and omega-phenylalkane carboxylic acids by the photosynthetic bacterium Rhodopseudomonas palustris: Evidence for a beta-oxidation mechanism. Arch Microbiol 157: 148–154 Elshahed MS and McInerney MJ (2001) Benzoate fermentation by the anaerobic bacterium Syntrophus aciditrophicus in the absence of hydrogen-using microorganisms. Appl Environ Microbiol 67: 5520–5525 Fina LR and Fiskin AM (1960) The anaerobic decomposition of benzoic acid during methane fermentation. II. Fate of carbon atoms one and seven. Arch Biochem Biophys 91: 163–165 Fisher K and Newton WE (2002) Nitrogen fixation—a general
Caroline S. Harwood overview. In: Leigh GJ (ed) Nitrogen Fixation in the Millennium, pp 1–34, Elsevier, Amsterdam Fu Z, Wang M, Paschke R, Rao KS, Frerman FE and Kim J-J P (2004) Crystal structures of human glutaryl-CoA dehydrogenase with and without an alternate substrate: Structural bases of dehydrogenation and decarboxylation reactions. Biochemistry 43: 9674–9684 Gallus C and Schink B (1994) Anaerobic degradation of pimelate by newly isolated denitrifying bacteria. Microbiology 140: 409–416 Geissler JF, Harwood CS and Gibson J (1988) Purification and properties of benzoate-coenzyme A ligase, a Rhodopseudomonas palustris enzyme involved in the anaerobic degradation of benzoate. J Bacteriol 170: 1709–1714 Gibson J and Harwood CS (1995) Degradation of aromatic compounds by nonsulfur purple bacteria. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2) , pp 991–1003. Kluwer Academic Publishers, Dordrecht Gibson J and Harwood CS (2002) Metabolic diversity in aromatic compound utilization by anaerobic microbes. Ann Rev Microbiol 56: 345–369 Gibson J, Dispensa M, Fogg GC, Evans DT and Harwood CS (1994) 4-Hydroxybenzoate-coenzyme A ligase from Rhodopseudomonas palustris: Purification, gene sequence, and role in anaerobic degradation. J Bacteriol 176: 634–641 Gibson J, Dispensa M and Harwood CS (1997) 4-Hydroxybenzoyl coenzyme A reductase (dehydroxylating) is required for anaerobic degradation of 4-hydroxybenzoate by Rhodopseudomonas palustris and shares features with molybdenum-containing hydroxylases. J Bacteriol 179: 634–642 Gibson KJ and Gibson J (1992) Potential early intermediates in anaerobic benzoate degradation by Rhodopseudomonas palustris. Appl Environ Microbiol 58: 696–698 Guyer M and Hegeman G (1969) Evidence for a reductive pathway for the anaerobic metabolism of benzoate. J Bacteriol 99: 906–907 Harrison FH and Harwood CS (2005) The pimFABCDE operon from Rhodopseudomonas palustris mediates dicarboxylic acid degradation and participates in anaerobic benzoate degradation. Microbiology 151: 727–736 Harwood CS and Gibson J (1988) Anaerobic and aerobic metabolism of diverse aromatic compounds by the photosynthetic bacterium Rhodopseudomonas palustris. Appl and Environ Microbiol 54: 712–717 Harwood CS, Burchhardt G, Herrmann H and Fuchs G (1999) Anaerobic metabolism of aromatic compounds via the benzoylCoA pathway. FEMS Microbiol Rev 22: 439–458 Hegeman GD (1967) The metabolism of p-hydroxybenzoate by Rhodopseudomonas palustris and its regulation. Arch Microbiol 59: 143–148 Heider J and Fuchs G (1997a) Microbial anaerobic aromatic metabolism. Anaerobe 3: 1–22 Heider J and Fuchs G (1997b) Anaerobic metabolism of aromatic compounds. Eur J Biochem 243: 577–596 Hirsch W, Schagger H and Fuchs G (1998) Phenylglyoxylate: NAD(+) oxidoreductase (CoA benzoylating), a new enzyme of anaerobic phenylalanine metabolism in the denitrifying bacterium Azoarcus evansii. Eur J Biochem 251: 907–915 Holden HM, Benning MM, Haller T and Gerlt JA (2001) The crotonase superfamily: Divergently related enzymes that catalyze
Chapter 29
Degradation of Aromatic Compounds
different reactions involving acyl coenzyme A thioesters. Acc Chem Res 34: 145–157 Härtel U, Eckel E, Koch J, Fuchs G, Linder D and Buckel W (1993) Purification of glutaryl-CoA dehydrogenase from Pseudomonas sp., an enzyme involved in the anaerobic degradation of benzoate. Arch Microbiol 159: 712–717 Imhoff JF (2001) Transfer of Rhodopseudomonas acidophila to the new genus Rhodoblastus as Rhodoblastus acidophilus gen. nov., comb. nov. Int J Syst Evol Microbiol 51: 1863–1866 Imhoff JF, Petri R and Suling J (1998) Reclassification of species of the spiral-shaped phototrophic purple non-sulfur bacteria of the alpha-Proteobacteria: description of the new genera Phaeospirillum gen. nov., Rhodovibrio gen. nov., Rhodothalassium gen. nov. and Roseospira gen. nov. as well as transfer of Rhodospirillum fulvum to Phaeospirillum fulvum comb. nov., of Rhodospirillum molischianum to Phaeospirillum molischianum comb. nov., of Rhodospirillum salinarum to Rhodovibrio salexigens. Int J Syst Bacteriol 48: 793–798 Kamal VS and Wyndham RC (1990) Anaerobic phototrophic metabolism of 3-chlorobenzoate by Rhodopseudomonas palustris WS17. Appl Environ Microbiol 56: 3871–3873 Kiley PJ and Beinert H (1998) Oxygen sensing by the global regulator, FNR: The role of the iron-sulfur cluster. FEMS Microbiol Rev 22: 341–352 Koch J, Eisenreich W, Bacher A and Fuchs G (1993) Products of enzymatic reduction of benzoyl-CoA, a key reaction in anaerobic aromatic metabolism. Eur J Biochem 221: 649–661 Korner H, Sofia HJ and Zumft WG (2003) Phylogeny of the bacterial superfamily of Crp-Fnr transcription regulators: Exploiting the metabolic spectrum by controlling alternative gene programs. FEMS Microbiol Rev 27: 559–592 Küver J, Xue JY and Gibson J (1995) Metabolism of cyclohexane carboxylic acid by the photosynthetic bacterium Rhodopseudomonas palustris. Arch Microbiol 164: 337–345 Laempe D, EisenreichW, Bacher A and Fuchs G (1998) Cyclohexa-1,5-diene-1-carbonyl-CoA hydratase [corrected], an enzyme involved in anaerobic metabolism of benzoyl-CoA in the denitrifying bacterium Thauera aromatica. Eur J Biochem 255: 618–627 [published erratum appears in Eur J Biochem (1998) 257:528] Larimer FW, Chain P, Hauser L, Lamerdin J, Malfatti S, Do L, Land ML, Pelletier DA, Beatty JT, Lang AS, Tabita FR, Gibson JL, Hanson TE, Bobst C, Torres JL, Peres C, Harrison FH, Gibson J and Harwood CS (2004) Complete genome sequence of the metabolically versatile photosynthetic bacterium Rhodopseudomonas palustris. Nat Biotechnol 22: 55–61 Madigan MT and Gest H (1988) Selective enrichment and isolation of Rhodopseudomonas palustris using trans-cinnamic acid as sole carbon source. FEMS Microbiol Ecol 53: 53–58 Möbitz H and Boll M (2002) A Birch-like mechanism in enzymatic benzoyl-CoA reduction: A kinetic study of substrate analogues combined with an ab initio model. Biochemistry 41: 1752–1758 Möbitz H, Friedrich T and Boll M (2004) Substrate binding and reduction of benzoyl-CoA reductase: Evidence for nucleotide-dependent conformational changes. Biochemistry 43: 1376–1385 Nogales J, Macchi R, Franchi F, Barzaghi D, Fernandez C, Garcia JL, Bertoni G and Diaz E (2007) Characterization of the last step of the aerobic phenylacetic acid degradation pathway. Microbiology 153: 357–365
593 Noh U, Heck S, Giffhorn F and Kohring GW (2002) Phototrophic transformation of phenol to 4-hydroxyphenylacetate by Rhodopseudomonas palustris. Appl Microbiol Biotechnol 58: 830–835 Oda Y, de Vries YP, Forney LJ and Gottschal JC (2001) Acquisition of the ability for Rhodopseudomonas palustris to degrade chlorinated benzoic acids as the sole carbon source. FEMS Microbiol Lett 38: 133–139 Oda Y, Meijer WG, Gibson JL, Gottschal JC and Forney LJ (2004) Analysis of diversity among 3-chlorobenzoate-degrading strains of Rhodopseudomonas palustris. Microbial Ecology 47: 68–79 Parke D and Ornston LN (2003) Hydroxycinnamate (hca) catabolic genes from Acinetobacter sp. strain ADP1 are repressed by HcaR and are induced by hydroxycinnamoyl-coenzyme A thioesters. Appl Environ Microbiol 69: 5398–5409 Pelletier DA and Harwood CS (1998) 2-Ketocyclohexanecarboxyl coenzyme A hydrolase, the ring cleavage enzyme required for anaerobic benzoate degradation by Rhodopseudomonas palustris. J Bacteriol 180: 2330–2336 Pelletier DA and Harwood CS (2000) 2-Hydroxycyclohexanecarboxyl coenzyme A dehydrogenase, an enzyme characteristic of the anaerobic benzoate degradation pathway used by Rhodopseudomonas palustris. J Bacteriol 182: 2753–2760 Peres CM and Harwood CS (2006) BadM is a transcriptional repressor and one of three regulators that control benzoyl coenzyme A reductase gene expression in Rhodopseudomonas palustris. J Bacteriol 188: 8662–8665 Perrotta JA and Harwood CS (1994) Anaerobic metabolism of cyclohex-1-ene-1-carboxylate, a proposed intermediate of benzoate degradation, by Rhodopseudomonas palustris. Appl Environ Microbiol 60: 1775–1782 Peters F, Rother, M and Boll M (2004) Selenocysteine-containing proteins in anaerobic benzoate metabolism of Desulfococcus multivorans. J Bacteriol 186: 2156–2163 Peters F, Shinoda Y, McInerney MJ and Boll M (2007) Cyclohexa-1,5-diene-1-carbonyl-coenzyme A (CoA) hydratases of Geobacter metallireducens and Syntrophus aciditrophicus: Evidence for a common benzoyl-CoA degradation pathway in facultative and strict anaerobes. J Bacteriol 189: 1055–1060 Pfenning N (1978) Rhodocyclus purpureus gen. nov. and sp. nov., a ring-shaped, vitamin B12-requiring member of the family Rhodospirillaceae. Int J Syst Bacteriol 28: 283–288 Pfennig N, Eimhjellen KE and Liaaen Jensen S (1965) A new isolate of the Rhodospirillum fulvum group and its photosynthetic pigments. Arch Mikrobiol 51: 258–266 Rabus R, Kube M, Heider J, Beck A, Heitmann K, Widdel F, and Reinhardt R (2005) The genome sequence of an anaerobic aromatic-degrading denitrifying bacterium, strain EbN1. Arch Microbiol 183: 27–36 Ramana Ch V, Sasikala C, Arunasri K, Anil Kumar P, Srinivas TN, Shivaji S, Gupta P, Suling J and Imhoff JF (2006) Rubrivivax benzoatilyticus sp. nov., an aromatic, hydrocarbon-degrading purple betaproteobacterium. Int J Syst Evol Microbiol 56: 2157–2164 Rhee SK and Fuchs G (1999) Phenylacetyl-CoA:acceptor oxidoreductase, a membrane-bound molybdenum-iron-sulfur enzyme involved in anaerobic metabolism of phenylalanine in the denitrifying bacterium Thauera aromatica. Eur J Biochem 262: 507–515 Roper DI, Fawcett T and Cooper RA (1993) The Escherichia coli
594 C homoprotocatechuate degradative operon: hpc gene order, direction of transcription and control of expression. Mol. Gen. Genet. 237: 241–250 Samanta SK and Harwood CS (2005) Use of the Rhodopseudomonas palustris genome sequence to identify a single amino acid that contributes to the activity of a coenzyme A ligase with chlorinated substrates. Mol Microbiol 55: 1151–1159 Schneider S, Mohamed ME and Fuchs G (1997) Anaerobic metabolism of L-phenylalanine via benzoyl-CoA in the denitrifying bacterium Thauera aromatica. Arch Microbiol 168: 310–320 Sharma V, Suvarna K, Meganathan R and Hudspeth MES (1992) Menaquinone (Vitamin K2) biosynthesis: Nucleotide sequence and expression of the menB gene from Escherichia coli. J. Bacteriol. 174: 5057–5062 Spiro S, Gaston KL, Bell AI, Roberts RE, Busby SJW and Guest JR (1990) Interconversion of the DNA-binding specificities of two related transcriptional regulators, CRP and FNR. Mol Microbiol 4: 1831–1838 Tavin D and Buswell AM (1934) The methane fermentation of organic acids and carbohydrates. J Am Chem Soc 56: 1751–1755 Trudgill PW (1986) Terpenoid metabolism by Pseudomonas. In: Sokatch JR and Ornston N (eds), The Bacteria (Vol. 10, The Biology of Pseudomonas), pp 483–525. Academic Press, New York Unciuleac M and Boll M (2001) Mechanism of ATP-driven electron transfer catalyzed by the benzene ring-reducing enzyme benzoyl-CoA reductase. Proc Natl Acad Sci USA 98: 13619–13624 Unciuleac M, Warkentin E, Page CC, Boll M, and Ermler U (2004) Structure of a xanthine oxidase-related 4-hydroxybenzoyl-CoA reductase with an additional [4Fe-4S] cluster and an inverted electron flow. Structure 12: 2249–2256 van der Woude BJ, de Boer M, van der Put NM, van der Geld FM, Prins RA and Gottschal JC (1994) Anaerobic degradation
Caroline S. Harwood of halogenated benzoic acids by photoheterotrphic bacteria. FEMS Microbiol Lett 119: 199–208 VerBerkmoes NC, Shah MB, Lankford PK, Pelletier DA, Strader MB, Tabb DL, McDonald WH, Barton JW, Hurst GB, Hauser L, Davison BH, Beatty JT, Harwood CS, Tabita FR, Hettich RL and Larimer FW (2006) Determination and comparison of the baseline proteomes of the versatile microbe Rhodopseudomonas palustris under its major metabolic states. J Proteome Res 5: 287–298 Whittle PJ, Lunt DO and Evans WC (1976.) Anaerobic photometabolism of aromatic compounds by Rhodopseudomonas sp. Biochem Soc Trans 4: 490–491 Winter J, Popoff MR, Grimont P and Bokkenheuser VD (1991) Clostridium orbiscindens sp. nov., a human intestinal bacterium capable of cleaving the flavonoid C-ring. Int J Syst Bacteriol 41: 355–357 Wischgoll S, Heintz D, Peters F, Erxleben A, Sarnighausen E, Reski R, Van Dorsselaer, A and Boll, M (2005) Gene clusters involved in anaerobic benzoate degradation of Geobacter metallireducens. Mol Microbiol 58: 1238–1252 Wright GE and Madigan MT (1991) Photocatabolism of aromatic compounds by the phototrophic purple bacterium Rhodomicrobium vannielii. Appl Environ Microbiol 57: 2069–2073 Yamanaka K, Moriyama M, Minoshima R and Tsuyuki Y (1983) Isolation and Characterization of a methanol-utilizing phototropic bacterium, Rhodopseudomonas acidophila M402 and its growth on vanillin derivatives. Agric and Biol Chem 47: 1257–1267 Zaar A, Gescher J, Eisenreich W, Bacher A and Fuchs G (2004) New enzymes involved in aerobic benzoate metabolism in Azoarcus evansii. Mol Microbiol 54: 223–238 Zengler K, Heider J, Rossello-Mora R and Widdel F (1999) Phototrophic utilization of toluene under anoxic conditions by a new strain of Blastochloris sulfoviridis. Arch Microbiol 172: 204–212
Chapter 30 Metabolism of Inorganic Sulfur Compounds in Purple Bacteria Johannes Sander and Christiane Dahl* Institut für Mikrobiologie und Biotechnologie, Rheinische Friedrich-Wilhelms-Universität Bonn, Meckenheimer Allee 168, D-53115 Bonn, Germany
Summary ............................................................................................................................................................... 596 I. Introduction..................................................................................................................................................... 596 II. Sulfur Oxidation Capabilities of Purple Bacteria............................................................................................. 596 A. Purple Non-Sulfur Bacteria............................................................................................................... 596 1. Phototrophic Alphaproteobacteria (Rhodospirillum rubrum and Relatives) ........................... 597 2. Betaproteobacteria (Rhodocyclus purpureus and Relatives) ................................................ 599 3. Aerobic Anoxygenic Phototrophic Bacteria ............................................................................ 599 B. Purple Sulfur Bacteria ...................................................................................................................... 600 1. Ectothiorhodospiraceae ......................................................................................................... 600 2. Chromatiaceae ....................................................................................................................... 601 III. Sulfur Oxidation Pathways ............................................................................................................................. 602 A. The Sox Multienzyme System .......................................................................................................... 602 1. Sox Systems in Purple Alpha- and Betaproteobacteria ........................................................ 605 2. Sox Systems in Purple Gammaproteobacteria ...................................................................... 605 B. Oxidation of H2S to S 0 ..................................................................................................................... 606 1. Flavocytochrome c ................................................................................................................. 606 2. Sulfide:Quinone Oxidoreductase ............................................................................................ 606 3. Sox Proteins and Sulfide Oxidation......................................................................................... 607 4. Other Proposed Sulfide-Oxidizing Enzymes ........................................................................... 607 C. Oxidation of Polysulfides .................................................................................................................. 607 D. Uptake of External Sulfur ................................................................................................................. 607 E. Sulfur Globules Formed by Purple Sulfur Bacteria........................................................................... 608 F. Oxidation of Stored Sulfur to Sulfite in Purple Sulfur Bacteria.......................................................... 609 G. Oxidation of Sulfite to Sulfate........................................................................................................... 610 1. Indirect Pathway via Adenylylsulfates ..................................................................................... 610 2. Direct Pathway ........................................................................................................................ 611 3. Remaining Questions .............................................................................................................. 611 H. Oxidation of Thiosulfate to Tetrathionate ......................................................................................... 612 IV. Sulfate Assimilation ........................................................................................................................................ 612 V. Conclusions .................................................................................................................................................... 615 Acknowledgments ................................................................................................................................................. 616 References ............................................................................................................................................................ 616
*Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 595–622. © 2009 Springer Science + Business Media B.V.
596
Johannes Sander and Christiane Dahl
Summary This chapter focuses on dissimilatory and assimilatory metabolism of inorganic sulfur compounds by the bacteriochlorophyll-containing purple bacteria. Many anoxygenic phototrophic purple bacteria use inorganic sulfur compounds (e.g., sulfide, elemental sulfur, polysulfides, thiosulfate, or sulfide) as electron donors for reductive carbon dioxide fixation during photolithoautotrophic growth. With regard to their sulfur metabolism, purple bacteria are characterized by a great variability of sulfur substrates used and pathways employed. We will therefore first give an overview about the sulfur oxidation capabilities of the various groups of purple bacteria. Comparison of genome sequence data provides additional insight into the sulfur oxidation pathways. Special attention is given to current knowledge on the biochemical details of the metabolic pathways employed. A variety of enzymes catalyzing sulfur oxidation reactions have been isolated from purple bacteria, and Allochromatium vinosum, a representative of the Chromatiaceae, has been especially well characterized also on a molecular genetic level. Comparative genomics in combination with older biochemical data results in a clear picture of sulfate assimilation and the enzymes involved in purple bacteria. I. Introduction Dissimilatory sulfur metabolism (i.e., use of sulfur compounds as sources or sinks of electrons, as opposed to assimilatory sulfur metabolism which uses sulfur compounds as biosynthetic substrates) has been most investigated in the purple sulfur bacteria of the families Chromatiaceae and Ectothiorhodospiraceae. Contrary to the misleading nomenclature even many of the purple non-sulfur bacteria are also able to use sulfur compounds as a source of electrons (Brune, 1989, 1995b; Imhoff et al., 2005). In addition to the purple bacteria, the green sulfur bacteria (Chlorobiaceae), some cyanobacteria and some members of the filamentous anoxygenic phototrophs (Chloroflexaceae) are able to grow phototrophically with reduced sulfur compounds as electron donors (Brune, 1995b; Dahl, 2008). Moreover, in the future it is expected that an increasing number of bacteria will be identified that show non-classical phototrophic growth (i.e., bacteria that harvest light energy with the help of rhodopsin-like molecules) and are able to use sulfur compounds as a source of electrons (Sabehi et al., 2005; Sander et al., 2006). In this chapter we will focus on dissimilatory and assimilatory metabolism of inorganic sulfur compounds by the bacteriochlorophyll-containing purple Abbreviations: Acd. – Acidiphilium; Alc. – Allochromatium; APS – adenosine-5´-phosphosulfate; DMSP – dimethylsulfoniopropionate; E. – Escherichia; EC – extracellular; Ect. – Ectothiorhodospira; Ers. – Ectothiorhodosinus; FccAB – flavocytochrome c; Hlr. – Halorhodospira; IC – intracellular; Pcs. – Paracoccus; Rba. – Rhodobacter; Rdv. – Rhodovulum; Rps. – Rhodopseudomonas; Rsp. – Rhodospirillum; SQR – sulfide:quinone oxidoreductase; Tca. – Thiocapsa; Trs. – Thiorhodospira; U – unknown
bacteria. With regard to their sulfur metabolism, purple bacteria are characterized by a great variability of sulfur substrates used and pathways employed. We will therefore first give an overview about the sulfur oxidation capabilities of the various groups of purple bacteria. In the second part of this chapter we will give special attention to current knowledge of the biochemical details of the metabolic pathways employed. Useful information is also gained by comparative analysis of complete purple bacterial genomes. Metabolism of organosulfur compounds in purple bacteria can be very complex. Regarding this topic we refer the reader to recent articles by Cook and coworkers (Denger et al., 2004, 2006; Baldock et al., 2007). II. Sulfur Oxidation Capabilities of Purple Bacteria A. Purple Non-Sulfur Bacteria The purple non-sulfur bacteria are an extremely heterogenic group of bacteria. Representatives are found within the Alpha- and the Betaproteobacteria (Imhoff, 2001; Imhoff et al., 2005). The species of this group vary not only with respect to their cell morphology, the structures of their intracytoplasmic membranes, the carotenoid composition and the carbon sources used but also with respect to electron donors used. Purple non-sulfur bacteria grow preferentially photoheterotrophically under anoxic conditions in the light and some species even lack the ability to grow photolithoautotrophically (Imhoff et al., 2005). For a long time purple non-sulfur bac-
Chapter 30
Sulfur Metabolism in Purple Bacteria
teria were thought to be unable to use reduced sulfur compounds as photosynthetic electron donors, since, compared with the purple sulfur bacteria, they are capable of tolerating toxic sulfur compounds such as sulfide to a much lesser extent. The first hint that purple non-sulfur bacteria are indeed capable of using sulfur compounds came from Hansen and van Gemerden (1972), who showed that Rhodospirillum (Rsp.) rubrum, Rhodopseudomonas (Rps.) palustris, Rhodobacter (Rba.) capsulatus and Rba. sphaeroides use sulfide for photolithoautotrophic growth when supplied continuously at low concentrations (0.5–2.0 mM). It is now well established that a number of purple non-sulfur bacteria are able to grow photolithoautotrophically with reduced sulfur compounds. Their metabolic capacities will be described in more detail in the following sections. The sulfur-oxidizing capabilities of the species in each family (order) are briefly discussed. As will become apparent, patterns of sulfur oxidation are rather complex. Purple nonsulfur bacteria vary considerably with respect to the intermediates and final products of sulfur compound oxidation. 1. Phototrophic Alphaproteobacteria (Rhodospirillum rubrum and Relatives) The phototrophic Alphaproteobacteria are found in three taxonomic groups. Within each group we find species able to use reduced sulfur compounds (Table 1). Within the Rhodospirillales, Rsp. rubrum is known to oxidize sulfide to sulfur of the oxidation state zero, i.e., elemental sulfur (Imhoff et al., 2005). Rhodopila globiformis is able to oxidize thiosulfate to tetrathionate, but the ability to grow photoautotrophically has never been shown (Then and Trüper, 1981). While Roseospirillum parvum exhibits anoxygenic phototrophic growth on sulfide or thiosulfate, Rhodospira trueperi cannot use thiosulfate. In both organisms, sulfur globules are formed outside the cell and cannot be further oxidized to sulfate (Glaeser and Overmann, 1999; Pfennig et al., 1997). Roseospira mediosalina is able to use sulfide as electron source (Kompantseva and Gorlenko, 1984). Within the Rhizobiales, Rhodomicrobium vannielii oxidizes sulfide to tetrathionate. This organism grows well as a photoautotroph. In sulfide-limited chemostat cultures it oxidizes sulfide entirely to tetrathionate. In batch cultures thiosulfate and elemental sulfur appear intermediarily, but since neither of these compounds
597 is used as an electron donor, they are probably side products resulting from a chemical reaction between sulfide and tetrathionate (reviewed in Brune, 1989). Rhodobium marinum grows only poorly on sulfide, but oxidizes it to sulfur and thiosulfate when grown photomixotrophically (Imhoff, 1983). In the presence of 2 mM thiosulfate Rhodobium orientis shows weak, but significant phototrophic growth. However the oxidation of thiosulfate was not confirmed! Tests with sulfide as electron donor did not show clear results (Hiraishi et al., 1995). The closely related Rhodobium marinum uses sulfide as an electron donor, whereas photoautotrophic growth does not occur with thiosulfate (Hiraishi et al., 1995). Rhodoplanes roseus and Rhodoplanes elegans use thiosulfate, but not sulfide as electron donors (Hiraishi and Ueda, 1994). Rps. palustris is very sensitive to sulfide. Nevertheless, it can oxidize sulfide (at low concentrations) and thiosulfate to sulfate: intermediates are not observed. With thiosulfate concentrations higher than 10 mM the end product of the oxidation process is tetrathionate rather than sulfate (Trüper, 1984). Rps. julia uses sulfide and sulfur as electron donors for photoautotrophic growth. The ability to grow on thiosulfate has never been investigated (Imhoff et al., 2005). Blastochloris sulfoviridis is able to tolerate sulfide and to photooxidize it completely to sulfate. During the oxidation process an unidentified intermediate at approximately the redox level of elemental sulfur (probably polysulfides) is accumulated. Thiosulfate is oxidized to sulfate without detectable intermediates. Within the phototrophic members of the Rhodobacterales the capability to use reduced sulfur compounds as electron donors appears more widespread than within the preceding two groups. Rba. capsulatus and Rba. sphaeroides oxidize sulfide to elemental sulfur, which is accumulated extracellularly (Brune, 1989). With respect to sulfur metabolism, Rba. azotoformans is most similar to Rba. sphaeroides (Hansen and Imhoff, 1985). In contrast to other species of this genus Rba. veldkampii is able to oxidize sulfide first to sulfur and then further to sulfate after sulfide depletion in batch cultures (Hansen and Imhoff, 1985). It resembles Blastochloris sulfoviridis except that Rba. veldkampii accumulates sulfur as an intermediate during sulfide oxidation. Thiosulfate is oxidized to sulfate without detectable intermediates (Neutzling et al., 1985; Brune, 1989). In this context it is interesting to note that Rba. sphaeroides ATCC17025 harbors a full sox gene cluster, suggesting that this strain,
Johannes Sander and Christiane Dahl
598
Table 1. Sulfur oxidation capabilities of phototrophic Alphaproteobacteria. Only those species are listed for which photolithotrophic growth with reduced sulfur compounds has been experimentally verified. Data from Brune (1989, 1995b) and Imhoff et al. (2005). EC, extracellular; nd, not determined Species Rhodospirillales Rhodospirillum rubrum Roseospirillum parvum Rhodopila globiformis
Substrates
End product(s)
Sulfide Sulfide, thiosulfate Thiosulfate, (sulfide nd)
Sulfur, EC Sulfur, EC Tetrathionate
Rhodospira trueperi Roseospira mediosalina
Sulfide Sulfide
Sulfur, EC
Rhizobales Rhodomicrobium vannielii
Sulfide
Tetrathionate
Rhodobium orientis Rhodobium marinum
Thiosulfate Sulfide
nd Sulfur, thiosulfate
Rhodoplanes roseus Rhodoplanes elegans Rhodopseudomonas palustris
Thiosulfate Thiosulfate Sulfide, thiosulfate
nd nd Sulfate, with more than 10 mM thiosulfate: tetrathionate
Rhodopseudomonas julia Blastochloris sulfoviridis
Sulfide, sulfur Sulfide, thiosulfate, sulfur
Sulfate
Rhodobacter capsulatus Rhodobacter sphaeroides
Sulfide Sulfide, thiosulfate
Sulfur, EC Sulfur, EC
Rhodobacter blasticus Rhodobacter veldkampii
Sulfide Sulfide, thiosulfate, sulfur Sulfur Sulfide, thiosulfate, sulfur
Sulfur Sulfate, sulfur
Sulfide, sulfur Sulfide, thiosulfate, sulfur Sulfide, thiosulfate, Sulfur Sulfide, thiosulfate, sulfur Sulfide, thiosulfate
Sulfate Sulfate
Comments, references
Autotrophic growth not shown (Then and Trüper, 1981)
Good autotrophic growth Poor autotrophic growth on sulfide, oxidizes sulfide when grown mixotrophically
Very sensitive to sulfide, no observable intermediates (Trüper, 1984)
No assimilatory sulfate reduction, sulfur as an intermediate (Neutzling and Trüper, 1982)
Rhodobacterales
Rhodobacter azotoformans Rhodovulum sulfidophilum
Rhodovulum euryhalinum Rhodovulum adriaticum Rhodovulum iodosum Rhodovulum robiginosum Rhodovulum strictum
in contrast to all the other strains of this species, is able to produce sulfate (Table 3). Nearly all known Rhodovulum species are able to oxidize sulfide and thiosulfate (Straub et al., 1999). The only known exception is Rhodovulum (Rdv.) marinum which is unable to grow photo- or chemoautotrophically (Srinivas et al., 2006). Sulfur is used by Rdv. iodosum, Rdv. robiginosum (Straub et al., 1999) and Rdv.
Sulfate, (thiosulfate)
Some strains able to use thiosulfate (Wang et al., 1993) No assimilatory sulfate reduction, (Neutzling et al., 1985) Sulfite is transiently released. Sulfite consumed only when sulfide or thiosulfate are present (Neutzling et al., 1985)
Sulfate
No assimilatory sulfate reduction, (Neutzling et al., 1985) (Straub et al., 1999)
Sulfate
(Straub et al., 1999)
Sulfate
(Hiraishi and Ueda, 1995)
sulfidophilum (Hansen and Veldkamp, 1973). The ability to use elemental sulfur has not been tested for all species of the genus (Straub et al., 1999). In rare cases extracellular sulfur is produced during sulfide oxidation (for instance some strains of Rdv. strictum (Hiraishi and Ueda, 1995) or Rdv. adriaticum (Brune, 1989). Rdv. sulfidophilum transiently releases sulfite into the culture medium during oxidation of sulfide
Chapter 30
Sulfur Metabolism in Purple Bacteria
and thiosulfate to sulfate (Neutzling et al., 1985). Surprisingly, sulfite by itself does not function as an electron donor, but is only consumed in the presence of sulfide and thiosulfate. For Rhodobaca bogoriensis Imhoff et al. (2005) indicated that sulfur is the final oxidation product of sulfide, whereas Milford et al. (2000) who described this species found no photoautotrophic growth, neither on sulfide nor on H2. 2. Betaproteobacteria (Rhodocyclus purpureus and Relatives) Phototrophic Betaproteobacteria (Rhodocyclales and Burkholderiales) have not been reported to use reduced sulfur compounds as electron donors and sulfide inhibits growth at low concentrations (Brune, 1995b; Imhoff et al., 2005). Nevertheless, sulfate can be reductively assimilated (Imhoff, 1982). Moreover, several species (Rhodocyclus purpureus, Rhodocyclus tenuis, Rubrivivax gelatinosus and Rhodoferax antarcticus) are at least able to derive electrons from molecular hydrogen (Pfennig, 1978; Imhoff et al., 2005). For Rhodoferax fermentans the ability to use hydrogen or sulfur as electron donors has never been investigated. 3. Aerobic Anoxygenic Phototrophic Bacteria The AAP bacteria are able to gain energy from their photosynthetic machinery, but do not grow solely at the expense of light energy (Chapter 3, Yurkov and Csotonyi). All currently known species of the AAP bacteria are Alphaproteobacteria, except Roseotales depolymerans which belongs to the Betaproteobacteria. AAP bacteria have been described as inhabitants of freshwater and marine microbial mats or as freefloating populations in sea water, meromictic lakes, warm and hot springs, acidic drainage waters, soil, and deep-ocean hydrothermal environments. Globally, they exist in high numbers and are important in the transformation of organic compounds (Yurkov, 2006). All AAP bacteria share aerobic chemoorganotrophy as the preferred mode of growth, the inability to use bacteriochlorophyll for anaerobic photosynthetic growth and the presence of photochemical reactions in cells only under aerobic conditions (Hiraishi and Shimada, 2001). None of the AAP bacteria are able to grow photolithotrophically with sulfur compounds as electron donors. However, the ability to oxidize inorganic sulfur compounds has been described for several representatives of this group. Examples are
599 Roseinatronobacter thiooxidans, a strictly aerobic obligately heterotrophic alkaliphile that can oxidize sulfide, thiosulfate, sulfite and elemental sulfur to sulfate in the presence of organic compounds. The notable increase in the efficiency of organic carbon utilization observed in the presence of thiosulfate suggests that the bacterium is a sulfur-oxidizing lithoheterotroph (Sorokin et al., 2000). In another study, Yurkov et al. (1994) showed that while none of six tested species oxidized sulfide, thiosulfateoxidizing activity was present in Erythromicrobium hydrolyticum, strain E4(1) and Roseococcus thiosulfatophilus, strain RB-7. Utilization of thiosulfate by both species was dependent on the presence of organic carbon substrate and aeration. The most pronounced oxidative sulfur metabolism among AAP bacteria is present in species of the genus Acidiphilium. A number of studies have demonstrated sulfur-dependent chemolithotrophy (but not photolithotrophy) and sulfur oxidation by Acidiphilium (Acd.) acidophilum (formerly Thiobacillus acidophilus) (Pronk et al., 1990; Meulenberg et al., 1992b; Hiraishi et al., 1995). In Acd. acidophilum the utilization of thiosulfate is initiated by the oxidative condensation of two molecules of thiosulfate yielding tetrathionate. This step is catalyzed by the periplasmic enzyme thiosulfate:cytochrome c oxidoreductase (Meulenberg et al., 1993). The details of the further oxidation of tetrathionate to sulfate are unclear. Meulenberg et al. (1993) obtained indications that tetrathionate oxidation takes place in the periplasm in Acd. acidophilum. Furthermore, a tetrathionate hydrolase (de Jong et al., 1997), a trithionate hydrolase (Meulenberg et al., 1992a) and a sulfite: cytochrome c oxidoreductase (de Jong et al., 2000) have been characterized from the organism. Interestingly, the phototrophic strain Roseobacter denitrificans Och 114 (Shiba, 1991) and Roseovarius (Rsv.) nubinhibens, for which the presence of bacteriochlorophyll has not been proven, but cannot be excluded (Gonzalez et al., 2003), both contain full sox-gene clusters (Table 3, see also Roseovarius sp. 217). In other organisms, sox-encoded proteins catalyze oxidation of thiosulfate to sulfate (Appia-Ayme et al., 2001; Friedrich et al., 2001, 2005). This is especially interesting because Gonzalez et al. (2003) claimed that the Roseovarius strain investigated was unable to oxidize thiosulfate. Rsv. nubinhibens is closely related to the bacteriochlorophyll-containing Rsv. tolerans (Labrenz et al., 1999). While virtually nothing is known about metabolism of inorganic
600 sulfur compounds in these organisms, it is well established that members of the Roseobacter group are able to degrade dimethylsulfoniopropionate (DMSP), an organic sulfur compound produced in abundance by marine algae. Some isolates have been reported to carry out the major DMSP transformations that have been observed in natural bacterial communities. Among these are two competing pathways for DMSP degradation and a pathway for incorporation of the sulfur moiety of DMSP into bacterial protein (Gonzalez et al., 2003; Moran et al., 2003). In contrast to sulfide and thiosulfate, DMSP and dimethyl sulfide are abundant in aerobic marine environments (Liss, 1999). B. Purple Sulfur Bacteria In contrast to purple non-sulfur bacteria, purple sulfur bacteria preferentially use reduced sulfur compounds as electron donors during photolithoautotrophic growth. These bacteria belong to the Gammaproteobacteria (order Chromatiales) and fall into two families (Ectothiorhodospiraceae and Chromatiaceae). The ability to tolerate sulfide extends to a concentration of up to 11 mM, as has been shown for Thiorhodococcus drewsii (Zaar et al., 2003). The most important difference between the two families of purple sulfur bacteria is the formation of intracellular sulfur globules within the confines of the cell wall in Chromatiaceae during growth on sulfide, polysulfides, thiosulfate or elemental sulfur, while sulfur accumulates extracellularly in members of the Ectothiorhodospiraceae (Brune, 1995b; Dahl, 2008), hence the name of the family: the Greek εκτος means ‘outside’ and θειον ‘sulfur’. An exception among the latter family is Thiorhodovibrio sibirica which deposits sulfur not only outside of the cell but also in the periplasm (Bryantseva et al., 1999). 1. Ectothiorhodospiraceae The Ectothiorhodospiraceae are halophilic and alkaliphilic bacteria. They comprise three phototrophic genera (Ectothiorhodospira, Halorhodospira, Thiorhodospira). The genus Ectothiorhodosinus (Gorlenko et al., 2004) has no standing in nomenclature. Not all bacteria grouped into the family Ectothiorhodospiraceae have a phototrophic life style. Exceptions are for example Alkalilimnicola (Yakimov et al., 2001), Alkalispirillum (Rijkenberg et al., 2001) and Arhodomonas aquaeolei (Adkins et al., 1993).
Johannes Sander and Christiane Dahl The non-phototrophic species might also be able to oxidize sulfur compounds, for example the genera Thioalkalivibrio and Thioalkalispira (Sorokin et al., 2001; Sorokin and Kuenen, 2005). On the other hand, some of the phototrophic species of the family such as Ectothiorhodospira (Ect.) haloalkaliphila and Ect. shaposhnikovii (Imhoff, 2005b) are also able to grow chemoautotrophically on sulfur compounds. When purple sulfur bacteria of the family Ectothiorhodospiraceae grow on sulfide as electron donor, extracellular sulfur globules are usually formed as intermediates (biphasic growth). Under alkaline conditions, polysulfides are stable intermediates and accordingly, polysulfides have been described as the first measurable intermediates during sulfide oxidation by Ectothiorhodospira species (Then and Trüper, 1983, 1984). Members of the genus Ectothiorhodospira are generally able to oxidize sulfide and sulfur (Table 2). Thiosulfate is not used by Ect. marismortui (Oren et al., 1989). The ability to grow on sulfite has not been determined for all species, but also appears to be widespread in the genus (Imhoff, 2005b). Members of the genus Halorhodospira oxidize sulfide to sulfur which may be oxidized further to sulfate (Table 2). Thiosulfate is only used by Halorhodospira (Hlr.) halophila (Raymond and Sistrom, 1969). Sulfur can also be used by some species (for instance: Hlr. halophila) and poorly by Hlr. neutriphila (HirschlerRea et al., 2003). The most restricted sulfur oxidation capacities are observed in Hlr. halochloris and Hlr. abdelmalekii. Both species appear to be unable to oxidize sulfur to sulfate (Then and Trüper, 1983, 1984). During growth on thiosulfate, sulfur globules are not detectable as an intermediate in all cases. In Ect. shaposhnikovii, for instance, sulfur is formed from sulfide but not with thiosulfate as electron donor (Kusche, 1985). In other cases the ability to form sulfur globules from thiosulfate is not explicitly mentioned, but also not excluded (Trüper, 1968). In this context it is interesting to note that in Ect. shaposhnikovii thiosulfate oxidation does not stimulate growth, but is accompanied by the formation of tetrathionate and, later, higher polythionates (Gogtova and Vainstein, 1981). Thiorhodospira (Trs.) sibirica uses sulfide and sulfur, but not thiosulfate as electron donors for photoautotrophic growth (Bryantseva et al., 1999). The final oxidation product is sulfate (Table 2). Ectothiorhodosinus (Ers.) mongolicum grows preferentially as a photoorganoheterotroph, but is able to grow photolithotrophically with sulfide. Thiosulfate is
Chapter 30
Sulfur Metabolism in Purple Bacteria
601
Table 2. Sulfur metabolism in purple sulfur bacteria of the family Ectothiorhodospiraceae. Data from Hirschler-Rea et al. (2003) and Imhoff (2005b). U, unknown; EC, extracellular; IC, intracellular Genus
Sulfide
Thiosulfate
Sulfur
Sulfite
End product
Sulfate assimilation
Chemolitho-trophic growth
Ectothiorhodospira Ect. mobilis Ect. vauolata Ect. shaposhnikovii Ect. marina Ect. haloalkaliphila Ect. marismortui
+ + + + + +
+ + + + + –
+ + + + + +
+ U + + U U
Sulfate Sulfate Sulfate Sulfate Sulfate Sulfate
+ – + (+) + –
– U + U + –
Halorhodospira Hlr. halophila Hlr. abdelmalekii Hlr. halochloris Hlr. neutriphila
+ + + (+)
+ – – –
– + – (+)
U U U U
Sulfate Sulfur? Sulfur Sulfate
– – + –
– – – –
Thiorhodospira Trs. sibirica
+
–
+
U
Sulfate
U
–
Ectothiorhodosinus Ers. mongolicum
+
+
_
U
Sulfate
U
U
only used in the presence of organic matter. The final oxidation product is sulfate (Gorlenko et al., 2004). 2. Chromatiaceae This family comprises an increasing number of genera. Closely related organisms may be non-phototrophic (for instance: Arsukibacterium ikkense (Schmidt et al., 2007) or members of the genus Rheinheimera (Brettar et al., 2002; Romanenko et al., 2003)). Here we focus on those members of the Chromatiaceae that grow obligately or facultatively photoautotrophically. Phototrophic growth always requires anoxic conditions in these organisms. All phototrophic members of the Chromatiaceae use sulfide and elemental sulfur as photosynthetic electron donors. Two major physiological groups can be distinguished, the versatile and the specialized species. The specialized species (e.g., Allochromatium warmingii, Isochromatium buderi and Thiospirillum jenense) typically have larger cells, need strictly anoxic conditions and are obligately phototrophic. Furthermore these species are unable to use thiosulfate, molecular hydrogen or organic compounds as electron donors (Imhoff et al., 1998). It should be noted, however, that some of the larger celled species have never been tested for their ability to use thiosulfate and sulfite, so that this metabolic capacity cannot be completely excluded. On the other hand, the versatile species, especially the small-celled ones, are
able to use several different sulfur compounds such as thiosulfate or sulfite. They photoassimilate a larger variety of organic substances. Some of them even can grow chemo(auto-/hetero-)(organo-/litho)trophically (Imhoff et al., 1998). In contrast to the large-celled Chromatiaceae most of the small-celled species are able to assimilate sulfate. The sulfur-metabolizing capacities of the different genera of the Chromatiaceae have recently been tabulated (Dahl, 2008). Those members of the Chromatiaceae that have been studied with respect to the utilization of externally added polysulfides (e.g., Allochromatium (Alc). vinosum and Thiocapsa (Tca.) roseopersicina) readily used these compounds as photosynthetic electron donors (van Gemerden, 1987; Steudel et al., 1990; Visscher et al., 1990). Utilization of polysulfides does not appear astonishing on the basis of the finding that polysulfides are found in the medium as intermediates of the oxidation of sulfide to internally stored sulfur by Alc. vinosum (Prange et al., 2004). Some organic sulfur compounds can also serve as electron donors for photosynthetic growth of Chromatiaceae: Tca. roseopersicina splits mercaptomalate to fumarate and H2S, and mercaptoproprionate to acrylate and H2S, and then uses the liberated H2S as electron donor (Visscher and Taylor, 1993). This organism furthermore oxidizes dimethyl sulfide to dimethyl sulfoxide (Visscher and van Gemerden, 1991). During fermentative dark metabolism of Chromatiaceae sulfur compounds (elemental sulfur) can serve as
602 acceptors of electrons liberated by the oxidation of stored carbon compounds (polyhydroxyalcanoic acid). III. Sulfur Oxidation Pathways As outlined above and also described in detail in the excellent reviews of Daniel C. Brune (Brune, 1989, 1995b), sulfur oxidation patterns by whole cells of purple bacteria vary considerably. Accordingly, a common pathway cannot be described. Here, we briefly describe enzymes or multienzyme systems involved in sulfur compound oxidation in phototrophic purple bacteria and summarize information on their occurrence, mainly on the basis of genome sequence data (Table 3). On a molecular genetic and biochemical level, sulfur oxidation is best characterized in the purple sulfur bacterium Alc. vinosum. The overview in Table 3 shows that purple nonsulfur and purple sulfur bacteria differ quite fundamentally in the extent to which they are equipped with genes potentially involved in the oxidation of reduced sulfur compounds: genes for the sulfide-oxidizing enzyme flavocytochrome c (FccAB) and sulfide:quinone oxidoreductase (SQR) occur in both groups. The thiosulfate-oxidizing Sox system is also found in both groups, however with an important difference. Genes encoding the hemomolybdoprotein SoxCD (Wodara et al., 1997; Friedrich et al., 2001) are not found in purple sulfur bacteria, one characteristic feature of which is the formation of intra- or extracellular sulfur deposits as an obligate intermediate during sulfide or thiosulfate oxidation. The dsr genes including dsrAB for dissimilatory sulfite reductase (Hipp et al., 1997; Dahl et al., 2005) are present only in the purple sulfur bacteria. We will now describe the two types of Sox multienzyme systems and then turn to sulfide-oxidizing enzymes. The following sections are focused on our current knowledge of the properties of sulfur deposited by purple sulfur bacteria, and the proteins involved in the further oxidation of sulfur to sulfate via sulfite. A. The Sox Multienzyme System There are several alternative ways to oxidize sulfur compounds, but the Sox pathway seems to be of major importance (Friedrich et al., 2001, 2005). The Sox complex, a periplasmic thiosulfate-oxidizing multienzyme complex was first found and characterized
Johannes Sander and Christiane Dahl in Paracoccus (Pcs.) versutus (Lu et al., 1985) and Pcs. pantotrophus (Friedrich et al., 2001; Rother et al., 2001). In Pcs. pantotrophus the Sox complex is essential for thiosulfate oxidation in vivo and catalyzes reduction of cytochrome c coupled to the oxidation of thiosulfate, sulfide, sulfite and elemental sulfur in vitro. The sox gene cluster of Pcs. pantotrophus comprises 15 genes (soxRSVWXYZABCDEFGH). SoxR is a repressor protein of the ArsR family. SoxS is a periplasmic thioredoxin and essential for the full expression of the sox gene cluster (Rother et al., 2005). SoxV is a membrane protein and SoxW is another periplasmic thioredoxin. The following seven genes (soxX to soxD) encode four periplasmic proteins (SoxXA, SoxYZ, SoxB and Sox(CD)2), which constitute the core Sox system (Fig. 1A). SoxXA is composed of two c-type cytochromes, the diheme SoxX and the monoheme SoxA. SoxYZ is free of cofactors and able to covalently bind sulfur compounds of various oxidation states (Quentmeier and Friedrich, 2001). The monomeric SoxB has been shown to interact with the SoxYZ complex (Quentmeier et al., 2003) and contains a dinuclear manganese cluster (Epel et al., 2005). The SoxB protein was proposed to function as a sulfate thiohydrolase. Sox(CD)2 is composed of the molybdoprotein SoxC and the diheme c-type cytochrome SoxD (Quentmeier et al., 2000). SoxF is a monomeric flavoprotein with sulfide dehydrogenase activity (Bardischewsky et al., 2006). The proposed mechanism for thiosulfate oxidation requires four different proteins: SoxB, SoxXA, SoxYZ and SoxCD (Friedrich et al., 2001) (Fig. 1A). SoxXA is reduced while oxidatively coupling thiosulfate to the sulfhydryl group of conserved Cys138 of SoxY. SoxB is believed to act as a sulfate thiol esterase and to be responsible for hydrolytic cleavage of a sulfate group from the bound sulfur substrate. Sox(CD)2 then oxidizes the remaining sulfane sulfur, acting as a sulfur dehydrogenase. Further action of SoxB releases a second sulfate molecule and thereby restoring the free sulfhydryl group of Cys138 of SoxYZ which can then accept another thiosulfate molecule in the next cycle. In the meantime, sox gene clusters have been found by cloning or genome sequencing approaches not only in a great number of chemotrophic but also in many phototrophic sulfur-oxidizing bacteria, including green sulfur bacteria as well as purple sulfur and non-sulfur bacteria (Friedrich et al., 2005; Hensen et al., 2006; Frigaard and Bryant, 2008) (see also Table 3).
sqr
fccAB
soxEF
soxXABYZ
soxCD
dsr
sorAB
apr
+/– RPE_3986: 3e–15 RPC_4096
–
3067 & 68: soxF 3069: soxE –
3073 to 3077
3071 & 3072
–
–
–
–
–
–
–
–
+/– RPD_1184: 7e–22 –
–
4253 to 4257
4251 & 4252
–
–
–
4465 to 4469
4463 & 4464
–
–
–
RPB_1326
–
4247 & 48: soxF 4249: soxE: RPA4459 & 60: soxF 4461: soxE 4363 & 64: soxF 4365: soxE
4369 to 4273
4367 & 4368
–
–
–
–
–
09505 to 09485
09480 & 09475
–
–
–
Roseovarius nubinhibens ISM §
–
–
ROS217_09470 & 09465 –
10426 & 10431
–
–
–
Roseobacter sp. MED193 §
–
–
–
05724 & 05729
–
–
–
Roseobacter denitrificans OCh 114 Dinoroseobacter shibae DFL_12§ Rhodobacter sphaeroides ATCC17025 §
–
–
RD1_1518 & 1519
ISM_10436 to 10456 MED193_ 05754 to 05734 1511 to 1515
1516 & 1517
–
–
–
DshiDRAFT_1066 –
– –
1341–1345 2112 to 2116
1339 & 1340 2110 & 2111
– –
– –
– –
Rsph17029DRAFT_2662
–
1337 & 1338 soxF Rsph17025DRAFT_ 2109, no soxE –
–
–
–
–
–
RSP_3562 RB2654_18453
– –
– 01700: soxF 01695: soxE
– 01660 to 01680
– 01690 & 01685
– –
– –
– –
AcryDRAFT_0907
–
–
2333 to 2337
2340 & 2339
–
2880 & 2879
–
+/– Rru_A2307: 8e–13
–
–
–
–
–
–
–
–
–
Rgel02001633 & 34 02003462 & 63
l02003466 to 70
–
–
–
Rfer_3759
–
–
–
l02003471 & 72, l02002864 & 2863, l02000559 & 558 –
–
–
Rhodobacter sphaeroides ATCC17029 § Rhodobacter sphaeroides 2.4.1 Rhodobacterales bacterium HTCC2654 § Acetobacteraceae Acidiphilium cryptum JF–5 § Rhodospirillaceae Rhodospirillum rubrum ATCC11170 Betaproteobacteria Comamonadaceae Rubrivivax gelatinosus PM1
Rhodoferax ferrireducens T118
–
–
Sulfur Metabolism in Purple Bacteria
Organism Alphaproteobacteria Bradyrhizobiaceae Rhodopseudomonas palustris BisA53 Rhodopseudomonas palustris BisB18 Rhodopseudomonas palustris BisB5 Rhodopseudomonas palustris CGA009 Rhodopseudomonas palustris HaA2 Rhodobacteraceae Roseovarius sp. 217 §
Chapter 30
Table 3. Occurrence of genes related to sulfur oxidation in purple bacteria. Three non–phototrophic species (underlined) that do not contain genes encoding reaction center polypeptides are included for comparative reasons.
603
604
Table 3. Continued. Organism
sqr
fccAB
soxEF
soxXABYZ
soxCD
dsr
sorAB
apr
Alkalilimnicola ehrlichei MLHE–1
+/–; Mlg_1494:5e–20
1673 & 1674*, 1266, & 1267*
*
–
1653 to 1666
–
–
Halorhodospira halophila SL1
+/–; Hhal_1665: 4e–18
1945 & 1946* 1162 &1163* 1330 & 1331*
*
No soxBXA 1269 (fusion of soxYZ)), 1681 & 1682 Hhal_1939, 1941, 1942, 1948 (soxAX fusion)
–
1951 to 1963
–
–
Enzyme activity
AAB86576, AAA23316
?
ABE01359, 60 & 61, ABE01369 & 70
–
U84760
–
U84759
Gammaproteobacteria Ectothiorhodospiracea
Chromatiaceae Allochromatium vinosum DSMZ 1802
§ draft genomes at the time of compilation of this table. * Relationships of the encoded proteins to FccAB are higher in all cases than that to SoxEF, however similarity to SoxEF is still significant. Genomes were analyzed by BLAST searches using the resources provided by Intgrated Microbial Genomes (DOE Joint Genomes Institute, http://img.jgi.doe.gov). Locus tags are given according to JGI nomeclature. Baits: FccAB from Allochromatium vinosum (AAA23316 and AAB86576); SQR from Rhodobacter capsulatus (CAA66112); SorAB from Starkeya novella (AAF64400 and AAF64401); Dsr proteins from Alc. vinosum (GenBank acc. no. U84760), Sox proteins from Paracoccus denitrificans (CAA55824: SoxB, CAA55829: SoxC, CAB94380: SoxY, CAB94381: SoxZ, CAA55827: SoxA, CAB94379: SoxX, CAA55826: SoxF, CAA55828: SoxE); APS reductase (AprMBA) from Alc. vinosum (GenBank acc. no. U84759)
Johannes Sander and Christiane Dahl
Chapter 30
Sulfur Metabolism in Purple Bacteria
605
Fig. 1. Sox systems in purple non-sulfur (A) and purple sulfur (B) bacteria. All reactions take place in the bacterial periplasm. The cycles show the transformation of sulfur species, as proposed to be catalyzed by Sox proteins. Cycle A reflects the model proposed by Friedrich et al. (2001, 2005) and involves SoxCD acting as a sulfur dehydrogenase. Cycle B is suggested to operate in purple sulfur bacteria lacking SoxCD (Hensen et al., 2006). The boxed names indicate Sox proteins catalyzing transformations of sulfur compounds bound to SoxYZ. SoxYZ — acting as a sulfur compound binding protein — is circled. Sulfur compounds are bound by a conserved cysteine residue in SoxY (cysteine138 in Pcs. pantotrophus SoxY (Quentmeier and Friedrich, 2001)). The sulfur atom of this cysteine residue is indicated by a capital S letter. Sulfur formed as an intermediate of thiosulfate oxidation and deposited as sulfur globules by purple sulfur bacteria is shown as Sn, with n indicating the number of sulfur atoms in chains.
It should be noted that the formation of sulfur globules as an intermediate of thiosulfate oxidation is an obligatory step in Alc. vinosum and probably also other purple sulfur bacteria (Pott and Dahl, 1998; Hensen et al., 2006) and has also been observed for at least some green sulfur bacteria (Steinmetz and Fischer, 1982; Brune, 1989). It therefore appears that the Sox system occurs in two physiologically different groups of thiosulfate-oxidizing phototrophic bacteria: those that form sulfur as an intermediate and those that perform a direct oxidation of thiosulfate without accumulation of sulfur (Hensen et al., 2006). The survey of currently sequenced purple bacterial genomes presented in Table 3 supports the presence of two subgroups of organisms containing sox gene clusters as proposed by Hensen et al. (2006): those containing genes for the potential sulfur dehydrogenase SoxCD and those lacking these genes. This conspicuous difference is well related to the intermediates formed in the organisms. A complete sox gene cluster occurs in organisms not forming deposits of elemental sulfur; these include the thiosulfate-oxidizing purple nonsulfur bacteria. The soxCD genes are not present in organisms forming sulfur deposits; these include the purple sulfur bacteria. 1. Sox Systems in Purple Alpha- and Betaproteobacteria Using reverse genetics, Berks and coworkers showed
that the Sox system in Rdv. sulfidophilum is very similar to that from Pcs. pantotrophus, and is essential for oxidation of not only thiosulfate but also sulfide (Appia-Ayme et al., 2001). The crystal structure of SoxXA of Rdv. sulfidophilum has been solved and found to be unique (Bamford et al., 2002a,b). The SoxXA active site contains a heme with unprecedented cysteine persulfide (cysteine sulfane) coordination. This unusual post-translational modification is also seen in sulfurtransferases such as rhodanese. 2. Sox Systems in Purple Gammaproteobacteria As shown in Table 3 neither the sequenced representative of the Ectothiorhodospiraceae nor Alc. vinosum contains soxCD genes while soxBXAYZ are clearly present. For Alc. vinosum it has been shown that this truncated Sox system is functional in and essential for thiosulfate oxidation, involving the formation of intracellular sulfur deposits as an obligate intermediate (Hensen et al., 2006). Additional proteins, the products of the so-called dsr (dissimilatory sulfite reductase) genes, are required for the oxidation of this intermediate to sulfite (Pott and Dahl, 1998; Dahl et al., 2005). On the basis of these results and the model suggested by Friedrich et al. (2001) for bacteria that do not store elemental sulfur (Fig. 1A), a model for thiosulfate oxidation in sulfur-storing organisms is
606 proposed (Fig. 1B). The initial oxidation and covalent binding of thiosulfate to SoxYZ is catalyzed by SoxXA and sulfate is then hydrolytically released by SoxB. Due to the lack of the ‘sulfur dehydrogenase’ SoxCD, the sulfane sulfur atom still linked to SoxY cannot be directly further oxidized in organisms like Alc. vinosum. Probably, the sulfur is instead transferred to growing sulfur globules. Such a suggestion is feasible, as the sulfur globules in Alc. vinosum and in many if not all other organisms forming intracellular sulfur deposits reside in the bacterial periplasm (Pattaragulwanit et al., 1998; Dahl and Prange, 2006), and therefore in the same cellular compartment as the Sox proteins. How the transfer of SoxY-bound sulfur to the sulfur globules is achieved is currently unclear, as the lack of the potential sulfur transferase encoded by the rhd gene immediately adjacent to soxXA in Alc. vinosum did not lead to a detectable phenotype. Possibly, other sulfur transferases present in the cells function as a backup system (Hensen et al., 2006). In members of the Ectothiorhodospiraceae such as Hlr. halophila additional steps may be necessary for formation of extracellular sulfur globules, i.e., sulfur deposited beyond the outer membrane. B. Oxidation of H2S to S 0 Several enzymes are candidates for sulfide oxidation. For example, SQR (Schütz et al., 1997), a flavocytochrome c sulfide dehydrogenase (FccAB) (Chen et al., 1994; Brune, 1995b; Kostanjevecki et al., 2000; Meyer and Cusanovich, 2003) and also the Sox system (Appia-Ayme et al., 2001). The occurrence of the respective genes in purple bacterial genomes is summarized in Table 3. From this survey and the following considerations, it appears that generalizations about the pathway of sulfide oxidation in purple bacteria are not possible. Obviously, different mechanisms are implemented in different organisms as summarized below. 1. Flavocytochrome c In a variety of sulfide-oxidizing species flavocytochrome c is present as a soluble protein in the periplasm (Brune, 1995b) or as a membrane-bound enzyme (Kostanjevecki et al., 2000). The protein typically consists of a larger flavoprotein (FccB) and a smaller hemoprotein (FccA) subunit. They are related to SoxF and SoxE, respectively. Flavocytochromes c (FccAB) show sulfide:cytochrome c oxidoreductase
Johannes Sander and Christiane Dahl activity in vitro (Davidson et al., 1985; Bosshard et al., 1986). However, the in vivo role of the protein is unclear. FccAB occurs in many purple and green sulfur bacteria but there are also many species that lack FccAB (Table 3). Moreover, an Alc. vinosum mutant deficient in FccAB exhibits sulfide oxidation rates similar to those of the wild type (Reinartz et al., 1998). 2. Sulfide:Quinone Oxidoreductase Other purple bacteria contain SQR, a membranebound flavoprotein that is able to transfer the electrons obtained from the oxidation of sulfide to quinones (Griesbeck et al., 2000). Related genes are fairly widespread within the purple bacteria (Table 3). In Rba. capsulatus SQR is absolutely essential for the oxidation of sulfide (Schütz et al., 1999). The enzyme is a peripherally membrane-bound flavoprotein with its active site located in the periplasm (Schütz et al., 1997, 1999). Interestingly, a classical N-terminal signal peptide is missing and co-transport with the products of neighboring genes was excluded by gene deletion. The 38 carboxy-terminal amino acids are necessary for the translocation, indicating a hithertho-unknown transport mechanism (Schütz et al., 1999). In Rhodoferax ferrireducens T118, a potential sqr gene appears to be the only gene present related to sulfur oxidative metabolism (Table 3). We would therefore predict that this strain is able to oxidize sulfide to sulfur as a macroscopically detectable end product and that this oxidation is dependent on SQR comparable to the situation in Rba. capsulatus. Although Alc. vinosum membranes exhibit SQR activity (Reinartz et al., 1998), our laboratory has so far neither been able to detect a sqr-related gene via Southern hybridization with heterologous probes or heterologous PCR, nor could we detect the protein with antibodies directed against the Rba. capsulatus protein (M. Reinartz and C. Dahl, unpublished). We therefore hypothesize that the enzyme from Alc. vinosum and possibly other purple sulfur bacteria has properties distinct from those of characterized SQRs. In accordance, the Hlr. halophila genome contains one only distantly related homolog of the SQR from Rba. capsulatus (Table 3). The primary product of the SQR reaction is soluble polysulfide whereas elemental sulfur does not appear to be formed in vitro (Griesbeck et al., 2002). Very probably, disulfide (or possibly a longer chain
Chapter 30
Sulfur Metabolism in Purple Bacteria
polysulfide) is the initial product of sulfide oxidation, which is released from the enzyme. Polysulfide anions of different chain lengths are in equilibrium with each other, and longer-chain polysulfides can be formed by disproportionation reactions from the initial disulfide (Steudel, 1996). When Rba. capsulatus grows with sulfide, elemental sulfur is formed as the final product. In principle, elemental sulfur can form spontaneously from polysulfides (Sn2– + H+ ↔ H-Sn– ↔ H-S– + Sn– (Steudel, 1996)). In experiments using isolated spheroplasts from Chlorobium vibrioforme and Allochromatium minutissimum, soluble polysulfides have also been detected as the product of sulfide oxidation (Blöthe and Fischer, 2000). Polysulfides were also detected as primary products of sulfide oxidation by whole cells of Alc. vinosum (Prange et al., 2004) and have been reported as intermediates of the oxidation of sulfide to extracellular sulfur by species of the purple sulfur bacterial family Ectothiorhodospiraceae (Trüper, 1978; Then and Trüper, 1983). While transient formation of polysulfide by the latter organism species has originally been attributed to chemical reaction between H2S and elemental sulfur promoted by the alkaline culture medium (Trüper, 1978), it now appears more likely that polysulfides are biochemically generated intermediates. 3. Sox Proteins and Sulfide Oxidation In Rdv. sulfidophilum, a member of the Rhodobacteraceae, the Sox enzyme system catalyzing the oxidation of thiosulfate to sulfate (see above) is also indispensable for the oxidation of sulfide in vivo (Appia-Ayme et al., 2001). The same might well be the case for other purple non-sulfur bacteria containing sox genes. However, in Alc. vinosum mutants deficient in either flavocytochrome c (Reinartz et al., 1998), sox genes or both (D. Hensen, B. Franz and C. Dahl, unpublished) sulfide oxidation proceeds with wildtype rates indicating that that SQR plays the main role in sulfide oxidation in this organism. 4. Other Proposed Sulfide-Oxidizing Enzymes In Alc. vinosum, a dissimilatory sulfite reductase (DsrAB) operating in reverse, i.e., in the direction of sulfite formation, was suggested to be involved in sulfide oxidation (Schedel et al., 1979). However, we have shown that this protein is not essential for sulfide oxidation but rather absolutely required for
607 oxidation of intracellularly stored sulfur (Pott and Dahl, 1998). This finding is consistent with the occurrence of dsr genes exclusively in sulfur-storing purple sulfur bacteria. It should be noted that cytochromes without flavin groups have also been proposed to mediate electron transfer from sulfide to the reaction center in some purple sulfur bacteria (Fischer, 1984; Brune, 1989; Leguijt, 1993). C. Oxidation of Polysulfides As outlined above, polysulfides appear to be the primary product of the oxidation of sulfide in a number of purple bacteria. Utilization of externally added polysulfides has been studied in Alc. vinosum and Tca. roseopersicina. Both readily used these compounds as photosynthetic electron donors (van Gemerden, 1987; Steudel et al., 1990; Visscher et al., 1990). It is currently unknown how polysulfides are converted into sulfur globules. Theoretically this could be a purely chemical, spontaneous process as longer polysulfides are in equilibrium with elemental sulfur (Steudel et al., 1990). However, we have shown that Alc. vinosum sulfur globules do not contain major amounts of sulfur rings but probably consist of long chains of sulfur with organic residues at one or both ends (Prange et al., 1999, 2002). Such organylsulfanes must eventually be formed by an unknown (enzymatic) mechanism. D. Uptake of External Sulfur Many purple sulfur bacteria including Alc. vinosum and also several purple non-sulfur bacteria especially those of the genus Rhodovulum (Table 1) are able to oxidize externally supplied solid, virtually insoluble elemental sulfur. Elemental sulfur is of zero valency and consists of S8 rings, traces of S7 rings which are responsible for the yellow color and varying amounts of polymeric sulfur. Cyclic, orthorhombic α-sulfur (α-S8) is the most stable form of elemental sulfur at ambient pressure and temperature (Steudel, 2000). Polymeric sulfur consists mainly of chain-like macromolecules (Steudel and Eckert, 2003). Uptake and oxidation of elemental sulfur includes binding and/or activation of the sulfur as well as transport inside of the cells. In principle, there could be either physical contact of the cells to their insoluble substrate via outer membrane proteins, or soluble low molecular weight substances could be excreted that
608 then act on the sulfur substrate distant from the cells. However, for Alc. vinosum intimate physical cell-sulfur contact is a prerequisite for uptake of elemental sulfur (Franz et al., 2007). For the chemotrophic acidophilic sulfur oxidizing species Acidithibacillus and Acidiphilium it was proposed that extracellular elemental sulfur is mobilized by thiol groups of special outer membrane proteins and transported into the periplasmic space as persulfide sulfur (Rohwerder and Sand, 2003). Indeed, there is evidence for the existence of an outer membrane protein involved in cell-sulfur adhesion in this organism (Ramírez et al., 2004). Comparable studies have not been performed on anoxygenic phototrophic bacteria. Recently, we demonstrated that Alc. vinosum uses only or at least strongly prefers the polymeric sulfur (sulfur chain) fraction of commercially available elemental sulfur and is probably unable to take up and form sulfur globules from cyclo-octasulfur (Franz et al., 2007). Evidence for the formation of intermediates like sulfide or polysulfides during uptake of elemental sulfur was not obtained. It may be speculated that ‘sulfur chains’ rather than the more stable ‘sulfur rings’ are the microbiologically preferred form of elemental sulfur also for other sulfur-oxidizing bacteria. While purple sulfur bacteria like Alc. vinosum contain dsr proteins that have been implicated in the oxidation of sulfur deposits formed by the organisms themselves (Dahl et al., 2005) it is absolutely unclear how purple non-sulfur bacteria lacking dsr genes would oxidize elemental sulfur. One likely possibility is the Sox system. The Sox proteins from Pcs. pantotrophus have been shown to oxidize elemental sulfur in vitro. However, so far, sox-negative purple bacterial mutants have not been tested for their ability to act on elemental sulfur. E. Sulfur Globules Formed by Purple Sulfur Bacteria In anoxygenic phototrophic sulfur bacteria, sulfur appears to be generally deposited outside of the cytoplasm. Green sulfur bacteria and purple sulfur bacteria of the family Ectothiorhodospiraceae form extracellular sulfur globules while the globules are located in the periplasmic space in members of the family Chromatiaceae (Pattaragulwanit et al., 1998). Despite the different site of deposition (outside or inside the confines of the cell) the sulfur appears to be of a similar speciation in the different groups of phototrophic sulfur bacteria: In situ investigations
Johannes Sander and Christiane Dahl of sulfur globules in or formed by whole cells using X-ray absorption near-edge structure (XANES) showed that the sulfur mainly consists of long sulfur chains very probably terminated by organic residues (mono-/bis-organyl polysulfanes) in members of the Chromatiaceae and Ectothiorhodospiraceae. The organic residue present at the end of the sulfur chains appears to be glutathione or very similar to glutathione (Prange et al., 2002). This finding supports earlier speculations that reduced glutathione (probably in its amidated form) could act as a carrier molecule of sulfur to and from the globules (Bartsch et al., 1996; Pott and Dahl, 1998). As in many chemotrophic sulfur-oxidizing bacteria that form intracellular sulfur globules the sulfur globules in the Chromatiaceae are enclosed by a protein envelope (Brune, 1995a; Dahl, 1999; Dahl and Prange, 2006). In Alc. vinosum this envelope is a monolayer of 2–5 nm consisting of three different hydrophobic ‘sulfur globule proteins’ (Sgps) of 10.5 kDa, 10.6 kDa (SgpA and SgpB) and 8.5 kDa (SgpC) (Brune, 1995a). The proteins are targeted to the bacterial periplasm by Sec-dependent signal peptides (Pattaragulwanit et al., 1998). In Alc. vinosum, the envelope is indispensable for the formation and deposition of intracellular sulfur (Prange et al., 2004). All three sulfur globule proteins are rich in glycine and aromatic amino acids, particularly tyrosine. The amino acid sequences contain tandem repeats typically found in cytoskeletal keratin or plant cell wall proteins suggesting that they are structural proteins rather than enzymes involved in sulfur metabolism (Brune, 1995a). A direct/covalent attachment of chains of stored sulfur to the proteins enclosing the globules is unlikely as none of the Sgp proteins sequenced so far contains cysteine residues. In Alc. vinosum SgpC appears to play an important role in sulfur globule expansion (Prange et al., 2004). SgpA and SgpB are partly but not fully competent to replace each other (Pattaragulwanit et al., 1998; Prange et al., 2004). Proteinaceous envelopes have never been reported for extracellular sulfur globules. Neither the complete genome sequence of Hlr. halophila nor those of several green sulfur bacteria (Frigaard and Bryant, 2008) contain potential sgp genes. As outlined above, the sulfur speciation in sulfur globules of anoxygenic phototrophic bacteria is nearly identical irrespective of whether it is accumulated in globules inside or outside the cells. It therefore appears that the Sgp proteins themselves are not responsible for keeping
Chapter 30
Sulfur Metabolism in Purple Bacteria
the sulfur in a certain chemical structure. Sulfur globules can also serve as an electron acceptor reserve that allows a rudimentary anaerobic respiration with sulfur. Under anoxic conditions in the absence of light purple sulfur bacteria such as Alc. vinosum can reduce stored sulfur back to sulfide (van Gemerden, 1968; Trüper, 1978). Nothing is known about the enzymatic mechanisms underlying these processes. F. Oxidation of Stored Sulfur to Sulfite in Purple Sulfur Bacteria The oxidative degradation of sulfur deposits in purple sulfur bacteria is a main subject of current research on purple bacterial sulfur metabolism but it is still not completely understood. In the case of extracellularly deposited sulfur, this process not only involves oxidation of the sulfur but must also include binding, activation and transport into the cells (see above). In Alc. vinosum several of the proteins encoded by the 15-gene dsr operon (dsrABEFHCMKLJOPNRS) are essential for the oxidation of sulfur stored in sulfur globules. A very similar gene cluster is found in Hlr. halophila (Table 3). Most of these genes are widespread not only in photo- and chemotrophic sulfur oxidizers but also in sulfate-reducing bacteria (Sander et al., 2006). The dsrAB products form the α2β2-structured cytoplasmic sulfite reductase, which is closely related to the dissimilatory sulfite reductase from sulfate reducing bacteria and archaea (Hipp et al. 1997). The prosthetic group of DsrAB is siro(heme)amide-[Fe4S4]. The dsrN-encoded protein resembles cobyrinic acid a,c diamide synthases and probably catalyzes the glutamine-dependent amidation of siroheme (Lübbe et al., 2006). We propose an involvement of DsrR in biosynthesis of siroamide on the basis of a fusion of dsrN and dsrR in the chemotrophic sulfur-oxidizing bacterial endosymbiont (Candidatus Ruthenia magnifica) of the bivalve Calyptogena magnifica (Newton et al., 2007). The products of the homologous dsrEFH genes form a soluble α2β2γ2-structured 75-kDa holoprotein (Dahl et al., 2005). DsrC is a small soluble cytoplasmic protein with a highly conserved C-terminus including two conserved cysteine residues. Proteins closely related to DsrEFH and DsrC have recently been shown to act as parts of a sulfur relay system involved in thiouridine biosynthesis at tRNA wobble positions in Escherichia (E.) coli (Ikeuchi et al., 2006; Numata et al., 2006). The dsrM-encoded protein is a membrane-bound
609 b-type cytochrome with similarities to a subunit of heterodisulfide reductases from methanogenic archaea. The cytoplasmic iron-sulfur protein DsrK exhibits relevant similarity to the catalytic subunit of heterodisulfide reductases. DsrP is another integral membrane protein. The periplasmic proteins DsrJ and DsrO are a triheme c-type cytochrome and an iron-sulfur protein, respectively. DsrK, J and O were co-purified from membranes, indicating the presence of a transmembrane electron-transporting complex consisting of DsrKMJOP (Dahl et al., 2005). Individual in-frame deletions of the dsrMKJOP genes led to the complete inability of the mutants to oxidize stored sulfur (Sander et al., 2006). DsrL is a cytoplasmic iron-sulfur flavoprotein with NADH: acceptor oxidoreductase activity (Y. Lübbe and C. Dahl, unpublished). In-frame deletion of dsrL completely inhibited the oxidation of stored sulfur (Lübbe et al., 2006). Since the proteins encoded at the dsr locus are either cytoplasmic or membrane-bound and cannot act directly on the extracytoplasmic sulfur globules, it is proposed that the sulfur is reductively activated, transported to and further oxidized in the cytoplasm. One possible model for the roles of the dsr-encoded proteins in such a scenario is shown in Fig. 2. We suggest that DsrL uses NADH as electron donor for reduction of a persulfidic compound. Such a function is feasible as the protein carries a CxxC thioredoxin motif immediately preceding the carboxy-terminal iron-sulfur cluster binding sites. DsrL could be involved in the reductive release of sulfide from a perthiolic organic carrier molecule transporting sulfur from the periplasmic sulfur globules to the cytoplasm (Dahl et al., 2005). Glutathione amide is a likely candidate for the carrier molecule: it was found mainly as a persulfide in cells of Alc. vinosum grown photoautotrophically on sulfide (Bartsch et al., 1996). The DsrL from Alc. vinosum co-purifies with sulfite reductase (Y. Lübbe and C. Dahl, unpublished). Sulfide released from the perthiol could therefore be directly passed to dsrAB-encoded sulfite reductase. On the other hand, sulfite reductase specifically interacts also with the membrane-bound Dsr proteins and DsrEFHC (Dahl et al., 2005). Electrons released from the oxidation of sulfide by sulfite reductase may therefore be fed into photosynthetic electron transport via DsrC and DsrKMJOP, which would be analogous to the pathway postulated for sulfate reducers (Pires et al., 2006), operating in the reverse direction. The function of DsrEFH remains unclear, but as it occurs
610
Johannes Sander and Christiane Dahl
Fig. 2. Schematic presentation of Dsr proteins from Allochromatium vinosum. The scheme is based on sequence analysis of the encoding genes and on biochemical information where available. The products of the dsrS and dsrR genes are not shown for clarity because biochemical information is not available and possible functions cannot be predicted on the basis of sequence homologies. Both proteins are predicted to be soluble and to reside in the cytoplasm. DsrN is also not shown as it does not participate in redox or sulfur transfer reaction but is involved in biosynthesis of siroamide. Siroamide-[Fe4S4] is a prosthetic group of sulfite reductase.
exclusively in sulfur oxidizers and shows some interaction with DsrC, it may be important for the pathway to operate in the sulfide oxidizing direction. On the other hand, evidence is increasing that DsrEFH has sulfur transferase activity just as the related TusBCD does from E. coli (Ikeuchi et al., 2006; Dahl et al., 2007; C. Dahl, F. Grimm, J. Sander, U. Selan and D. H. Shin, unpublished). G. Oxidation of Sulfite to Sulfate Two fundamentally different pathways for sulfite oxidation have been rather well characterized in a number of chemotrophic and phototrophic sulfur oxidizers (Kappler and Dahl, 2001): (a) direct oxidation by a sulfite dehydrogenase (EC 1.8.2.1) which probably contains molybdenum; and (b) indirect, AMP-dependent oxidation via the intermediate adenylylsulfate (adenosine-5´-phosphosulfate, APS). So far, there is no evidence for an occurrence of the sulfite-oxidizing form of the APS reductase pathway in Alpha- and Betaproteobacteria or in Ectothiorhodospiraceae (Table 3). A more detailed account of sulfite oxidation in purple sulfur bacteria is given in Dahl (2008).
1. Indirect Pathway via Adenylylsulfates Within the purple bacteria, the indirect oxidation of sulfite via APS currently appears to be restricted to the Chromatiaceae (Table 3). During indirect sulfite oxidation, APS is formed from sulfite and AMP by APS reductase (EC 1.8.99.2). In a second step the AMP moiety of APS is transferred either to pyrophosphate by ATP sulfurylase (ATP:sulfate adenylyltransferase, EC 2.7.7.4), or to phosphate by adenylysulfate:phosphate adenylyltransferase (APAT, formerly ADP sulfurylase (Brüser et al., 2000)), resulting in the formation of ATP or ADP, respectively. The APS pathway is also used in the reverse direction during dissimilatory and assimilatory sulfate reduction (Bick et al., 2000; Leustek et al., 2000; Matias et al., 2005). The APS reductases from sulfur oxidizers resemble the enzymes found in dissimilatory sulfate reducers (Hipp et al., 1997). APS reductases functioning in assimilatory sulfate reduction are completely different enzymes related to 3´-phosphoadenosine-5´phosphosulfate (PAPS) reductases (Bick et al., 2000; Kopriva et al., 2001). APS reductases in purple sulfur bacteria are either membrane-bound (e.g., in many Chromatiaceae) or soluble cytoplasmic enzymes (Brune, 1995b). All dissimilatory APS reductases
Chapter 30
Sulfur Metabolism in Purple Bacteria
have been characterized as heterodimers with one
α-subunit of 70–75-kDa (1 FAD) and one β-sub-
unit of 18–23 kDa (2 [4Fe-4S] centers) (Fritz et al., 2000). Additional subunits mediating membrane association may be present (Hipp et al., 1997). A catalytic mechanism has been proposed in which sulfite initially forms a complex with the flavin (Brune, 1995b, and references therein). This then reacts with AMP to yield APS, releasing two electrons that are transferred via the flavin to the iron-sulfur centers. In Alc. vinosum the genes for ATP sulfurylase (sat) and APS reductase (aprMBA, with aprM encoding a putative membrane anchor) form an operon (Hipp et al., 1997; A. Wynen, H. G. Trüper, C. Dahl unpublished, GenBank No. U84759). In the sulfate reducer Desulfovibrio vulgaris, the membrane-bound qmo complex is thought to deliver electrons from quinol to APS reductase (Pires et al., 2003). In green sulfur bacteria, APS reductase and qmo genes are clustered, indicating that the Qmo complex accepts electrons from APS reductase operating in the sulfite-oxidizing direction. However, qmo genes have not been described for purple sulfur bacteria and we currently assume that the membrane protein AprM serves an analogous function in Alc. vinosum. The best characterized ATP sulfurylase (Sat) from a sulfur-oxidizing bacterium is the enzyme from the endosymbiont of the hydrothermal vent worm Riftia pachyptila (Renosto et al., 1991; Beynon et al., 2001). The ATP sulfurylase from Alc. vinosum is isolated as a monomer with an apparent molecular mass of 45 kDa (A. Wynen, C., Dahl, H. G. Trüper, unpublished). An APAT does not appear to be present in Alc. vinosum while significant activity was found in strains of Tca. roseopersicina (Dahl and Trüper, 1989; Alguero et al., 1988). The only well characterized APAT is that of the chemotrophic sulfur oxidizer Thiobacillus denitrificans (Brüser et al., 2000). The in vivo role of APAT is especially difficult to assign because all organisms with significant APAT activity (>100 mU mg–1 in crude extracts) also contain ATP sulfurylase. It has been hypothesized that APAT may serve to ensure a high turnover of APS under pyrophosphate limiting conditions as this enzyme is independent of the energy-rich pyrophosphate molecule (Brüser et al., 2000). 2. Direct Pathway All sulfite dehydrogenases characterized to date belong to the sulfite oxidase family of molybdoenzymes
611 comprising established sulfite-oxidizing enzymes and related proteins as well as assimilatory nitrate reductases from plants (Hille, 1996). The active site is formed by a single molydopterin cofactor. Additional redox active centers may be present. The periplasmic SorAB protein from Starkeya novella is currently the best characterized bacterial sulfiteoxidizing enzyme (Kappler et al., 2000; Feng et al., 2003; Kappler and Bailey, 2005; Raitsimring et al., 2005; Doonan et al., 2006). SorAB is a heterodimer of a large MoCo-dimer domain (40.2 kDa) and a small cytochrome c subunit (8.8. kDa). The sorAB genes form an operon. Characterized related proteins also appear to be localized in the periplasm and to contain a heme c-binding subunit (Myers and Kelly, 2005). Genes closely related to sorAB do not occur in the currently available genomes of anoxygenic phototrophic purple bacteria (Table 3). Genes encoding proteins belonging into the sulfite oxidase family are present in many bacterial genomes (Kappler, 2007). While the well-characterized bacterial sulfite dehydrogenases are soluble periplasmic proteins, membrane-bound bacterial sulfite-oxidizing enzymes have also been reported in the literature (reviewed in Kappler and Dahl, 2001; Kappler 2007). Most of the established or predicted soluble members of the sulfite oxidase family are periplasmic enzymes; however, some of the proteins belonging to this group (but without a biochemically characterized function) are predicted to reside in the bacterial cytoplasm (Kappler, 2007). Direct oxidation of sulfite to sulfate in the bacterial cytoplasm can, therefore, not generally be excluded. 3. Remaining Questions The overview given in Table 3 shows that sulfite oxidation in purple bacteria is still an essentially unresolved question: The oxidative APS reductase pathway occurs exclusively in members of the Chromatiaceae. In addition, the pathway is dispensable for sulfite oxidation in Alc. vinosum (Dahl, 1996). Further experiments on Alc. vinosum indicated the involvement of a molybdoenzyme in sulfite oxidation. However, so far all attempts have failed to prove the existence of sorAB-related genes or the respective protein (U. Kappler and C. Dahl, unpublished). In addition genes homolgous to those encoding proteins of the sulfite oxidase family are not present in the genome of Hlr. halophila. Alternative means for sulfite oxidation must therefore exist. For green sulfur
612 bacteria Frigaard and Bryant (2008) presented the attractive speculation that a potential cytoplasmically located protein encoded by three genes resembling those for polysulfide reductase from Wolinella succinogenes (Krafft et al., 1992) could play this role. On the other hand, we have some indications that a soxY-deficient mutant of Alc. vinosum is severely impaired in the oxidation of sulfite (D. Hensen, B. Franz and C. Dahl, unpublished). The Sox system is also the most likely candidate for sulfite oxidation in Alpha- and Betaproteobacteria, as sulfite is accepted as a substrate in vitro for the reconstituted Sox system of Pcs. denitrificans (Friedrich et al., 2001). Clearly, the question of sulfite oxidation in purple bacteria will require special attention in the future. H. Oxidation of Thiosulfate to Tetrathionate The formation of tetrathionate from thiosulfate has been studied mainly in chemoorganotrophic bacteria that use thiosulfate as a supplemental but not as the sole energy source (Jørgensen, 1990; Podgorsek and Imhoff, 1999; Sorokin et al., 1999). The pathway occurs only in a few purple bacteria including Alc. vinosum (Smith and Lascelles, 1966; Hensen et al., 2006) and Rhodopila globiformis (Then and Trüper, 1981). In Alc. vinosum the ratio between tetrathionate and sulfate formed from thiosulfate is strongly pHdependent with more tetrathionate as the product under slightly acidic conditions (Smith, 1966). In Alc. vinosum thiosulfate dehydrogenase is a periplasmic 30-kDa monomer with an isoelectric point of 4.2. The enzyme contains heme c and is reduced by thiosulfate at pH 5.0 but not at pH 7.0. In accordance, the pH optimum of the enzyme was determined to be 4.25 (Hensen et al., 2006). The properties of Alc. vinosum thiosulfate dehydrogenase described by Hensen et al. (2006) are compatible with older data presented by Smith (1966) and Fukumori and Yamanaka (1979). In both reports a tetrathionate-forming activity with a pH optimum in the acidic range was described. We have been unable to detect the presence of any tetrathionate-forming enzyme operating at pH 8.0 in Alc. vinosum, as was claimed earlier (Knobloch et al., 1981; Schmitt et al., 1981). Thiosulfate dehydrogenases from other sources show remarkable heterogeneity with respect to structural properties and catalytic characteristics (Kusai and Yamanaka, 1973; Then and Trüper, 1981; Brune, 1989; Visser et al., 1996) which has been interpreted as indicating con-
Johannes Sander and Christiane Dahl vergent rather than divergent evolution (Visser et al., 1996). A gene sequence encoding a heme-containing thiosulfate dehydrogenase has not yet been reported. A BLAST search with the amino-terminal sequence of the enzyme from Alc. vinosum yielded only one significantly related sequence, a hypothetical c-type cytochrome from Cupriavidus (Ralstonia, Wautersia) metallidurans (Hensen et al., 2006). IV. Sulfate Assimilation Assimilatory sulfate reduction occurs in a variety of phototrophic purple bacteria, although the need for a pathway for sulfate assimilation is not obvious for those species thriving in environments rich in reduced sulfur compounds. Indeed, within the Chromatiaceae and Ectothiorhodospiraceae, the more specialized species are unable to grow photoorganotrophically and accordingly none of these species assimilates sulfate as a sulfur source (Imhoff, 2005a; Imhoff, 2005b). Even some purple non-sulfur bacteria for instance Rdv. iodosum, Rdv. robiginosum, Rdv. adriaticum and Rdv. euryhalinum are unable to assimilate sulfate and depend on reduced sulfur compounds (Neutzling et al., 1984; Kompantseva, 1985; Straub et al., 1999). Assimilatory sulfate reduction commences with the uptake of extracellular sulfate (Fig. 3). In E. coli three membrane-bound components of a periplasmic substrate-binding transport system are encoded by the contiguous genes cysTWA (Hryniewicz and Kredich, 1991). The preceding gene cysP was shown to encode a periplasmic thiosulfate binding protein distinct from the sulfate binding protein of Salmonella typhimurium. In plants as well as many lower eukaryotes, sulfate transport is mediated by both high- and low-affinity systems. Both function in active processes which are thought to be driven by proton motive force, through a H+/SO42− co-transport (Mendoza-Cózatl et al., 2005). The characterized high affinity transporters sul1 from Saccharomyces cerevisiae (Cherest et al., 1997) and hst1At from Arabidopsis thaliana (Vidmar et al., 2000) are prominent members of the major facilitator superfamily of sulfate permeases and related transporters (Pao et al., 1998; Kertesz, 2001). This family (Cluster of Orthologous Groups family COG0659, http://www. ncbi.nlm.nih.gov/COG/) also comprises a number of bacterial proteins. Once inside the cell, the low reactivity of sul-
Chapter 30
Sulfur Metabolism in Purple Bacteria
Fig. 3. Pathways of assimilatory sulfate reduction. Two alternative pathways occurring in purple bacteria are shown. Direct reduction of APS is more common in purple bacteria than reduction of PAPS (Table 4). Gene names are boxed. All indicated steps are cytoplasmic processes.
fate must first be overcome by the formation of a phosphate-sulfate anhydride bond in the compound adenosine-5´-phosphosulfate ( adenylylsulfate, APS). This reaction is catalyzed by the enzyme ATP sulfurylase (EC 2.7.7.4) (Leustek and Saito, 1999). Assimilatory ATP sulfurylases occur in two different forms: A heterodimeric CysDN type from E. coli (Leyh, 1993) and an unrelated homooligomeric type found in bacteria, plants and fungi (Foster et al., 1994; MacRae et al., 2001). The latter polypeptides are related to the sat-encoded dissimilatory ATP sulfurylases of sulfate-reducing and chemotrophic sulfur-oxidizing prokaryotes (Sperling et al., 1998; Beynon et al., 2001). In both heterodimeric and homooligomeric ATP sulfurylases, carboxy-terminal APS kinase domains may be present. For incorporation of sulfur into biomolecules, e.g., amino acids, the sulfate in APS is reduced to sulfite and finally into sulfide. This process may occur through two different pathways, depending on the organism (Fig. 3). One of them involves the phosphorylation of APS by an APS kinase (EC 2.7.1.25, gene cysC) using ATP to produce phosphoadenosine phosphosulfate (PAPS) and ADP. In the following
613 reaction, PAPS reductase (EC 1.8.99, gene cysH) generates free sulfite using reduced thioredoxin as electron donor. The other pathway involves the direct reduction of APS by APS reductase (EC 1.8.99.x, also cysH) which uses glutathione as an electron source to produce sulfite (Bick et al., 2000). Phosphorylation of APS to PAPS is also necessary in organisms that use APS as the substrate for reduction because PAPS serves as a donor molecule in the synthesis of sulfonated compounds e.g., sulfolipids. Assimilatory APS reductases and PAPS reductases are related on the amino acid sequence level, but the enzymes using APS contain additional iron-sulfur clusters and differentiation on the sequence level is possible via the presence/absence of the characteristic cluster binding cysteine residues (Kopriva et al., 2002). Sulfite is finally reduced to sulfide by an assimilatory sulfite reductase (EC 1.8.7.1). Again, different types can be differentiated: The NADPH-dependent assimilatory sulfite reductases of Enterobacteria are composed of two different subunits: a siroheme-[Fe4S4]-containing protein (CysI) and a flavoprotein (CysJ). In contrast the ferredoxin-dependent sulfite reductases from cyanobacteria, algae and higher plants are much simpler homooligomers of just a siroheme-[Fe4S4]containing protein. These enzymes are only very distantly related to their dissimilatory counterparts (Dhillon et al., 2005). Depending on the organism, there are two different ways by which sulfide is finally incorporated into a carbon backbone to produce cysteine: In bacteria like E. coli and in plants (Kredich, 1996; Hell et al., 2002), sulfide is finally incorporated in O-acetyl-L-serine by O-acetyl-L-serine-(thiol)-lyase yielding cysteine. The same enzyme also catalyzes the condensation of sulfide with O-acetylhomoserine to form homocysteine. The latter is transformed into cysteine by trans-sulfuration. During this process, homocysteine is combined with serine yielding cystathionine which is finally dissociated into cysteine, α-ketobutyrate and ammonia by cystathionine γ-lyase. The latter pathway is present in some but not all fungi (Ono et al., 1999; Mendoza-Cózatl et al., 2005). The genes related to assimilatory sulfate reduction in the currently available genomes of purple bacteria are tabulated in Table 4. A survey of the data reveals that the sulfate assimilation pathway is not uniform in purple bacteria but may vary even within the same family (see Alc. vinosum and Tca. roseopersicina). Most of the organisms listed in Table 4 encode genes related to cysPTWA, indicating that they use a sulfate
Organism Alphaproteobacteria Bradyrhizobiaceae Rhodopseudomonas palustris BisA53 Rhodopseudomonas palustris BisB18 Rhodopseudomonas palustris BisB5 Rhodopseudomonas palustris CGA009 Rhodopseudomonas palustris HaA2 Rhodobacteraceae Roseovarius sp. 217 § Roseovarius nubinhibens ISM §
Rhodospirillaceae Rhodospirillum rubrum ATCC11170 Betaproteobacteria Comamonadaceaee Rubrivivax gelatinosus PM1 Rhodoferax ferrireducens T118
sat
cysDN
cysH (APS)
cysH (PAPS)
cysJ
cysI
RPE_1758–60 + 1768–71 RPC_4018–20 + 4007–4010 RPD_1376–78 + 1156–59 RPA_0747–50
–
1767
–
–
1765
4011
–
–
4013
–
0334 & 0333 (+ APS kinase) , 1762 & 1761 (+ APS kinase) 0063 & 0064 (+ APS kinase) 4016 & 4017 (+ APS kinase) 1154 & 1153 (+ APS kinase)
1155
–
–
1381
–
0752 & 0753 (+ APS kinase)
0751
–
–
4213
RPB_1396–98 + 1045–48
–
1043 & 1042 (+ APS kinase)
1044
–
–
1401
sul1: ROS217_09535 no cysP, no sul1
11496 (+ APS kinase) ISM_11065 (+ APS kinase) 20284 (+ APS kinase) –
– –
19182 15740
– –
– –
19187 15735
– –
06644 2967
– –
– –
06639 2966
Rsph17025DRAFT_ 0200 (+ APS kinase) Rsph17029DRAFT_ 3256 (+ APS kinase) 1575 (+ APS kinase) RB2654_19213 (+ APS kinase) 1196
–
3015
–
–
3014
–
–
–
–
2931
– –
1941 04651
– –
– –
1942 04646
–
3289
–
–
3290
sul1: MED193_04017 sul1: RD1_0783 3 copies 2 copies RSP_3696–99 sul1: RB2654_01730 sul1: DshiDRAFT_ 1056
–
no cysP, no sul1
–
AcryDRAFT_1556 & 1557 (+ APS kinase)
1277
–
1276
1278
Rru_A3400–03
–
A2289 & A2290 (+ APS kinase)
A1929
–
–
A1931
Rgel02001952–56 Rfer1759–62
– –
l02001658 & l02001658 2459 & 2460
l02001659 2458
– –
– –
l02001661 2456
Johannes Sander and Christiane Dahl
Roseobacter sp. MED193 § Roseobacter denitrificans OCh 114 Rhodobacter sphaeroides ATCC17025 § Rhodobacter sphaeroides ATCC17029 § Rhodobacter sphaeroides 2.4.1 Rhodobacterales bacterium HTCC2654 § Dinoroseobacter shibae DFL_12 § Acetobacteraceae Acidiphilium cryptum JF–5 §
cysPTWA*
614
Table 4. Occurrence of genes related to sulfate assimilation in purple bacteria. Three non–phototrophic species (underlined) that do not contain genes encoding reaction center polypeptides are included for comparative reasons.
+ + + – unknown unknown
unknown
+ (dissimilatory) unknown
draft genomes at the time of compilation of this table, *, either locus tags for cysPTWA–related genes are given or the tag of a sul1–related gene (encoding a high affinity sulfate transporter of yeast belonging to COG0659) is indicated in the absence of cysPTWA, 1Neumann et al., 2000, 2Haverkamp and Schwenn, 1999.
§
+ – – +
– sul1: Hhal_1967
+
– – – 1777
– sul1: Mlg_1307, 1668
1260 & 1261, 2344 & 2345 (+ APS kinase) Hhal_2353 & 2354 (+ APS kinase)
1263
2114
–
1264
Sulfur Metabolism in Purple Bacteria
Gammaproteobacteria Ectothiorhodospiracea Alkalilimnicola ehrlichei MLHE–1 Halorhodospira halophila SL1 Chromatiaceae Allochromatium vinosum DSMZ 1801 Thiocapsa roseopersicina M12
Organism
Table 4. Continued.
cysPTWA*
sat
cysDN
cysH (APS)
cysH (PAPS)
cysJ
cysI
Chapter 30
615 transport system related to that of Enterobacteria. In other purple bacteria, including Alkalimnicola ehrlichei and Hlr. halophila related genes do not appear to be present. Instead, a gene is found encoding for a potential sulfate permease related to the high affinity sulfate transporter of Saccharomyces cerevisiae (Cherest et al., 1997). It should be pointed out, however, that this gene is located in immediate vicinity of clusters of genes involved in oxidative sulfur metabolism in Hlr. halophila, Rhodobacterales bacterium HTCC2654, and Roseovarius sp. 217 (see also Table 3). In at least some of these cases the protein may also be involved in export of sulfate produced in the cytoplasm as an end product of the oxidation of reduced sulfur compounds. The Sat-type of ATP sulfurylase appears to be present in Rhodobacteraceae while all other analyzed genomes contain cysDN genes. In all organisms except Alc. vinosum, cysN is fused with an APS kinase gene. Older analyses by Imhoff claiming that phototrophic Alphaproteobacteria preferentially use the PAPS pathway (Imhoff et al., 1983) and that Betaproteobacteria in contrast favor the APS pathway (Imhoff, 1982) are not congruent with our comparative analysis of genome sequence data. Our analysis indicates that reduction of APS is far more widespread among purple bacteria than reduction of PAPS: Of all organisms analyzed only the nonphototrophic Alkalimnicola ehrlichii contains a bona fide PAPS reductase gene albeit in addition to a gene encoding an assimilatory APS reductase. For Tca. roseopersicina the presence of a PAPS reductase gene and a product dedicated to the reduction of PAPS has been shown biochemically (Haverkamp and Schwenn, 1999). It should, however, be kept in mind that the strains investigated by Imhoff are not identical with those for which the genome sequences are now available. The data presented in Table 4 indicates that the ferredoxin-dependent sulfite reductase is more common in purple bacteria than the NADPH-dependent enzyme consisting of two different polypeptides. Tca. roseopersicina is currently the only purple bacterium for which the existence of a cysIJ-encoded sulfite reductase has been firmly established (Haverkamp and Schwenn, 1999). V. Conclusions Purple bacteria exhibit a great variability of sulfur
616 substrates used and sulfur oxidation pathways employed. In combination with available biochemical information the survey of genome sequence data presented here points at an especially important role of sulfide:quinone oxidoreductase for the oxidation of sulfide. Flavocytochrome c appears less widespread. At least in some alphaproteobacterial purple nonsulfur bacteria (among them Rdv. sulfidophilum), sulfide oxidation is dependent on the presence of the Sox multienzyme system. Sox systems occur in purple non-sulfur as well as in purple sulfur bacteria and are responsible for the oxidation of thiosulfate. In thiosulfate-oxidizing purple non-sulfur bacteria seven core genes encoding components of this complex are present: soxYZ, soxXA, soxB and soxCD. In most of these organisms thiosulfate is oxidized to sulfate without formation of intermediates. In Rdv. sulfidophilum sulfite is excreted. In the gammaproteobacterial purple sulfur bacteria, elemental sulfur appears as an intermediate of thiosulfate oxidation. This characteristic difference to the purple non-sulfur bacteria coincides with the lack of soxCD genes in purple sulfur bacteria. Siroamide-sulfite reductase (DsrAB) and proteins encoded by several other dsr genes are essential for the oxidation of sulfur formed as an intermediate during sulfide or thiosulfate oxidation in purple sulfur bacteria. Sulfite is the product of this process. The question of sulfite oxidation in purple bacteria will require special attention in the future, as the APS reductase pathway is restricted to some members of the Chromatiaceae. It has even been shown that this pathway is not essential in Alc. vinosum. In addition, genes for a classical sulfite: acceptor oxidoreductase are not present in most purple bacteria. Assimilatory sulfate reduction occurs in many purple bacteria. Two different types of sulfate transporters appear to be used. While sulfate is activated to adenosine-5´-phosphosulfate via a heterodimeric cysDN-encoded ATP sulfurylase in members of the Bradyrhizobiaceae, Acetobacteraceae, Rhodospirillaceae and Comamonadaceae, the homooligomeric sat-encoded type of this enzyme occurs in Rhodobacteraceae and is also present in the purple sulfur bacterium Alc. vinosum. In most if not all sulfateassimilating purple bacteria APS is immediately reduced without further phosphorylation to PAPS. The sulfite generated is reduced further to sulfide by assimilatory sulfite reductase. This enzyme appears to be ferredoxin-dependent in almost all cases, a flavoprotein subunit able to use NADPH as an electron
Johannes Sander and Christiane Dahl donor — as is the case in E. coli — is not encoded in the sequenced purple bacterial genomes. Acknowledgments Support by the Deutsche Forschungsgemeinschaft to CD is gratefully acknowledged. We thank Birgitt Hüttig for excellent technical assistance. References Adkins JP, Madigan MT, Mandelco L, Woese CR and Tanner RS (1993) Arhodomonas aquaeolei gen. nov., sp. nov., an aerobic, halophilic bacterium isolated from a subterranean brine. Int J Syst Bacteriol 43: 514–520 Alguero M, Dahl C and Trüper HG (1988) Partial purification of ADP sulfurylase from the purple sulfur bacterium Thiocapsa rosoepersicina. Microbiologia SEM 4: 149–160 Appia-Ayme C, Little PJ, Matsumoto Y, Leech AP and Berks BC (2001) Cytochrome complex essential for photosynthetic oxidation of both thiosulfate and sulfide in Rhodovulum sulfidophilum. J Bacteriol 183: 6107–6118 Baldock MI, Denger K, Smits THM and Cook AM (2007) Roseovarius sp. strain 217: Aerobic taurine dissimilation via acetate kinase and acetate-CoA ligase. FEMS Microbiol Lett 271: 202–206 Bamford VA, Berks BC and Hemmings AM (2002a) Novel domain packing in the crystal structure of a thiosulphate-oxidizing enzyme. Biochem Soc Trans 638–642 Bamford VA, Bruno S, Rasmussen T, Appia-Ayme C, Cheesman MR, Berks BC and Hemmings AM (2002b) Structural basis for the oxidation of thiosulfate by a sulfur cycle enzyme. EMBO J 21: 5599–5610 Bardischewsky F, Quentmeier A and Friedrich CG (2006) The flavoprotein SoxF functions in chemotrophic thiosulfate oxidation of Paracoccus pantotrophus in vivo and in vitro. FEMS Microbiol Lett 258: 121–126 Bartsch RG, Newton GL, Sherrill C and Fahey RC (1996) Glutathione amide and its perthiol in anaerobic sulfur bacteria. J Bacteriol 178: 4742–4746 Beynon JD, MacRae IJ, Huston SL, Nelson DC, Segel IH and Fisher AJ (2001) Crystal structure of ATP sulfurylase from the bacterial symbiont of the hydrothermal vent tubeworm Riftia pachyptila. Biochemistry 40: 14509–14517 Bick JA, Dennis JJ, Zylstra GJ, Nowack J and Leustek T (2000) Identification of a new class of 5´-adenylylsulfate (APS) reductases from sulfate-assimilating bacteria. J Bacteriol 182: 135–142 Blöthe, M and Fischer, U (2000) New insights in sulfur metabolism of purple and green phototrophic sulfur bacteria and their spheroplasts. BIOspektrum, Special edition 1st Joint Congress of DGHM, ÖGHMP and VAAM: ‘Microbiology 2000,’ 12 bis 16 März, München : 62 Bosshard HR, Davidson MW, Knaff DB and Millett F (1986) Complex formation and electron transfer between mitochondrial cytochrome c and flavocytochrome c552 from Chromatium
Chapter 30
Sulfur Metabolism in Purple Bacteria
vinosum. J Biol Chem 261: 190–193 Brettar I, Christen R and Höfle MG (2002) Rheinheimera baltica gen. nov., sp. nov., a blue-coloured bacterium isolated from the central Baltic Sea. Int J Syst Evol Microbiol 52: 1851–1857 Brune DC (1989) Sulfur oxidation by phototrophic bacteria. Biochim Biophys Acta 975: 189–221 Brune DC (1995a) Isolation and characterization of sulfur globule proteins from Chromatium vinosum and Thiocapsa roseopersicina. Arch Microbiol 163: 391–399 Brune DC (1995b) Sulfur compounds as photosynthetic electron donors. In: Blankenship RE, Madigan, MT and Bauer, CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 847–870. Kluwer Academic Publishers, Dordrecht Brüser T, Selmer T and Dahl C (2000) ‘ADP sulfurylase’ from Thiobacillus denitrificans is an adenylylsulfate:phosphate adenylyltransferase and belongs to a new family of nucleotidyltransferases. J Biol Chem 275: 1691–1698 Bryantseva IA, Gorlenko VM, Kompantseva EI, Imhoff JF, Süling J and Mityushina L (1999) Thiorhodospira sibirica gen. nov., sp. nov., a new alkaliphilic purple sulfur bacterium from a Siberian Soda lake. Int J Syst Bacteriol 49: 697–703 Chen ZW, Koh M, van Driessche G, van Beeumen JJ, Bartsch RG, Meyer TE, Cusanovich MA and Mathews FS (1994) The structure of flavocytochrome c sulfide dehydrogenase from a purple phototrophic bacterium. Science 266: 430–432 Cherest H, Davidian JC, Thomas D, Benes V, Ansorge W and Surdin-Kerjan Y (1997) Molecular characterization of two high affinity sulfate transporters in Saccharomyces cerevisiae. Genetics 145: 627–635 Dahl C (1996) Insertional gene inactivation in a phototrophic sulphur bacterium: APS-reductase-deficient mutants of Chromatium vinosum. Microbiology 142: 3363–3372 Dahl C (1999) Deposition and oxidation of polymeric sulfur in prokaryotes. In: Steinbüchel A (ed) Biochemical Principles and Mechanisms of Biosynthesis and Biodegradation of Polymers, pp 27–34. Wiley-VCH, Weinheim Dahl C (2008) Inorganic sulfur compounds as electron donors in purple sulfur bacteria. In: Hell R, Dahl, C, Leustek, T and Knaff, DB (eds) Sulfur in Phototrophic Organisms (Advances in Photosynthesis and Respiration, Vol 27). Springer, Dordrecht, in press Dahl C and Prange A (2006) Bacterial sulfur globules: Occurrence, structure and metabolism. In: Shively JM (ed) Inclusions in Prokaryotes, pp 21–51. Springer, Heidelberg Dahl C and Trüper HG (1989) Comparative enzymology of sulfite oxidation in Thiocapsa roseopersicina strains 6311, M1 and BBS under chemotrophic and phototrophic conditions. Z Naturforsch 44c: 617–622 Dahl C, Engels S, Pott-Sperling AS, Schulte A, Sander J, LübbeY, Deuster O and Brune DC (2005) Novel genes of the dsr gene cluster and evidence for close interaction of Dsr proteins during sulfur oxidation in the phototrophic sulfur bacterium Allochromatium vinosum. J Bacteriol 187: 1392–1404 Dahl C, Schulte A and Shin DH (2007) Cloning, expression, purification, crystallization and preliminary X-ray diffraction analysis of DsrEFH from Allochromatium vinousm. Acta Cryst F 63: 890–892 Davidson MW, Gray GO and Knaff DB (1985) Interaction of Chromatium vinosum flavocytochrome c-552 with cytochromes c studied by affinity chromatography. FEBS Lett
617 187: 155–159 de Jong GAH, Hazeu W, Bos P and Kuenen JG (1997) Isolation of the tetrathionate hydrolase from Thiobacillus acidophilus. Eur J Biochem 243: 678–683 de Jong GAH, Tang JA, Bos P, de Vries S and Kuenen GJ (2000) Purification and characterization of a sulfite:cytochrome c oxidoreductase from Thiobacillus acidophilus. J Mol Catal B 8: 61–67 Denger K, Smits THM and Cook AM (2006) Genome-enabled analysis of the utilization of taurine as sole source of carbon or of nitrogen by Rhodobacter sphaeroides 2.4.1. Microbiology 152: 3197–3206 Denger, K, Weinitschke, S, Hollemeyer, K and Cook, AM (2004) Sulfoacetate generated by Rhodopseudomonas palustris from taurine. Arch Microbiol 182: 254–258 Dhillon A, Goswami S, Riley M, Teske A and Sogin M (2005) Domain evolution and functional diversification of sulfite reductases. Astrobiology 5: 18–29 Doonan CJ, Kappler U and George GN (2006) Structure of the active site of sulfite dehydrogenase from Starkeya novella. Inorg Chem 45: 7488–7492 Epel B, Schäfer KO, Quentmeier A, Friedrich C and Lubitz W (2005) Multifrequency EPR analysis of the dimanganese cluster of the putative sulfate thiohydrolase SoxB of Paracoccus pantotrophus. J Biol Inorg Chem 10: 636–642 Feng C, Kappler U, Tollin G and Enemark JH (2003) Intramolecular electron transfer in a bacterial sulfite dehydrogenase. J Am Chem Soc 125: 14696–14697 Fischer U (1984) Cytochromes and iron sulfur proteins in sulfur metabolism of phototrophic sulfur bacteria. In: Müller A and Krebs, B (eds) Sulfur, Its Significance for Chemistry, for the Geo-, Bio- and Cosmosphere and Technology, pp 383–407. Elsevier Science Publishers, Amsterdam Foster BA, Thomas SM, Mahr JA, Renosto F, Patel HC and Segel IH (1994) Cloning and sequencing of ATP sulfurylase from Penicillium chrysogenum: Identification of a likely allosteric domain. J Biol Chem 269: 19777–19786 Franz B, Lichtenberg H, Hormes J, Modrow H, Dahl C and Prange A (2007) Utilization of solid ‘elemental’ sulfur by the phototrophic purple sulfur bacterium Allochromatium vinosum: A sulfur K-edge XANES spectroscopy study. Microbiology 153: 1268–1274 Friedrich CG, Rother D, Bardischewsky F, Quentmeier A and Fischer J (2001) Oxidation of reduced inorganic sulfur compounds by bacteria: emergence of a common mechanism? Appl Environ Microbiol 67: 2873–2882 Friedrich CG, Bardischewsky F, Rother D, Quentmeier A and Fischer J (2005) Prokaryotic sulfur oxidation. Curr Opin Microbiol 8: 253–259 Frigaard NU and Bryant DA (2008) Genomic insights into the sulfur metabolism of phototrophic green sulfur bacteria. In: Hell R, Dahl, C, Knaff, DB and Leustek, T (eds) Sulfur Metabolism in Phototrophic Organisms (Advances in Photosynthesis and Respiration, Vol 27). Springer, Dordrecht, in press Fritz G, Buchert T, Huber H, Stetter KO and Kroneck PMH (2000) Adenylylsulfate reductases from archaea and bacteria are 1:1 alpha beta-heterodimeric iron-sulfur flavoenzymes — high similarity of molecular properties emphasizes their central role in sulfur metabolism. FEBS Lett 473: 63–66 Fukumori Y and Yamanaka T (1979) A high-potential nonheme iron protein (HiPIP)-linked, thiosulfate-oxidizing enzyme de-
618 rived from Chromatium vinosum. Curr Microbiol 3: 117–120 Glaeser J and Overmann J (1999) Selective enrichment and characterization of Roseospirillum parvum, gen. nov. and sp. nov., a new purple nonsulfur bacterium with unusual light absorption properties. Arch Microbiol 171: 405–416 Gogtova GI and Vainstein MB (1981) Thiosulfate oxidation in the dark by the purple sulfur bacterium Ectothiorhodospira shaposhnikovii. Mikrobiologiya 55: 960–963 (in Russian) Gonzalez JM, Covert JS, Whitman WB, Henriksen JR, Mayer F, Scharf B, Schmitt R, Buchan A, Fuhrman JA, Kiene RP and Moran MA (2003) Silicibacter pomeroyi sp. nov. and Roseovarius nubinhibens sp. nov., dimethylsulfoniopropionatedemethylating bacteria from marine environments. Int J Syst Evol Microbiol 53: 1261–1269 Gorlenko VM, Bryantseva IA, Panteleeva EE, Tourova TP, Kolganova TV, Makhneva ZK and Moskalenko AA (2004) Ectothiorhodosinus mongolicum gen. nov., sp. nov., a new purple bacterium from a soda lake in Mongolia. Microbiology 73: 66–73 Griesbeck C, Hauska G and Schütz M (2000) Biological sulfide oxidation: Sulfide:quinone reductase (SQR), the primary reaction. In: Pandalai SG (ed) Recent Research Developments in Microbiology, Vol 4, pp 179–204. Research Signpost, Trivandrum, India Griesbeck C, Schütz M, Schödl T, Bathe S, Nausch L, Mederer N, Vielreicher M and Hauska G (2002) Mechanism of sulfidequinone oxidoreductase investigated using site-directed mutagenesis and sulfur analysis. Biochemistry 41: 11552–11565 Hansen TA and Imhoff JF (1985) Rhodobacter veldkampii, a new species of phototrophic purple nonsulfur bacteria. Int J Syst Bacteriol 35: 115–116 Hansen TA and van Gemerden H (1972) Sulfide utilization by purple sulfur bacteria. Arch Mikrobiol 86: 49–56 Hansen TA and Veldkamp H (1973) Rhodopseudomonas sulfidophila, nov. spec., a new species of the purple nonsulfur bacteria. Arch Mikrobiol 92: 45–58 Haverkamp T and Schwenn JD (1999) Structure and function of a cysBJIH gene cluster in the purple sulfur bacterium Thiocapsa roseopersicina. Microbiology 145: 115–125 Hell R, Jost R, Berkowitz O and Wirtz M (2002) Molecular and biochemical analysis of the enzymes of cysteine biosynthesis in the plant Arabidopsis thaliana. Amino Acids 22: 245–257 Hensen D, Sperling D, Trüper HG, Brune DC and Dahl C (2006) Thiosulfate oxidation in the phototrophic sulfur bacterium Allochromatium vinosum. Mol Microbiol 62: 794–810 Hille R (1996) The mononuclear molybdenum enzymes. Chem Rev 96: 2757–2816 Hipp WM, Pott AS, Thum-Schmitz N, Faath I, Dahl C and Trüper HG (1997) Towards the phylogeny of APS reductases and sirohaem sulfite reductases in sulfate-reducing and sulfur-oxidizing prokaryotes. Microbiology 143: 2891–2902 Hiraishi A and Shimada K (2001) Aerobic anoxygenic photosynthetic bacteria with zinc-bacteriochlorophyll. J Gen Appl Microbiol 47: 161–180 Hiraishi A and Ueda Y (1994) Rhodoplanes gen. nov., a new genus of phototrophic bacteria including Rhodopseudomonas rosea as Rhodoplanes roseus comb. nov. and Rhodoplanes elegans sp. nov. Int J Syst Bacteriol 44: 665–673 Hiraishi A and Ueda Y (1995) Isolation and characterization of Rhodovulum strictum sp. nov. and some other purple nonsulfur bacteria from colored blooms in tidal and seawater pools. Int J
Johannes Sander and Christiane Dahl Syst Bacteriol 45: 319–326 Hiraishi A, Urata K and Satoh T (1995) A new genus of marine budding phototrophic bacteria, Rhodobium gen. nov., which includes Rhodobium orientis sp. nov. and Rhodobium marinum comb. nov. Int J Syst Bacteriol 45: 226–234 Hirschler-Rea A, Matheron R, Riffaud C, Moune S, Eatock C, Herbert RA, Willison JC and Caumette P (2003) Isolation and characterization of spirilloid purple phototrophic bacteria forming red layers in microbial mats of Mediterranean salterns: Description of Halorhodospira neutriphila sp. nov. and emendation of the genus Halorhodospira. Int J Syst Evol Microbiol 53: 153–163 Hryniewicz M and Kredich NM (1991) The cysP promoter of Salmonella typhimurium: Characterization of two binding sites for CysB protein, studies of in vivo transcription initiation and demonstration of the anti-inducer effects of thiosulfate. J Bacteriol 173: 5876–5886 Ikeuchi Y, Shigi N, Kato J, Nishimura A and Suzuki T (2006) Mechanistic insights into sulfur relay by multiple sulfur mediators involved in thiouridine biosynthesis at tRNA wobble positions. Mol Cell 21: 97–108 Imhoff JF (1982) Occurrence and evolutionary significance of two sulfate assimilation pathways in the Rhodospirillaceae. Arch Microbiol 132: 197–203 Imhoff JF (1983) Rhodopseudomonas marina sp. nov., a new marine phototropic purple bacterium. Syst Appl Microbiol 4: 512–521 Imhoff JF (2001) The anoxygenic phototrophic purple bacteria. In: Boone DR, Castenholz, RW and Garrity GM (eds) Bergey’s Manual of Systematic Bacteriology, Vol 1, pp 631–637. Springer, New York Imhoff JF (2005a) Family I. Chromatiaceae Bavendamm 1924, 125AL emend. Imhoff 1984b, 339. In: Brenner DJ, Krieg, NR, Staley, JT and Garrity GM (eds) Bergey’s Manual of Systematic Bacteriology, Vol 2, part B, pp 3–40. Springer, New York Imhoff JF (2005b) Family II. Ectothiorhodospiraceae Imhoff 1984b, 339VP. In: Brenner DJ, Krieg, NR, Staley, JT and Garrity GM (eds) Bergey’s Manual of Systematic Bacteriology, Vol 2, part B, pp 41–57. Springer, New York Imhoff JF, Kramer M and Trüper HG (1983) Sulfate assimilation in Rhodopseudomonas sulfidophila. Arch Microbiol 136: 96–101 Imhoff JF, Süling J and Petri R (1998) Phylogenetic relationships among the Chromatiaceae, their taxonomic reclassification and description of the new genera Allochromatium, Halochromatium, Isochromatium, Marichromatium, Thiococcus, Thiohalocapsa, and Thermochromatium. Int J Syst Bacteriol 48: 1129–1143 Imhoff JF, Hiraishi A and Süling J (2005) Anoxygenic phototrophic purple bacteria. In: Brenner DJ, Krieg, NR, Staley, JT and Garrity GM (eds) Bergey’s Manual of Systematic Bacteriology, Vol 2, part A, pp 119–132. Springer, New York Jørgensen BB (1990) The sulfur cycle of freshwater sediments: Role of thiosulfate. Limnol Oceanogr 35: 1329–1342 Kappler U (2007) Bacterial sulfite-oxidizing enzymes — enzymes for chemolithotrophy only? In: Dahl C and Friedrich CG (eds) Microbial Sulfur Metabolism, pp 151–169. Springer, Heidelberg Kappler U and Dahl C (2001) Enzymology and molecular biology of prokaryotic sulfite oxidation (minireview). FEMS Microbiol Lett 203: 1–9
Chapter 30
Sulfur Metabolism in Purple Bacteria
Kappler, U and Bailey, S (2005) Molecular basis of intramolecular electron transfer in sulfite-oxidizing enzymes is revealed by high resolution structure of a heterodimeric complex of the catalytic molybdopterin subunit and a c-type cytochrome subunit. J Biol Chem 280: 24999–245007 Kappler U, Bennett B, Rethmeier J, Schwarz G, Deutzmann R, McEwan AG and Dahl C (2000) Sulfite:cytochrome c oxidoreductase from Thiobacillus novellus — Purification, characterization, and molecular biology of a heterodimeric member of the sulfite oxidase family. J Biol Chem 275: 13202–13212 Kertesz, MA (2001) Bacterial transporters for sulfate and organosulfur compounds. Res Microbiol 152: 279–290 Knobloch K, Schmitt W, Schleifer G, Appelt N and Müller H (1981) On the enzymatic system thiosulfate-cytochrome coxidoreductase. In: Bothe H and Trebst A (eds) Biology of Inorganic Nitrogen and Sulfur, pp 359–365. Springer, Berlin Kompantseva EI (1985) New halophilic purple bacteria Rhodobacter euryhalinus sp. nov. Mikrobiologiya 54: 974–981 (in Russian) Kompantseva EI and Gorlenko VM (1984) A new species of moderately halophilic purple bacterium Rhodospirillum mediosalinum. Microbiology 53: 775–781 Kopriva S, Büchert T, Fritz G, Suter M, Weber M, Benda R, Schaller J, Feller U, Schürmann P, Schünemann V, Trautwein AX, Kroneck PMH and Brunold C (2001) Plant adenosine 5´-phosphosulfate reductase is a novel iron-sulfur protein. J Biol Chem 276: 42881–42886 Kopriva S, Büchert T, Fritz G, Suter M, Benda R, Schünemann V, Koprivova A, Schürmann P, Trautwein AX, Kroneck PMH and Brunold C (2002) The presence of an iron-sulfur cluster in adenosine 5´-phosphosulfate reductase separates organisms utilizing adenosine 5´-phosphosulfate and phosphoadenosine 5´-phosphosulfate for sulfate assimilation. J Biol Chem 277: 21786–21791 Kostanjevecki V, Brige A, Meyer TE, Cusanovich MA, Guisez Y and van Beeumen J (2000) A membrane-bound flavocytochrome c-sulfide dehydrogenase from the purple phototrophic sulfur bacterium Ectothiorhodospira vacuolata. J Bacteriol 182: 3097–3103 Krafft T, Bokranz M, Klimmek O, Schröder I, Fahrenholz F, Kojro E and Kröger A (1992) Cloning and nucleotide sequence of the psrA gene of Wolinella succinogenes polysulphide reductase. Eur J Biochem 206: 503–510 Kredich NM (1996) Biosynthesis of cysteine. In: Neidhardt FC (ed) Escherichia coli and Salmonella typhimurium. Cellular and Molecular Biology, pp 514–527. American Society for Microbiology, Washington D.C. Kusai K and Yamanaka T (1973) The oxidation mechanisms of thiosulphate and sulphide in Chlorobium thiosulphatophilum: roles of cytochrome c-551 and cytochrome c-553. Biochim Biophys Acta 325: 304–314 Kusche, WH (1985) Untersuchungen an Elektronentransportproteinen und zum Schwefelstoffwechsel in Ectothiorhodospiraceae. PhD dissertation, University of Bonn, Germany Labrenz M, Collins MD, Lawson PA, Tindall BJ, Schumann P and Hirsch P (1999) Roseovarius tolerans gen. nov., sp. nov., a budding bacterium with variable bacteriochlorophyll a production from hypersaline Ekho Lake. Int J Syst Bacteriol 49: 137–147 Leguijt T (1993) Photosynthetic electron transfer in Ectothiorhodospira. PhD Dissertation, University of Amsterdam
619 Leustek, T, Martin, MN, Bick, JA and Davies, JP (2000) Pathways and regulation of sulfur metabolism revealed through molecular and genetic studies. Annu Rev Plant Physiol Plant Mol Biol 51: 141–165 Leustek T and Saito K (1999) Sulfate transport and assimilation in plants. Plant Physiol 120: 637–643 Leyh TS (1993) The physical biochemistry and molecular genetics of sulfate activation. Crit Rev Biochem Mol Biol 28: 515–542 Liss P (1999) Biogeochemistry — Take the shuttle-from marine algae to atmospheric chemistry. Science 285: 1217–1218 Lu W-P, Swoboda EP and Kelly DP (1985) Properties of the thiosulfate-oxidizing multi-enzyme system from Thiobacillus versutus. Biochim Biophys Acta 828: 116–122 Lübbe YJ, Youn H-S, Timkovich R and Dahl C (2006) Siro(haem)amide in Allochromatium vinosum and relevance of DsrL and DsrN, a homolog of cobyrinic acid a,c diamide synthase for sulfur oxidation. FEMS Microbiol Lett 261: 194–202 MacRae IJ, Segel IH and Fisher AJ (2001) Crystal structure of ATP sulfurylase from Penicillium chrysogenum: Insights into the allosteric regulation of sulfate assimilation. Biochemistry 40: 6795–6804 Matias PM, Pereira IAC, Soares CM and Carrondo MA (2005) Sulphate respiration from hydrogen in Desulfovibrio bacteria: A structural biology overview. Prog Biophys Molec Biol 89: 292–329 Mendoza-Cózatl, D, Loza-Tavera, H, Hernández-Navarro, A and Moreno-Sánchez, R (2005) Sulfur assimilation and glutathione metabolism under cadmium stress in yeast, protists and plants. FEMS Microbiol Rev 29: 653–671 Meulenberg R, Pronk JT, Frank J, Hazeu W, Bos P and Kuenen JG (1992a) Purification and partial characterization of a thermostable trithionate hydrolase from the acidophilic sulphur oxidizer Thiobacillus acidophilus. Eur J Biochem 209: 367–374 Meulenberg R, Pronk JT, Hazeu W, Bos P and Kuenen JG (1992b) Oxidation of reduced sulphur compounds by intact cells of Thiobacillus acidophilus. Arch Microbiol 157: 161–168 Meulenberg R, Pronk JT, Hazen W, van Dijken JP, Frank J, Bos P and Kuenen JG (1993) Purification and partial characterization of thiosulphate dehydrogenase from Thiobacillus acidophilus. J Gen Microbiol 139: 2033–2039 Meyer TE and Cusanovich MA (2003) Discovery and characterization of electron transfer proteins in the photosynthetic bacteria. Photosynth Res 76: 111–126 Milford AD, Achenbach LA, Jung DO and Madigan MT (2000) Rhodobaca bogoriensis gen. nov. and sp. nov., an alkaliphilic purple nonsulfur bacterium from African Rift Valley soda lakes. Arch Microbiol 174: 18–27 Moran MA, Gonzalez JM and Kiene RP (2003) Linking a bacterial taxon to sulfur cycling in the sea: Studies of the marine Roseobacter group. Geomicrobiol J 20: 375–388 Myers JD and Kelly DJ (2005) A sulphite respiration system in the chemoheter otrophic human pathogen Campylobacter jejuni. Microbiology 151: 233–242 Neumann S, Wynen A, Trüper HG and Dahl C (2000) Characterization of the cys gene locus from Allochromatium vinosum indicates an unusual sulfate assimilation pathway. Mol Biol Reports 27: 27–33 Neutzling O and Trüper HG (1982) Assimilatory sulfur metabolism in Rhodopseudomonas sulfoviridis. Arch Microbiol
620 133: 145–148 Neutzling O, Imhoff JF and Trüper HG (1984) Rhodopseudomonas adriatica sp. nov., a new species of the Rhodospirillaceae, dependent on reduced sulfur compounds. Arch Microbiol 137: 256–261 Neutzling O, Pfleiderer C and Trüper HG (1985) Dissimilatory sulphur metabolism in phototrophic ‘non-sulphur’ bacteria. J Gen Microbiol 131: 791–798 Newton ILG, Woyke T, Auchtung TA, Dilly GF, Dutton RJ, Fisher MC, Fontanez KM, Lau E, Stewart FJ, Richardson PM, Barry KW, Saunders E, Detter JC, Wu D, Eisen JA and Cavanaugh CM (2007) The Calyptogena magnifica chemoautotrophic symbiont genome. Science 315: 998–1000 Numata T, Fukai S, Ikeuchi Y, Suzuki T and Nureki O (2006) Structural basis for sulfur relay to RNA mediated by heterohexameric TusBCD complex. Structure 14: 357–366 Ono B, Hazu T, Yoshida S, Kawato T, Shinoda S, Brzvwczy J and Paszewski A (1999) Cysteine biosynthesis in Saccharomyces cerevisiae: A new outlook on pathway and regulation. Yeast 15: 1365–1375 Oren A, Kessel M and Stackebrandt E (1989) Ectothiorhodospira marismortui sp. nov., an obligately anaerobic, moderately halophilic purple sulfur bacterium from a hypersaline sulfur spring on the shore of the Dead Sea. Arch Microbiol 151: 524–529 Pao SS, Paulsen IT and Saier MH (1998) Major facilitator superfamily. Microbiol Mol Biol Rev 62: 1–34 Pattaragulwanit K, Brune DC, Trüper HG and Dahl C (1998) Molecular genetic evidence for extracytoplasmic localization of sulfur globules in Chromatium vinosum. Arch Microbiol 169: 434–444 Pfennig N (1978) General physiology and ecology of photosynthetic bacteria. In: Clayton RK and Sistrom WR (eds) The Photosynthetic Bacteria, pp 3–18. Plenum Press, New York Pfennig N, Lunsdorf H, Süling J and Imhoff JF (1997) Rhodospira trueperi gen. nov., spec. nov., a new phototrophic Proteobacterium of the alpha group. Arch Microbiol 168: 39–45 Pires RH, Lourenco AI, Morais F, Teixeira M, Xavier AV, Saraiva LM and Pereira IAC (2003) A novel membrane-bound respiratory complex from Desulfovibrio desulfuricans ATCC 27774. Biochim Biophys Acta 1605: 67–82 Pires, RH, Venceslau, SS, Morais, F, Teixeira, M, Xavier, AV and Pereira, IAC (2006) Characterization of the Desulfovibrio desulfuricans ATCC 27774 DsrMKJOP complex — a membrane-bound redox complex involved in the sulfate respiratory pathway. Biochemistry 45: 249–262 Podgorsek L and Imhoff JF (1999) Tetrathionate production by sulfur oxidizing bacteria and the role of tetrathionate in the sulfur cycle of Baltic Sea sediments. Aquat Microb Ecol 17: 255–265 Pott AS and Dahl C (1998) Sirohaem-sulfite reductase and other proteins encoded in the dsr locus of Chromatium vinosum are involved in the oxidation of intracellular sulfur. Microbiology 144: 1881–1894 Prange A, Arzberger I, Engemann C, Modrow H, Schumann O, Trüper HG, Steudel R, Dahl C and Hormes J (1999) In situ analysis of sulfur in the sulfur globules of phototrophic sulfur bacteria by X-ray absorption near edge spectroscopy. Biochim Biophys Acta 1428: 446–454 Prange A, Chauvistre R, Modrow H, Hormes J, Trüper HG and Dahl C (2002) Quantitative speciation of sulfur in bacterial sulfur globules: X-ray absorption spectroscopy reveals at least three
Johannes Sander and Christiane Dahl different speciations of sulfur. Microbiology 148: 267–276 Prange A, Engelhardt H, Trüper HG and Dahl C (2004) The role of the sulfur globule proteins of Allochromatium vinosum: Mutagenesis of the sulfur globule protein genes and expression studies by real-time RT PCR. Arch Microbiol 182: 165–174 Pronk JT, Meulenberg R, Hazeu W, Bos P and Kuenen JG (1990) Oxidation of reduced inorganic sulphur compounds by acidophilic thiobacilli. FEMS Microbiol Rev 75: 293–306 Quentmeier A and Friedrich CG (2001) The cysteine residue of the SoxY protein as the active site of protein-bound sulfur oxidation of Paracoccus pantotrophus GB17. FEBS Lett 503: 168–172 Quentmeier A, Hellwig P, Bardischewsky F, Grelle G, Kraft R and Friedrich CG (2003) Sulfur oxidation in Paracoccus pantotrophus: Interaction of the sulfur-binding protein SoxYZ with the dimanganese SoxB protein. Biochem Biophys Res Commun 312: 1011–1018 Quentmeier A, Kraft R, Kostka S, Klockenkamper R and Friedrich CG (2000) Characterization of a new type of sulfite dehydrogenase from Paracoccus pantotrophus GB17. Arch Microbiol 173: 117–125 Raitsimring AM, Kappler U, Feng CJ, Astashkin AV and Enemark JH (2005) Pulsed EPR studies of a bacterial sulfite-oxidizing enzyme with pH-invariant hyperfine interactions from exchangeable protons. Inorg Chem 44: 7283–7285 Ramírez, P, Guiliani, N, Valenzuela, L, Beard, S and Jerez, CA (2004) Differential protein expression during growth of Acidithiobacillus ferrooxidans on ferrous iron, sulfur compounds, or metal sulfides. Appl Enivron Microbiol 70: 4491–4498 Raymond JC and Sistrom WR (1969) Ectothiorhodospira halophila — a new species of the genus Ectothiorhodospira. Arch Mikrobiol 69: 121–126 Reinartz M, Tschäpe J, Brüser T, Trüper HG and Dahl C (1998) Sulfide oxidation in the phototrophic sulfur bacterium Chromatium vinosum. Arch Microbiol 170: 59–68 Renosto F, Martin RL, Borrell JL, Nelson DC and Segel IH (1991) ATP sulfurylase from trophosome tissue of Riftia pachyptila (hydrothermal vent tube worm). Arch Biochem Biophys 290: 66–78 Rijkenberg MJA, Kort R and Hellingwerf KJ (2001) Alkalispirillum mobile gen. nov., spec. nov., an alkaliphilic non-phototrophic member of the Ectothiorhodospiraceae. Arch Microbiol 175: 369–375 Rohwerder T and Sand W (2003) The sulfane sulfur of persulfides is the actual substrate of the sulfur-oxidizing enzymes from Acidithiobacillus and Acidiphilium spp. Microbiology 149: 1699–1709 Romanenko LA, Uchino M, Falsen E, Zhukova NV, Mikhailov VV and Uchimura T (2003) Rheinheimera pacifica sp. nov., a novel halotolerant bacterium isolated from deep sea water of the Pacific. Int J Syst Evol Microbiol 53: 1973–1977 Rother D, Heinrich HJ, Quentmeier A, Bardischewsky F and Friedrich CG (2001) Novel genes of the sox gene cluster, mutagenesis of the flavoprotein SoxF, and evidence for a general sulfur-oxidizing system in Paracoccus pantotrophus GB17. J Bacteriol 183: 4499–4508 Rother D, Orawski G, Bardischewsky F and Friedrich CG (2005) SoxRS-mediated regulation of chemotrophic sulfur oxidation in Paracoccus pantotrophus. Microbiology 151: 1707–1716 Sabehi G, Loy A, Jung K-H, Partha R, Spudich JL, Isaacson T, Hirschberg J, Wagner M and Béjà O (2005) New insights into
Chapter 30
Sulfur Metabolism in Purple Bacteria
metabolic properties of marine bacteria encoding proteorhodopsins. PLoS Biology 3: 1409–1417 Sander J, Engels-Schwarzlose S and Dahl C (2006) Importance of the DsrMKJOP complex for sulfur oxidation in Allochromatium vinosum and phylogenetic analysis of related complexes in other prokaryotes. Arch Microbiol 186: 357–366 Schedel M, Vanselow M and Trüper HG (1979) Siroheme sulfite reductase from Chromatium vinosum. Purification and investigation of some of its molecular and catalytic properties. Arch Microbiol 121: 29–36 Schmidt M, Priemé A and Stougaard P (2007) Arsukibacterium ikkense gen. nov., sp. nov., a novel alkaliphilic, enzyme-producing gamma-Proteobacterium isolated from a cold and alkaline environment in Greenland. Syst Appl Microbiol 30: 197–201 Schmitt W, Schleifer G and Knobloch K (1981) The enzymatic system thiosulfate:cytochrome c oxidoreductase from photolithoautotrophically grown Chromatium vinosum. Arch Microbiol 130: 334–338 Schütz M, Shahak Y, Padan E and Hauska G (1997) Sulfidequinone reductase from Rhodobacter capsulatus. J Biol Chem 272: 9890–9894 Schütz M, Maldener I, Griesbeck C and Hauska G (1999) Sulfidequinone reductase from Rhodobacter capsulatus: Requirement for growth, periplasmic localization, and extension of gene sequence analysis. J Bacteriol 181: 6516–6523 Shiba T (1991) Roseobacter litoralis gen. nov., sp. nov., and Roseobacter denitrificans sp. nov., aerobic pink-pigmented bacteria which contain bacteriochlorophyll a. Syst Appl Microbiol 14: 140–145 Smith AJ (1966) The role of tetrathionate in the oxidation of thiosulphate by Chromatium sp. strain D. J Gen Microbiol 42: 371–380 Smith AJ and Lascelles J (1966) Thiosulphate metabolism and rhodanese in Chromatium sp. strain D. J Gen Microbiol 42: 357–370 Sorokin DY and Kuenen JG (2005) Haloalkaliphilic sulfuroxidizing bacteria in soda lakes. FEMS Microbiol Rev 29: 685–702 Sorokin DY, Teske A, Robertson LA and Kuenen JG (1999) Anaerobic oxidation of thiosulfate to tetrathionate by obligately heterotrophic bacteria, belonging to the Pseudomonas stutzeri group. FEMS Microbiol Ecol 30: 113–123 Sorokin DY, Tourova TP, Kuznetsov BB, Bryantseva IA and Gorlenko VM (2000) Roseinatronobacter thiooxidans gen. nov., sp. nov., a new alkaliphilic aerobic bacteriochlorophyll acontaining bacterium isolated from a soda lake. Microbiology 69: 75–82 Sorokin DY, Kuenen JG and Jetten MSM (2001) Denitrification at extremely high pH values by the alkaliphilic, obligately chemolithoautotrophic, sulfur oxidizing bacterium Thioalkalivibrio denitrificans strain ALJD. Arch Microbiol 175: 94–101 Sperling D, Kappler U, Wynen A, Dahl C and Trüper HG (1998) Dissimilatory ATP sulfurylase from the hyperthermophilic sulfate reducer Archaeoglobus fulgidus belongs to the group of homo-oligomeric ATP sulfurylases. FEMS Microbiol Lett 162: 257–264 Srinivas TNR, Kumar PA, Sasikala C, Ramana CV, Süling J and Imhoff JF (2006) Rhodovulum marinum sp. nov., a novel phototrophic purple non-sulfur alphaproteobacterium from marine tides of Visakhapatnam, India. Int J Syst Evol Microbiol 56: 1651–1656
621 Steinmetz MA and Fischer U (1982) Cytochromes of the green sulfur bacterium Chlorobium vibrioforme f. thiosulfatophilum. Purification, characterization and sulfur metabolism. Arch Microbiol 131: 19–26 Steudel R (1996) Mechanism for the formation of elemental sulfur from aqueous sulfide in chemical and microbiological desulfurization processes. Ind Eng Chem Res 35: 1417–1423 Steudel R (2000) The chemical sulfur cycle. In: Lens P and Hulshoff Pol, W (eds) Environmental Technologies to Treat Sulfur Pollution, pp 1–31. IWA Publishing, London Steudel R and Eckert B (2003) Solid sulfur allotropes. In: Steudel R (ed) Elemental Sulfur and Sulfur-Rich Compounds, pp 1–79. Springer, Berlin Steudel R, Holdt G, Visscher PT and van Gemerden H (1990) Search for polythionates in cultures of Chromatium vinosum after sulfide incubation. Arch Microbiol 155: 432–437 Straub KL, Rainey FA and Widdel F (1999) Rhodovulum iodosum sp. nov., and Rhodovulum robiginosum sp. nov., two new marine phototrophic ferrous-iron-oxidizing purple bacteria. Int J Syst Bacteriol 49: 729–735 Then J and Trüper HG (1981) The role of thiosulfate in sulfur metabolism of Rhodopseudomonas globiformis. Arch Microbiol 130: 143–146 Then J and Trüper HG (1983) Sulfide oxidation in Ectothiorhodospira abdelmalekii. Evidence for the catalytic role of cytochrome c-551. Arch Microbiol 135: 254–258 Then J and Trüper HG (1984) Utilization of sulfide and elemental sulfur by Ectothiorhodospira halochloris. Arch Microbiol 139: 295–298 Trüper HG (1968) Ectothiorhodospira mobilis Pelsh, a photosynthetic sulfur bacterium depositing sulfur outside the cells. J Bacteriol 95: 1910–1920 Trüper HG (1978) Sulfur metabolism. In: Clayton RK and Sistrom, WR (eds) The Photosynthetic Bacteria, pp 677–690. Plenum, New York Trüper HG (1984) Phototrophic bacteria and their sulfur metabolism. In: Müller A and Krebs, B (eds) Sulfur, Its Significance for Chemistry, for the Geo-, Bio-, and Cosmosphere and Technology, pp 367–382. Elsevier Science Publishers, Amsterdam van Gemerden, H (1968) On the ATP generation by Chromatium in the dark. Arch Mikrobiol 64: 118–124 van Gemerden H (1987) Competition between purple sulfur bacteria and green sulfur bacteria: role of sulfide, sulfur and polysulfides. In: Lindholm T (ed) Ecology of Photosynthetic Prokaryotes With Special Reference to Meromictic Lakes and Coastal Lagoons, pp 13–27. Abo Academy Press, Abo Vidmar JJ, Tagmount A, Cathala N, Touraine B and Davidian JCE (2000) Cloning and characterization of a root specific high-affinity sulfate transporter from Arabidopsis thaliana. FEBS Lett 475: 65–69 Visscher PT and Taylor BF (1993) Organic thiols as organolithotrophic substrates for growth of phototrophic bacteria. Appl Environ Microbiol 59: 93–96 Visscher PT and van Gemerden H (1991) Photoautotrophic growth of Thiocapsa roseopersicina on dimethyl sulfide. FEMS Microbiol Lett 81: 247–250 Visscher PT, Nijburg JW and van Gemerden H (1990) Polysulfide utilization by Thiocapsa roseopersicina. Arch Microbiol 155: 75–81 Visser JM, de Jong GAH, Robertson LA and Kuenen JG (1996) Purification and characterization of a periplasmic thiosulfate
622 dehydrogenase from the obligately autotrophic Thiobacillus sp. W5. Arch Microbiol 166: 372–378 Wang X, Modak HV and Tabita FR (1993) Photolithoautotrophic growth and control of CO2 fixation in Rhodobacter sphaeroides and Rhodospirillum rubrum in the absence of ribulose bisphosphate carboxylase-oxygenase. J Bacteriol 175: 7109–7114 Wodara C, Bardischewsky F and Friedrich CG (1997) Cloning and characterization of sulfite dehydrogenase, two c-type cytochromes, and a flavoprotein of Paracoccus denitrificans GB17: essential role of sulfite dehydrogenase in lithotrophic sulfur oxidation. J Bacteriol 179: 5014–5023 Yakimov MM, Guiliano L, Chernikova TN, Gentile G, Abraham WR, Lunsdorf H, Timmis KN and Golyshin PN (2001) Alcalilimnicola halodurans gen. nov., sp. nov., an alkaliphilic,
Johannes Sander and Christiane Dahl moderately halophilic and extremely halotolerant bacterium, isolated from sediments of soda-depositing Lake Natron, East Africa Rift Valley. Int J Syst Evol Microbiol 51: 2133–2143 Yurkov VV (2006) Aerobic phototrophic proteobacteria. In: Dworkin M, Falkow, S, Rosenberg, E, Schleifer, K-H and Stackebrandt E (eds) The Prokaryotes, Vol 5, pp 562–584. Springer, New York Yurkov VV, Krasil’nikova EN and Gorlenko VM (1994) Thiosulfate metabolism in the aerobic bacteriochlorophyll-a-containing bacteria Erythromicrobium hydrolyticum and Roseococcus thiosulfatophilus. Microbiology 63: 91–94 Zaar A, Fuchs G, Golecki JR and Overmann J (2003) A new purple sulfur bacterium isolated from a littoral microbial mat, Thiorhodococcus drewsii sp. nov. Arch Microbiol 179: 174–183
Chapter 31 Dissimilatory and Assimilatory Nitrate Reduction in the Purple Photosynthetic Bacteria James P. Shapleigh* Department of Microbiology, Wing Hall, Cornell University, Ithaca, NY 14853, U.S.A.
Summary ............................................................................................................................................................... 623 I. Introduction..................................................................................................................................................... 624 II. Denitrification.................................................................................................................................................. 624 A. Background ...................................................................................................................................... 624 B. Occurrence of Denitrification In the Purple Photosynthetic Bacteria ................................................ 626 C. Other Genes Associated With Denitrification Gene Clusters ........................................................... 629 D. Physiological Role of Denitrification ................................................................................................. 630 E. Impacts of Denitrification on Cell Phenotype.................................................................................... 631 F. Regulation of Denitrification.............................................................................................................. 632 1. Nitrate Reductase ................................................................................................................... 632 2. Nitrite Reductase..................................................................................................................... 633 3. Nitrous Oxide Reductase ........................................................................................................ 634 4. Is Roseobacter denitrificans an Exception? ........................................................................... 634 G. Other Factors Affecting Denitrification Activity ................................................................................. 634 H. The Enzymes of Denitrification ........................................................................................................ 635 III. Assimilation of Nitrogen.................................................................................................................................. 636 A. Nitrate Assimilation........................................................................................................................... 636 B. Assimilation of N2 Produced Via Denitrification ................................................................................ 637 C. Regulatory Overlap in Nitrogen Metabolism .................................................................................... 637 IV. Conclusion...................................................................................................................................................... 638 Acknowledgments ................................................................................................................................................. 639 References ............................................................................................................................................................ 639
Summary Nitrate reduction can be either a dissimilatory or assimilatory process. Nitrate reduction to nitrogen gas via a series of nitrogen oxide intermediates is a dissimilatory process termed denitrification. Denitrification is common among the purple photosynthetic bacteria. While some reduce nitrate to nitrogen gas many are missing components of the denitrification pathway. In the complete denitrifiers, denitrification can be used as an alternative form of respiration when oxygen levels are low. Denitrification can also be used as a mechanism to dispose of excess reducing equivalents. In partial denitrifiers, it is unlikely that denitrification serves a respiratory function. In these bacteria it is likely that the enzymes that are present are used for redox balancing, or to mitigate the toxicity of certain nitrogen oxide intermediates. Available genome sequences demonstrate that closely related bacteria can have different denitrification *Email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 623–642. © 2009 Springer Science + Business Media B.V.
James P. Shapleigh
624
capacities. For example, analysis of three strains of Rhodobacter sphaeroides revealed that one is a complete denitrifier, while one of the other strains has two of the four nitrogen oxide reductases enzymes, and the other strain has only one. This suggests that denitrification is selectively modified to best fit each bacterium’s environmental niche and the entire pathway does not have to be present for dissimilatory nitrogen oxide reduction to be beneficial. Optimal expression of the nitrogen oxide reductases requires the presence of nitrogen oxides and low oxygen. Nitrate along with nitric oxide and nitrous oxide, two of the denitrification intermediates, are effector molecules. Denitrification has also been shown to be under control of the global Reg/Prr regulatory system. This may coordinate expression of denitrification with other energy conservation and redox dissipation processes. Some photosynthetic bacteria can also reduce nitrate to ammonia, which is then used for assimilatory purposes. As with denitrification, the capacity for nitrate assimilation does not follow any obvious phylogenetic pattern. The genes for nitrate assimilation are expressed when ammonia and other forms of fixed nitrogen are limiting and nitrate is available. I. Introduction Given their metabolic diversity it is not surprising that some members of the purple photosynthetic bacteria are capable of carrying out denitrification or nitrate assimilation. These two processes are important reactions of the nitrogen cycle. The nitrogen cycle can most easily be viewed as the reduction of N2 to NH4+ via nitrogen fixation, conversion of NH4+ to nitrate (NO3–) via nitrification, and the return of the fixed nitrogen to the atmosphere by reduction of nitrate and other nitrogen oxides to N2 via denitrification (Fig. 1). The complexity of the nitrogen cycle has increased with the discovery of anaerobic ammonia oxidation (ANAMMOX). During ANAMMOX nitrite and ammonia are converted to N2 (Strous et al., 2006). This process appears to be carried out by specialized bacteria and consequently is not found to occur among the purple photosynthetic bacteria (Strous et al., 1999). While denitrification and nitrate assimilation are parts of the same elemental cycle, and both utilize nitrate and nitrite, the two processes are not physiologically related (Fig. 1). Denitrification is a dissimilatory respiratory process in which nitrogen oxides (N-oxides) are used as terminal oxidants when oxygen is limiting. Assimilatory nitrate reduction allows cells to use nitrate as a nitrogen source for biosynthetic purposes. Denitrification will be discussed first with Abbreviations: cNor – cytochrome c dependent nitric oxide reductase; Nap – periplasmic nitrate reductase ; Nar – nitrate reductase; Nas – assimilatory nitrate reductase; Nir – nitrite reductase; NO – nitric oxide; Nor – nitric oxide reductase; Nos – nitrous oxide reductase; N-oxides – nitrogen oxides; qNor – quinol oxidizing nitric oxide reductase; Rba. – Rhodobacter; Rps. – Rhodopseudomonas; Rsb. – Roseobacter
an emphasis on the physiological role of denitrification since not all photosynthetic denitrifiers contain the entire complement of denitrification enzymes. Nitrate assimilation will then be discussed; however, because this process has only been studied in a few purple bacteria the discussion will necessarily focus on a limited number of species. II. Denitrification A. Background Denitrification is the reduction of nitrate to nitrogen gas (Payne, 1981). There are four enzymes required for the generation of N2 from nitrate, which produce three intermediates: nitrite (NO2–), nitric oxide (NO) and nitrous oxide (N2O) (Fig. 2A). Each intermediate is obligatory and freely diffusible. The names of the enzymes required for each step are given by combining the enzyme’s substrate with the term reductase. Therefore, denitrification involves nitrate reductase (Nar), nitrite reductase (Nir), nitric oxide reductase (Nor) and nitrous oxide reductase (Nos). With the exception of the ultimate reaction, multiple biochemically distinct enzymes have been described that catalyze each reductive step. All of these enzymes are found either in the inner membrane or in the periplasmic space of bacteria (Fig. 2B). It should be remembered that denitrification is a dissimilatory process, meaning that except under unusual circumstances none of the nitrate reduced is incorporated into cell biomass. Assimilatory nitrate reduction involves nitrate reduction to ammonia, which is used as nitrogen source (Fig. 1). Another related process is ammonification, which is the dissimilatory reduction
Chapter 31
Nitrogen Oxide Reduction
Fig. 1. The reactions of the nitrogen cycle. Arrows indicate direction of transformation.
of nitrate to ammonia and is carried out by bacteria such as Escherichia (E.) coli, but this process is not found in the purple photosynthetic bacteria. Nitrate reductase catalyzes the initial reductive step in denitrification, the two-electron reduction of nitrate to nitrite (Fig. 2A). This reaction has a midpoint potential of + 420 mV. This is the only denitrification enzyme that is also found in nondenitrifying bacteria. For example, E. coli contains two nitrate reductases that are used for the dissimilatory reduction of nitrate to nitrite (Richardson et al., 2001). The dissimilatory nitrate reductases that are found in denitrifiers are either membrane bound and referred to as Nar, or periplasmic and referred to as Nap (Fig. 2B). These two enzymes are related and both utilize a molybdenum containing cofactor for the site of nitrate reduction. Both utilize quinol as a source of electrons. Electrons are transferred from quinol to the Mo-site via [Fe-S] centers and cytochromes (Richardson et al., 2001). Nitrite reductase catalyzes the one electron reduction of nitrite to nitric oxide (Fig. 2A). The mid-point potential of this reaction is +370 mV. This is the defining reaction of denitrification as it is the first step in which a gaseous compound is produced (Payne, 1981). There are two unrelated NO producing nitrite reductases, both of which are found in the periplasm. One is a copper-containing enzyme while the other is a heme-containing enzyme (Watmough et al., 1999). Both receive electrons from c-type cytochromes or copper proteins that receive electrons from the cytochrome bc1 complex, and are found in the periplasm.
625
Fig. 2. (a) The enzymes of denitrification and assimilatory nitrate reduction and their designations. Reaction 1 is catalyzed by Nar – membrane bound nitrate reductase, Nap – periplasmic nitrate reductase or Nas – assimilatory nitrate reductase; 2 is catalyzed by Cu-Nir – copper nitrite reductase or cd1-Nir – heme nitrite reductase; 3 by qNor – quinol oxidizing NO reductase or cNor – cytochrome c oxidizing NO reductase; 4 is catalyzed by Nos – nitrous oxide reductase; and 5 is catalyzed by ANir – assimilatory nitrite reductase. (b) The topological location of the enzymes found in photosynthetic bacteria that are involved in reduction of nitrate and other nitrogen oxides. UQH2 indicates ubiquinol and cytochrome c indicates periplasmic c-type cytochrome.
Nitric oxide reductase is found exclusively in the membrane (Fig. 2B). This is not too surprising since NO, like oxygen, is lipid soluble and will concentrate in membranes (Liu et al., 1998). For many years, there was a controversy over whether NO was an obligatory intermediate (LeGall et al., 1979; Garber and Hollocher, 1981; Averill and Tiedje, 1982). It was eventually found that inactivation of the genes encoding the NO reductase was lethal due to the accumulation of NO, ruling out the possibility that NO was a minor product generated by side reactions of the other reductases (Braun and Zumft, 1991). NO reductase is related to the heme copper oxidase superfamily (Saraste and Castresana, 1994). A critical difference between Nor and heme copper oxidases is that the copper at the active site of the latter is replaced by iron in the former. Nitric oxide reductases receive electrons from either quinol or c-type cytochrome (Zumft, 2005) (Fig. 2A and 2B). The form receiving electrons from cytochromes is a multi-subunit enzyme referred to as cNor (Kastrau et al., 1994). The other, referred to as qNor, is a single subunit enzyme. Photosynthetic denitrifiers only have cNor. The single subunit in qNor is related to the active site containing subunit in cNor, but has an extension of about 300 amino acids (Cramm et al., 1997). The midpoint potential of the NO to reduction to N2O is about +1100 mV.
626 The final step in denitrification is the reduction of N2O to nitrogen (Fig. 2A). Nitrous oxide reductase is a novel copper-containing enzyme (Zumft and Kroneck, 2006). Like NO reductase, it has a link to oxygen reducing heme copper oxidases since the unusual binuclear CuA site found in aa3-type oxidases also is found in nitrous oxide reductase (Farrar et al., 1991). Nitrous oxide reductase is found in the periplasm, and in most denitrifiers is part of the cytochrome bc1 complex dependent respiratory chain (Fig. 2B). The midpoint potential for the N2O to N2 reduction is +1350 mV. The regulation of the four denitrification enzymes is complex and involves multiple transcriptional regulators. Denitrification is an alternative to oxygen respiration so the proteins are not expressed unless oxygen levels are low (Zumft, 1997). However, nitrate reductase and nitrous oxide reductase are expressed at higher oxygen levels than the other two reductases (Korner and Zumft, 1989; Coyne and Tiedje, 1990). Besides a decrease in oxygen, some form of nitrogen oxide must also be present for optimal expression. Nitrate, NO and nitrous oxide all appear to be signal molecules (Zumft, 1997). NO appears to be a key signal for expression of both nitrite and nitric oxide reductase and may activate the transcriptional regulatory protein designated NNR for nitrite and nitric oxide reductase regulator (Zumft, 2002). N2O will induce expression of nitrous oxide reductase (Korner and Zumft, 1989), but in some cases NO may induce nos expression (Arai et al., 2003). B. Occurrence of Denitrification In the Purple Photosynthetic Bacteria The capacity for denitrification can be found in Gram positive and Gram negative bacteria, archaea, fungi and recently there has been a report of a denitrifying foraminifer (Risgaard-Petersen et al., 2006). The phylogenetic group with the most denitrifiers is the proteobacteria with denitrification being particularly prominent among the α-proteobacteria (Shapleigh, 2007). Therefore, it is not surprising that many purple photosynthetic bacteria are denitrifiers. While many of the early isolates were complete denitrifiers, the so called partial denitrifiers, which contain a subset of the four N-oxide reductases required to produce N2, are also wide spread among this group. In general, purple photosynthetic bacteria most commonly use Nap, the copper containing Nir and cNor. A few contain Nar and the heme containing Nir but none
James P. Shapleigh contain qNor. Since purple sulfur photosynthetic bacteria are strict anaerobes there are no reports of denitrification in this group. Comparison of the denitrification capacities of the three Rhodobacter (Rba.) sphaeroides and five Rhodopseudomonas (Rps.) palustris strains whose genomes are available is an effective way of demonstrating the variability in the distribution of denitrification components among photosynthetic denitrifiers (http://img.jgi.doe.gov/cgi-bin/pub/main.cgi). Denitrification has been studied extensively in Rba. sphaeroides. Rba. sphaeroides forma sp. denitrificans was the first photosynthetic denitrifier isolated but other strains of Rba. sphaeroides were soon found to be denitrifiers (Satoh et al., 1976; Michalski and Nicholas, 1988). Another strain, designated 2.4.3 (also known as ATCC 17025), was characterized as a complete denitrifier and has been used as a model organism for the study of the regulation of denitrification in this group. Rba. sphaeroides strains that are complete denitrifiers utilize Nap, Cu-Nir, and cNor (http://img.jgi.doe.gov/cgi-bin/pub/main.cgi). Phenotypic analysis of Rba. sphaeroides 2.4.1, the Rba. sphaeroides type strain, suggested it was not a denitrifier since it was unable to grow as a denitrifier and did not have nitrite reductase activity (Michalski and Nicholas, 1988). A more thorough analysis found, however, that 2.4.1 contains Nap and cNor making it a partial denitrifier (Kwiatkowski et al., 1997). Genome sequencing has confirmed that Rba. sphaeroides 2.4.1 is a partial denitrifier lacking both Nir and Nos (http://img.jgi.doe.gov/cgi-bin/pub/main.cgi). Sequencing of the genome of Rba. sphaeroides 2.4.9 (ATCC 17029) showed that this strain has only nor (http://img.jgi.doe.gov/cgi-bin/pub/main.cgi). The region of the genome containing nor is conserved among the Rba. sphaeroides strains 2.4.1, 2.4.3 and 2.4.9, indicating that nor was present before they diverged. Interestingly, a large photosynthetic gene cluster is adjacent to nor and may account for this region having complete synteny in all three genomes. The nap genes in Rba. sphaeroides 2.4.1 are located on a plasmid (Mackenzie et al., 2001), and the loss of this plasmid may explain the loss of nap in strain 2.4.9. Strain 2.4.3 contains two nap clusters. It is uncertain if either or both of the clusters are on a chromosome or a plasmid since the genome sequence is not complete. However, one of the clusters is found on the largest assembled contig, which is about 275 kb in length. A fragment this size is unlikely to come from a plasmid (Michalski and Nicholas, 1988).
Chapter 31
Nitrogen Oxide Reduction
627
Table 1. Content and type of N-oxide reductases found in purple photosynthetic bacterium as revealed by analysis of available genomes Bacterium Nitrate reductase Nitrite reductase NO reductase N2O reductase cNor + Bradyrhizobium sp. BTAi1 Nap, Nas Cu-Nir, ANirb,c cNor + Dinoroseobacter shibae DFL Nap, Nas cd1-Nir, ANir 12 Rhodobacter capsulatus – – – + SB1003 Rhodobacter sphaeroides 2.4.1 Nap – cNor – a Cu-Nir cNor + Rhodobacter sphaeroides ATCC Nap (x2) 17025 Rhodobacter sphaeroides ATCC – – cNor – 17029 Rhodobacterales bacterium Nas Cu-Nir, ANir cNor + HTCC2654 Rhodopseudomonas palustris cNor + Nas Cu-Nir, ANirc BisA53 Rhodopseudomonas palustris Nap – – + BisB18 Rhodopseudomonas palustris – – – – BisB5 Rhodopseudomonas palustris cNor + – Cu-Nir (x2)a CGA009 Rhodopseudomonas palustris – + Nas Anirc HaA2 Roseobacter denitrificans OCh cNor + Nar, Nas cd1-Nir, Anir 114 Roseovarius sp. 217 Nar Cu-Nir cNor + a Genome contains two copies; bAssimilatory nitrite reductase; cContains a truncated form of assimilatory nitrite reductase in the Nas cluster
The Rba. sphaeroides 2.4.3 strain is the only strain among the sequenced Rba. sphaeroides strains that has nitrite reductase (Table 1). One explanation for this is that strain 2.4.3 acquired nirK after the strains diverged, but available evidence is inconsistent with this interpretation. All three Rba. sphaeroides strains contain a gene encoding pseudoazurin in a locally conserved region of DNA. In strain 2.4.3, nirK clusters with the pseudoazurin gene (Jain and Shapleigh, 2001). Pseudoazurin is a copper protein that donates electrons to Nir in many denitrifiers (Pearson et al., 2003). It would seem unlikely that nirK would randomly insert next to a gene whose product interacts with Nir. Therefore, it is more likely that all Rba. sphaeroides strains contain pseudoazurin because at one time they all contained Nir. Then, as the strains diverged, nirK was lost from most via deletion. Interestingly, the pseudoazurin genes in all three Rba. sphaeroides strains contain identical mutations that prevent copper from binding to this protein (Jain and Shapleigh, 2001). The significance of this mutation is
unclear. The Rba. sphaeroides 2.4.3 strain contains a second copper protein that is not found in the other two strains that could be involved in transferring electrons to Nir. Why is NO reductase the only denitrification protein common to all three strains of Rba. sphaeroides? NO produced by denitrifiers is freely diffusible and has been found to activate expression of genes in the NO stimulon of other bacteria (Choi et al., 2006). NO reductase has been shown to help reduce toxicity of exogenous NO (Choi et al., 2006). This suggests that Nor is used to mitigate NO toxicity. Since the majority of Rba. sphaeroides strains that have been isolated seem to be partial denitrifiers, the question becomes why do some strains retain the complete denitrification pathway? The obvious answer is that it provides a bioenergetic benefit, but this benefit may not be sufficient to compensate for the hazards associated with Nir turnover. A diverse range of denitrification potential has also been revealed in the five Rps. palustris strains
628
James P. Shapleigh
Fig. 3. Gene maps indicating organization and transcriptional direction of some of the denitrification gene clusters discussed in the text. Capital letters indicate genes whose designation is derived by combining the letter with the cluster designation. For example, K in the Rsb. denitrificans nar cluster is narK. Genes left blank indicate there is no known function for the designated open reading frame. Arrows indicate direction of transcription. If a designation cannot fit into the space available it is placed immediately below the gene location. Genes encoding proteins of defined function are: narK - nitrate/nitrite antiporter, narG – catalytic subunit, narH – electron transfer subunit, narI – electron transfer subunit, napC – membrane anchored electron transfer subunit, napB – electron transfer subunit, napA – catalytic subunit; nnrR – nitrite and nitric oxide reductase regulatory protein; nirK(1 or 2) – Cu-nitrite reductase catalytic subunit; hemN- oxygen independent coproporphyrinogen oxidase; nirS – cd1 nitrite reductase catalytic subunit, nirC – c-type cytochrome, nirD/F/G/H/J – involved in heme biosynthesis; norC – electron transfer subunit, norB – catalytic subunit; nosZ – catalytic subunit; nosL – outer membrane copper chaperone.
whose genomes have been sequenced (http://img. jgi.doe.gov/cgi-bin/pub/main.cgi) (Table 1). There have only been a few studies on denitrification in this bacterium so little is known about the role of N-oxide reductases in the growth of Rps. palustris (Preuss and Klemme, 1983; McEwan et al., 1985). Only one of the five, strain BisB18 has a dissimilatory nitrate reductase as it contains the nap gene cluster. Two of the strains of Rps. palustris, CGA009 and BisA53, encode copper type nitrite reductases (Fig. 3). Rps. palustris CGA009 is somewhat unique in having two copies of nirK.
Not surprisingly, the two strains with nirK also contain the nor operon while none of the others have these genes. This suggests that Nir is functional in these species and Nor is used to limit NO accumulation. However, preliminary results indicate that NO dependent expression of neither of the two nirK was found in the Rps. palustris CGA009 strain (D. Y. Lee and J. P. Shapleigh, unpublished). The nirK designated nirK2 (RPA4145) in the annotated genome of Rps. palustris was expressed in Rba. sphaeroides 2.4.3, but nirK1 (RPA3306) was not (D. Y. Lee and J. P. Shapleigh, unpublished). The nos genes are the most
Chapter 31
Nitrogen Oxide Reduction
common denitrification genes in the Rps. palustris strains, with only Rps. palustris BisB5 lacking them. While there is a high level of identity between the primary sequences of the N-oxide reductases present in the various strains there is little obvious conservation of their location in the chromosome. For example, there is no synteny in the regions flanking the nor clusters in the Rps. palustris strains CGA009 and BisA53. Moreover, the nor cluster in Rps. palustris strain BisA53 contains a gene encoding a putative transposase protein that disrupts the four-gene nor operon, which is conserved among almost all denitrifiers (Zumft, 2005). This is in contrast to Rba. sphaeroides, where there is synteny in the region of the genome containing the nor operon. This lack of synteny in Rps. palustris might reflect sporadic acquisition of the denitrification genes by the strains. However, pairwise comparisons of denitrification proteins from the various Rps. palustris strains show they are all closely related. Pair-wise comparisons of the Rps. palustris N-oxide reductases yielded levels of identity similar to values obtained in pair-wise comparisons of the same proteins from the Rba. sphaeroides strains. This high level of identity suggests the denitrification genes were not acquired independently by each strain of Rps. palustris, but as in Rba. sphaeroides, they have been present in the genome of Rps. palustris since before the strains diverged. If this is the case, then it seems that the chromosomes of the Rps. palustris strains undergo frequent rearrangements leading to genes being scattered or lost, particularly if they are not essential or strongly beneficial, as is likely the case with the denitrification genes. Analysis of the available genomes suggests that partial denitrification is as common as complete denitrification in purple photosynthetic bacteria. Another notable example of this situation is Rba. capsulatus SB1003, whose genome contains only nos. Biochemical evidence has suggested that NO reductase is present in some strains of Rba. capsulatus although DNA sequence evidence for this is lacking (Bell et al., 1992). Some Rba. capsulatus strains also contain an assimilatory nitrate reductase, which is also absent in the sequenced strain (Pino et al., 2006). If denitrification provides only limited selectable benefits, it is possible that some of the denitrification genes could be lost during laboratory domestication. However, recent isolates such as the Rhodobacterales bacterium HTCC2654 are partial denitrifiers. Available genome sequence indicates this bacterium lacks a dissimilatory nitrate reductase. Our repeated
629 transfer of Rba. sphaeroides strain 2.4.3 without loss of any of the denitrification reductases also argues against the denitrification genes being unusually prone to deletion. C. Other Genes Associated With Denitrification Gene Clusters Given that the genes required for synthesis and assembly of each individual N-oxide are found in clusters, and in many cases these clusters are grouped together, it is informative to determine if there are other genes that appear with some frequency in these regions. One interesting association is the occurrence of hemN in denitrification related clusters. hemN encodes an oxygen-independent coproporphyrinogen III oxidase (Coomber et al., 1992). In all Rba. sphaeroides strains hemN is found downstream of, and in the same orientation as, the nor gene cluster (Bartnikas et al., 1997) (Fig. 3). Inactivation of hemN renders the cells unable to grow photosynthetically but does not affect aerobic respiration (Coomber et al., 1992). There is a paralog of hemN in Rba. sphaeroides, designated hemZ that clusters with the genes encoding the cbb3-type cytochrome c oxidase (Oh et al., 2000). This grouping seems unlikely to be coincidental since the latter and Nor are members of the same family of enzymes (Saraste and Castresana, 1994). In Bradyrhizobium sp. BTAi1 hemN clusters with nirK (Fig. 3). In this bacterium a second copy of hemN does not appear to be present. In Roseobacter (Rsb.) denitrificans, which utilizes a heme-type Nir, hemN clusters with nnrS and nnrU (see below). In Rhodobacterales bacterium HTCC2654 hemN does not cluster with either nor or the cco gene cluster. This is also the case with the Rps. palustris strains. It is not immediately obvious why a gene essential for photosynthetic growth would be found associated with a denitrification gene cluster. However, in several other non-photosynthetic α-proteobacterial denitrifiers such as Sinorhizobium meliloti, Rhizobium etli CFN 42 and B. japonicum USDA strain 110 hemN clusters with either nirK or nor. In these bacteria denitrification may be the primary means of growth under microoxic or anoxic conditions, so the clustering of hemN with a denitrification cluster fits its physiological function. The clustering observed in photosynthetic denitrifiers may indicate that the hemN used to support anoxic growth was acquired via horizontal transfer from non-photosynthetic denitrifiers.
630 Another gene almost always found in the denitrification cluster but of unknown function is nnrS. The nnrS product has been shown to be an integral membrane protein that contains two b-type cytochromes and a type 2 copper (Bartnikas et al., 2002). This organization is similar to that of the subunit I proteins of heme copper oxidases, which contain the site where oxygen is reduced (Hosler et al., 2006). However, there has been no oxygen reducing or NO reducing activity detected in purified NnrS (Bartnikas et al., 2002). Genome sequencing has revealed NnrS occurs in many bacteria and can be easily identified by a highly conserved WHxHEM motif. In Rba. sphaeroides nnrS clusters with nor and nnrR (the NNR ortholog in this bacterium), and is a member of the NnrR regulon (Bartnikas et al., 2002) (Fig. 3). In Rhodobacterales bacterium HTCC2654 nnrS also clusters with nor and nnrR, and is likely a member of the NNR regulon in this bacterium. In Bradyrhizobium species BTAi1 nnrS clusters with nnrR and nirK (Fig. 3). Orthologs of the Rba. sphaeroides NnrS are found in all five strains of Rps. palustris, even the non-denitrifying BisB5 strain. This suggests the function of NnrS is not obligately linked to denitrification. In Rba. capsulatus SB1003, a gene encoding a protein that is a member of the NnrS family is found to cluster with the genes encoding the assimilatory nitrate reductase genes (Haselkorn et al., 2001). This is interesting since assimilatory nitrate reductase is a potential source of NO (Klepper, 1987). Some bacteria such as Rba. sphaeroides contain paralogs of nnrS. The paralogs are not found in association with denitrification genes, and their genomic location does not provide any obvious clues to the function of this protein. Another gene found frequently in the denitrification gene clusters has been termed nnrU. The product of this gene, like that of nnrS, is a membrane protein of unknown function. In Rba. sphaeroides strains, nnrU is downstream of the nor operon and immediately upstream of hemN (Fig. 3). Little is known about this protein, but its occurrence in most denitrifiers suggests that it is important under certain environmental conditions. It is found in all Rps. palustris strains and clusters with nnrS, suggesting a physiological connection between the two, but this remains to be determined. D. Physiological Role of Denitrification Denitrification is typically considered to be used as
James P. Shapleigh a process that allows ATP generation when oxygen concentrations become limiting. It is important to remember, however, that denitrification is not a strictly anoxic process. The nitrogen oxide reductases are expressed when some oxygen is present (Korner and Zumft, 1989). Growth under low oxygen conditions in nitrate-supplemented medium is not always due to denitrification though, as the presence of high affinity oxidases allows oxygen dependent ATP generation to occur under extremely low oxygen concentrations (Preisig et al., 1996). For example, we find that Rba. sphaeroides strain 2.4.3 will grow quite well in the dark in an atmosphere that has been flushed with commercial grade, high purity N2. This growth occurs with or without nitrate, but if nitrate is present the denitrification genes are expressed (A. K. Hartsock and J. P. Shapleigh, unpublished). Inactivation of ccoN, encoding a structural subunit of the high O2 affinity cbb3-type cytochrome c oxidase, severely restricts growth under the N2 atmosphere indicating O2 respiration is the primary means of energy generation under this condition. Therefore, it can be difficult to determine if growth is actually nitrogen oxide dependent, since trace oxygen is almost always present in experiments done to assess the contribution of nitrate reduction to growth. The most reliable method for confirming if growth is nitrogen oxide dependent is to compare growth on medium with and without nitrate. If cells in media without nitrate show much slower rates of growth this indicates that denitrification is being used to support growth in some way. When working with photosynthetic bacteria nitrate respiration can be assessed in either light or dark conditions. Under photosynthetic conditions it has been found that the addition of nitrate reduces doubling times (Satoh et al., 1976). However, this may not be due to nitrate respiration, but instead, to a process known as redox balancing, which is discussed below. In some cases it has been hard to demonstrate growth via denitrification under dark, strictly anoxic conditions. In the initial paper describing Rba. sphaeroides strain 2.4.3 as a denitrifying strain it was found that it, as well as the denitrifying Rba. sphaeroides strain 81–3 had doubling times of about 40 h under anoxic conditions in the dark (Michalski and Nicholas, 1988). Our efforts to grow Rba. sphaeroides strain 2.4.3 under strictly anoxic conditions have yielded similar results (Laratta et al., 2002). There is no obvious bioenergetic reason why growth should be this slow given that doubling time under aerobic conditions is
Chapter 31
Nitrogen Oxide Reduction
about 1 h. These strains obviously can grow under anoxic conditions since they will grow with relatively rapid doubling times when light is present. In Rba. sphaeroides f. sp. denitrificans IL106 growth has been reported under anaerobic conditions although how the anaerobic conditions were achieved was not described (Sabaty et al., 1999). More significantly, loss of nitrate reductase activity prevented growth under anaerobic conditions as would be expected if nitrate were being respired. This variability in the ability of denitrification to support growth under conditions where it would be the primary mode of respiration might reflect the fact that denitrification may fill a supplementary role in some photosynthetic denitrifiers. This may explain the prevalence of partial denitrifiers among the purple photosynthetic bacteria. Implicit in this conclusion is that partial denitrification is as beneficial as complete denitrification. There is evidence demonstrating that partial denitrification is beneficial to some photosynthetic bacteria. Rba. sphaeroides DSM158 contains Nap but lacks Nir activity so the nitrite produced cannot be further reduced. Nap activity by itself does not support anoxic growth, likely due to the inefficient generation of a proton motive force (McEwan et al., 1984). However, it seems likely that in Rba. sphaeroides Nap activity is beneficial since it allows balancing of the cellular redox pool (Gavira et al., 2002). Under certain situations, particularly in the presence of reduced carbon sources, the redox pool can become over-reduced and processes that allow cells to rapidly dispose of these excess reducing equivalents become beneficial. In Rba. sphaeroides DSM158 nap expression does not change in response to changes in carbon source but Nap activity under photosynthetic conditions does increase as the carbon source becomes more reduced (Gavira et al., 2002). This increase in activity under photosynthetic conditions, where disposal of reducing equivalents is important for growth, indicates Nap activity is beneficial, allowing excess reducing equivalents to be used in nitrate reduction. Another example of the role of Nap for redox balancing is provided by Rba. capsulatus strain N22DNAR+, which has Nap but lacks Nir enzymes (McEwan et al., 1982). The presence of this Nap activity was shown to benefit cells transitioning between light and dark growth conditions in continuous cultures (Ellington et al., 2003). Under dark conditions the cells reducing nitrate maintained higher viable cell numbers. In co-cultures of strains with and without
631 Nap, cells with Nap out-competed those without Nap under cycles of light and dark (Ellington et al., 2003). These results suggest that the electron transfer to Nap aids cell survival by maintaining optimal redox balance and a minimal transmembrane electrochemical gradient. Although not directly tested, nitrous oxide reduction could play a similar role in strains such as Rba. capsulatus SB1003 that cannot make nitrous oxide but contain nitrous oxide reductase (McEwan et al., 1985). While Nor could play a role in redox balancing it may also play an important role in mitigating the toxicity of NO. NO is a diffusible, reactive free radical that has both direct and indirect effects on cellular components (Wink and Mitchell, 1998). In complete denitrifiers the loss of Nor activity is conditionally lethal due to the accumulation of NO under denitrifying conditions (Braun and Zumft, 1991). Strains unable to make NO, such as Rba. sphaeroides 2.4.1, retain an effective mechanism for reducing exogenous NO since they have an NO inducible Nor activity (Kwiatkowski et al., 1997). The source of NO in most environments is likely to be other denitrifiers. It has been shown in co-cultures that the NO produced by a denitrifier can activate expression of nor in Rba. sphaeroides 2.4.1 (Kwiatkowski et al., 1997). The observation that all Rba. sphaeroides strains examined so far contain nor suggests that these strains are affected by the NO produced by other denitrifiers, and that Nor activity is beneficial for growth even though it is not part of a complete denitrification pathway. E. Impacts of Denitrification on Cell Phenotype Denitrifying cells of Rba. sphaeroides 2.4.3 and Rba. sphaeroides IL106 can be readily identified by their unique color (Satoh et al., 1976). When cells of either strain are grown without nitrate under limiting oxygen they are reddish brown. Cells grown with nitrate under the same conditions are greenish brown. Pigmentation of Rba. sphaeroides comes from a variety of compounds including heme, bacteriochlorophyll and carotenoids. Cells of Rba. sphaeroides IL106 grown photosynthetically in the presence of nitrate have markedly lower concentrations of certain carotenoids and bacteriochlorophyll a (Michalski and Nicholas, 1985; Michalski and Nicholas, 1987). Since addition of nitrite to the medium of cells growing photosynthetically also causes a change in pigment levels this suggests the
James P. Shapleigh
632 phenotypic changes are either due to nitrite or nitrite reduction (Michalski and Nicholas, 1985). Cells of a Nir-deficient mutant of Rba. sphaeroides 2.4.3 do not take on the appearance of denitrifying cells (J. P. Shapleigh, unpublished). This indicates either Nir activity or NO is required for the changes observed in wild type cells. If it is Nir activity per se then the phenotypic changes might be due to the availability of an alternate electron acceptor influencing the overall redox balance of the cell. However, Rba. sphaeroides 2.4.1 growth with nitrate does not take on the same color as Rba. sphaeroides 2.4.3 suggesting that it is not the redox balancing but rather the Nir activity that causes the color change. It is not unreasonable to speculate that NO might be responsible for affecting pigment synthesis, as NO could affect the enzymes responsible for the biosynthesis of carotenoids and bacteriochlorophyll. In addition to changes in pigmentation, Rba. sphaeroides cells grown photosynthetically in the presence of nitrate have lower numbers of reaction center complexes (Michalski and Nicholas, 1985). A decrease in the LH2/LH1 ratio with a significant decrease in the production of LH2 has also been reported (Michalski and Nicholas, 1985). The presence of nitrate also results in the production of long tube-like structures in cells (Sabaty et al., 1994). Similar structures have been observed in cells lacking LHII, so their formation may be a consequence of changes in levels of light harvesting complexes. It is unknown if denitrification causes similar shifts in the photosynthetic apparatus of other denitrifiers. F. Regulation of Denitrification As with most aspects of denitrification in purple photosynthetic bacteria much of the work on the regulation of denitrification has been carried out in various strains of Rba. sphaeroides. Given the variability in content and organization of the denitrification genes in the photosynthetic bacteria it is doubtful that the specific regulatory mechanisms identified in Rba. sphaeroides occur in all other photosynthetic denitrifiers. Even the basic tenet that oxygen represses expression of the denitrification genes may not be true in all members of this group (see below). Nonetheless, it seems likely that in most of these bacteria regulation of denitrification involves mechanisms to sense oxygen or redox status as well as nitrogen oxides. Each N-oxide reductase seems to require its own regulatory factors and so each
is discussed separately. However, it is important to remember that in complete denitrifiers the reduction of nitrate or subsequent intermediates, with the exception of nitrous oxide, leads to the endogenous production of nitrogen oxide effectors which induce expression of multiple steps in the pathway. This is not the case in partial denitrifiers since steps are missing so in many cases critical signal molecules must be provided exogenously. 1. Nitrate Reductase The most detailed studies of expression of nitrate reductase have been carried out in Rba. sphaeroides. In strain Rba. sphaeroides DSM158, a partial denitrifier lacking Nir (Martinez-Luque et al., 1991), the nap genes are expressed under both aerobic and anoxic conditions with the highest expression occurring under oxic conditions (Gavira et al., 2002). Nitrate does not appear to be an inducer of nap expression. While there is detectable Nap in aerobically grown cells it has little activity since nitrite does not accumulate during growth. However, under anoxic photosynthetic conditions nitrite does accumulate. Cells grown photosynthetically without nitrate rapidly gain Nap activity when nitrate is added to the medium, indicating Nap is present but requires an unknown posttranslational modification to gain activity (Gavira et al., 2002). In Rba. sphaeroides IL106 optimal nap expression was found to be nitrate dependent (Tabata et al., 2005). Expression of nap genes in Rba. sphaeroides IL106 was similar under oxic and anoxic conditions in nitrate amended medium. There was a decrease in nap expression in a nap mutant background under photosynthetic conditions (Tabata et al., 2005). Mutation of nap had no impact on aerobic nap expression indicating that there is a difference in the regulatory mechanisms under the two conditions. In Rba. sphaeroides IL106, expression of the nap operon under photosynthetic conditions increased as the carbon source became more reduced (Tabata et al., 2005). This trend was not observed in Rba. sphaeroides DSM158 since expression was not influenced by the nature of the carbon source (Gavira et al., 2002). The differences in the regulatory patterns of Rba. sphaeroides strains DSM158 and IL106 probably reflect the fact that the former is a partial denitrifier while the latter is a complete denitrifier. In Rba. sphaeroides DSM158 Nap would primarily be used for redox balancing but in Rba. sphaeroides IL106 Nap activity
Chapter 31
Nitrogen Oxide Reduction
633
is coupled to growth since the nitrite produced can be further reduced. The transcriptional factors involved in regulating nap have not been identified. 2. Nitrite Reductase Members of the DNR/NNR family of transcriptional regulators appear to be critical for expression of nitrite reductase in the photosynthetic bacteria. These proteins are members of the CRP/FNR family of transcriptional activators (Korner et al., 2003). Most purple bacteria have a member of the NNR family, because this group of proteins is found in bacteria with copper containing nitrite reductases (Korner et al., 2003). In those cases where it has been studied NNR also regulates expression of nitric oxide reductase (Zumft, 2002). This coordinated expression is useful in ensuring that Nir expression is coupled to Nor expression to minimize the accumulation of nitric oxide. NNR appears to be a NO sensor although this has not been directly shown. The conclusion that NNR is a NO sensor is based on the observation that nirK is poorly expressed in cells lacking Nir activity, and that addition of NO donors or NO-generating denitrifiers to cultures of Nir deficient strains increases expression of nirK in the Nir deficient strains (Tosques et al., 1997; Yin et al., 2003). These results suggest that nirK expression is dependent on NO suggesting that NNR is an NO sensor. The mechanism by which NNR senses NO is unclear since purified protein will not bind DNA. The inability to isolate functional protein also makes it difficult to definitively identify the region of DNA targeted by NNR. However, genes regulated by NNR have a sequence upstream of the translation start site which is a variation of the sequence TTGN(8)CAA. Since this sequence is similar to the FNR consensus motif, it is the likely NNR binding site. Evidence in support of this hypothesis has been obtained in Rba. sphaeroides, where the NNR ortholog is termed NnrR, since elimination of this sequence prevents expression of both nirK and nor (Bartnikas et al., 1997; Tosques et al., 1997). In addition, mutation of the putative NnrR binding site to a perfect FNR consensus sequence, shifted nnrS from the NnrR regulon to the FNR regulon (Bartnikas et al., 2002). By using a consensus sequence based on the NnrR sites upstream of nirK and nor the size of the NnrR regulon in Rba. sphaeroides 2.4.3 can be estimated to be about six genes (Fig. 4). This includes nirK, nor, nnrS as well as nnrR itself, which is negatively
Fig. 4. Regulation of NnrR dependent genes in Rba. sphaeroides 2.4.3. Dashed lines indicate an uncertain interaction. cbb3 oxidase is the high affinity cytochrome c oxidase that is used when O2 concentrations are limiting. PrrBA is the two-component sensor regulator so named because they are photosynthetic response regulators (called RegAB in Rba. capsulatus). The other gene designations are described in the text.
auto regulated (Kwiatkowski et al., 1997). In addition, there is norEF and a gene encoding a putative pseudoazurin that is likely to be an electron donor to Nir. These latter two genes are absent from Rba. sphaeroides 2.4.1 genome. Interestingly, a perfect NnrR site is found in the regulatory region of RSP2952 of Rba. sphaeroides 2.4.1, which encodes a protein similar to BolA of E. coli. Analysis of the expression of RSP2952 found that its expression was not NnrR dependent (R. Zordan and J.P. Shapleigh, unpublished). The orthologous gene in Rba. sphaeroides 2.4.3 does not have a NNR site. In Rba. sphaeroides nirK expression is also regulated by the two-component sensor-regulator PrrB-PrrA (called RegB-RegA in Rba. capsulatus), respectively (Laratta et al., 2002) (Fig. 4). PrrB is the sensor that interacts with the cbb3-type cytochrome c oxidase (Oh and Kaplan, 2001). High oxygen levels inhibit the phosphorylation by PrrB of the regulator PrrA (Oh et al., 2001). (For a more thorough description of PrrBA, see Chapter 35, Bauer et al.). Binding sites for PrrA are poorly conserved and consequently hard to identify by sequence analysis. Truncations of the nirK regulatory region indicate that the PrrA binding is within about 100 bp of the NnrR binding site (Laratta et al., 2002). Available evidence suggests that PrrA is not directly involved in the regulation of the other members of the NnrR regulon in Rba. sphaeroides 2.4.3 (Laratta et al., 2002). However, NnrR activation requires Nir activity, and PrrBA is required for nirK transcription, suggesting that PrrBA function is critical for expression of all genes in the NnrR regulon. Little is known about the expression of nirK and nor in other purple photosynthetic bacteria. The nirK in Rps. palustris strain 1a1 is maximally expressed under low oxygen but it does not seem to require
634 nitrate or nitrite for expression (Preuss and Klemme, 1983). This could indicate that nirK is controlled by FNR. However, examination of the available Rps. palustris genomes indicates that those strains that have nirK also possess putative NNR orthologs. Using lacZ fusions we have found that both nirK genes in Rps. palustris CGA009 show little expression under limiting oxygen, even if nitrite or a NO generating compound such as sodium nitroprusside is present (D. Y. Lee and J. P. Shapleigh, unpublished). This is somewhat surprising since nirK2 and nnrR are adjacent (Fig. 3), and putative NNR binding sites are present upstream of nirK2. Similar sites can be found upstream of norEF and the norCBQD cluster in Rps. palustris CGA009. Insight into the reason for the poor expression of nirK in Rps. palustris awaits further investigation. 3. Nitrous Oxide Reductase Regulation of nos has been poorly studied in photosynthetic denitrifiers. In a nitrate reductase deficient strain Rba. sphaeroides IL106 expression of nos is lost if nitrate is the sole nitrogen oxide present (Sabaty et al., 1999). Nitrous oxide was able to induce nos expression in this strain suggesting it is an important signal molecule. In Rba. capsulatus strain N22 Nos activity has been reported when grown on nitrate but this strain is suggested not to have Nir activity (McEwan et al., 1982; McEwan et al., 1985). The signal for Nos expression is unclear when this strain is grown on nitrate. There have been no detailed molecular studies on the regulation of nos in photosynthetic bacteria. In some denitrifiers the DNR/NNR proteins appear to be required for expression of nos genes (Vollack and Zumft, 2001; Arai et al., 2003). However, N2O by itself can induce expression, explaining how bacteria with nos but unable to make NO can induce nos expression (Korner and Zumft, 1989; Sabaty et al., 1999). Along with the DNR/NNR proteins a protein designated NosR has been found to be essential for nos expression (Wunsch and Zumft, 2005). This protein is found in all the genomes of purple photosynthetic bacteria that contain the nos structural genes. In Rba. sphaeroides 2.4.3, however, nosR contains a frameshift, which may explain the poor Nos activity of this strain (Choi et al., 2006). NosR is a membrane-associated protein that contains a flavin and [Fe-S] centers (Wunsch and Zumft, 2005). The function of NosR is unknown but is required for expression of the nos cluster and is
James P. Shapleigh also likely required for maintaining enzyme stability (Zumft and Kroneck, 2006). 4. Is Roseobacter denitrificans an Exception? In most denitrifiers oxygen is the preferred electron acceptor and represses the expression of genes encoding N-oxide reductases. One bacterium may prove to be an exception to this general rule. Rsb. denitrificans is unique among most photosynthetic bacteria in that genes required for photosynthesis, such as puf, show high levels of expression under aerobic conditions (Nishimura et al., 1996). Aerobically grown cells have also been found to have nitrogen oxide reductase activity (Doi and Shioi, 1991). In these experiments the cells were grown aerobically but the denitrification assays were done under anaerobic conditions. The production of N2O from nitrate was immediate and occurred in the presence of chloramphenicol indicating the denitrification proteins were expressed before the shift to anaerobic conditions. A copper-containing Nor was purified from aerobically grown cells of Rsb. denitrificans providing further evidence of aerobic expression of nitrogen oxide reductases. Exposure to light increased the denitrification activity of cells grown under aerobic conditions (Doi and Shioi, 1991). The presence of nitrate significantly decreases pigment production irrespective of the presence of oxygen (Shioi et al., 1988). Since nitrate will support growth under dark, anaerobic conditions it seems likely denitrification is an anaerobic form of respiration in Rsb. denitrificans, but expression of the denitrification genes are unaffected by oxygen. This is unexpected since Rsb. denitrificans contains a member of the DNR family in the denitrification cluster, and it has been reported that oxygen inhibits members of this class of proteins (Lee et al., 2006). G. Other Factors Affecting Denitrification Activity Denitrification is also regulated by light. Isolated cells of Rba. sphaeroides IL106, which were grown photosynthetically in the presence of nitrate, stop nitrogen oxide reduction when exposed to light (Sabaty et al., 1993). Inhibition could be relieved by the addition of artificial electron donors such as TMPD. This indicates that for the cytochrome c dependent steps, light-dependent inhibition is likely a response to photosynthesis-dependent oxidation of
Chapter 31
Nitrogen Oxide Reduction
the cytochrome c pool. Since Nap receives electrons from quinol, which has a lower midpoint potential than TMPD, it is not as obvious why TMPD restores Nap activity. However, it is likely that this is due to the exogenous electron source preventing the oxidation of the electron transport chain. Similar inhibition of nitrate and nitrous oxide reductase by light has been observed in Rba. capsulatus (Satoh, 1977; McEwan et al., 1982; Richardson et al., 1989). While this light dependent inhibition certainly plays some role in limiting electron flow through the denitrification pathway this inhibition cannot be complete since denitrification proteins are expressed under photosynthetic conditions (Michalski and Nicholas, 1984; Michalski et al., 1986). Moreover, it has been shown that nitrate will increase the growth rate of Rba. sphaeroides 2–3 fold, demonstrating that denitrification and photosynthesis are concurrent (Satoh et al., 1976). Denitrification is also inhibited when oxygen is introduced to cells of Rba. sphaeroides IL106 that are actively denitrifying (Sabaty et al., 1993). The cause of the inhibition is similar to that in photosynthesis, i.e., a diversion of electrons to the oxygen dependent oxidases. H. The Enzymes of Denitrification This section will only include enzymes that have been purified from purple photosynthetic bacteria. There are a number of reviews available for a more complete description of studies on the N-oxide reductases (Richardson et al., 2001; Wherland et al., 2005; Zumft, 2005; Zumft and Kroneck, 2006). The NapAB from Rba. sphaeroides is the only heterodimeric Nap that has been crystallized (Arnoux et al., 2003). NapA is the catalytic subunit that contains molybdopterin, the cofactor directly involved in nitrate reduction. NapB is a smaller subunit that contains two c-type hemes that transfer electrons to NapA. One of these hemes is near an iron-sulfur [Fe-S] center in NapA, indicating a likely route of electron transfer between the two subunits. The other heme is near the surface, and is likely the acceptor of electrons from NapC, which is membrane associated and receives electrons from quinol. Analysis of the redox midpoint potential and spatial arrangement of the metal centers suggests a flow of electrons through a ‘slightly bent molecular wire’ that is likely to involve all of the cofactors (Arnoux et al., 2003). A Nar type nitrate reductase has not yet been purified from a photosynthetic denitrifier.
635 Copper containing nitrite reductases have been crystallized from several denitrifiers including Rba. sphaeroides 2.4.3 (Jacobson et al., 2005). The structure of the oxidized and reduced form of this enzyme were resolved to <2.0 Å. As expected, the structure of this enzyme was similar to that of related denitrifiers. Copper nitrite reductases are homotrimers with two coppers per monomer, referred to as Type 1 and 2 (Moura and Moura, 2001). The ligands for the Type 1 center are supplied by one monomer but residues from two different monomers supply the ligands for the Type 2 center. In the Nir from Rba. sphaeroides the Type I site has a midpoint potential of about 250 mV while that of the Type 2 center is less than 200 mV (Olesen et al., 1998). This suggests that nitrite binding to the oxidized Type 2 center allows electron transfer from the Type 1 center. Once the Type 2 center is reduced, the active site must facilitate removal of one of the oxygen atoms in nitrite. ENDOR analysis indicates there is an exchangeable proton near the Type 2 center that could be used to facilitate water formation (Zhao et al., 2002). Once the oxygen is removed the electronic state of the Cu and the NO is uncertain. NO bound to the Type 2 center of the Nir from Alcaligenes faecalis was reported to be bound side-on, but NO bound to the same site in Nir from Rba. sphaeroides was reported to be bound end-on (Tocheva et al., 2004; Usov et al., 2006). To facilitate release of NO from the Cu it is likely that Cu is oxidized, however the catalytic cycle leading to the formation of a Cu(II)-NO center is unclear. The only denitrifier among the photosynthetic denitrifiers known to have the cd1 type nitrite reductase is Rsb. denitrificans (formerly Erythrobacter species strain OCh114). This enzyme has been purified from aerobically grown cells of this bacterium (Doi et al., 1989). The enzyme is found in two spectroscopically distinguishable forms. While these enzymes have the same molecular mass they have different pIs. The genome sequence of Rsb. denitrificans indicates that it contains only a single copy of nirS, the Nir structural gene, suggesting that the differences in the enzymes are post-translational (Swingley et al., 2006). The role of this enzyme is unclear since it has greater oxygen reducing activity than nitrite reducing activity. As noted above Nor is related to the heme copper family of terminal oxidases. Almost all NO reductases purified from denitrifiers have a non-heme Fe at the active site but a Nor from Rsb. denitrificans has been purified that contains a copper in the active site (Matsuda et al., 2002). This enzyme was isolated
James P. Shapleigh
636 from aerobically grown cells. The protein had very limited NO reductase activity but was able to reduce oxygen to water, albeit at a relatively low rate. One explanation proposed for this result was that Rsb. denitrificans has two nor clusters, with one encoding the iron Nor and one the copper Nor (Matsuda et al., 2002). Genome sequencing indicates there is only one nor with high identity to iron containing nor genes, so the Nor protein being expressed is not obviously atypical. It is not clear what role this enzyme would play. Analysis of the Rsb. denitrificans genome suggests that it contains an aa3-type cytochrome c oxidase and a high oxygen affinity cbb3-type cytochrome c oxidase. In most photosynthetic denitrifiers these are the major oxygen reducing enzymes utilized during oxic growth so the Cu-Nor is unlikely to serve as an oxygen respiring enzyme. It is also not clear why copper is inserted in the active site instead of iron. The two ancillary genes in the nor operon, norD and norQ, which might be involved in metal insertion, are present in the Rsb. denitrificans genome (Bartnikas et al., 1997). Nor has not been purified from any other denitrifier, but FTIR analysis of membranes of denitrifying cells of Rba. sphaeroides found the active site binding pocket was similar to orthologs from related denitrifiers, and gave no evidence for copper in the binuclear center (Mitchell et al., 1998; Moenne-Loccoz and de Vries, 1998). Nos has been purified from at least one photosynthetic denitrifier (Sato et al., 1999). As found in most other denitrifiers the purified enzyme was a coppercontaining enzyme that is unstable in the presence of oxygen (Zumft and Kroneck, 2006). The Nos from two different non-photosynthetic denitrifiers has been crystallized helping to resolve much of the uncertainty about the structure of the copper centers (Brown et al., 2000; Haltia et al., 2003). One of the copper centers in Nos, termed CuA, is also found in the aa3-type cytochrome c oxidases used for oxygen respiration, further demonstrating the interrelationship of the enzymes involved in oxygen and nitrogen oxide respiration. This is a dinuclear site, and as in the aa3-type cytochrome c oxidase this site is used for electron transfer. The enzyme has a second copper center, referred to as the CuZ site that is the catalytic site. The CuZ site contains four coppers coordinated by histidines that are arranged in a novel butterfly cluster (Brown et al., 2000). The cluster is bridged by a sulfide. Nos requires a large complement of accessory enzymes for assembly and function. Besides nosZ (the
Nos structural gene) and nosR, the nos gene clusters typically contain several others genes including nosD, nosF, nosY and nosL (Fig. 3). A gene designated as nosX is found also on the cluster of many α-proteobacteria. NosD, F and Y encode an ABC transporter that is likely used to transport a component, perhaps copper or sulfur, necessary for Nos assembly. NosL is predicted to be an outer membrane protein, and has been shown to bind copper (McGuirl et al., 2001; Taubner et al., 2006). NosX encodes a flavoprotein essential for N2O respiration in some α-proteobacteria (Chan et al., 1997; Saunders et al., 2000). Its function is unknown but also likely plays a role in assembly of copper centers. Nir, Nor and Nos receive electrons via the cytochrome bc1 complex branch of the respiratory chain (Zumft, 1997) (Fig. 2B). There is some evidence that electrons can flow to Nos via a cytochrome bc1 complex independent branch (Richardson et al., 1989). In Rba. sphaeroides and Rba. capsulatus, it has been shown that cytochrome c2 is a direct donor to the nitrogen oxide reductases (Sawada et al., 1978; Richardson et al., 1991; Laratta et al., 2006). Cytochrome c2 is a major component of both the respiratory and photosynthetic parts of the electron transport systems in these bacteria (Oh and Kaplan, 2001; Axelrod and Okamura, 2005). In related α-proteobacteria electron transfer to Nir is through both a c-type cytochrome and a copper protein (Pearson et al., 2003). A dependence on both cytochrome and copper proteins has not been shown in photosynthetic denitrifiers. However, it is not unreasonable to suggest that copper proteins are also involved in transferring electrons to the reductases of Rba. sphaeroides 2.4.3 since the genome contains a gene encoding a pseudoazurin with the necessary copper binding residues (Laratta et al., 2006). None of the Rps. palustris genomes encode either a pseudoazurin or a plastocyanin. This may correlate with the low Nir activity seen in strains of this bacterium (Preuss and Klemme, 1983). III. Assimilation of Nitrogen A. Nitrate Assimilation Some photosynthetic bacteria also have the ability to reduce nitrate to ammonia for assimilatory purposes. This is an 8-electron reduction in which nitrate is reduced to nitrite, and nitrite is reduced directly to ammonia (Richardson et al., 2001) (Fig. 1). The en-
Chapter 31
Nitrogen Oxide Reduction
zyme that reduces nitrate to nitrite is termed Nas for assimilatory nitrate reductase (Fig. 2A). This enzyme has structural and functional similarity with Nar and Nap, but is found in the cytoplasm (Fig. 2B). Electrons for reduction of the substrate can come from either NAD(P)H or ferredoxin, and consequently not directly from the respiratory chain. There are two forms of nitrite reductase. One form accepts electrons from ferredoxin while the other accepts electrons from NAD(P)H. Both enzymes contain [Fe-S] centers and siroheme, but the NAD(P)H accepting form also contains a flavin moiety (Richardson et al., 2001). The ability to use nitrate as a nitrogen source is not widespread among purple photosynthetic bacteria (Pino et al., 2006). Like the capacity for dissimilatory nitrate reduction the capacity for nitrate assimilation is distributed in a seemingly random manner. For example, genome analysis suggests two of the strains of Rps. palustris, HaA2 and BisA53, can assimilate nitrate. Neither of these strains contains Nap or Nar but Rps. palustris BisA53 contains all the other denitrification enzymes. Genome analysis indicates Nas and the assimilatory nitrite reductase is also found in Rsb. denitrificans but in none of the Rba. sphaeroides strains. In the α-proteobacteria the nitrite reductases are typically NAD(P)H reductases that contain flavin (Olmo-Mira et al., 2006). This is somewhat unexpected since the ferredoxin form predominates in other photosynthetic organisms. The best-studied photosynthetic bacterium capable of assimilatory nitrate reduction is Rba. capsulatus strain E1F1 (Pino et al., 2006). In this bacterium nasA, the nitrate reductase structural gene clusters with the nitrite reductase as well as the genes required for the synthesis of the ABC transporter required for nitrate transport across the inner membrane (Pino et al., 2006). In addition, there is a series of potential regulatory genes including a member of the Rrf2 family and a putative transcription antitermination protein NasT. NasS is a potential sensor protein found in the cluster that could work together with NasT. However, the role of NasS is unclear since the nasS gene contains a frame shift that affects the terminal third of the reading frame. These proteins are similar to the NasRT shown to regulate the assimilatory nitrate reductase in Azotobacter vinelandii (Gutierrez et al., 1995). Putative orthologs of these regulatory proteins are found in other photosynthetic denitrifiers. They are also found in bacteria such as Rps. palustris CGA009, which does not contain nas. Nas expression in Rba. capsulatus is induced if
637 nitrate is present as the sole source of fixed nitrogen (Dobao et al., 1994). Ammonia rapidly inhibits nitrate assimilation. This appears to be achieved by inhibiting nitrate transport. Expression is also repressed if carbon is limiting as would be expected of an assimilatory process (Dobao et al., 1993). In addition to the use of nitrate as a nitrogen source, Rba. capsulatus can also use hydroxylamine as a nitrogen source. A gene that is adjacent to the nas cluster appears to be required for hydroxylamine reduction (Cabello et al., 2004). The gene, termed hcp, for hybrid cluster protein, encodes an enzyme that contains [Fe-S] clusters and reduces hydroxylamine to ammonia. The significance of this reaction is unclear, but it may be required to deal with toxic hydroxylamine that is produced during the reduction of nitrite to ammonia. B. Assimilation of N2 Produced Via Denitrification Since many purple photosynthetic bacteria are unable to utilize nitrate as a nitrogen source this can lead to the situation where nitrate or nitrite is available but the cells are starved for fixed nitrogen. Denitrification can potentially be used to release the nitrogen in nitrate for assimilation. Using 15N-labeled nitrate it has been found that Rba. sphaeroides f. sp. denitrificans can assimilate the nitrogen produced by complete denitrification of nitrate (Dunstan et al., 1982). While not directly demonstrated a similar form of assimilation may occur in Rps. palustris (Klemme et al., 1980). In Rba. sphaeroides the assimilation of denitrification produced nitrogen gas only occurred under photosynthetic conditions (Dunstan et al., 1982). This likely reflects that the energy available for nitrogen fixation under dark, denitrifying conditions is somewhat limited. Photosynthesis would provide sufficient ATP for growth and nitrogen fixation. The environmental significance of this N2 recycling is unclear since it is doubtful cells would ever be N2-limited. However, it is possible that under conditions where gas exchange is difficult N-recycling could supply some of the N required for growth. C. Regulatory Overlap in Nitrogen Metabolism There does not seem to be any overlap in the regulation of nitrate assimilation and denitrification although there remains much to be learned about regulation
638 of nitrate reduction in photosynthetic bacteria. Like nitrate reduction to ammonia, nitrogen fixation is an assimilatory process and is regulated to a large degree by availability of fixed nitrogen (Chapter 38, Masepohl and Kranz). This is in contrast to denitrification where optimal expression of the denitrification genes requires nitrogen oxides (Zumft, 1997). Not surprisingly, no gene involved in detecting nitrogenous compounds is utilized by both the denitrification and nitrogen fixation regulatory pathways. There is one important similarity in both pathways though, which is that both are only functional under low oxygen conditions. Nitrogen fixation requires low oxygen because nitrogenase is sensitive to oxygen. Low oxygen is required for denitrification because oxygen respiration is more energetically efficient. Oxygen regulation of nitrogen fixation and denitrification is complex and the oxygen dependent function and regulation of the enzymes in these pathways involves direct and indirect effects (Zumft, 1997; Zumft, 2002; Dixon and Kahn, 2004). The one regulatory mechanism known to regulate both, PrrAB/RegAB, indirectly senses oxygen levels through the redox state of the respiratory chain (Chapter 35, Bauer et al.). As discussed above, it has been shown that PrrBA regulates nirK expression in Rba. sphaeroides (Laratta et al., 2002). Mutations in PrrBA also impact expression of nitrogen fixation in Rba. sphaeroides (Joshi and Tabita, 1996). Why then are assimilatory and dissimilatory processes both regulated by the Prr/Reg system? An important common factor is that both processes consume electrons. Denitrification is well established as a process that can be used for redox balancing. Under certain conditions nitrogen fixation has been shown to serve the same role (Joshi and Tabita, 1996). The relatively high frequency of occurrence of N-oxide reductases in purple photosynthetic bacteria is likely connected to this requirement for multiple mechanisms for disposing of excess reducing equivalents. IV. Conclusion The ability to denitrify and assimilate nitrate is found in diverse photosynthetic bacteria. However, the distribution of these metabolic traits among the members of this group does not follow any obvious phylogenetic pattern. Even among members of the same species there can be significant differences in nitrate reduction capacity. This variation suggests that the genes for denitrification are beneficial to some
James P. Shapleigh strains in certain environments but are not beneficial or are perhaps even detrimental in other environments. Interestingly, partial denitrification is widespread among this group with some bacteria having only a single nitrogen oxide reductase. This indicates that the cell can benefit from having only a portion of the pathway. Experiments have shown that a single enzyme can be beneficial either through redox balancing, or as a means for dealing with toxic oxides of nitrogen. It is clear though that complete denitrification can be beneficial for some photosynthetic bacteria since it can be used to support growth under dark, anoxic conditions. The diversity of denitrification capacity among the photosynthetic bacteria makes clear that denitrification should not only be thought of as an obligatorily connected set of reactions but also as a modular group of enzymes that can provide a selectable benefit separately or in groups. Even though the genes for all or parts of the denitrification transport chain are found in photosynthetic denitrifiers, and roles for the enzymes can be postulated, much remains to be known about the metabolic function of these enzymes. For example, N-oxide reductases are common in Rps. palustris but there is no data indicating that they play any role in supporting growth. Given their high frequency of occurrence in this bacterium it is likely nitrogen oxide reduction is beneficial, but the conditions under which this occurs have not been identified. Part of this uncertainty has to do with our lack of insight into the regulation of many of the denitrification genes. Significant progress has been made in understanding how genes such as nirK and nor or regulated in some bacteria, but other denitrifiers likely use different regulatory mechanisms. Currently, we have little understanding about how other genes such as nap and nos are regulated in any photosynthetic denitrifier. To more fully understand the physiological role of denitrification in photosynthetic bacteria we need to better understand how both global and denitrification specific regulators control expression of nitrogen oxide reductases to support growth and survival in varying environments. Nitrate assimilation is also relatively unstudied in this group of bacteria. While the regulation that has been described is consistent with its assimilatory role, it would be of interest to determine if it could also play a role in redox balancing. As with denitrification, we need to have a more complete understanding of the regulatory mechanisms if we are to understand the physiological function of this process.
Chapter 31
Nitrogen Oxide Reduction
Acknowledgments I would like to thank the Department of Energy for funding. I would like to thank Angela Hartsock for help in editing the manuscript. The availability of genome sequences from numerous sources has been an invaluable resource. These include the Department of Energy Joint Genome Institute, the J. Craig Venter Institute in collaboration with the Gordon and Betty Moore Foundation Marine Microbial Genome Sequencing Project and the Phototrophic Genome Project sponsored by NSF. References Arai H, Mizutani M and Igarashi Y (2003) Transcriptional regulation of the nos genes for nitrous oxide reductase in Pseudomonas aeruginosa. Microbiol 149: 29–36 Arnoux P, Sabaty M, Alric J, Frangioni B, Guigliarelli B, Adriano JM and Pignol D (2003) Structural and redox plasticity in the heterodimeric periplasmic nitrate reductase. Nat Struct Biol 10: 928–934 Averill BA and Tiedje JM (1982) The chemical mechanism of microbial denitrification. FEBS Lett 138: 8–12 Axelrod HL and Okamura MY (2005) The structure and function of the cytochrome c2:reaction center electron transfer complex from Rhodobacter sphaeroides. Photosynth Res 85: 101–114 Bartnikas TB, Tosques IE, Laratta WP, Shi J and Shapleigh JP (1997) Characterization of the region encoding the nitric oxide reductase of Rhodobacter sphaeroides 2.4.3. J Bacteriol 179: 3534–3540 Bartnikas TB, Wang Y, Bobo T, Veselov A, Scholes CP and Shapleigh JP (2002) Characterization of a member of the NnrR regulon in Rhodobacter sphaeroides 2.4.3 encoding a heme-copper protein. Microbiol 148: 825–833 Bell LC, Richardson DJ and Ferguson SJ (1992) Identification of nitric oxide reductase activity in Rhodobacter capsulatus: The electron transport pathway can either use or bypass both cytochrome c2 and the cytochrome bc1 complex. J Gen Micro 138: 437–443 Braun C and Zumft WG (1991) Marker exchange of the structural genes for nitric oxide reductase blocks the denitrification pathway of Pseudomonas stutzeri at nitric oxide. J Biol Chem 266: 22785–22788 Brown K, Prudencio M, Pereira AS, Besson S, Moura JJG, Moura I, Tegoni M and Cambillau C (2000) A novel type of catalytic copper cluster in nitrous oxide reductase. Nat Struct Biol 7: 191–195 Cabello P, Pino C, Olmo-Mira MF, Castillo F, Roldan MD and Moreno-Vivian C (2004) Hydroxylamine assimilation by Rhodobacter capsulatus E1F1. Requirement of the hcp gene (hybrid cluster protein) located in the nitrate assimilation nas gene region for hydroxylamine reduction. J Biol Chem 279: 45485–45494 Chan YK, McCormick WA and Watson RJ (1997) A new nos gene downstream from nosdfy is essential for dissimilatory reduction of nitrous oxide by Rhizobium (Sinorhizobium) meliloti.
639 Microbiol 143: 2817–2824 Choi PS, Naal Z, Moore C, Casado-Rivera E, Abruna HD, Helmann JD and Shapleigh JP (2006) Assessing the impact of denitrifier-produced nitric oxide on other bacteria. Appl Environ Microbiol 72: 2200–2205 Coomber SA, Jones RM, Jordan PM and Hunter CN (1992) A putative anaerobic coproporphrinogen iii oxidase in Rhodobacter sphaeroides. I. Molecular cloning, transposon mutagenesis and sequence analysis of the gene. Mol Microbiol 6: 3159–3169 Coyne MS and Tiedje JM (1990) Induction of denitrifying enzymes in oxygen-limited Achromobacter cycloclastes continuous culture. FEMS Microbiol Ecol 73: 263–270 Cramm R, Siddiqui RA and Friedrich B (1997) Two isofunctional nitric oxide reductases in Alcaligenes eutrophus H16. J Bacteriol 179: 6769–6777 Dixon R and Kahn D (2004) Genetic regulation of biological nitrogen fixation. Nat Rev Microbiol 2: 621–631 Dobao MM, Martínez-Luque M and Castillo F (1993) Nitrate reductase activity in spheroplasts from Rhodobacter capsulatus E1F1 requires a periplasmic protein. Arch Microbiol 160: 471–476 Dobao MM, Martinez-Luque M, Moreno VC and Castillo F (1994) Effect of carbon and nitrogen metabolism on nitrate reductase activity of Rhodobacter capsulatus E1F1. Can J Microbiol 40: 645–650 Doi M and Shioi Y (1991) Enhancement of denitrifying activity in cells of Roseobacter denitrificans grown aerobically in the light. Plant Cell Physiol 32: 365–370 Doi M, Shioi Y, Morita M and Takamiya K (1989) Two types of cytochrome cd1 in the aerobic photosynthetic bacterium, Erythrobacter sp. Och 114. Eur J Biochem 184: 521–527 Dunstan RH, Kelley BC and Nicholas DJ (1982) Fixation of dinitrogen derived from denitrification of nitrate in a photosynthetic bacterium, Rhodopseudomonas sphaeroides forma sp. denitrificans. J Bacteriol 150: 100–104 Ellington MJK, Richardson DJ and Ferguson SJ (2003) Rhodobacter capsulatus gains a competitive advantage from respiratory nitrate reduction during light–dark transitions. Microbiol 149: 941–948 Farrar JA, Thompson AJ, Cheesman MR, Dooley DM and Zumft WG (1991) A model of the copper centres of nitrous oxide reductase (Pseudomonas stutzeri). Evidence from optical, EPR and MCD spectroscopy. FEBS Lett 294: 11–15 Garber EA and Hollocher TC (1981) 15N tracer studies on the role of NO in denitrification. J Biol Chem 256: 5459–5465 Gavira M, Roldan MD, Castillo F and Moreno-Vivian C (2002) Regulation of nap gene expression and periplasmic nitrate reductase activity in the phototrophic bacterium Rhodobacter sphaeroides DSM158. J Bacteriol 184: 1693–1702 Gutierrez JC, Ramos F, Ortner L and Tortolero M (1995) nasST, two genes involved in the induction of the assimilatory nitritenitrate reductase operon (nasAB) of Azotobacter vinelandii. Mol Microbiol 18: 579–591 Haltia T, Brown K, Tegoni M, Cambillau C, Saraste M, Mattila K and Djinovic-Carugo K (2003) Crystal structure of nitrous oxide reductase from Paracoccus denitrificans at 1.6 Å resolution. Biochem J 369: 77–88 Haselkorn R, Lapidus A, Kogan Y, Vlcek C, Paces J, Paces V, Ulbrich P, Pecenkova T, Rebrekov D, Milgram A, Mazur M, Cox R, Kyrpides N, Ivanova N, Kapatral V, Los T, Lykidis A, Mikhailova N, Reznik G, Vasieva O and Fonstein M (2001) The
640 Rhodobacter capsulatus genome. Photosynth Res 70: 43–52 Hosler JP, Ferguson-Miller S and Mills DA (2006) Energy transduction: Proton transfer through the respiratory complexes. Annu Rev Biochem 75: 165–187 Jacobson F, Guo H, Olesen K, Okvist M, Neutze R and Sjolin L (2005) Structures of the oxidized and reduced forms of nitrite reductase from Rhodobacter sphaeroides 2.4.3 at high pH: Changes in the interactions of the type 2 copper. Acta Cryst D 61: 1190–1198 Jain R and Shapleigh JP (2001) Characterization of nirV and a gene encoding a novel pseudoazurin in Rhodobacter sphaeroides 2.4.3. Microbiol 147: 2505–2515 Joshi HM and Tabita FR (1996) A global two component signal transduction system that integrates the control of photosynthesis, carbon dioxide assimilation, and nitrogen fixation. Proc Natl Acad Sci U S A 93: 14515–14520 Kastrau DHW, Heiss B, Kroneck PMH and Zumft WG (1994) Nitric oxide reductase from Pseudomonas stutzeri, a novel cytochrome bc complex. Eur J Biochem 222: 293–303 Klemme J-H, Chyla I and Preuss M (1980) Dissimilatory nitrate reduction by strains of the facultative phototrophic bacterium Rhodopseudomonas palustris. FEMS Microbiol Lett 9: 137–140 Klepper LA (1987) Nitric oxide emissions from soybean leaves during in vivo nitrate reductase assays. Plant Physiol 85: 96–99 Korner H and Zumft WG (1989) Expression of denitrification enzymes in response to the dissolved oxygen and respiratory substrate in continuous culture of Pseudomonas stutzeri. Appl Environ Microbiol 55: 1670–1676 Korner H, Sofia HJ and Zumft WG (2003) Phylogeny of the bacterial superfamily of CRP-FNR transcription regulators: Exploiting the metabolic spectrum by controlling alternative gene programs. FEMS Microbiol Rev 27: 559–592 Kwiatkowski AV, Laratta WP, Toffanin A and Shapleigh JP (1997) Analysis of the role of the nnrR gene product in the response of Rhodobacter sphaeroides 2.4.1 to exogenous nitric oxide. J Bacteriol 179: 5618–5620 Laratta WP, Choi PS, Tosques IE and Shapleigh JP (2002) Involvement of the Prrb/Prra two-component system in nitrite respiration in Rhodobacter sphaeroides 2.4.3: Evidence for transcriptional regulation. J Bacteriol 184: 3521–3529 Laratta WP, Nanaszko MJ and Shapleigh JP (2006) Electron transfer to nitrite reductase of Rhodobacter sphaeroides 2.4.3: Examination of cytochromes c2 and cY Microbiol 152: 1479–1488 Lee YY, Shearer N and Spiro S (2006) Transcription factor Nnr from Paracoccus denitrificans is a sensor of both nitric oxide and oxygen: Isolation of nnr* alleles encoding effector-independent proteins and evidence for a haem-based sensing mechanism. Microbiol 152: 1461–1470 LeGall J, Payne WJ, Morgan TV and DerVartanian D (1979) On the purification of nitrite reductase from Thiobacillus denitrificans and its reaction with nitrite under reducing conditions. Biochem Biophys Res Commun 87: 355–362 Liu X, Miller MJ, Joshi MS, Thomas DD and Lancaster JR, Jr. (1998) Accelerated reaction of nitric oxide with O2 within the hydrophobic interior of biological membranes. Proc Natl Acad Sci USA 95: 2175–2179 Mackenzie C, Choudhary M, Larimer FW, Predki PF, Stilwagen S, Armitage JP, Barber RD, Donohue TJ, Hosler JP, Newman
James P. Shapleigh JE, Shapleigh JP, Sockett RE, Zeilstra-Ryalls J and Kaplan S (2001) The home stretch, a first analysis of the nearly completed genome of Rhodobacter sphaeroides 2.4.1. Photosynth Res 70: 19–41 Martinez-Luque M, Dobao MM and Castillo F (1991) Characterization of the assimilatory and dissimilatory nitrate-reducing systems in Rhodobacter: A comparative study. FEMS Micro Lett 83: 329–334 Matsuda Y, Inamori K, Osaki T, Eguchi A, Watanabe A, Kawabata S, Iba K and Arata H (2002) Nitric oxide-reductase homologue that contains a copper atom and has cytochrome c-oxidase activity from an aerobic phototrophic bacterium Roseobacter denitrificans. J Biochem (Tokyo) 131: 791–800 McEwan AG, George CL, Ferguson SJ and Jackson JB (1982) A nitrate reductase activity in Rhodopseudomonas capsulata linked to electron transfer and generation of a membrane potential. FEBS Lett 150: 277–280 McEwan AG, Jackson JB and Ferguson SJ (1984) Rationalisation of properties of nitrate reductases in Rhodopseudomonas capsulata. Arch Microbiol 137: 344–49 McEwan AG, Greenfield AJ, Wetzstein HG, Jackson BJ and Ferguson SJ (1985) Nitrous oxide reduction by members of the family Rhodospirillaceae and the nitrous oxide reductase of Rhodopseudomonas capsulata. J Bacteriol 164: 823–830. McGuirl MA, Bollinger JA, Cosper N, Scott RA and Dooley DM (2001) Expression, purification, and characterization of nosl, a novel Cu (I) protein of the nitrous oxide reductase (nos) gene cluster. J Biol Inorg Chem 6: 189–195 Michalski WP and Nicholas DJD (1984) The adaptation of Rhodopseudomonas sphaeroides f. sp. denitrificans for growth under denitrifying conditions. J Gen Micro 130: 155–165 Michalski WP and Nicholas DJD (1985) Effects of nitrate, nitrite and diphenylamine on the photosynthetic apparatus of Rhodopseudomonas sphaeroides f. sp. denitrificans. J Gen Micro 131: 1951–1961 Michalski WP and Nicholas DJD (1987) Inhibition of bacteriochlorophyll synthesis in Rhodobacter sphaeroides subsp. denitrificans grown in light under denitrifying conditions. J Bacteriol 169: 4651–4659 Michalski W and Nicholas DJD (1988) Identification of two new denitrifying strains of Rhodobacter sphaeroides. FEMS Microbiol Lett 52: 239–244 Michalski WP, Miller DJ and Nicholas DJD (1986) Changes in cytochrome composition of Rhodopseudomonas sphaeroides f. sp. denitrificans. Biochim Biophys Acta 849: 304–315 Mitchell DM, Wang Y, Alben JO and Shapleigh JP (1998) Fourier transform infrared analysis of membranes of Rhodobacter sphaeroides 2.4.3 grown under microaerobic and denitrifying conditions. Biochim Biophys Acta 1409: 99–105 Moenne-Loccoz P and de Vries S (1998) Structural characterization of the catalytic high-spin heme b of nitric oxide reductase: A resonance Raman study. J Am Chem Soc 120: 5147–5152 Moura I and Moura JJG (2001) Structural aspects of denitrifying enzymes. Curr Op Chem Biol 5: 168–175 Nishimura K, Shimada H, Ohta H, Masuda T, Shioi Y and Takamiya K (1996) Expression of the puf operon in an aerobic photosynthetic bacterium, Roseobacter denitrificans. Plant Cell Physiol 37: 153–159 Oh J-I and Kaplan S (2001) Generalized approach to the regulation and integration of gene expression. Mol Microbiol 39: 1116–1123
Chapter 31
Nitrogen Oxide Reduction
Oh JI, Eraso JM and Kaplan S (2000) Interacting regulatory circuits involved in orderly control of photosynthesis gene expression in Rhodobacter sphaeroides 2.4.1. J Bacteriol 182: 3081–3087 Oh J-I, Ko I-J and Kaplan S (2001) The default state of the membrane-localized histidine kinase PrrB of Rhodobacter sphaeroides 2.4.1 is in the kinase-positive mode. J Bacteriol 183: 6807–6814 Olesen KO, Veselov A, Zhao Y, Wang Y, Danner B, Scholes CP and Shapleigh JP (1998) Spectroscopic, kinetic and electrochemical characterization of heterologously expressed wild type and mutant forms of copper-containing nitrite reductase from Rhodobacter sphaeroides 2.4.3. Biochem 37: 6086–6094 Olmo-Mira MF, Cabello P, Pino C, Martinez-Luque M, Richardson DJ, Castillo F, Roldan MD and Moreno-Vivian C (2006) Expression and characterization of the assimilatory NADHnitrite reductase from the phototrophic bacterium Rhodobacter capsulatus E1F1. Arch Microbiol 186: 339–344 Payne WJ (1981) Denitrification. John Wiley & Sons, New York Pearson IV, Page MD, van Spanning RJ and Ferguson SJ (2003) A mutant of Paracoccus denitrificans with disrupted genes coding for cytochrome c550 and pseudoazurin establishes these two proteins as the in vivo electron donors to cytochrome cd1 nitrite reductase. J Bacteriol 185: 6308–6315 Pino C, Olmo-Mira F, Cabello P, Martinez-Luque M, Castillo F, Roldan MD and Moreno-Vivian C (2006) The assimilatory nitrate reduction system of the phototrophic bacterium Rhodobacter capsulatus E1F1. Biochem Soc Trans 34: 127–129 Preisig O, Zufferey R, Thony-Meyer L, Appleby CA and Hennecke H (1996) A high-affinity cbb3-type cytochrome oxidase terminates the symbiosis-specific respiratory chain of Bradyrhizobium japonicum. J Bacteriol 178: 1532–1538 Preuss M and Klemme J-H (1983) Purification and characterization of a dissimilatory nitrite reductase from the phototrophic bacterium Rhodopseudomonas palustris. Z Naturforsch C 38: 933–938 Richardson DJ, McEwan AG, Jackson JB and Ferguson SJ (1989) Electron transport pathways to nitrous oxide in Rhodobacter species. Eur J Biochem 185: 659–669 Richardson DJ, Bell LC, McEwan AG, Jackson JB and Ferguson SJ (1991) Cytochrome c2 is essential for electron transfer to nitrous oxide reductase from physiological substrates in Rhodobacter capsulatus and can act as an electron donor to the reductase in vitro. Eur J Biochem 199: 677–683 Richardson DJ, Berks BC, Russell DA, Spiro S and Taylor CJ (2001) Functional, biochemical and genetic diversity of prokaryotic nitrate reductases. Cell Mol Life Sci 58: 165–178 Risgaard-Petersen N, Langezaal AM, Ingvardsen S, Schmid MC, Jetten MS, Op den Camp HJ, Derksen JW, Pina-Ochoa E, Eriksson SP, Nielsen LP, Revsbech NP, Cedhagen T and van der Zwaan GJ (2006) Evidence for complete denitrification in a benthic foraminifer. Nature 443: 93–96 Sabaty M, Gans P and Verméglio A (1993) Inhibition of nitrate reduction by light and oxygen in Rhodobacter sphaeroides forma sp. denitrificans. Arch Microbiol 159: 153–159 Sabaty M, Jappe J, Olive J and Verméglio A (1994) Organization of electron transport components in Rhodobacter sphaeroides forma sp. denitrificans whole cells. Biochim Biophys Acta 1187: 313–323 Sabaty M, Schwintner C, Cahors S, Richaud P and Verméglio A
641 (1999) Nitrite and nitrous oxide reductase regulation by nitrogen oxides in Rhodobacter sphaeroides f. sp. denitrificans IL106. J Bacteriol 181: 6028–6032 Saraste M and Castresana J (1994) Cytochrome oxidase evolved by tinkering with denitrification enzymes. FEBS Lett 341: 1–4 Sato K, Okubo A and Yamazaki S (1999) Anaerobic purification and characterization of nitrous oxide reductase from Rhodobacter sphaeroides f. sp. denitrificans IL106. J Biochem (Tokyo) 125: 864–868 Satoh T (1977) Light-activated, -inhibited and -independent denitrification by a denitrifying phototrophic bacterium. Arch Microbiol 115: 293–298 Satoh T, Hoshino Y and Kitamura H (1976) Rhodopseudomonas sphaeroides forma sp. denitrificans a denitrifying strain as a subspecies of Rhodopseudomonas sphaeroides. Arch Microbiol 108: 265–269 Saunders NF, Hornberg JJ, Reijnders WN, Westerhoff HV, de Vries S and van Spanning RJ (2000) The nosX and nirX proteins of Paracoccus denitrificans are functional homologues: Their role in maturation of nitrous oxide reductase. J Bacteriol 182: 5211–5217 Sawada E, Satoh T and Kitamura H (1978) Purification and properties of a dissimilatory nitrite reductase of a denitrifying phototrophic bacterium. Plant Cell Physiol 19: 1339–1351 Shapleigh JP (2007) The denitrifying prokaryotes. In: Dworkin M (ed) The Prokaryotes: A Handbook on the Biology of Bacteria, Vol 2: Ecophysiology and Biochemistry, pp 769–792. Springer-Verlag, New York Shioi Y, Doi M, Arata H and Takamiya K (1988) A denitrifying activity in an aerobic photosynthetic bacterium, Erythrobacter sp. strain och 114. Plant Cell Physiol 29: 861–865 Strous M, Fuerst JA, Kramer EH, Logemann S, Muyzer G, van de Pas-Schoonen KT, Webb R, Kuenen JG and Jetten MS (1999) Missing lithotroph identified as new planctomycete. Nature 400: 446–449 Strous M, Pelletier E, Mangenot S, Rattei T, Lehner A, Taylor MW, Horn M, Daims H, Bartol-Mavel D, Wincker P, Barbe V, Fonknechten N, Vallenet D, Segurens B, Schenowitz-Truong C, Medigue C, Collingro A, Snel B, Dutilh BE, Op den Camp HJ, van der Drift C, Cirpus I, van de Pas-Schoonen KT, Harhangi HR, van Niftrik L, Schmid M, Keltjens J, van de Vossenberg J, Kartal B, Meier H, Frishman D, Huynen MA, Mewes HW, Weissenbach J, Jetten MS, Wagner M and Le Paslier D (2006) Deciphering the evolution and metabolism of an anammox bacterium from a community genome. Nature 440: 790–794 Swingley WD, Gholba S, Mastrian SD, Matthies HJ, Hao J, Ramos H, Acharya CR, Conrad AL, Taylor HL, Dejesa LC, Shah MK, O’Huallachain M E, Lince MT, Blankenship RE, Beatty JT and Touchman JW (2007) The complete genome sequence of Roseobacter denitrificans reveals a mixotrophic as opposed to photosynthetic metabolism. J Bacteriol 2007 189 :683–90 Tabata A, Yamamoto I, Matsuzaki M and Satoh T (2005) Differential regulation of periplasmic nitrate reductase gene (napKEFDABC) expression between aerobiosis and anaerobiosis with nitrate in a denitrifying phototroph Rhodobacter sphaeroides f. sp. denitrificans. Arch Microbiol 184: 108–116 Taubner LM, McGuirl MA, Dooley DM and Copie V (2006) Structural studies of apo NosL, an accessory protein of the nitrous oxide reductase system: Insights from structural homology with MerB, a mercury resistance protein. Biochem 45: 12240–12252
642 Tocheva EI, Rosell FI, Mauk AG and Murphy ME (2004) Sideon copper-nitrosyl coordination by nitrite reductase. Science 304: 867–870 Tosques IE, Kwiatkowski AV, Shi J and Shapleigh JP (1997) Characterization and regulation of the gene encoding nitrite reductase in Rhodobacter sphaeroides 2.4.3. J Bacteriol 179: 1090–1095 Usov OM, Sun Y, Grigoryants VM, Shapleigh JP and Scholes CP (2006) EPR-ENDOR of the Cu(I)NO complex of nitrite reductase. J Am Chem Soc 128: 13102–13111 Vollack K and Zumft W (2001) Nitric oxide signaling and transcriptional control of denitrification genes in Pseudomonas stutzeri. J Bacteriol 183: 2516–2526 Watmough NJ, Butland G, Cheesman MR, Moir JWB, Richardson DJ and Spiro S (1999) Nitric oxide in bacteria: Synthesis and consumption. Biochim Biophys Acta 1411: 456–474 Wherland S, Farver O and Pecht I (2005) Intramolecular electron transfer in nitrite reductases. Chem Phys Chem 6: 805–812 Wink D and Mitchell J (1998) Chemical biology of nitric oxide: Insights into regulatory, cytotoxic, and cytoprotective mechanisms of nitric oxide. Free Radic Biol Med 25: 434–456 Wunsch P and Zumft WG (2005) Functional domains of NosR, a novel transmembrane iron-sulfur flavoprotein necessary for
James P. Shapleigh nitrous oxide respiration. J Bacteriol 187: 1992–2001 Yin S, Fuangthong M, Laratta WP and Shapleigh JP (2003) Use of a green fluorescent protein-based reporter fusion for detection of nitric oxide produced by denitrifiers. Appl Environ Microbiol 69: 3938–3944 Zhao Y, Lukoyanov DA, Toropov YV, Wu K, Shapleigh JP and Scholes CP (2002) Catalytic function and local proton structure at the type 2 copper of nitrite reductase: The correlation of enzymatic pH dependence, conserved residues and proton hyperfine structure. Biochem 41: 7464–7474 Zumft WG (1997) Cell biology and molecular basis of denitrification. Microbiol Mol Biol Rev 61: 533–616 Zumft WG (2002) Nitric oxide signaling and NO dependent transcriptional control in bacterial denitrification by members of the FNR-CRP regulator family. J Mol Microbiol Biotechnol 4: 277–286 Zumft WG (2005) Nitric oxide reductases of prokaryotes with emphasis on the respiratory, heme-copper oxidase type. J Inorg Biochem 99: 194–215 Zumft WG and Kroneck PM (2006) Respiratory transformation of nitrous oxide (N(2)O) to dinitrogen by bacteria and archaea. Adv Microb Physiol 52: 107–227
Chapter 32 Swimming and Behavior in Purple Non-Sulfur Bacteria Judith P. Armitage* Department of Biochemistry, University of Oxford, South Parks Road, Oxford OX1 3QU, U.K.
Summary ............................................................................................................................................................... 643 I. Introduction..................................................................................................................................................... 644 II. Swimming ....................................................................................................................................................... 644 A. Regulation of Expression ................................................................................................................. 645 B. Patterns of Swimming ...................................................................................................................... 646 III. Behavioral Responses ................................................................................................................................... 646 A. The Escherichia coli ‘Paradigm’ ....................................................................................................... 646 B. Chemotaxis in Rhodobacter sphaeroides ...................................................................................... 647 1. Gene Organization .................................................................................................................. 647 2. Phosphotransfer between Chemosensory Proteins ................................................................ 649 3. Localization of the Chemosensory Proteins ............................................................................ 649 4. What Do the Receptors Sense?.............................................................................................. 649 5. Comparison with Other Purple Bacterial Species ................................................................... 650 C. Phototaxis and Aerotaxis ................................................................................................................. 651 1. Aerotaxis ................................................................................................................................. 651 2. Phototaxis ............................................................................................................................... 652 Acknowledgments ................................................................................................................................................. 653 References ........................................................................................................................................................... 653
Summary Many species of purple bacteria swim, and they direct that swimming towards carbon and nitrogen sources, light and/or oxygen using a complex set of chemosensory pathways. The review will illustrate how genomics, molecular biology, biochemistry, behavioral analysis and in vivo imaging have been used to try and dissect the pathways, identifying common features and the variations on those themes. Work on two purple species in particular has provided insights into these complexities. Rhodobacter (Rba.) sphaeroides shows that different pathways can be localized to different parts of the cell, revealing a complex intracellular organization allowing cells to tune their responses to metabolic need, while in Rhodospirillum centenum not all genes annotated as chemosensory turn out to be involved in chemotaxis, rather they are involved in regulating development. The complexity and variation of the sensory pathways identified to date in the purple bacteria illustrate the modification of the inputs and outputs of the related systems, reflecting the different growth environments and requirements of the different species.
*Email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 643–654. © 2009 Springer Science + Business Media B.V.
Judith P. Armitage
644 I. Introduction Except perhaps for the open oceans, where short steep gradients may be rare, being able to sense and swim to a better environment for growth is important for survival for the majority of bacterial species. In chemoheterotrophic species this is relatively straightforward; the cells sense gradients of metabolites through a limited number of chemoreceptors and move in that direction. Things are more complex for facultative photosynthetic bacteria as the optimum environment varies greatly from, for example, day to night. How purple photosynthetic bacteria respond to different carbon sources, to oxygen and/or to light depends on their current metabolic state: what does Rhodobacter do at night when the lights go out? Does it sleep or switch to respiratory growth, in which case it must be able to sense terminal electron acceptors (or changes in electron transport) and the optimum carbon source may change. The complexity of metabolism is mirrored by the complexity of the chemosensory system evolved to maintain or move the cells to their optimum environment. Escherichia (E.) coli has only four chemoreceptors, whereas purple bacterial species may have over 40, at any one time expressing only those relevant to current environments. They may have multiple chemosensory pathways, and alter their expression to match the prevailing conditions. This allows the purple bacteria to tune their responses to current conditions. Since the last article (Armitage et al., 1995) the genomes of an increasing number of purple bacteria have been sequenced and this has identified an unexpected variation in complexity between apparently related species. Studies of purple bacteria in particular have led the way in characterizing some of the different complexities of the bacterial chemosensory systems, identifying that related sensory Abbreviations: Aer – aerotaxis sensor in E. coli; CheA – chemotaxis histidine protein kinase; CheB – chemotaxis methyl esterase; CheR – chemotaxis methyl transferase; CheW – chemotaxis linker protein; CheZ – chemotaxis CheY-P phosphatase; C-ring – ring at cytoplasmic face of motor interacting with chemotaxis machinery; E. – Escherichia; HAMP – linker domain between sensory domain and output domain in MCPs; MCP – methyl accepting chemotaxis protein; MotA/MotB – protein components of stator ring; PAS – fold in redox sensing proteins (PER), vertebrate aryl hydrocarbon receptor nuclear translocator (ARNT) and Drosophila single minded protein (SIM); PYP – photoactive yellow protein, contains PAS domain; Rba. – Rhodobacter; Rps. – Rhodopseudomonas; Rsp. – Rhodospirillum; Tlp – transducer like protein (cytoplasmic MCP); UQ – ubiquinone
pathways can be physically separate in the cell, showing that bacteria can tune their behavior to different environments and importantly showing that operons identified as encoding homologs of chemosensory genes might in fact encode proteins regulating very different functions. The flagellum is the most complex structure in a bacterial cell and expression of genes encoding flagellar proteins needs to be tightly regulated and ordered. In Rba. sphaeroides expression is regulated by a specific σ54 in addition to the flagella-specific σ28 identified in most species, possibly linking expression and flagellar number to growth rate. Genome sequencing has shown that the majority of bacterial species have multiple homologs of chemosensory genes. II. Swimming I will deal only very briefly with the swimming behavior of purple bacteria, as this is a very highly conserved feature in bacteria. However, interesting differences in the regulation of expression of the flagellar genes provide new insights, and perhaps suggest swimming is more tightly environmentally regulated in purple bacteria than in other species. Swimming depends on transmembrane rotary motors driving semi-rigid helical flagella (Berg, 2003). The flagellum is one of the most complex structures in a bacterial cell, and it has protein components in every cellular compartment and extracellularly. Thousands of flagellin monomers polymerize to form a flagellum. Flagellin genes are therefore generally only translated when a mechanism is in place to export and assemble the monomers into extracellular flagella, preventing the damaging accumulation of flagellin in the cytoplasm. The expression and assembly of the motor requires the coordinated and sequential expression of over 50 genes and the motor complex is therefore expressed and assembled before the extracellular components. The flagellar motor comprises a rotor and a stator. The rotor spans the cytoplasmic and the outer membranes, and rotation is driven by protons (or in some cases by sodium ions) moving down the electrochemical proton gradient. The rotors rotate at about 300 Hz, driving the extracellular flagellar helix. A ring of about 11 stator protein complexes is assembled around the rotor (Leake et al., 2006). Each stator complex consists of a dimer of the peptidoglycan binding protein MotB embedded in a tetramer of membrane spanning MotA proteins. MotB has
Chapter 32
Swimming and Behavior
a single transmembrane helix and is the site of ion translocation. MotB has the only amino acid essential for rotation in the complete motor complex, Asp32. The MotB dimer is the site of ion translocation, and is thought to change the conformation of associated MotA proteins. MotA proteins probably make electrostatic contact with the rotor protein FliG and the change in MotA conformation as the ions move through MotB drives motor rotation. The ring of 26 FliG rotor proteins may link the M- and C-rings of the rotor. Between 32 and 39 FliN and FliM proteins form the C-ring. The C-ring is the site of interaction with the chemosensory machinery (Thomas et al., 2006). The chemosensory machinery regulates the switching frequency of the motor. When a bacterium is moving in a positive direction, the switching frequency of the motor decreases, while when moving in a direction where the environment is getting worse, the motor switches direction more frequently. This biases the normally random swimming pattern of a bacterium towards a favorable environment for growth. While different species have different numbers and arrangements of flagella, from the single randomly positioned flagellum of Rba. sphaeroides
645 to the multiple peritrichously positioned flagella of surface swarming Rsp. centenum, genome analysis suggests that the mechanism of motor rotation is very similar across all species (Armitage and Macnab, 1987; Ragatz et al., 1994). A. Regulation of Expression In most species there is a flagellar regulon with operons encoding proteins assembled at different stages being expressed at different times. A master operon in enteric species is regulated by the CRP/cAMP system and drives expression of a flagellar-specific sigma factor, σ28 and an anti sigma factor FlgM. In enteric species this system regulates a cascade with the assembly of the transmembrane motor components allowing the export of an anti-sigma factor which then frees the σ28 to express the next class of genes (Macnab, 2000). In the purple bacteria there appears to be an additional level of control, also identified in other α-proteobacteria. The expression of class II and class III flagellar genes is regulated by a specific σ54, distinct from that essential for regulation of nitrogen metabolism. This has been confirmed in Rba, sphaer-
Fig. 1. Hierarchy of expression of flagellar genes in Rba. sphaeroides. (Left) Swimming of Rba. sphaeroides. (Right) Regulation of flagellar gene expression: FleQ, FleT and enhancer binding proteins. Low concentrations of FleQ allow the expression of the initial components of the flagellar motor, while increasing concentration allows oligomerization and the expression of σ54 dependent promoters, including the σ28 protein FliA and the anti-sigma factor FlgM. Once the hook and basal body (HBB) is built FlgM is exported allowing expression of the Class IV genes including the filament proteins (FliC). Diagram from Poggio et al (1995) with permission.
Judith P. Armitage
646 oides, Rsp. centenum and Rhodopseudomonas (Rps.) palustris, with the hierarchy analyzed in detail in the case of Rba. sphaeroides (Fig. 1) (Poggio et al., 2005). The master operon fleQ appears to be constitutive, regulating the expression of the Class II and Class III operons, which include two σ54 enhancer binding proteins, FleQ and FleT. Their mechanism of operation is unclear, but these proteins seem to need to oligomerize before binding the sequences upstream of the σ54 dependent promoters. The timing of interaction may depend on the concentration of FleT. The Class III genes include the structural genes for σ28 and the anti sigma factor FlgM, so that when a basal body is successfully assembled these σ28 dependent genes can allow expression of the filament proteins after the export of FlgM. Promoter sequences upstream of the flagellar and the nitrogen fixation genes have been compared and mutated, and the differences in the sequences that allow differential recognition of the two promoters identified. Why this extra level of regulation is required is unclear, but it may link the synthesis of new flagella to the growth rate. Interestingly, there are two additional σ54 homologs in Rba. sphaeroides, and recent work has identified that one of these proteins regulates one of the chemosensory operons in this species (Poggio et al., 2002; Martin et al., 2006). Why the two chemosensory operons are both differentially regulated, and also regulated independently from the flagellum is unclear, but it indicates increased complexity in these species. Intriguingly, Rsp. centenum has yet another layer of complexity. It has three sets of chemosensory genes (see later); one of these controls not behavioral responses but flagella, with mutants in the operon being either hyper- or hypo-flagellate (Berleman and Bauer, 2005a). Why an additional sensory transduction pathway is required to regulate flagellation in addition to the already complex pathway is unclear. However, the data suggest that regulation is posttranscriptional, possibly implying that it may be the flagellar stability that is being affected
regulate flagella synthesis to optimize the swarmswim differentiation in response to the local environment (McClain et al., 2002). Analysis of the two sets of flagellar genes suggests that the switch and rotor component genes probably arose by gene duplication. However, the genes for the MotA and MotB stator proteins used by the polar flagellum were acquired by lateral gene transfer, with the MotA for the polar flagellum being more closely related to Vibrio parahemolyticus MotA than the MotA found in the lateral flagella (which show good alignment with MotA sequences from other α-proteobacteria). Rba. sphaeroides has a single randomly positioned flagellum, and rather than changing direction by switching the direction of rotation of the motor, it transiently stops the motor, allowing Brownian motion to re-orient the cell. While there are some structural differences between the hook and flagellar proteins of Rba. sphaeroides and enteric species, there is no indication from the motor proteins as to why the motors would stop rather than switch (Kobayashi et al., 2003). Chimeras between the N-terminal region of the E. coli rotor protein FliG and the C-terminal region of Rba. sphaeroides support bidirectional rotation in E. coli (Morehouse et al., 2005). A chimeric motor that combines the stator protein of Rba. sphaeroides MotA with the sodium driven motor of Vibrio alginolyticus was also functional although only in the presence of sodium ions (Asai et al., 2000). These data suggest that the differences between motors are subtle and may lie more in interaction with the chemosensory machinery than in the rotational apparatus. Genome sequencing suggested that Rba. sphaeroides encodes a second set of flagellar proteins, but these are not expressed under laboratory conditions. Recent research has shown that these can encode a polar bundle of flagella, but the conditions under which they are normally expressed have not been identified (Poggio et al., 2007) III. Behavioral Responses
B. Patterns of Swimming A. The Escherichia coli ‘Paradigm’ Rsp. centenum swims in liquid using a constitutive sheathed polar flagellum. However on a surface, large numbers of additional lateral unsheathed flagella are induced that allow colony movement over that surface (Ragatz et al., 1994). The different flagella have different motor and switch proteins, and the chemotaxis-like sensory transduction pathway may
Chemotaxis in E. coli is probably the best understood behavioral system in biology. All of the components are known and, with the exception of the transmembrane domains of the chemoreceptors, there are highresolution structures of the proteins both with and without ligand and interacting with partner proteins.
Chapter 32
Swimming and Behavior
The copy number of each protein and the in vitro kinetics of each phosphotransfer reaction are known (Wadhams and Armitage, 2004). The proteins have been mutated, cross-linked to each other, and studied by in vivo fluorescence resonance energy transfer experiments (Sourjik and Berg, 2001; Sourjik and Berg, 2002b). This has led to a number of predictive models to describe responses in changing environments. Can this be extended to purple bacteria? The purple bacteria illustrate how much more complex bacteria tend to be that live in varied environments, and the challenge at the moment is to see whether the models developed in E. coli can be extended to these species. In E. coli transmembrane dimers of chemoreceptors, methyl accepting chemotaxis proteins (MCPs) sense changes in a limited number of compounds in the external environment. Any change in concentration results in a signal which is transmitted across the membrane to a histidine protein kinase associated with the cytoplasmic domain. In E. coli there are several thousand MCPs clustered at the cell poles sensing serine, aspartate, galactose and maltose. This receptor clustering is dependent on the presence of the kinase, CheA and the linked CheW. The chemoreceptors are 30 nm long proteins with the 20 nm cytoplasmic regions composed of coiled coils, connected to the transmembrane domain via a HAMP linker domain that is essential for signaling (Appleman et al., 2003; Ma et al., 2005). The C-terminal domain interacts with a methyl transferase, CheR, which can add methyl groups to conserved glutamate residues on the MCP — a posttranslational modification essential for adaptation. When the concentration of a chemo-attractant decreases, the change in binding to the periplasmic domain is signaled across the membrane to the histidine kinase CheA, the CheA dimer autophosphorylates at the P4 kinase domain and the phosphoryl group is transferred to the conserved histidine on the P1 domain of the partner CheA (Hirschman et al., 2001). There is competition between two response regulators, CheY and CheB for the P2 domain of CheA, with the two proteins being phosphorylated by CheA at different rates. CheY is a single domain response regulator which, once phosphorylated, shows reduced affinity for the P2 domain of CheA and increased affinity for FliM on the flagellar motor (Stewart and Van Bruggen, 2004). When the phosphorylated CheY-P binds the motor, it causes switching from counterclockwise to clockwise, resulting in the cell tumbling and reorienting
647 for the next period of swimming (Sourjik and Berg, 2002a). A typical response time is about 100 ms. Because bacterial cells both counter rotate in response to flagellar rotation, and are subject to Brownian motion, they must change direction about every second to be able to sense gradients. The dephosphorylation time of E. coli CheY is greater than 1 sec and therefore dephosphorylation is increased by a protein CheZ, primarily localized with the receptors at the poles (Zhao et al., 2002). CheB is a methylesterase whose activity is increased about 100-fold when phosphorylated by CheA. CheB-P demethylates the chemoreceptors, resetting the signaling state, reducing CheA activity and hence its own and CheY-P concentrations. CheB and CheR activities therefore work in concert to reset the signaling state of the MCPs and CheAs, allowing future changes in attractant to be measured and responded to (Stewart et al., 1988; Sourjik and Berg, 2002b). The MCPs probably form a lattice of hundreds of mixed trimers of dimers, together with the CheW and CheA proteins, at the poles. The activity of the CheR and CheB proteins and the methylation state of the receptors are thought to alter the strength of lattice interactions, and this allows an E. coli cell to respond to a change of two or three molecules across a background concentrations of five to six orders of magnitude (Levit et al., 1998; Bray, 2002; Sourjik and Berg, 2004). The sensitivity and gain of the system, along with its apparent robustness has made it ideal for modeling sensory transduction (Alon et al., 1999). B. Chemotaxis in Rhodobacter sphaeroides What are the similarities and differences between the well studied E. coli paradigm and the purple bacteria? Not many of these latter species have been examined in detail, with probably Rba. sphaeroides being the best studied. Major differences have been identified and genomic analysis suggests these differences might be very common in other species. 1. Gene Organization Rba. sphaeroides has three loci which could encode homologs for complete chemosensory pathways, plus a number of unlinked sites encoding putative MCP chemoreceptors (Porter et al., 2002)(Fig. 2). In total the sites could encode 13 chemoreceptor, four CheA, four CheW, three CheR, two CheB and six CheY proteins. As with all non-enteric species
648
Judith P. Armitage
Fig. 2. Chemotaxis operon organization and protein localization in Rba. sphaeroides. Organization of the chemosensory genes showing the three chemotaxis operons, plus a locus with a gene encoding a putative chemsensory fusion, and an unlinked locus on the small chromosome CII encoding an MCP and an essential CheY. Cartoon showing the cellular localization of the proteins encoded by chemotaxis operons 2 and 3 identified using functional GFP fusions.
there is no gene encoding a CheZ homolog. Two of these operons, CheOp2 and CheOp3, have been shown to both be essential for chemotaxis. The third, and first to be identified, CheOp1, gives no major behavioral phenotype when components are deleted under laboratory conditions. While the conditions under which the operon is expressed remain unknown, recent data indicate that the CheY’s encoded in this operon can interact with the motor proteins of the second flagellar system (Martinez et al., 2007). Overexpression of genes from this operon in E. coli can complement appropriate E. coli mutants, suggesting that it can encode active proteins. However, whether the complementation of chemotaxis mutants is the result of the overexpression, allowing crosstalk between related pathways, is unclear (Hamblin et al., 1997). Putative chemosensory operons have now been shown to regulate other activities, as first seen in Rsp. centenum. In addition to the large number of possible chemotaxis proteins, there are other major differences. Of the 13 possible chemoreceptor genes, four encode proteins with no predicted transmembrane domain, but rather are predicted to be cytoplasmic (a feature that is now being identified in receptor homologs in many recently sequenced genomes). These putative cytoplasmic chemoreceptors (named Tlp for Trans-
ducer like proteins) are encoded within the Che operons while the genes for membrane spanning MCPs are, in general, found encoded at unlinked sites. Each putative Che operon contains genes for a complete chemosensory pathway. There are some interesting modifications that may provide clues to differences in sensory signaling. In CheOp3, rather than a single gene to encode the kinase CheA, there are two genes, each encoding part of the histidine protein kinase. CheA4 encodes a protein with the P5 receptor binding, P4 kinase and P3 dimerization domains, but no P1 domain that should have the conserved His residue. Instead this is encoded by a separate gene, cheA3 encoding a second protein also with a P5 receptor binding domain but a P1 domain with the conserved histidine (Porter and Armitage, 2004). CheA4 is a dimer and forms a functional phosphotransfer complex with CheA3. Why CheOp3 should contain this unusual histidine protein kinase hetero-complex while CheOp2 encodes a CheA2 homologous to that of E. coli is unclear. However one key requirement of all sensory pathways is signal termination. In E. coli this is carried out by CheZ, which increases the dephosphorylation rate of the motor binding CheY-P. Only enteric species contain CheZ, other species use a range of different mechanisms. In Rba. sphaeroides, CheY6-P is the major motor binding response
Chapter 32
Swimming and Behavior
regulator, and is dephosphorylated by an unusual phosphatase domain formed between the P1 and P5 domains of CheA3. 2. Phosphotransfer between Chemosensory Proteins Why should Rba. sphaeroides have so many putative chemosensory proteins, and how could that number rationally regulate the stop-start behavior of a single flagellar motor? All of the proteins have been purified and phosphotransfer analyzed in vitro. This suggested a complex possible set of interactions, with CheA2 from CheOp2 being able to phosphotransfer to all the response regulators encoded by the different operons (6 CheY proteins and 2 CheB proteins) (Porter and Armitage, 2002). CheA1 however could only phosphotransfer in vitro to CheY1 and CheY2, encoded in CheOp1 and CheA3/CheA4 (a phosphotransfer hetero-complex) could only transfer to CheY6 and CheB2. How and why does this complexity and discrimination come about? 3. Localization of the Chemosensory Proteins All the che genes were fused to the yfp or cfp gene at either their 3´ or 5´ ends, and the wild type genes were replaced in the genome with the corresponding fusion genes. This allowed the expression of a fluorescent fusion protein from the native promoter along with the other proteins in the operon. This replacement also allows two or more proteins to be expressed as fusions in the same cell. All fusions were tested for normal behavior and the location of the proteins identified. All of the proteins, with the exception of TlpC, expressed from CheOp2 localized to the poles of the cells with the membrane spanning MCP chemoreceptors (Wadhams et al., 2003) while all of the proteins encoded by CheOp3 localize to a tight cluster in the middle of the cell (Fig. 2), with clusters moving to ¼ and ¾ prior to cell division (Wadhams et al., 2002). This means that the two chemosensory pathways are physically separate in the cell, preventing the cross-talk measured in vitro. This finding leads to two other major questions: (1) How do the two pathways regulate motor behavior — are they operating in parallel or in sequence? And (2) how does the cell regulate localization of several thousand proteins in the cytoplasmic cluster? We are still a long way from answering question 1.
649 However, despite earlier findings that all six CheYs can bind a fragment of the FliM switch in both the phosphorylated and unphosphorylated state in vitro (Ferre et al., 2004), in vivo studies show that only CheY6 and either of CheY3 and CheY4 are required for chemotaxis, with CheY6 encoded by CheOp3, being the dominant motor binding response regulator (Porter et al., 2006). Current data suggest that the MCPs sense changes in the extracellular environment while the Tlps sense components of the metabolic state. The cell only responds to the signals through the MCPs if one component of the metabolic state is low relative to other signals. This makes the cytoplasmic sensory cluster in Rba. sphaeroides the dominant ‘decision making’ center, and allows the cells to respond only when it will be a benefit metabolically—a ‘decision’ E. coli cannot make. How the cytoplasmic cluster localizes led to a finding which may have implications beyond purple bacteria and chemotaxis. CheOp3 encodes an essential chemoreceptor, TlpT and a small protein, PpfA, with homology to a ParA chromosome partitioning protein. PpfA mutants only show a small decrease in chemosensory ability but when the chemosensory cluster was measured in a PpfA mutant, rather than being in the middle of the cell, it was randomly positioned. When cell division was inhibited to cause the cells to elongate, wild type cells have clusters evenly positioned along the length of the cell, at one cell length intervals. The PpfA mutant filaments had only one large cluster (Wadhams et al., 2005; Thompson et al., 2006) (Fig. 3). This suggests that an apparatus similar to that used to partition plasmids may be involved in partitioning protein complexes where the proteins are required in specific stoichiometries. The partner proteins in PpfA dependent movement are unclear, but PpfA does interact with TlpT, which is the only soluble chemoreceptor with both a HAMP domain and a proline rich N-terminal domain, suggesting it may be key to regulating both localization and chemosensory signaling. 4. What Do the Receptors Sense? We now know a fair amount about the biochemistry of phosphotransfer and protein:protein interaction in the chemosensory pathways of Rba. sphaeroides, but what do the MCPs and Tlps sense? Unlike E. coli there is little evidence that specific MCPs sense individual
Judith P. Armitage
650 (i) (a)
(ii)
(iii)
(b)
(c) (i)
(ii)
Fig. 3. Localization of cytoplasmic chemosensory proteins and loss of localization and position in PpfA mutant. (a) Rba. sphaeroides with a soluble chemoreceptor TlpC-GFP fusion and McpG-GFP, expressed from genomic replacements. (b) Fluorescence images of cells of Rba. sphaeroides expressing both (i) CheW2-YFP and (ii) CheW4-CFP from genomic replacements showing separation of the two chemotaxis pathways; (iii) overlay image. (c)(i) Cephalexin treated Rba. sphaeroides with genomically expressed TlpT-CFP showing localization at single cell distances through the filament. (ii) Cephalexin treated Rba. sphaeroides ∆PpfA showing loss of localization of the chemosensory clusters. The scale bars are 2 µM. See also Color Plate 14, Fig. 22.
chemo-effectors. The major chemo-effectors for Rba. sphaeroides are organic acids such as propionate and succinate, with amino acids only inducing a small response and sugars other than fructose being to a large extent ignored. It was shown many years ago that inhibiting transport and metabolism reduced the chemotactic response (Poole et al., 1993), possibly because of the role of the (then unknown) cytoplasmic cluster. Deletion of each chemoreceptor leads to a general reduction in responses to all the chemoeffectors tested, possibly suggesting that the lattice of interactions between the trimers of dimers in the membrane is required for a concerted response to all chemoreceptors over a wide range of concentrations. Interestingly an analysis of MCPs from the sequence databases put chemoreceptors into a range of different subgroups, dependent on length and conserved glutamates. Only a few Rba. sphaeroides MCPs and Tlps have the conserved glutamates so important for adaptation in E. coli, and are also the same length as the E. coli MCPs, but these include TlpT. In E. coli it has been shown that adaptation of one receptor can have an effect on neighbors (Ames et al., 2002) — is
that what happens in Rba. sphaeroides? Recent transcriptomic studies suggest that of the nine MCPs a number are constitutively expressed while a number show maximum expression under aerobic and others maximum expression under photoheterotrophic conditions (unpublished data from the Armitage and Kaplan laboratories). Expression of CheOp2 is itself regulated by light and oxygen while that of CheOp3 is regulated by an alternative σ54 (Martin et al., 2006). This σ54 is not involved in either nitrogen metabolism or flagellar gene expression. This complex pattern of environmentally regulated expression of receptors and chemosensory pathways means that more subtle approaches to analysis need to be undertaken, rather than simple gene deletion. 5. Comparison with Other Purple Bacterial Species What about the other purple bacterial species? There has been little investigation into their behavior, but analysis of the sequence databases suggests there are similarities. Rba. capsulatus has two putative chemo-
Chapter 32
Swimming and Behavior
taxis operons with good homology to CheOp1 and CheOp2 of Rba. sphaeroides. Rps. palustris and Rsp. rubrum both have three chemotaxis operons. The Rps. palustris genome encodes 42 putative MCP while Rsp. rubrum could encode a staggering 48. Many years ago it was suggested that E. coli had four MCP types because of (a) membrane space and that (b) it only needs to respond to a few compounds as these will be representative of any nutrient gradient. If the chemoreceptors are only expressed individually under certain growth conditions, or the gene products are expressed at much lower amounts, this may still be true, however the finding that many species encode tens of MCPs suggests that at least this is unlikely to be the explanation for species with very flexible metabolisms. Rps. palustris shows chemotaxis to a range of aromatic compounds, including pollutants and plant derived compounds, however identification of specific chemoreceptors has proved elusive. It has now been shown that at least one of the 42 MCPs is induced by growth on p-coumarate, an aromatic, plant lignin monomer (C. Harwood, personal communication). This suggests that Rps. palustris does really tune its behavior to the complex environments in which it can grow. Recently the genomes of a number of marine Roseobacter species, which include aerobic anoxygenic phototrophs, have been sequenced (see Chapter 3, Yurkov and Csotonyi for a review of these bacteria). Although these species are hard to make motile under laboratory conditions, their genomes suggest that they are mainly motile and chemotactic, again using complex chemosensory pathways (Robert Belas, personal communication). Other than Rba. sphaeroides, the most studied purple bacterial species from the point of view of behavior has been Rsp. centenum. This bacterium also encodes three putative chemosensory pathways, but it has been shown that only one pathway is required for chemotaxis; the other two putative pathways are not involved in chemotaxis but one is involved in regulating the lateral to polar flagellar location modifications in response to growth conditions, while the other regulates cyst development (Berleman and Bauer, 2005a,b). Both of these pathways would be annotated in any sequenced genomes as chemosensory pathways, and the findings that the Rsp. centenum putative pathways are really involved in developmental regulation led the way for a re-examination of multiple pathways in other species, with some similar findings.
651 This highlights the weaknesses of annotations based on homology rather than experiment. C. Phototaxis and Aerotaxis 1. Aerotaxis E. coli and several other facultative species have an aerotaxis sensor. In E.coli this is Aer, composed of a modified chemoreceptor with the sensor domain containing an FAD binding PAS domain (Bibikov et al., 1997; Taylor et al., 2001; Ma et al., 2005). In other species, the MCP signaling domain might be fused to a cytochrome-like domain, but in all cases the receptor senses electron flow and signals the change to the chemosensory pathway. Obligate anaerobes have been identified with the signaling domain fused to a heme. However, no Aer homolog or modified MCP chemoreceptor has been identified in purple bacteria. Independent deletions of all genes encoding proteins with possible PAS domain did not result in any change in aerotaxis behavior (unpublished data). Purple bacteria do however show an aerotactic response. This indeed can be positive or negative depending on whether the cells have been grown under aerobic or photoheterotrophic conditions. When cultures are drawn into capillary tubes they tend to form bands at specific concentrations of oxygen, the position of the band depending on growth conditions and changing if oxygen or nitrogen is blown across the meniscus (Romagnoli et al., 2002). In Rba. sphaeroides aerotactic responses have been shown to depend on active electron transfer, inhibitors of electron transfer preventing aerotactic responses in the presence of oxygen (Gauden and Armitage, 1995). Data suggest that changes in the rate of electron transfer are sensed, rather than the electron donor or acceptor itself. The signal is then processed through the chemosensory pathway, as mutations in the Che pathways cause a loss of aerotaxis as well as chemotaxis (Romagnoli and Armitage, 1999; Romagnoli et al., 2002). What links electron transfer rate sensing to the chemosensory pathway is unclear. In Rba. sphaeroides the PrrA/B histidine kinase system (RegA/B in Rba. capsulatus) is a global regulator controlling expression of the photosynthetic system and the nitrogen fixation system. This system also regulates the expression of CheOp2, suggesting that a different copy number of the chemosensory proteins in the polar cluster is required under different conditions, with the highest level of expression under aerobic conditions and the
652 lowest under low light photoheterotrophic conditions (unpublished data). Deletion of PrrA also causes a loss of aerotaxis, suggesting that PrrA may in some way interact with the chemosensory system, although it cannot be ruled out that the effect is an indirect result of a change in CheOp2 protein copy number. 2. Phototaxis Colonies of Rsp. centenum move over hard agar surfaces towards infrared light and away from visible light. For the colonies to move over hard agar surfaces they need to become multiflagellate, a developmental feature regulated by a pathway homologous to a chemosensory pathway (see above). In Rsp. centenum a photosensory protein has been identified which controls the photosensory behavior of colonies of bacteria moving using lateral flagella over a solid surface. Deletion of the gene ptr results in the loss of all phototactic responses (Jiang et al., 1997; Jiang and Bauer, 2001). The encoded protein, Ptr, has the conserved cytoplasmic domain of an MCP but shows no conservation in the periplasmic domain, although it may contain a heme group. Again the signals are transmitted through the proteins of the chemosensory machinery to control motor behavior. Despite much mutagenesis, a specific photosensor has not been clearly identified in Rba. sphaeroides. PYP (photoactive yellow protein), a PAS domain containing protein, has been implicated in blue light negative photoresponses in Ectothiorhodospira halophila, the spectrum causing negative photoresponses matching that of PYP (Sprenger et al., 1993). However, deletion of all the genes including PAS domains individually in Rba. sphaeroides has not identified a clear photosensor. In some strains of Rba. sphaeroides the deletion of PYP produced a reduction in photoresponses to blue light, but the light intensities required to cause a response may suggest a secondary effect. PYP may be involved in the expression of a photosensor, and indeed not all Rba. sphaeroides strains encode PYP (Kort et al., 2000). Interestingly, the PYP in Rba. capsulatus is encoded close to the gas vacuole genes, and the difference in the photocycle kinetics of PYP from Rhodobacter species has lead to the suggestion they might be involved in regulating gas vacuole activity (Kyndt et al., 2004). Deletion of AppA, a blue light sensor involved in photosynthetic gene expression, has no effect on swimming behavior, nor does deletion of bacteriophytochrome gene (unpublished
Judith P. Armitage data). It is therefore unclear whether there is a true negative response to blue light in species such as Rba. sphaeroides. The dominant light-dependent behavioral response is to a drop in light intensity; this response serves to trap bacteria in actinic light. As a bacterium swims over a dark → light boundary there is no obvious response, but when it swims over a light → dark boundary there is a tumble, reversal or a stop (depending on the species) and when smooth swimming resumes it is often back into the light. This results in trapping in the light. The reversal/tumbling response when moving over a light → dark boundary depends on the electron transfer chain, as in aerotaxis, with inhibitors of electron transfer having the same effect as a drop in light intensity; indeed a pulse of oxygen will have the same response as there is a transient reduction in electron flow through the photosynthetic system (Grishanin and Bibikov, 1997; Grishanin et al., 1997). It therefore seems likely that the system sensing electron flow through the shared photosynthetic and respiratory electron transfer chain sends a signal to the chemosensory apparatus. Recent results show that autophosphorylation of the histidine kinase RegB (PrrB) is altered by binding to ubiquinone (UQ), suggesting there may be a direct redox sensing mechanism through the redox state of UQ, regulating gene expression (Swem et al., 2006). Given the possible role of PrrB (RegB) in the activity of the chemosensory system, it remains to be identified whether the signal to the chemosensory pathway is through the UQ pool via the Prr regulatory system. We are only just starting to understand the environmentally regulated behavior in species living in complex environments. These species can tune their responses to their current needs and these are hard to replicate in a laboratory environment. Motility and its tactic control is involved in not only moving towards nutrients, but moving towards light and oxygen and probably changing the response to different nutrients under different growth conditions is also essential. Recent work has also shown that motility and taxis may be essential for reaching surfaces and developing biofilms. Most research is undertaken to generate fast results, but normally purple bacteria often grow slowly and grow in changing environments, not least the dial rhythm. There has been no research on the effects of light/dark cycles on bacterial behavior, or indeed metabolic activity, although recent research and genome sequencing has shown that Rba. sphaeroides contains homologs of the cyanobacterial clock
Chapter 32
Swimming and Behavior
genes and gene expression show different patterns and different cycle periods under different growth conditions (Min et al., 2005). We also now realize that the chemosensory pathways are just a part of the network of histidine kinase dependent two-component systems operating in bacteria. There are on average about 50 different two-component pathways in most species, but there may be over 200 in some species. Bacterial cells therefore measure and respond to changes in their environment by altering gene expression in addition to altering swimming behavior. All these systems must be balanced to produce the optimal response for the current environment and understanding these interactive networks is the challenge of the next decade. Acknowledgments I would like to thank Carrie Harwood and Bob Belas for unpublished data, and the UK Biotechnology and Biological Sciences Research Council for funding my research over the years. References Alon U, Surette MG, Barkai N and Leibler S (1999) Robustness in bacterial chemotaxis. Nature 397: 168–171 Ames P, Studdert CA, Reiser RH and Parkinson JS (2002) Collaborative signaling by mixed chemoreceptor teams in Escherichia coli. Proc Natl Acad Sci USA 99: 7060–7065 Appleman JA, Chen LL and Stewart V. (2003) Probing conservation of HAMP linker structure and signal transduction mechanism through analysis of hybrid sensor kinases. J Bacteriol 185: 4872–4882 Armitage JP, Kelly, DJ and Sockett RE (1995) Flagellate motility, behavioral responses and active transport in purple nonsulfur bacteria. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 1005–1028. Kluwer Academic Publishers, Dordrecht Armitage JP and Macnab RM (1987) Unidirectional intermittent rotation of the flagellum of Rhodobacter sphaeroides. J Bacteriol 169: 514–518 Asai Y, Kawagishi I, Sockett RE and Homma M. (2000) Coupling ion specificity of chimeras between H(+)- and Na(+)-driven motor proteins, MotB and PomB, in Vibrio polar flagella. EMBO J 19: 3639–3648. Berg HC (2003) The rotary motor of bacterial flagella. Ann Rev Biochem 72: 19–54 Berleman JE and Bauer CE (2005a) A che-like signal transduction cascade involved in controlling flagella biosynthesis in Rhodospirillum centenum. Mol Microbiol 55: 1390–1402 Berleman JE and Bauer CE (2005b) Involvement of a Che-like
653 signal transduction cascade in regulating cyst cell development in Rhodospirillum centenum. Mol Microbiol 56: 1457–1466 Bibikov SI, Biran R, Rudd KE and Parkinson JS (1997) A signal transducer for aerotaxis in Escherichia coli. J Bacteriol 179: 4075–407 Bray D (2002) Bacterial chemotaxis and the question of gain. Proc Natl Acad Sci USA 99: 7–9 Ferre A, de la Mora J, Ballado T, Camarena L and Dreyfus G (2004) Biochemical Study of Multiple CheY Response Regulators of the Chemotactic Pathway of Rhodobacter sphaeroides. J Bacteriol 186: 5172–5177 Gauden DE and Armitage JP (1995) Electron transport-dependent taxis in Rhodobacter sphaeroides. J Bacteriol 177: 5853–5859 Grishanin RN and Bibikov SI (1997) Mechanisms of oxygen taxis in bacteria. Biosci Rep 17: 77–83 Grishanin RN, Gauden DE and Armitage JP (1997) Photoresponses in Rhodobacter sphaeroides: role of photosynthetic electron transport. J Bacteriol 179: 24–30 Hamblin PA, Bourne NA and Armitage JP (1997) Characterization of the chemotaxis protein CheW from Rhodobacter sphaeroides and its effect on the behavior of Escherichia coli. Mol Microbiol 24: 41–51 Hirschman A, Boukhvalova M, VanBruggen , Wolfe AJ and Stewart RC (2001) Active site mutations in CheA, the signal-transducing protein kinase of the chemotaxis system in Escherichia coli. Biochemistry 40: 13876–13887 Jiang ZY and Bauer CE (2001) Component of the Rhodospirillum centenum photosensory apparatus with structural and functional similarity to methyl-accepting chemotaxis protein chemoreceptors. J Bacteriol 183: 171–177 Jiang ZY, Gest H and Bauer CE (1997) Chemosensory and photosensory perception in purple photosynthetic bacteria utilize common signal transduction components. J Bacteriol 179: 5720–5727 Kobayashi K, Saitoh T, Shah DSH, Ohnishi K, Goodfellow IG, Sockett RE and Aizawa SI (2003) Purification and characterization of the flagellar basal body of Rhodobacter sphaeroides. J Bacteriol 185: 5295–5300 Leake MC, Chandler JH, Wadhams GH, Bai F, Berry RM and Armitage JP (2006) Stoichiometry and turnover in single, functioning membrane protein complexes. Nature 443: 355–358 Levit MN, Liu Y and Stock JB (1998) Stimulus response coupling in bacterial chemotaxis: receptor dimers in signalling arrays. Mol Microbiol 30: 459–466 Ma Q, Johnson MS and Taylor BL (2005) Genetic Analysis of the HAMP domain of the Aer aerotaxis sensor localizes flavin adenine dinucleotide-binding determinants to the AS-2 helix. J Bacteriol 187: 193–201 Macnab RM (2000) Microbiology. Action at a distance — bacterial flagellar assembly. Science 290: 2086–2087 Martin AC, Gould M, Byles E, Roberts MAJ and Armitage JP (2006) Two chemosensory operons of Rhodobacter sphaeroides are regulated independently by sigma 28 and sigma 54. J Bacteriol 188: 7932–7940 Martínez del Campo A, Ballado T, de la Mora J, Poggio S, Camarena L and Dreyfus G (2007) Chemotactic control of the two flagellar systems of Rhodobacter sphaeroides is mediated by different sets of CheY and FliM proteins. J Bacteriol 189: 8397–8401 McClain J, Rollo DR, Rushing BG and Bauer CE (2002) Rho-
654 dospirillum centenum utilizes separate motor and switch components to control lateral and polar flagellum rotation. J Bacteriol 184: 2429–2438 Morehouse KA, Goodfellow IG and Sockett RE (2005) A chimeric N-terminal Escherichia coli-C-terminal Rhodobacter sphaeroides FliG rotor protein supports bidirectional E. coli flagellar rotation and chemotaxis. J Bacteriol 187: 1695–1701 Poggio S, Osorio A, Dreyfus G and Camarena L (2002) The four different sigma54 factors of Rhodobacter sphaeroides are not functionally interchangeable. Mol Microbiol 46: 75–85 Poggio S, Osorio A, Dreyfus G and Camarena L (2005) The flagellar hierarchy of Rhodobacter sphaeroides is controlled by the concerted action of two enhancer-binding proteins. Mol Microbiol 58: 969–983 Poggio S, Abreu-Godger C, Fabela S, Osorio A, Dreyfus G, Vincesa P and Camarena L (2007) A complete set of flagellar genes acquired by horizontal gene transfer coexist with the endogeous flagellar system of Rhodobacter sphaeroides. J Bacteriol 189: 3208–3216 Poole PS, Smith MJ and Armitage JP (1993) Chemotactic signalling in Rhodobacter sphaeroides requires metabolism of attractants. J Bacteriol 175: 291–294 Porter SL and Armitage JP (2002) Phosphotransfer in Rhodobacter sphaeroides chemotaxis. J Mol Biol 324: 35–45 Porter SL and Armitage JP (2004) Chemotaxis in Rhodobacter sphaeroides requires an atypical histidine protein kinase. J Biol Chem 279: 54573–54580 Porter SL, Warren AV, Martin AC and Armitage JP (2002) The third chemotaxis locus of Rhodobacter sphaeroides is essential for chemotaxis. Mol Microbiol 46: 1081–1094 Porter SL, Wadhams GH, Martin AC, Byles ED, Lancaster DE and Armitage JP (2006) The CheYs of Rhodobacter sphaeroides. J Biol Chem 281: 32694–32704 Ragatz L, Jiang Z-Y, Bauer CE and Gest H (1994) Phototactic purple bacteria. Nature 370: 104 Romagnoli S and Armitage JP (1999) Role of the chemosensory pathways in transient changes in swimming speed of Rhodobacter sphaeroides induced by changes in photosynthetic electron transport. J Bacteriol 181: 34-39 Romagnoli S, Packer HL and Armitage JP (2002) Tactic responses to oxygen in the phototrophic bacterium Rhodobacter sphaeroides WS8N. J Bacteriol 184: 5590–5598 Sourjik V and Berg HC (2001) Using fluorescent resonance energy transfer (FRET) to study interactions of cytoplasmic and motor proteins in chemotaxis of E. coli. Biophys J 80: 773 Sourjik V and Berg HC (2002a) Binding of the Escherichia coli response regulator CheY to its target measured in vivo by
Judith P. Armitage fluorescence resonance energy transfer. Proc Natl Acad Sci USA 99: 12669–12674 Sourjik V and Berg HC (2002b) Receptor sensitivity in bacterial chemotaxis. Proc Natl Acad Sci USA 99: 123–127 Sourjik V and Berg HC (2004) Functional interactions between receptors in bacterial chemotaxis. Nature 428: 437–441 Sprenger WW, Hoff WD, Armitage JP and Hellingwerf KJ (1993) The eubacterium Ectothiorhodospira halophila is negatively phototactic, with a wavelength dependence that fits the absorption spectrum of the photoactive yellow protein. J Bacteriol 175: 3096–3104 Stewart RC, Russell CB, Roth AF and Dahlquist FW (1988) Interaction of CheB with chemotaxis signal transduction components in Escherichia coli: Modulation of the methylesterase activity and effects on cell swimming behavior. Cold Spring Harbor Symp Quant Biol 53: 27–40 Stewart RC and Van Bruggen R (2004) Association and dissociation kinetics for CheY interacting with the P2 domain of CheA. J Mol Biol 336: 287–301 Swem LR, Gong X, Yu CA and Bauer CE (2006) Identification of a ubiquinone-binding site that affects autophosphorylation of the sensor kinase RegB. J Biol Chem 281: 6768–6775 Taylor BL, Rebbapragada A and Johnson MS (2001) The FADPAS domain as a sensor for behavioral responses in Escherichia coli. Antioxid Redox Signal 3: 867–879 Thomas DR, Francis NR, Xu C and DeRosier DJ (2006) The threedimensional structure of the flagellar rotor from a clockwiselocked mutant of Salmonella enterica Serovar Typhimurium. J Bacteriol 188: 7039–7048 Thompson SR, Wadhams GH and Armitage JP (2006) The positioning of cytoplasmic protein clusters in bacteria. Proc Natl Acad Sci USA 103: 8209–8214 Wadhams GH and Armitage JP (2004) Making sense of it all: Bacterial chemotaxis. Nat Rev Mol Cell Bio 5: 1024–1037 Wadhams GH, Martin AC, Porter SL, Maddock JR, Mantotta JC, King HM and Armitage JP (2002) TlpC, a novel chemotaxis protein in Rhodobacter sphaeroides, localizes to a discrete region in the cytoplasm. Mol Microbiol 46: 1211–1221 Wadhams GH, Warren AV, Martin AC and Armitage JP (2003) Targeting of two signal transduction pathways to different regions of the bacterial cell. Mol Microbiol 50: 763–770 Wadhams GH, Martin AC, Warren AV and Armitage JP (2005) Requirements for chemotaxis protein localization in Rhodobacter sphaeroides. Mol Microbiol 58: 895–902 Zhao R, Collins EJ, Bourret RB and Silversmith RE (2002) Structure and catalytic mechanism of the E. coli chemotaxis phosphatase CheZ. Nat Struct Biol 9: 570–575
Chapter 33 Metals and Metalloids in Photosynthetic Bacteria: Interactions, Resistance and Putative Homeostasis Revealed by Genome Analysis Francesca Borsetti1, Pier Luigi Martelli2, Rita Casadio2 and Davide Zannoni1* Department of Biology, 1General Microbiology and 2Bioinformatics Units, University of Bologna, 42 Irnerio, 40126 Bologna, Italy Summary ............................................................................................................................................................... 656 I. Introduction..................................................................................................................................................... 656 II. Classification of Metals and Metalloids by their Toxicity or Essentiality to Bacterial Cells (Groups I, II and III). ....................................................................................................................................... 656 III. Metal Toxicity, Tolerance and Resistance: Generalities ................................................................................ 657 IV. General Features on Microbial Metal Resistance/Tolerance Mechanisms .................................................... 658 A. Active Transport /Efflux Systems .................................................................................................... 658 1. CBA Efflux Pumps................................................................................................................... 658 2. Cation Diffusion Facilitators (CDF) Family .............................................................................. 661 3. P-type ATPases ...................................................................................................................... 661 B. Enzymatic Detoxification .................................................................................................................. 662 C. Intracellular and Extracellular Sequestration.................................................................................... 665 D. Accessory Mechanisms (Metal Exclusion by Permeability Barrier, Mutation of the Target Site) ..... 665 V. On the Bacterial Interactions with Metals and Metalloids ............................................................................... 666 A. Group I Metals (Fe, Mn, Mo) ............................................................................................................ 666 B. Group II Metals (Zn, Ni, Cu, Co and Cr) ........................................................................................... 667 C. Group III Metals and Metalloids (Cd, Hg, Pb, As, Te, Se)................................................................ 668 1. Cd, Hg, Pb, As ....................................................................................................................... 668 2. Se and Te................................................................................................................................ 669 a. Selenium........................................................................................................................ 669 b. Tellurium ........................................................................................................................ 671 3. Mechanism(s) of Metalloid Toxicity ......................................................................................... 673 4. Methylation .............................................................................................................................. 674 5. Other Metalloids ...................................................................................................................... 674 VI. Metal(loid)s Homeostasis in Phototrophs as Revealed by Genome Analysis ................................................ 675 A. Model Microorganisms for the Study of Metal(loid)s/Bacterial Interactions...................................... 675 B. Database Selection .......................................................................................................................... 676 1. The ‘Bait and Prey’ Virtual Strategy ........................................................................................ 676 a. Clustering of Prokaryotic Genes in Similarity Sets ........................................................ 676 b. Genome Screening with Known Baits: Looking for Preys ............................................. 679 VII. Concluding Remarks ...................................................................................................................................... 679 Acknowledgments ................................................................................................................................................. 682 References ............................................................................................................................................................ 682
*Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 655–689. © 2009 Springer Science + Business Media B.V.
656
Francesca Borsetti, Pier Luigi Martelli, Rita Casadio and Davide Zannoni
Summary Microbial metabolism of metals and metalloids has been a topic of interest since the early seventies when it was recognized that bacteria are involved in the transformation of metal compounds in the environment. For this reason, bacterial processing of inorganic compounds has been reviewed several times over the past decade. However, this is the first time that the metal(loid)s metabolism of photosynthetic bacteria has been considered in detail as compared to non-phototrophs. Another aspect touched on for the first time in this chapter is the analysis of genomes of representative phototrophs in an attempt to reveal common or unique features of the interactions of these bacteria with metal(loid)s. This work not only identified new genes linked to metal resistance, but also contributed to unify the nomenclature used among the genomes of different photosynthetic species. Based on our analysis, similarities and differences can be used more efficiently in future work as new ‘preys’ that have been either hypothesized or described generically in the past have also been uncovered. I. Introduction Microorganisms have coexisted with metals since the beginning of life. Indeed, a wide range of divalent or transition metals are present in the active sites of many enzymes linked to chemotrophic and photosynthetic metabolisms as the chemical properties of metals have been recruited to catalyze key reactions, or maintain protein structure. Under physiological metabolic conditions, small amounts of metals are required so that their uptake is regulated by fine homeostatic mechanisms ensuring sufficient, but not excessive, acquisition. On the other hand, dozens of metals have no relevant biological function, and even might cause serious damage to cells in the event of abnormal accumulation. Interactions between the bacteria and metals must be accurately considered for at least two main reasons: a) bacteria are interesting models to better understand metal speciation in higher organisms, and b) bacteria can be used in environmental biotechnology. Indeed, microorganisms play an important role in establishing the environmental fate of an element, by modifying its mobility and bioavailability, and hence its intrinsic toxicity (White et al., 1997). Emerging technologies, such as biosorption of metals into biomass and precipitation of ions, take cues from metal-related bacterial metabolisms (Gadd, 2000). Purple photosynthetic bacteria are a very versatile group of microorganisms that can grow under difAbbreviations: CDF – Cation Diffusion Facilitator; COX – cytochrome c oxidase; E. – Escherichia; MIC – minimal inhibitory concentration; P. – Pseudomonas; R. – Ralstonia; Rba. – Rhodobacter; RND – Resistance-Nodulation-Cell division; ROS – reactive oxygen species; Rps. – Rhodopseudomonas; Rsp. – Rhodospirillum; S. – Staphylococcus; SAM – S-adenosylmethionine; Trx – thioredoxin
ferent conditions: photoautotrophically or photoheterotrophically in the light, and chemotrophically in darkness under aerobic conditions. Some species are also able to grow by anaerobic respiration in the dark. This versatility allows purple photosynthetic bacteria to easily adapt to environmental changes. Thanks to these metabolic features, facultative photosynthetic microorganisms could be successfully employed in bioremediation processes for the treatment of polluted area. This chapter summarizes the available information on the response of photosynthetic bacteria to toxic metals and metalloids (non-metals). The first parts (I to IV) of this chapter will describe briefly the mechanisms by which microorganisms become tolerant or resistant to specific metal(loid) elements. Then, attention will be turned to the interaction of specific metals and metalloids with a few phototrophs (Section V). For the first time, a genomic survey of several representative species, including Chlorobium tepidum, Rhodospirillum (Rsp.) rubrum, Rhodobacter (Rba.) sphaeroides, Rhodopseudomonas (Rps.) palustris and Roseobacter denitrificans, will be reported (Section VI) in an attempt to reveal the existence of common, or unique, mechanisms for metal(loid) reduction, resistance and homeostasis. II. Classification of Metals and Metalloids by their Toxicity or Essentiality to Bacterial Cells (Groups I, II and III). The main factor for the effect of a metal(loid) on a bacterial cell is, by definition, its actual role in the cell. Some metals, such as iron, potassium, magnesium, manganese, molybdenum, cobalt, chromium, copper, zinc and nickel are required nutrients, and are essential
Chapter 33
Metals and Metalloids in Photosynthetic Bacteria
in trace amounts for regulation of osmotic pressure, redox processes and stabilization of macromolecules through electrostatic interactions. Most of them (Cr, Mn, Fe, Co, Ni, Cu, Zn and Mo) are transition metals with a strong ability to form complex compounds. Thus, as soon as the accumulation of these elements exceeds the biological need, heavy-metal ions can form unspecific and potentially toxic complexes inside the cell. Other elements, such as silver, cadmium, lead, antimony and mercury have no biological roles, and are non-essential. However once in the cell, they can interact with essential cellular components through covalent or ionic bonds (Nies, 1992a). From a physiological point of view, elements fall into three groups, namely: a) group I, which includes essential metals characterized by their low toxicity (i.e., Fe, Mo, Mn); b) group II, which includes metals with moderate importance as trace elements but which could be potentially very toxic (Zn, Ni, Cu, V, Co, W and Cr), and finally c) group III, which includes metals without any biological role and which are toxic per se (Ag, Hg, Cd, Sb and Pb). Among the non-metal (or metalloid) elements, there are both essential (i.e., selenium, Se) and non-essential (i.e., arsenic, As, and tellurium, Te) elements. In general, a metal(loid) must enter the bacterial cell in order to exert its physiological effect. For a given microorganism, the cell permeability to an element depends also on the chemical speciation of that element, and therefore on the chemical-physical proprieties, such as the pH and osmolarity, of its environment (Mukhopadhyay et al., 2002). Furthermore, the chemical energy source and the presence of a complex microbial community can also influence the actual bioavailability of the metal(loid) (Hassen et al., 1998; Sandrin and Maier, 2003). Once inside the cell, metals might exert their toxicity via several mechanisms. Most of the heavymetal ions are able to bind to SH groups, affecting the activity of sensitive enzymes or the role of structural proteins (Kachur et al., 1998). Cations may substitute for physiologically essential cations within the enzymes, modifying their functionality or causing an imbalance of the redox state of the cell leading to oxidative stress (Stohs and Bagchi, 1995). Metalloid oxyanions, as well, may be used in place of an essential anion (i.e., such as arsenate instead of phosphate) (Mukhopadhyay et al., 2002). Indeed, the toxicity of a metal element is strongly dependent on its oxidation state (as it determines the chemical proprieties of the ion) and on the role of other struc-
657
turally or functionally related chemical species in the cell. Thus, the need to acquire or reject metal(loid)s has led to the selection of a complex repertoire of interaction mechanisms ensuring bacterial adaptation to changing environments. Hereafter, we will refer to this adaptation as tolerance or resistance mechanisms. III. Metal Toxicity, Tolerance and Resistance: Generalities The entry of metals or metalloids into cells is due to the relatively low specificity of several physiological transport systems. Indeed, under ‘housekeeping’ conditions, fast and non-specific transporters allow abundant accumulation of intracellular ions. These systems, usually driven by the chemiosmotic gradient, consume less energy and are effective. In Gram negative bacteria, for example, several cations such as Cd2+, Co2+, Ni2+ and Zn2+ can be transported by the magnesium uptake system (Nies and Silver, 1989a; Nies, 1992a), which is constitutively expressed. Thus, in contaminated environments, accumulation of non-desired elements is mediated by this ‘opengate’ mechanism (Nies and Silver, 1995). A second uptake mechanism, which is usually induced under metal ion starvation, is highly specific but less efficient, and requires extra energy (often provided by ATP hydrolysis) (Nies, 1992a; Silver and Walderhaug, 1992; Mukhopadhyay et al., 2002). The easiest way by which a microbial population can limit metal toxicity is through selection of mutations affecting the uptake in some way; the resulting mutants are called metal-tolerant (Nies, 1999). This kind of tolerance is quite frequent (Park et al., 1976; Rosen, 1996; Sanders et al., 1997), but it is also easily lost when the selection pressure (the metal or metalloid) is relieved. Conversely, under conditions in which the metal ions are in excess, more specific mechanisms have to be developed, including active efflux, transformation to less toxic chemical species or sequestration. These systems are either plasmid- or chromosome-encoded (see Section IV). The response of a microbial community to metals or metalloids in ‘naturally’ polluted environments is unpredictable, because it depends on multiple factors related to the specific bacterial microenvironments. Therefore, the conventional way to define the tolerance or resistance of a microorganism to a given chemical is by experimental determination of
658
Francesca Borsetti, Pier Luigi Martelli, Rita Casadio and Davide Zannoni
MIC in a medium that best supports the growth of the microorganism or a group of microorganisms. It is worth noting that the activity of a free metal ion rarely parallels the total amount of metal added to the medium. Indeed, the metal-binding capacity of the external cell surface, along with chelation to components of the growth medium, can strongly modify the actual bioavailability of the free metal (Angle and Chaney, 1989; Chang et al., 1993). An alternative way to define the toxic effect of a metal(loid) is to determine the concentration (EC50 value) of that metal(oid) required to reduce by 50% the growth rate observed in a medium without any supplement (Giotta et al.,2006). This parameter can also be used for elements that are not intrinsically toxic, but that can have deleterious effects if present in excess. IV. General Features on Microbial Metal Resistance/Tolerance Mechanisms A. Active Transport /Efflux Systems Although there are a number of ways by which cells can become resistant to a metal, the most frequent ones are those that control the transport of ionic species to decrease the internal concentration of the toxic element. This process can be done by either the loss of an uptake system, or by the modification of an efflux system (Dey and Rosen, 1995). The former strategy is often transient, and the components of the latter systems are chromosomal or plasmid-mediated. These systems may or may not be ATP linked, and are highly specific for the cation or anion transported (Nies and Silver, 1995). Three main efflux systems have been described: 1. CBA Efflux Pumps These chemiosmotic ion/proton exchanger systems are composed of three polypeptides: (i) an inner membrane protein (subunit A) of over 1000 amino acid residues in length; (ii) an outer membrane protein (subunit C) and, (iii) a coupling protein (subunit B) connecting the two subunits in the periplasmic space, also called membrane fusion protein. The inner membrane subunit A belongs to the ‘Resistance-Nodulation-Cell division’ (RND) superfamily, a large group of proteins linked to transport of heavy metals, solutes and nodulation factors in bacteria and archaea, as well as sterols in Eukaryotes (Nies, 2003). This protein
complex may export the substrate from the cytoplasm, the cytoplasmic membrane or the periplasm across the outer membrane to the exterior of the cell, and is driven by protons that are extruded by the membrane redox systems. The RND protein confers substrate specificity, whereas the outer membrane factors are often interchangeable as members of the outer membrane factor family could functionally substitute for each other (Poole et al., 1993; Andersen et al., 2001). The first member of the CBA efflux pumps was the CzcCBA system of Ralstonia (R.) metallidurans CH34. In this species, the czc resistance determinant is harbored by the megaplasmid pMOL30, and it contains the structural genes for the outer membrane factor CzcC, the membrane fusion protein CzcB and the CzcA protein of the RND family. CzcA, located in the inner membrane, pumps heavy metals out of the cytoplasm, CzcB then shuttles them from the inner to the outer membrane, and CzC completes the process by exporting metals to the external environment (Diels et al., 1995). The czcCBA genes form an operon, and are flanked by regulatory genes involved in metal-dependent regulation of the czcCBA expression (Mergeay et al., 2003). The Czc system confers resistance to Co2+, Zn2+ and Cd2+. Analogous to Czc is the Cnr system encoded by the plasmid pMOL28 and confering resistance to Co2+ and Ni2+ in R. metallidurans CH34. The Cnr resistance determinant is composed of the cnrCBA structural region, and a predicted cnrYHZ regulatory region. Another CBA efflux permease has been found on plasmid pTOM9 of the related bacterium R. metallidurans 31A. This system, called Ncc, mediates resistance to Cd2+ in addition to Co2+ and Ni2+ (Mergeay et al., 2003; Nies, 2003). The czrSRCBA locus, which confers Zn2+ and Cd2+ resistance in Pseudomonas (P.) aeruginosa CMG103, shows significant similarity to czc, cnr and ncc determinants of Ralstonia species, where czrS and czrR code for the sensor and the response regulator, respectively, of a two-component regulatory system (Hassan et al., 1999). A CBA efflux system (Cus) has also been found on the Escherichia (E.) coli chromosome, and it confers resistance to Cu+/Cu2+ and Ag+. The structural genes cusCFBA form an operon, and encode the RND protein CusA, the periplasmic protein CusB, the outer membrane protein CusC, and an additional small factor CusF (Rensing and Grass, 2003) acting as a periplasmic copper chaperone. The cus determinant is induced by copper (and to a lesser extent by silver), and is regulated by the CusRS two-component regulatory system. To
Chapter 33
Metals and Metalloids in Photosynthetic Bacteria
659
Table 1. Genes and predicted functions associated with metal(loid)s resistance in selected microorganisms Metal(loid) resistance system mer
ars
aso
Location
Metal(loid)
Protein name
Protein function
References
Tn21, Shigella flexneri; pDU1358, S. marcescens (merBDA); Tn501, P. aeruginosa (merAD); pPB, P. stuzteri (merRBTPCA) Tn4378 on pMOL28 and Tn4380 on pMOL30 of R. metallidurans CH34 (merRTPADE)
Hg2+
MerA
Mercuric reductase
MerB
Organomercurial lyase
MerC
Transport protein
Silver and Phung (1996) Osborn et al. (1997) Qian et al. (1998) Silver and Phung (2005a)
MerD
Regulatory protein
MerR
Trans-acting protein
MerP
Transport protein
MerT
Transport protein
ArsA ArsB
ATPase Membrane protein involved in transport Arsenate reductase Trans-acting co-repressor Trans-acting repressor Arsenite oxidase and transport
R773, E. coli
Alcaligenes faecalis chromosome
arr
czc
pMOL30, R. metallidurans CH34
AsO43–, As(OH)3, Sb(III), TeO32–
As(OH)3
ArsC ArsD ArsR AsoAB
AsO43–
ArrAB
Respiratory arsenate reductase
Cd2+, Zn2+, Co2+
CzcCBA
CBA (RND) family efflux permease CDF chemiosmotic efflux pump Two components regulatory systems P-type efflux ATPase Metal binding protein with putative regulatory role (ArsR family) CBA (RND) family efflux permease Regulatory systems CBA (RND) family efflux permease Regulatory systems
CzcD
cad
pI258, S. auerus
Cd2+
cnr
pMOL28, R. metallidurans CH34
Co2+ , Ni2+
ncc
pTOM9, R. metallidurans 31A
Ni2+ , Co2+ , Cd2+
nre
pTOM9, R. metallidurans 31A
Ni2+
CzcR, CzcS CadA CadC
CnrCBA CnrYXH NccCBA NccYXH NreCBA
Silver et al. (1981) Rosen et al. (1988) Turner et al. (1992a) Bruins et al. (2000) Mukhopadhyay et al. (2002) Mukhopadhyay et al. (2002) Silver and Phung (2005b) Mukhopadhyay et al. (2002) Silver and Phung (2005b) Nies (1992b) Diels et al. (1995) Anton et al. (1999) Mergeay et al. (2003) Nies (2003) Novick et al. (1979) Nies and Silver (1995) Bruins et al. (2000) Nies (1999) Mergeay et al. (2003) Nies (2003) Mergeay et al. (2003) Nies (2003) Schmidt and Schlegel (1994) Mergeay et al. (2003) Nies (2003)
Francesca Borsetti, Pier Luigi Martelli, Rita Casadio and Davide Zannoni
660 Table 1. Continued Metal(loid) resistance system czr
Location
Metal(loid)
Protein name
Protein function
References
P. aeruginosa CMG103 chomosome
Cd2+, Zn2+
CzrCBA
CBA (RND) family efflux permease Two components regulatory systems P-type efflux ATPase Outer membrane lipoprotein Signal peptidase Intracellular Pb2+ binding protein Pb2+ uptake permease Regulatory protein (MerR family) Cu+ transporting ATPases
Hassan et al. (1999)
CzrR, CzrS pbr
pMOL30, R. metallidurans CH34
Pb2+
PbrA PbrB PbrC PbrD PbrT PbrR
cop
cus
Cu+
E. coli chromosome
Cu+
CopA, CopB CopY CopZ CopA
E. coli chromosome
Cu+/Cu2+, Ag+
CusCBA
Enterococcus hirae chromosome
CusF
cue
E. coli chromosome
Cu+, Ag+
CusR, CusS CueR
CueO pco
plasmid pRJ1004, E. coli (pcoABCDRSE)
Cu+/Cu2+
PcoA PcoB PcoC PcoD PcoE
chr
smt
pMOL28, R. metallidurans CH34
CrO42–
Synechococcus chromosome
Cd2+, Zn2+, Cu2+
ChrA
Cu-responsive repressor Intracellular chaperon Cu+ transporting ATPase CBA (RND) family efflux permease Periplasmic Cu+ binding protein Two components regulatory systems Regulatory protein (MerR family) activated by intracellular Cu+ and Ag+ levels Multicopper oxidase under O2 control Multicopper oxidase family Outer membrane protein Periplasmic Cu-binding protein Cu-transport protein (uptake into the cytoplasm) Periplasmic Cu-binding protein Membrane efflux protein
ChrB
Putative regulatory protein
SmtA SmtB
Metallothionein Trans-acting transcriptional repressor
Borremans et al. (2001) Mergeay et al. (2003) Chen et al. (2005) Silver and Phung (2005a)
Solioz and Stoyanov (2003)
Rensing and Grass (2003) Nies (2003) Rensing and Grass (2003)
Rensing and Grass (2003)
Tetaz and Luke (1983) Rensing and Grass (2003)
Nies and Silver (1989b) Nies et al. (1990) Juhnke et al. (2002) Mergeay et al. (2003) Gupta et al. (1992) Turner et al. (1993) Canovas et al. (2003)
Chapter 33
Metals and Metalloids in Photosynthetic Bacteria
661
Table 1. Continued Metal(loid) resistance system
Location
Metal(loid)
Protein name
Protein function
References
ter (terZABCDEF)
pMER610 and R478, E. coli; Proteus mirabilis chromosome E. coli chromosome
TeO32–
TerA, TerD, TerE TerC TerD TehA
Taylor (1999) Toptchieva et al. (2003)
Tantalean et al. (2003)
kilAtelAtelB (cryptic)
RP4, RK2 plasmids of E. coli; Rb. sphaeroides chromosome (telA)
TeO32–
KilA (or KlaA) TelA (or KlaB) TelB
tmp
P. syringae (pathovar pisi) chromosome Rb. sphaeroides chromosome Geobacillus stearothermophilus chromosome
TeO32–, SeO32–
TmpT
TeO32–
CysK
Putative stress response proteins Putative transporter cAMP binding protein Integral membrane homologues to C4 dicarboxylate transporter SAM-dependent methyltransferase Cytoplasmic protein (unknown function) Cytoplasmic protein (unknown function) Integral membrane protein (unknown function) SAM-dependent thiopurine methyltransferase Cysteine syntase
TeO32–
IscS
Cysteine desulfurase
teh (tehAB)
TeO32–
TehB
cysK isc
date, the Cus system is believed to transport copper directly across the outer membrane, and is one of the multiple copper-homeostasis mechanisms shown to be present in E. coli. 2. Cation Diffusion Facilitators (CDF) Family This group of transporters occurs in all three domains of life and, unlike the CBA system, is composed of a single polypeptide (Paulsen and Saier, 1997). It was first described along with the Zn2+ and Cd2+ CzcD efflux system of R. metallidurans (Nies, 2003; Haney et al., 2005) as a regulator of the CzcCBA resistance system (Nies, 1992b). However, it was shown later to mediate Co2+, Zn2+ and Cd2+ resistance even in the absence of CzcCBA, lowering the cytoplasmic concentration of these ions (Anton et al., 1999). The CDF system works as a divalent cation/proton exchanger in which the small CDF proteins are present as homodimers (Chao and Fu, 2004; Haney et al., 2005). Additional members of this family include the Zn2+ efflux system ZitB of E. coli and ZntA of Staphylococcus (S.) aureus (Xiong and Jayaswal, 1998; Anton et al., 2004).
Turner et al. (1995) Dyllick-Brenzinger et al. (2000) Liu et al. (2000) Turner et al. (1994a, b) O’Gara et al. (1997)
Cournoyer et al. (1998) O’Gara et al. (1997)
3. P-type ATPases The P-type ATPases form a superfamily of transport proteins driven by ATP hydrolysis (Fagan and Saier, 1994). As the P-type ATPases show a conserved proline residue near to a cysteine moiety, they are also called ‘CPx-type’ ATPases. Members of this family are very important for metal homeostasis and resistance, because they can work in both directions. Indeed, a P-type ATPase can either transport its substrate from the outside (or the periplasm) to the cytoplasm, or export it from the cytoplasm to the outside (or the periplasm). This group of transporters has a particular importance to determine the resistance to Cd2+, Cu+ and Pb2+ in several microorganisms. In S. aureus high concentrations of Cd2+ induce the expression of the CadA/CadC resistance system, encoded by the plasmid pI258 (Novick et al., 1979). The system is formed from the transporter protein CadA that mediates the ATP-dependent efflux of the cation, and the CadC metal-binding protein with a putative regulatory role (Tynecka et al., 1981; Nucifora et al., 1989; Tsai and Linet, 1993). In Enterococcus (E.) hirae two P-type ATPases are involved in copper homeostasis: CopA is thought to be a transporter important for the uptake of Cu+, while CopB medi-
662
Francesca Borsetti, Pier Luigi Martelli, Rita Casadio and Davide Zannoni
ates the efflux of the ion from the cytoplasm (Solioz and Stoyanov, 2003). It is worth noting that CopA exhibits only 32% sequence identity to CopB, but both proteins are indeed more related to CadA of S. aureus than to each other (Odermatt et al., 1993; Silver et al., 1993). The cop operon of Enterococcus hirae is regulated by the copper responsive repressor CopY and the intracellular chaperone CopZ (Solioz and Stoyanov, 2003). In E. coli the Cu+- translocating P-type ATPase CopA is required for copper resistance under both aerobic and anaerobic conditions. In this bacterium, CopA is involved in the transport of Cu+ from the cytoplasm to the periplasm, and its expression is regulated by CueR, a Mer-like activator (Solioz and Stoyanov, 2003). It is important to note that in E. coli the CopA system coexists with the Cus system belonging to the CBA family, and both systems participate in to copper homeostasis. It is believed that, while the latter system transports copper directly from the periplasm across the outer membrane, the CopA system does so across the cytoplasmic membrane (Rensing and Grass, 2003). The multi-copper oxidase CueO, also regulated by CueR, is thought to be involved in detoxification of the periplasm, specifically under anaerobic conditions (Rensing and Grass, 2003). In E. coli additional copper resistance can also be conferred by pco operon, carried by the plasmid pRJ1004 (Tetaz and Luke, 1983). 64Cu uptake experiments suggest that an energy-dependent copper efflux mechanism is associated with this system (Brown et al., 1995). R. metallidurans contains several genes for P-type ATPases. One of them is pbrA, a Pb2+ resistance determinant that mediates ion efflux (Solioz and Vulpe, 1996; Borremans et al., 2001). The pbr operon of this species includes other structural and regulatory genes. The PbrT product seems to be a membrane uptake-permease (although the accumulation of Pb2+ is counter-intuitive), PbrD could be an intracellular Pb2+ binding protein, and PbrB and PbrC are thought to be an outer membrane lipoprotein and a signal peptidase, respectively (Borremans et al., 2001; Silver and Phung, 2005). The pbr operon also codes for PbrR, a Mer-like regulatory protein that controls the expression of this operon (Borremans et al., 2001; Chen et al., 2005). Other metal(loid)s export systems that confer resistance to bacterial cells have also been described, like the members of the CHR family, which can lead to chromate (CrO42–) resistance. The R. metallidurans strain CH34 harbors two chr loci, with one being
carried by the plasmid pMOL28 (chr1) and the other one by the chromosome (chr2 ). It has been shown that these loci mediate resistance to the chromium oxyanion through a mechanism decreasing the accumulation of chromate (Nies and Silver, 1989b; Nies et al., 1990). The ChrA protein of R. metallidurans is probably a chromate efflux pump linked to the chemiosmotic gradient (Nies et al.,1990; Juhnke et al, 2002), and related to ChrA protein from a Pseudomonas plasmid (Cervantes et al., 1990). Interestingly, the two chromate resistance determinants of R. metallidurans CH34 may be used under different growth conditions in order to optimize the growth capacity of this organism (Juhnke et al, 2002). Resistance to arsenate (AsO42–) is widespread among both Gram-negative and Gram-positive bacteria, and is often linked to arsenite/antimonite resistance. The initial step of detoxification of the arsenic compounds is the reduction of As(V) to As(III) in the form of arsenite. However it is important to note that the ars determinants that mediate resistance to this oxyanion involve the efflux of arsenic compound as well. This transport is an energy-dependent process (Silver et al., 1981; Silver and Keach, 1982) driven by ATP hydrolysis in E. coli, and by the membrane potential in Staphylococcus species (Rosen and Borbolla, 1984; Bröer et al., 1993). The S. aureus ArsB is responsible for chemiosmotic-dependent arsenite efflux, while the ArsA protein encoded by the E. coli plasmid R773 is an arsenite (antimonite)-dependent ATPase, working in association with ArsB (Rosen et al., 1988). Because the membrane potential is positive on the periplasmic face of the cytoplasmic membrane, the anion export does not require energy, although an additional ATPase may be useful to increase the rate of arsenite efflux and the level of arsenic resistance (Nies and Silver, 1995; Rosen, 2002b). It was reported that the E. coli R773 ars operon also confers resistance to tellurite, and to date it is the only determinant that is known to cause a lower uptake (or a higher efflux) of the oxyanion (Turner et al., 1992a). All of the efflux systems described here are summarized in Table 1 together with the determinants that are discussed in the following sections. B. Enzymatic Detoxification When intracellularly accumulated ionic species cannot be directly exported to the outside of the cytoplasm, they can be modified enzymatically to decrease their intrinsic toxicity. Unlike the biodegradation of
Chapter 33
Metals and Metalloids in Photosynthetic Bacteria
organic compounds, inorganic ions can be reduced, oxidized, precipitated or volatilized, but cannot be degraded. In the most cases, the ion is converted into a less toxic form, which can be accumulated inside the cell or expelled. Sometimes conversion of an element to another form may change its affinity for structural components of the cells and enhance its egress. The most common modification process that decreases toxicity of a metal(loid) is its reduction. In general, metals or metalloids can be reduced under aerobic conditions if the ensuing redox couple has a redox potential between –421 and + 808 mV (H2/H+ and O2/H2O redox potentials, respectively). Thus, ions such as CrO42–, Hg2+ and AsO42– can be reduced in the cell, but those like Zn2+, Cd2+, Co2+ Ni2+, whose redox potentials are lower than - 421 mV (H2/H+) cannot (Nies, 1999). Among the phototrophic bacteria, the reduction of selenium, tellurium, rhodium and chromium oxyanions by members of Rhodospirillaceae has been reported (Moore and Kaplan, 1994; Nepple et al., 2000). Chromate (Cr, VI) is the most toxic form of this element, and is considered to be a 1000-fold more mutagenic than Cr (III) (Lloyd, 2003). A wide range of aerobic and anaerobic bacteria such as Ralstonia, Bacillus, Desulfovibrio, Escherichia and Pseudomonas are able to reduce Cr(VI) to Cr (III) (Nies, 1999; Cervantes et al., 2001; Lloyd, 2003). Thus, chromate resistance is probably mediated by both efflux process, as described above, and by enzymatic detoxification. The most studied metal resistance system based on enzymatic detoxification is encoded by the mer operon (Table 1), which can be located on plasmids, chromosomes or transposons of various microorganisms (Diels et al., 1985; Brown et al., 1986; Griffin et al., 1987; Inoue et al., 1991; Reniero et al., 1995). Various mer determinants can confer resistance to either inorganic mercury salts only (narrow-spectrum determinants) or to organomercurials compound as well (broad-spectrum determinants). Hg2+ is transported into the cell via specific transport systems. Due to the strong affinity of Hg2+ toward the thiol groups, and the absence of a physiological function for the metal, mercury is highly toxic to cells. Its transport into the cytoplasm allows mercury to be enzymatically modified by the cytosolic mercuric reductase. In Gram-negative bacteria, Hg2+ is first bound by a periplasmic protein MerP (Qian et al., 1998) to prevent periplasmic damages, and then delivered to the mercury transporters MerT, MerC or MerF (Hamlett
663
et al., 1992; Sahlmann et al., 1997; Wilson et al., 2000). Several lines of evidence support the concept that mercury translocation is driven by the membrane potential (Wilson et al., 2000). Once inside the cell, Hg2+ is reduced by the mercuric reductase MerA, a glutathione reductase related enzyme, converting the cation to metallic mercury (Schiering et al., 1991). Hg0 is volatile, and leaves the cell by passive diffusion (Silver and Phung, 1996). The mer operon also self-regulates the resistance to Hg2+. MerR functions as a positive regulator in the presence of Hg2+, and as a negative regulator in its absence, binding to the mer operon between the RNA polymerase binding sites of the promoter region (Silver and Phung, 1996; Nascimento and Chartone-Souza, 2003). Broad-spectrum Hg resistance determinants of some microorganisms also contain the merB gene, which codes for an organomercurial lyase (Silver and Phung, 1996). Organomercurials are then hydrolyzed into a thiolate adduct to become a substrate for MerA. Arsenate (As, V) enters the cell via the phosphate transporters, belonging to the Pit or Pst family, expressed differently depending on the phosphate concentrations (Nies and Silver, 1995). The arsenate anion cannot be excreted directly by the cell unless it is converted to arsenite (III) by the activity of ArsC, the arsenate reductase encoded by ars operon (Ji and Silver, 1992), despite the intrinsic high toxicity of the As(III) anion (Table 1). Conversion of a less toxic compound to a more toxic form seems counter-productive, but it is effective if ArsC activity is closely coupled to an efficient efflux system. Arsenate reductases from the plasmids pI258 and R773 have been purified and studied (Ji and Silver, 1992; Gladysheva et al., 1994; Ji et al., 1994). Both enzymes confer resistance to arsenate, although they show only 20% sequence homology and they are reduced by different electron donors such as thioredoxin or glutaredoxin (Ji and Silver, 1992; Shi et al., 1999). Oxyanions of the metalloids selenium and tellurium can also be detoxified by microbial cells via reduction. In both cases the +IV and +VI chemical forms (selenite/selenate and tellurite/tellurate) are reduced to their elemental state (Se(0) and Te(0)) in the cytoplasm or the periplasm in a wide range of microorganisms. This modification converts the very reactive and water-soluble compounds into relatively non-toxic and highly insoluble species. This conversion may be useful in the decontamination of polluted waters as the elemental Se(0) and Te(0) can be removed by filtration (Mattison, 1992).
664
Francesca Borsetti, Pier Luigi Martelli, Rita Casadio and Davide Zannoni
Various mechanisms have been proposed to explain the formation of deposits of Se(0) and Te(0) in bacterial cells, and we will describe them in sections 2.1 and 2.2 below. These reductive processes are not related to various dissimilatory processes allowing bacterial growth under anaerobiosis. Indeed, nitrate, sulfate and carbonate are good examples of non-metallic anions that can be used as alternative terminal electron acceptors. Furthermore, arsenic and selenium oxyanions can also used as alternative electron acceptors. For example, the Shewanella sp. ANA-3 contains two distinct systems for reducing As(V). The first one belongs to the ars detoxification family (Saltikov et al., 2003) while the second one is required for respiratory arsenate reduction (Saltikov and Newman, 2003). This system is encoded by the arr operon, comprised of the arrA and arrB genes, which is contiguous, but divergently transcribed in respect to the ars operon. Microorganisms able to use selenium oxyanions as terminal electron acceptors are widespread among the prokaryotes, and this feature is often used to distinguish between closely related species (Stolz and Oremland, 1999). Frequently, the capacity to use selenate as an alternative acceptor is associated with the use of arsenate (Stolz et al., 2002). Only a few species have been isolated for their ability to use selenite as an electron acceptor, namely the haloalkaliphilic Bacillus selenitireducens and three strains of an Aquificales species (Switzer Blum et al., 1998; Takai et al., 2002). Reduction of selenate and selenite to elemental selenium, which is insoluble and less toxic, may influence the mobility and hence the bioavailability of the element in the environment (Oremland et al., 1990, 1991; Steinberg and Oremland, 1990). Although the remobilization of selenium by oxidation is possible, this biotic process is very slow compared to dissimilatory reduction (Dowdle and Oremland, 1998). Overall, it appears that microbial metabolism is mainly responsible for the biogeotransformation of selenium oxyanions in the environment. A wide range of bacteria and archaea are able to conserve energy through the reduction of Fe(III) to Fe(II) or that of Mn(VI) to Mn(II). Environmental occurrence of these processes has been well documented (Lovley, 1991), and the reduction of Fe(III) was proposed as an early form of respiration on Earth (Vargas et al., 1998). These bacteria are likely to affect the biogeochemical cycling of carbon, hydrogen as well as several trace metals and metalloids (Nealson and Saffarini, 1994), and in the case of Fe(II)-reduc-
tion, they are also useful for bioremediation processes (Lloyd et al., 2003). Metal(loid)s can also be transformed by microbial oxidation. This modification is rarely associated with a resistance mechanism, but is mainly linked to a dissimilatory process (i.e., oxidation of Fe2+ to Fe3+). Microorganisms that oxide Fe(II) are widespread in nature and belong to different genera and species (Straub et al., 2001; Ehrlich, 2002). Some of them are anaerobic Fe(II)-oxidizing phototrophs and members of the purple non-sulfur group of bacteria (Widdel et al., 1993; Heising and Schink, 1998). On the other hand, an oxidative process for metalloid detoxification is the oxidation of arsenite (As, III) to arsenate (As, V). The arsenite oxidation is considered to be a resistance mechanism, converting the highly toxic As(III) to relatively less toxic arsenate (Silver and Phung, 2005b). Studies on several strains has led to the discovery of two genes required for As(III) oxidation and resistance. They are called asoA and asoB (Table 1), and they seem to be members of the DMSO reductase family. It is worth noting that although both arsenite oxidase and arsenate reductase contain homologous molybdopterin centers and [Fe/S] clusters, their positions along the respiratory chain are quite different; indeed, the oxidase is an initial electron donor whereas the reductase is the terminal electron acceptor under anaerobic processes. Additional information on these enzymes can be found in Mukhopadhyay et al. (2002) and Silver and Phung (2005b). Microorganisms can also modify the toxicity of metal(loid)s by alkylation. In particular, they can methylate a large number of elements, producing mono-, di-, tri- and tetra-methylated derivates, with some of these products being more toxic than the original compounds (Dopp et al., 2004). For example, divalent mercury can be methylated to mono- and dimethylmercury, which are highly toxic but also volatile, allowing for a rapid decrease of Hg(II) (Nies, 1999; Dopp et al., 2004). On the other hand, arsenic methylation can hardly be considered a detoxification mechanism, because the trivalent methylated arsenic derivates (mono-methyl-arsinous acid and di-methyl-arsinous acid in particular) have strong biological activities (Kitchin, 2001). Selenium, tellurium and their oxyanions can also be transformed into alkylated and volatile forms, as discussed in section IV. Some evidence suggests that biomethylation might have a role in conferring bacterial resistance to tellurium. The tmp
Chapter 33
Metals and Metalloids in Photosynthetic Bacteria
gene, which encodes a SAM-dependent thiopurine methyl-transferase, isolated from the tellurite resistant Pseudomonas (P.) syringae pathovar pisi, confers tellurite resistance (TeR) to this species (Table 1; Cournoyer et al., 1998). Thus, TeR might occur through a volatilization of tellurite/selenite as dimethyl-telluride/selenide, and Tmp could also be involved in the detoxification of thiopurines and other analogs. Another putative determinant able to confer tellurite resistance to E. coli is the tehAB gene pair (Table 1). The tehA gene product shows homology to C4-dicarboxylate transporter/malic acid transport proteins, however, its direct role on an efflux mechanism has been ruled out (Turner et al., 1995, 1997). TehB is similar to SAM-dependent non–nucleic acid methyl tranferases that are able to undergo a conformational change upon SAM and tellurite binding (Liu et al., 2000). Early studies have suggested that for a full resistance, E. coli cells must have functional cysteine and ubiquinone biosynthetic pathways, as well as an active nicotinamide metabolism and a complete thioredoxin/glutathione/ glutaredoxin system (Turner et al., 1995). Cysteine residues are also functionally important for both TehA and TehB (Dyllick-Brenzinger et al., 2000). Taken together these data suggest that TehB may be a methylase which is not involved in the production of (CH3)nTe. As TehAB requires glutathione to mediate resistance, it is possible that glutathione might take part in a reaction catalyzed by TehB leading to GSTeCH3. As in eukaryotes (Deeley, and Cole, 2006), this glutathione derivative might then be extruded by TehA via a proton antiport mechanism (Zannoni et al., 2007). C. Intracellular and Extracellular Sequestration Some researchers classify intra- and extra-cellular sequestration as additional strategies that allow bacteria to overcome the toxicity of metals. Intracellular sequestration is the accumulation of metals within the cytoplasm, bound to specific accessory proteins, to avoid excessive exposure to cellular components. This strategy is less common and is mediated by metallothioneins. Bacterial and eukaryotic metallothioneins have separate phylogenetic origins but common features, as they are small and thiolate-rich metal binding proteins (Higham et al., 1984; Trevors et al., 1986; Silver and Phung, 1996). The Synechococcus smtA gene codes for a 58 amino acid-long polypeptide containing nine cysteine residues clustered in
665
two groups along the molecule. These residues can bind divalent cations such as Zn2+, Cd2+ and Cu2+ independently, and cells that lack metallothionein are hypersensitive to these ions (Turner et al., 1993). In this cyanobacterium, the resistance mechanism is induced by high concentrations of cations, and is regulated at the transcriptional level in three ways: (i) SmtA synthesis is negatively regulated by the repressor protein SmtB (Morby et al., 1993); (ii) the smtB gene is specifically and irreversibly inactivated by high concentrations of cations (Gupta et al., 1993); and iii) the copy number of the smtA gene is amplified when this cyano-species is grown in the presence of high concentrations of Cd2+, increasing metallothionein production (Gupta et al., 1992). Alignment of the Synechococcus metallothione in SmtA with those in the pseudomonad homologs P. putida KT2440 and P. aeruginosa has shown that the in the latter case SmtA is longer, and that all cysteine and histidine residues, except His69, involved in metal binding are conserved (Canovas et al., 2003). In bacteria, metal resistance based on extracellular sequestration has only been hypothesized, while it is known in several species of fungi and yeast. Saccharomyces cerevisiae may restrict the absorption of Ni2+ by excreting large amounts of glutathione, which has the capacity to bind heavy metals. The resulting metal-glutathione complex is then excluded from the cell (Murata et al., 1985). A similar mechanism exists in Cu2+ resistant fungi, but in this case the microorganisms secrete large amounts of oxalate to form metal-oxalate complexes (Murphy and Levy, 1983). External precipitation of metals in bacteria can also occur as phosphate or sulfide complexes (McEntee et al., 1986, Scott and Palmer, 1990). D. Accessory Mechanisms (Metal Exclusion by Permeability Barrier, Mutation of the Target Site) The mechanisms described in this section are often associated with high levels of resistance, and the related genes are found in microorganisms isolated from metal(loid) polluted environments. However, bacteria that have not been exposed to toxic elements can also acquire some natural resistance called ‘tolerance.’ Natural resistance may be in the form of a mutation(s) that prevent(s) either the interaction of the cell with metals, or change the composition of cell membrane, and they are all examples of metal exclusion by permeability barrier changes. These al-
666
Francesca Borsetti, Pier Luigi Martelli, Rita Casadio and Davide Zannoni
terations may also include changes in the synthesis of outer membrane porins (Rouch et al., 1995), or in the production of extracellular polysaccharide coatings enhancing bio-adsorbation of metal ions (Beveridge and Murray, 1976; Hoyle and Beveridge, 1983; Scott and Palmer, 1990). Another example of natural protection to toxic elements can be achieved by altering the sensitivity of essential components to a defined metal(loid) (Rouch et al., 1995). This might result from activation of enzymatic DNA-repair systems, or induction of alternative biochemical pathway(s) bypassing the production of metal(loid) sensitive components (Mergeay, 1991; Rouch et al., 1995). Increased synthesis of antioxidant molecules such as glutathione, carotenoids, vitamin E or stress-response enzymes can also account for enhanced metal(loid) tolerance (Stohs and Bagchi, 1995). V. On the Bacterial Interactions with Metals and Metalloids A. Group I Metals (Fe, Mn, Mo) Iron is probably the most important heavy metal element in biological systems, because it is present in heme-structures, non-heme iron-sulfur (Fe/S) proteins involved in numerous cellular processes, and other reactive oxygen species (ROS)-scavenger enzymes. In oxic environments the ferric state (Fe3+) is more stable, while in anoxic conditions the reduction to the ferrous state (Fe2+) occurs either chemically or biologically thanks to the work of dissimilatoryreducing bacteria (Lovley, 1991). The most studied dissimilatory Fe(III) reducing bacteria are Shewanella and Geobacter spp. (Croal et al., 2004). Since at neutral pH, Fe(III) is mostly insoluble, Fe(III)-reducing bacteria have evolved several strategies to gain energy from this scarcely available substrate. Up to now, no phototrophic microorganism able to use Fe(III) as electron acceptor, has been isolated. Microbial Fe(II) oxidation is a more common phenomenon, and most of the available literature on this topic comes from studies on the obligately autotrophic and acidophilic bacterium Acidithiobacillus ferrooxidans (Croal et al., 2004). However, in recent years a new type of phototrophic metabolism, which is the anaerobic oxidation of Fe2+ by purple bacteria, has also been described (Ehrenreich and Widdel, 1994; Heising and Schink, 1998). The anaerobic
Fe(II)-oxiding phototrophs isolated for their ability to use ferrous ion as the sole electron donor for photosynthesis, belong to the genera Chromatium, Rhodobacter and Rhodomicrobium. Unfortunately, studies on this type of metabolism are scarce, since the species able to perform this process have a low growth-yield (Croal et al., 2004). Indeed, although electron transfer to the photosynthetic reaction center of anoxygenic phototrophs is energetically feasible, at neutral pH the mid-point redox potential of the couple Fe3+/Fe2+ (E0´ is + 0.32 V) is close to that of the reaction centers in purple bacteria (E´0 ≅ + 0.45 V) (Dutton and Prince, 1978; Ehrenreich and Widdel, 1994). The products formed by iron oxidation [Fe(OH)3 and FeOOH] have very low solubility, and are thought to precipitate on the outside of the cytoplasmic membrane (Ehrenreich and Widdel, 1994). Since the ability to use ferrous ion as electron donor is not widespread among photosynthetic bacteria, this might depend on the membrane location of specific electron carriers. On the other hand, it is worth noting that ferrous ion oxidation cannot support growth over prolonged periods of time and it has to be considered as a side activity of a few phototrophic species (Heising and Schink, 1998). Manganese exists in various oxidation states, from +II to +VII, but in vivo it can cycle between the +II and +III oxidation states, like iron. The biological roles of this element are likely to derive from its chemical analogy to magnesium. Indeed manganese can work as redox-active cofactor in the photosynthetic water-splitting complex of oxygenic phototrophs and in radical detoxifying enzymes (Brudvig, 1987). Mn2+-dependent superoxide dismutase (SodA) is considered to be the most phylogenetically widespread Mn2+-enzyme (Whittaker, 2000). The anoxygenic phototroph Rba. capsulatus contains a unique superoxide dismutase, which is defined as cambialistic as it displays a significant activity with both iron and manganese ions bound at the same active site (Tabares et al., 2003). A cambialistic superoxide dismutase is also present in the thermophilic photosynthetic bacterium Chloroflexus aurantiacus, but the enzyme is more efficient with manganese as cofactor than with iron (Lancaster et al., 2004). In bacteria, Mn2+ can be taken up by cells via the Mg2+ uptake system, such as for example the fast and unspecific CorA system of R. metallidurans (Nies and Silver, 1989a), or the more specific but slow transporters MgtA and MgtB of Salmonella typhimurium (Snavely et al., 1989), and under Mn2+ starvation, ABC-type permeases are
Chapter 33
Metals and Metalloids in Photosynthetic Bacteria
induced (Claverys, 2001). In purple photosynthetic bacteria, manganese transport is poorly understood. Early reports suggested that two distinct cation transport systems exist for Mn2+ and Mg2+ in aerobically and photosynthetically grown cells of Rba. capsulatus (Jasper and Silver, 1978). Interestingly Mn2+ has important roles in the regulation of some metabolic processes in Rhodospirillaceae, such as the nitrogen fixation. Rsp. rubrum and Rba. capsulatus can regulate their nitrogenase activity by a Mn2+-dependent activating factor. In addition to the normal repression-derepression system (Yoch, 1979; Hallenbeck, 1992). More recently it has been shown that in Rba. sphaeroides, Mn(II) can suppress the expression of the puc operon encoding polypeptides of the light-harvesting system II (LH2). This effect, which also decreases the level of bacteriochlorophyll and carotenoid pigments, is probably linked to the activity of PpsR, the central repressor of the photosynthesis genes expression (Oh and Kaplan, 2004). Mn2+ may indeed act as a co-repressor of the PpsR regulation system (Horne et al., 1998). Molybdenum occurs mostly as molybdate (MoO42–), which enters the cells either by the sulfate uptake system, or by inducible ABC transporters (Grunden and Shanmugam, 1997). In photosynthetic microorganisms this element participates to a wide range of metabolic processes, such as nitrification and denitrification and anaerobic respiration using DMSO or TMAO. The EC50 parameter (instead of MIC value) has been used by Giotta et al. (2006) to assess the tolerance of Rba. sphaeroides R26.1 to various heavy metals. This parameter allows evaluation of the toxicity of metals (like Fe and Mo) that do not completely inhibit bacterial growth even at high concentrations. The EC50 value for Fe2+ and MoO42– was particularly high (21 mM and 10 mM respectively), reflecting the essential roles of these elements in proteobacterial physiology. B. Group II Metals (Zn, Ni, Cu, Co and Cr) Zn2+ is a fundamental component of several enzymes and DNA-binding proteins, although in E. coli the toxicity level of this element is similar to that of Cu2+, Ni2+ and Co2+ (Nies, 1999). In Rba. sphaeroides, Zn2+ is essential for the activity of a periplasmic cuprozinc superoxide dismutase (Cu/Zn-SOD) (Kho et al., 2004). The negative effects of Zn2+ might be due to the fast and unspecific transport of this cation into
667
the cell via diverse Mg2+ transporters. Detoxification processes in bacteria are based on the efflux of the cation by P-type ATPases or by the RND system (Table 1) (Nies, 1999). Nickel occurs in biological systems (including phototrophic bacteria) as a cofactor of important enzymes, such as urease and hydrogenases. In Rsp. rubrum this element is inserted into the active site of a CO-dehydrogenase, responsible for a CO-oxidative process in response to CO exposure (Watt and Ludden, 1999). Nickel levels are closely regulated by both uptake and efflux mechanisms. Ni2+ uptake can be mediated by several mechanisms that can be energy-dependent such as the HoxN of R. metallidurans and the NikABCD transport systems of Helicobacter pylori, but also energy-independent as in Azotobacter and Bradyrhizobium spp. (Watt and Ludden, 1999). In Rsp. rubrum transport of this cation seems to be driven by the proton gradient, is quite specific for the element, and it does not depend on the presence of CO. Ni2+ efflux occurs via the CBA family complexes, such as the Ncc and Cnr systems of R. metallidurans CH34 and A31, associated with Ni2+ resistance (Silver and Phung, 2005). Zadvorny et al. (2006) have recently demonstrated that hydrogenases from the purple sulfur bacteria Thiocapsa roseopersicina and Lamprobacter modestohalophilus are able to reduce Ni2+ and other metallic ions (Pt4+, Pd2+ and Ru3+) to their metallic forms under H2 atmosphere. Nickel reduction is also evident in vivo, where the transformation yields a black precipitate in the bacterial cells and at the cell surfaces, according to the cellular localizations of the hydrogenases. On the other hand, hydrogenases purified from Thiocapsa roseopersicina and Lamprobacter modestohalophilus are able to oxidize Fe, Zn, Cd and Ni with simultaneous H2 evolution. Cobalt is an essential trace element because it occurs in the structure of vitamin B12, which catalyzes carbon-oxygen, carbon-carbon and carbon-nitrogen rearrangements (Chapter 5, Warren and Deery). Co2+ is taken up by cells through unspecific transporters, such as the CorA system of Salmonella typhimurium (Snavely et al., 1989). In Gram-negative bacteria, resistance to cobalt is based on efflux mechanisms mediated by transporters of the RND family (Table 1). The EC50 determined in Rba. sphaeroides R26.1 cells for Ni2+ and Co2+ (0.30 and 0.8 mM, respectively) is significantly lower than Fe2+ and MoO42– values, reflecting the intrinsic toxicity of these metals. Interestingly Ni2+ and Co2+ were found to decrease the
668
Francesca Borsetti, Pier Luigi Martelli, Rita Casadio and Davide Zannoni
cellular content of light-harvesting complexes of Rba. sphaeroides R26.1, probably as a consequence of the inhibition of bacteriochlophyll synthesis (Giotta et al., 2006). Due to the electrochemical potential of the Cu2+/ Cu+ pair, copper can interact easily with molecular oxygen and oxygen radicals. This feature is in line with the two main functions of copper proteins in aerobic microrganisms, which are electron transfer and dioxygen transport and activation. Also in photosynthetic microorganisms, copper is the most important cofactor for the activity and assembly of heme-copper oxidases and for nitrite/nitrous oxide reductases. Some phototrophic microorganisms, like Rba. sphaeroides, also posses a Cu/Zn superoroxide dismutase to scavenge toxic radical ions (Kho et al., 2004). Aerobic microorganisms require a Cuhomeostasis system, because this essential element is highly toxic even at low concentrations due to its redox-reactive proprieties, which can cause intracellular generation of ROS (Kimura and Nishioka, 1997). On the other hand, under anaerobic conditions, copper shifts to the Cu+ state that can easily diffuse through the cytoplasmic membrane and accumulate in the cell (Beswick et al., 1976). Both Gram-negative and Gram-positive bacteria are equipped with multiple systems to ensure appropriate copper homeostasis (Table 1). Mechanisms of resistance in E. coli involve the activity of CopA, which is responsible for cytoplasmic copper homeostasis, the Cus system, which accounts for the extrusion of the cation outside the cell, and the periplasmic multi-copper oxidase CueO, which protects periplasmic enzymes from copperinduced damage. The actual function of CueO is still unknown although several hypotheses have been proposed (Rensing and Grass, 2003). In some strains of E. coli, the plasmid-encoded pco determinant also confers additional copper resistance. This determinant detoxifies copper in the periplasm (Rensing and Grass, 2003), and in the Gram-positive bacterium Enterococcus hirae two P-type ATPases are involved in copper homeostasis (Solioz and Stoyanov, 2003). In the case of Rba. sphaeroides, the EC50 for Cu2+ is particularly low (0.08 mM) (Giotta et al., 2006). In trace amounts, chromium is an essential component for animal nutrition, involved mainly in glucose and fat metabolism. It is also required by some microbes possibly as a cofactor for specific enzymes (Losi et al., 1994). Chromium is present in the environment as either Cr (III) or Cr (VI). The oxyanion chromate (VI) can easily cross cell membranes and
be reduced to the less soluble trivalent chromium (Ohtake et al., 1990). In Rba. sphaeroides R26.1 cells, CrO42– toxicity is similar to that of nickel as indicated by its EC50 value (0.34 mM) (Giotta et al., 2006). However, unlike Ni2+, the chromate oxyanion can undergo reduction, converting hexavalent chromium to its trivalent form (Nepple et al., 2000). NADH dependent chromate reductase activity found in the cytoplasmic fractions of Rba. sphaeroides R26.1 is also present in other phototrophic bacteria such as Rba. capsulatus, Rsp. rubrum, Rhodocyclus tenuis and Rba. blasticus (Nepple et al., 2000). C. Group III Metals and Metalloids (Cd, Hg, Pb, As, Te, Se) 1. Cd, Hg, Pb, As Among the elements of group III, Hg2+, Cd2+ and Pb2+ are the most important ions of environmental concern. Because of their strong toxicity, they have no beneficial functions in the cell. Hg2+, highly reactive towards thiol groups, is the most toxic of the ions tested by Giotta et al. (2006) in Rba. sphaeroides cells, as judged by its very low EC50 value (0.03 mM). Cd2+ toxicity in microorganisms is less defined, and multiple mechanisms might be involved, including thiol-binding ability, protein denaturation and membrane damage (Nies, 1999). No information is available on the tolerance of purple bacterial species to this cation, although Rsp. rubrum is able to accumulate Cd2+, and its uptake is competitively inhibited by Zn2+ (Smiejan et al., 2003). In cyanobacteria the resistance to cadmium seems to be related to the production of metallothioneins. Indeed, amplification of the smt genes increases cadmium resistance, while their deletion decreases it (Gupta et al., 1992; Turner et al., 1993). Due to the low solubility of lead phosphate, concentrations of Pb2+ in environmental samples are usually lower than those of mercury, limiting the bioavailability of this metal. Microrganisms resistant to Pb2+ harbor the pbr locus conferring resistance based on ion efflux, as described above in section IVA. The strong toxicity of cadmium and lead has led researchers to study biotechnological approaches to remove them from contaminated aqueous environments. Several kind of biomass have been investigated for a biosorption strategy including the hydrogen bacteria Alcaligenes eutrophus H16, the photosynthetic bacteria Rba. sphaeroides and Rhodovulum PS88,
Chapter 33
Metals and Metalloids in Photosynthetic Bacteria
and the marine microalga Heterosigma akashiwo (Seki et al., 1998; Seki and Suzuki, 2002; Watanabe et al., 2003). 2. Se and Te This section is focused on the bioreduction of Se and Te oxyanions by bacteria, including phototrophs. This biological process has a history of application to clinical microbiology as well as to electron microscopy, and may also be important in the biogeochemical cycling of minerals. A frustrating limitation of the data presented here is that these data cannot yet resolve whether there is a true correlation between the metalloid reduction and the mechanism(s) of resistance.
669
reduced glutathione (GSH) (Turner et al., 2001), 2) enzymatic reduction of selenium oxyanions by periplasmic and cytosolic oxidoreductases, 3) inorganic reactions with bacterial metabolites, and 4) reduction-oxidation reactions of Se oxyanions involving the siderophore pyridine-2,6-bisthiocarboxylic acid. Four putative chemical reactions that yield reduced Se(0) are listed below: 4 RSH + H2SeO3 → RS-Se-SR + RSSR + 3H2O 6 GSH + 3 H2SeO3 → 3 GS-Se-SG + O2– GS-Se-SG + NADPH → GSH + GS-Se– + NADP+ GS-Se– + H+ → GSH + Se(0)
a. Selenium The biological toxicity of several selenium compounds representing the four different oxidation states of this element were originally evaluated in rats by Franke and Painter (1938), and in humans by Vinceti et al. (2001). The soluble oxyanions selenate (SeO42–) and selenite (SeO32–) were poisonous at concentrations of parts per million (ppm). In contrast, elemental selenium Se(0) is highly insoluble and relatively non-toxic, and is the prevalent chemical species under anoxic conditions (Barceloux, 1999). Selenide, Se2–, is both highly reactive and highly toxic, but is readily oxidized to Se(0) through several energetically favorable inorganic and biochemical reactions (Turner et al., 1998). A variety of soil and aquatic bacteria can reduce Se(VI) and Se(IV) oxyanions to insoluble Se(0). Representative genera include Wolinella, Pseudomonas, Sulfurospirillum, Enterobacter, Thaurea, Bacillus, Citrobacter, Rhodobacter, Rhodospirillum (Kessi and Hanselmann, 2004; Zhang and Frankenberger, 2005; Kessi, 2006; Sidique et al., 2006). Deposition of selenium particles may occur in the extracellular milieu for some microorganisms (Klonowska et al., 2005). For others, bioaccumulation of reduced selenium is intracellular, frequently in association with the cell wall or the cytoplasmic membrane (Gerrard et al., 1974). Four different types of biochemical mechanisms have been proposed to account for the formation of nanoparticles of elemental selenium in cultures supplemented with Se(VI) or Se(IV). These are the 1) Painter-type reactions of SeO42– or SeO32– with reduced thiols, in particular with
It is important to note that the first probable step in metabolic processing of SeO42– is an enzymatic or abiotic reduction-oxidation reaction to form SeO32–. The latter undergoes a slow but energetically favorable reaction with glutathione (Shamberger, 1985). In this manner, some cells may process SeO42– by the same pathways as SeO32–. As an interesting aside, the formation of selenotrisulfides by E. coli has since been confirmed using in vivo 77Se NMR to examine bacterial cultures supplemented with SeO32– (Rabenstein and Tan, 1988). Ganther (1968) identified that an analogous Painter-type reaction occurs between SeO32– and the tripeptide glutathione. Recent evidence in facultative phototrophs suggests that when glutathione functions as an electron donor, the reduction of SeO32– also leads to the formation of superoxide anions (O2–) (Kessi and Hanselmann, 2004). In biological systems, O2– may be removed by the combined enzymatic activity of superoxide dismutase and catalase. Generation of ROS such as O2– and H2O2 during this process may account for the observed oxidative stress response of bacterial cells exposed to selenium oxyanions (Bebien et al., 2002). As a terminal step in this biochemical pathway, elemental selenium may be produced by an inorganic reaction between the unstable glutathione selenopersulfide (GS-Se–) and a proton (H+), regenerating a single molecule of glutathione in the process. Thioredoxin is a ubiquitous protein with a redoxactive dithiol/disulfide in the active site. Work with E. coli extracts containing thioredoxin reductase suggests that thioredoxin (Trx) may reduce selenodiglutathione (Bjornstedt et al., 1992). Oxidized Trx can
670
Francesca Borsetti, Pier Luigi Martelli, Rita Casadio and Davide Zannoni
in turn be reduced by thioredoxin in a NADPH-dependent manner to regenerate reduced Trx (Bjornstedt et al., 1992; Kumar et al., 1992), as represented by the following two reactions: Trx-(SH)2 + GS-Se-SG → Trx-S2 + GSH + GS-Se– Trx-S2 + NADPH + H+ → Trx-(SH)2 + NADP+ The selenopersulfide product may then undergo a spontaneous dismutation reaction that generates Se(0). It is worth noting that the mechanism of biological reduction of SeO32– differs from the inorganic reduction-oxidation reaction that produces Se0, particularly with respect to the generation of ROS (Kessi and Hanselmann, 2004). Shortly, the likely first step in biochemical mechanism for generating elemental selenium in bacterial cultures involves a reaction between Se oxyanions and reduced thiols, followed by the subsequent action of glutathione reductase or thioredoxin reductase. Other biomolecules that contribute to the biological process of metalloid reduction are redox active enzymes, many of which are components of bacterial electron transport chains. Various enzymatic systems, such as nitrate (NO3–) and nitrite (NO2–) reductases as well as sulfate (SO42–) and sulfite (SO3–) reductases, are suspected to be involved in the overall reduction of SeO42– and SeO32– to Se(0). For example, the reduction of SeO42– to SeO32– may be carried out by the E. coli periplasmic nitrate reductase NapA, or through the action of the cytoplasmic nitrate reductases NarGHIJ or NarZUWV (Avazeri et al., 1997). More recently, the work of Kessi (2006) has demonstrated that there is metabolic interference between selenite and sulfite as well as selenite and nitrite metabolism in exponentially growing Rba. capsulatus cells. However, stationary phase cells of Rba. capsulatus that can no longer reduce nitrite or sulfite still may metabolize selenite, suggesting that nitrite, sulfite and selenite reduction may be catalyzed by independent pathways in this microorganism (Kessi, 2006). Overall, these studies suggest that the catalytic specificity of oxidoreductases for SeO42– and SeO32– may be different or even absent from certain classes of these enzymes. The biochemistry of selenium in microbes has been discussed extensively (Heider and Böck, 1993; Turner et al., 1998; Birringer et al., 2002; Stolz and Oremland, 1999; Stolz et al., 2002; 2006). As men-
tioned above, Se is a trace element incorporated into several proteins in bacteria, archaea and eukaryotes as selenocysteine and selenomethionine. To date, hundreds of microbial selenoproteins have been identified although this number is destined to increase thanks to the development of innovative bionformatic tools that allow the identification of new classes of selenoproteins (Zhang et al., 2005). Analysis of the composition of selenoproteomes revealed that most members are redox proteins, which use selenocysteine to coordinate redox-active metals (Mo, Ni, or W), or are involved in selenocysteine-thiol redox catalysis (Kryukov and Gladyshev, 2004; Zhang et al., 2005). At present, very little information exists concerning the metabolic processes responsible for assimilation of inorganic selenium (such as selenate and selenite) into these selenoproteins. It was noted that in selenateresistant E. coli strains sulfate uptake was inhibited by selenate. This occurs to a lesser extent in wild type strains, suggesting a connection between sulfate and selenate transport (Springer and Huber, 1973). Due to the similar chemical properties of selenium and sulfur, it was proposed that the two elements were assimilated through the same pathway (Shrift, 1969; Stadtman, 1974). In E. coli, selenate uptake is considered to follow the sulfate uptake system for the most part, for incorporation into selenocysteine. However, early studies suggested that there might be an alternative uptake pathway as well, as inhibitors of the sulfur incorporation pathway do not completely stop selenite assimilation (Brown and Shrift, 1982). In Rba. sphaeroides a polyol ABC transporter has been indicated as the possible carrier of selenite into the cytoplasm (Bebien et al., 2001). It is worth noting that uptake of the oxyanions arsenite [As(III)] and antimonite [Sb(III)] in E. coli AW3110 is facilitated by the glycerol facilitator GlpF, an aquaglyceroporin that facilitates movement of neutral substrates but not of ions (Sanders et al., 1997; Meng et al., 2004). Once inside the cell, selenium derived from selenate or selenite may be incorporated into polypeptides as selenocysteine and selenomethionine. In order for this to occur, selenium oxyanions must be reduced to selenide. Selenite is reduced to selenide by the Painter reaction with glutathione, which is the most abundant reduced thiol in the cytoplasm of the cells (Painter, 1941; Fahey et al., 1978). Although several Se-glutathione intermediates may be produced (such as GSSeO2– and GSOSeSR), the principal adduct formed, and shown by 77SeNMR, is selenodigluta-
Chapter 33
Metals and Metalloids in Photosynthetic Bacteria
thione GS-Se-SG (Milne et al., 1994). Selenate may also react with glutathione, albeit slowly (Shamberger, 1985), through selenate reduction to selenite catalyzed by periplasmic or membrane-associated nitrate reductases, which may be the first step for further incorporation of selenium into polypeptides. For detailed information on the biochemistry of selenium we refer to Zannoni et al. (2007) Phototrophic bacteria have been shown to grow in the presence of 0.1 to 10 mM selenate or selenite. When phototrophic bacteria are grown in the presence of selenium oxyanions small amounts of the oxyanions are reduced or methylated. A study in Rba. sphaeroides indicated that volatilization of selenite or selenate occurs only at low levels and is an insignificant fate for the selenium oxyanions taken up by this organism (Van Fleet-Stalder et al., 2000). Selenite was processed more efficiently than selenate in this organism. Although volatilization is not a significant path for detoxification, Rba. sphaeroides cells grown in the light produced more reduced volatile selenium than the cultures kept in the dark (Van Fleet-Stalder et al., 1997). The selenium uptake system in Rba. sphaeroides operates at very low concentrations of selenium oxyanions. Its poor initial affinity for selenite, and even lower affinity for selenate, seems to be compensated for by a very effective subsequent reduction to trap any selenium that enters the cell (Van Fleet-Stalder et al., 2000). The physiology of the organism appears to initially convert any excess Se into a substrate which might be selenomethionine or a form very similar to it. Any further Se excess appears to be completely reduced to the detoxified red elemental form of Se(0), which has a very low bioavailability (Combs et al., 1996). It has been suggested that members of the Rhodospirillaceae family utilize oxidized compounds, including Te and Se oxyanions, to get rid of the excess electrons produced during anaerobic photosynthesis (Moore and Kaplan, 1992, 1994). This proposal has recently gained strong support by the data indicating that tellurite reduction is likely mediated by the thiol: disulfide oxidoreductase DsbB by extracting reducing equivalents from the ubiquinone-pool (Borsetti et al., 2007). Selenite reduction is observed in Rba. sphaeroides f. sp. denitrificans under phototrophic conditions after approximately 100 hours lag time as a result of the induction of a molybdenum dependent enzyme (Pierru et al., 2006). This group concluded that there are several pathways of selenite reduction in
671
this organism, and that at least one of these pathways involves this enzyme. b. Tellurium Similar to selenium, tellurium has four inorganic oxidation states: the –II, 0, +IV and +VI valence states. Te(II) is chemically reactive and is naturally incorporated into organic tellurides. In fact, reduction/oxidation to dimethyl telluride is responsible for the hallmark garlic breath of acute tellurium toxicity in animals (Hollins, 1969; Taylor, 1996) and humans (Blackadder and Manderson, 1975; Yarema and Curry, 2005). Biomethylation of Te is further discussed below. The tellurium oxyanions, tellurite and tellurate, have been considered in the literature as strong oxidizers, and this chemical attribute is considered to be the explanation for their toxicity in vivo (Taylor, 1999). Gram-negative bacteria are especially sensitive to tellurium oxyanions. Hence, historically potassium tellurite (K2TeO3) has been used as a selective agent in microbiological growth media for the isolation of pathogenic bacterial species from food, clinical and environmental samples (Zadik et al., 1993; Donovan and van Netten, 1995). Klett (1900) was the first to note the biochemical transformation of TeO32– into a black, insoluble precipitate that, at the time, was presumed to be metallic tellurium. This metalloid is relatively non-toxic in its elemental state [Te(0)], although there are no published data explicitly addressing this assumption. The capacity to reduce tellurium is not restricted to resistant microorganisms, nor is it unique to pathogens (Harrison et al., 2005). A variety of bacterial aerobic and anaerobic phototrophs (Moore and Kaplan, 1992), hydrothermal vent heterotrophs (Rathgeber et al., 2002), eukaryotes such as fungi (Kuhn and Jerchel, 1941) and plants (Schreiner and Sullivan, 1911), and the mitochondria in animal tissues (Barrnett and Palade, 1957) may carry out various reactions leading to black Te(0) precipitates. In contrast to selenium, bacterial deposition of tellurium crystallites is almost exclusively intracellular (van Iterson and Leene, 1964a, b; Lloyd-Jones et al., 1994; Klonowska et al., 2005). Transmission electron microscopy (TEM) indicates that metalloid precipitation usually occurs in close physical proximity to the cell wall and membrane lipids. Reduction of tellurium oxyanions may also occur through four documented processes: 1) a Painter-type reaction with glutathione
672
Francesca Borsetti, Pier Luigi Martelli, Rita Casadio and Davide Zannoni
(Turner et al., 2001), 2) catalytic reduction by periplasmic and cytoplasmic oxidoreductases (Avazeri et al., 1997), 3) a reduction-oxidation reaction involving the iron siderophore pyridine-2,6-bisthiocarboxylic acid (PDTC) (Zawadzka et al., 2006), and 4) direct or indirect reduction by electrons siphoned from the membrane bound respiratory chain (Trutko et al., 2000). Tellurium and selenium chemistry are similar in many regards, as the first three mechanisms of tellurite reduction are similar to those mentioned above for selenium. However, tellurium oxyanions may differ in their site-specific interaction(s) with components of the bacterial respiratory chain. It has recently been shown that the redox state of several electron transport redox components can be affected by tellurite (Borsetti et al., 2007). Addition of tellurite to membrane fragments isolated from Rba. capsulatus cells induces acceleration of the QH2:Cyt c oxidoreductase activity. This effect is specifically inhibited by QH2:Cyt c oxidoreductase inhibitor antimycin A, and depends on the presence of the membrane-associated thiol:disulfide oxidoreductase DsbB. These results not only blur the proposal by Trutko et al. (2000) that membrane-bound oxidases are involved in tellurite reduction, but they also exclude the possibility that the oxyanion has a general oxidizing effect on the membrane redox components. Unlike selenite and selenate, no microorganism has been isolated for its ability to use tellurite as a terminal electron acceptor for growth. Tellurate, which is less toxic than tellurite, has recently been shown to sustain the anaerobic growth of a strain, EC-Te-48, isolated from hyperthermophilic vents (Csotonyi et al. 2006). Furthermore, it has been observed that nitrate reductases A and Z from E. coli present tellurite and selenate reduction activities, leading to the deposition of Te(0) and Se(0). A soluble nitrate reductase is also able to reduce tellurite in anaerobically grown cells, though this activity does not allow the growth of the microorganism under anaerobic conditions without nitrate as a terminal electron acceptor. Interestingly, E. coli is able to utilize selenate and tellurite for anaerobic respiration when the nitrate reductase A is induced in large amounts (Avazeri et al., 1997). This suggests that nitrate reductase activity on Se and Te oxyanions is an unfavorable reaction for both the enzyme and the cell. Periplasmic and membrane-bound nitrate reductases from R. eutropha, Paracoccus denitrificans, Paracoccus pantotrophus and the phototrophic bacterium Rba. sphaeroides have shown the ability
to reduce tellurite and selenite in vitro as well, suggesting that tellurite and selenate reducing activities are a general feature of different denitrifying species. However, the catalytic activity of purified periplasmic nitrate reductase from Rba. sphaeroides is too low to justify the high level of resistance of this bacterium to tellurite, suggesting that other mechanisms may contribute to the resistance phenotype (Sabaty et al., 2001). Indeed, tellurite resistance without metal accumulation has been observed in some obligately aerobic photosynthetic bacteria, showing that tellurite resistance does not strictly depend on reduction to Te(0) (Yurkov et al., 1996). Trutko et al. (2000) have proposed that components of the respiratory chain of Gram-negative bacteria are involved in the reductive process. They postulate that the location of tellurium deposits is dependent on the cytoplasmic membrane position of the active site of terminal membrane-bound oxidases. However, they have also shown that the rate of tellurite reduction does not always correlate with the strength of respiration. Indeed, in cells of P. aeruginosa PAO ML 4262, stimulation of the cytochrome c oxidase (COX) activity by the addition of ascorbate-dichlorophenolindophenol drastically lowers Te(0) deposition in cells (Trutko et al., 2000). This finding compromises the hypothesis that COX plays a direct role in reducing tellurite but is in line with other reports indicating that COX activities in membranes from P. pseudoalcaligenes KF707 and Rba. capsulatus cells grown in the presence of tellurite drop concurrently with a drastic decrease of the soluble c-type heme content (Di Tomaso et al., 2002; Borsetti et al., 2003a). In addition, despite the polarity of the respiratory COX in the membrane, Rba. capsulatus and P. pseudoalcaligenes KF707 accumulate elementary tellurium in the cytosol only (Di Tomaso et al., 2002; Borghese et al., 2004), suggesting that reduction of tellurite to Te(0) is unlikely to be performed by respiratory cytochrome c oxidases. On the other hand, whether the modifications observed in the cytoplasmic membrane redox chains of P. pseudoalcaligenes KF707 and Rba. capsulatus are specifically required to survive in the presence of tellurite, or they simply reflect toxic effects of the anion on the electron transport system, remains a matter for debate. This problem has recently been challenged in isolated plasma membrane vesicles of Rba. capsulatus (Borsetti et al., 2007) showing that tellurite (0.25–2.5 mM) alters the redox equilibrium of the Q/QH2-bc1-c2/cy segment of the redox chain. This effect is blocked by antimycin A, which
Chapter 33
Metals and Metalloids in Photosynthetic Bacteria
is a specific cytochrome bc1 complex inhibitor, and is absent in membranes of Rba. capsulatus MD22, a mutant lacking the thiol:disulfide oxidoreductase DsbB. The latter finding is particularly important because it suggests for the first time a possible molecular mechanism by which tellurite can perturb the cytoplasmic membrane redox components facing the periplasmic space. Little is known about the entry of tellurium oxyanions into bacterial cells. The observation of a ‘white’ tellurite resistance variant (Burian et al., 1998) is an important factor to consider when analyzing the location of reduction of Te(IV) to Te(0) in the cell, and the existence of specific transporters. It also addresses the question of how uptake is related to toxicity and resistance. A first report suggested that tellurite may be transported into E. coli cells by the phosphate transporter (Tomas and Kay, 1986). This conclusion was derived from two observations. First, TeO32– is a strong competitive inhibitor of phosphate transport in a wild type strain, and second, some mutants defective in phosphate transport are also resistant to high levels of tellurite. Indeed, sensitivity to the anion is restored by a plasmid carrying the phoB region, which is involved in phosphate transport. Likewise, tellurite uptake in Rba. capsulatus cells is a ∆pH-dependent process strongly repressed by the K+/H+ exchanger nigericin, and by the sulfydryl reagent NEM (Borsetti et al., 2003b). These observations support the idea that Rba. capsulatus imports tellurite by a phosphate transporter belonging to the PiT family, which catalyzes the transport of H2PO4–/HPO42– across the inner membrane in an electro-neutral manner, working as a H+/solute symport system (van Veen, 1997; Harris et al., 2001). However, these data do not exclude the existence of additional mechanisms for the uptake of the oxyanion. Indeed, recent results in aerobically grown cells of Rba. capsulatus suggest that tellurite may enter the cells by exploiting other carriers, such as an as yet uncharacterized monocarboxylate transporter in a ∆pH-dependent manner, as previously shown by Borsetti et al. (2003b) and Borghese et al., (2007). Tellurite transport experiments are rather challenging because non-toxic radio-isotopes of tellurite do not exist. Even so, a few studies have utilized this approach to demonstrate that tellurite is actively transported into cells (Lloyd-Jones et al., 1991, 1994). With the advent of a spectroscopic method to examine free tellurite concentration by using the chelator diethyldithiocarbamate, tellurite transport can be more easily explored (Turner et al.,
673
1992b). Using this assay, the Te resistance determinants, ter, kilA-telAB and tehAB were shown not to mediate any change in the uptake rate (Turner et al., 1995). On the other hand, it was observed that absence of the arsenite/arsenate/antimonite resistance determinant arsABC, an ATP dependent efflux pump, does reduce tellurite accumulation, suggesting that ars is a general chalcogen efflux transporter (Turner et al., 1992a). Finally, it cannot be ruled out that along with other oxyanions, GlpF (discussed above regarding Se oxyanion transport) may also facilitate uptake of tellurocompounds. 3. Mechanism(s) of Metalloid Toxicity While different elements have different toxicity levels towards different bacterial groups, in general, the toxicity levels of the different metalloid oxyanions, from most toxic to least toxic, are: TeO32– > TeO42– >> AsO2– > AsO42–, SbO2– > SeO32– >> SeO42–. Tellurite is toxic at MICs on the order of 0.006 – 0.8 mM, whereas selenite ranges from 8–28 mM for organisms such as E. coli, S. aureus, and P. aeruginosa (Harrison et al., 2004). Importantly, the toxicity of Se depends on the chemical form of the element, as this determines its bioavailability and ability to enter the cells. In addition, the toxic effects of the metalloid can be altered by its interactions with other substances, such as sulfate, methionine, cysteine, heavy metals (As, Cd, Cu, Pb, Hg, Ag, Zn) as well as vitamins C and E (Fan, 1990). Hence, a selenium effect is a result of a balance between antioxidant and pro-oxidant abilities in the cells. Superoxide production may be the major mechanism of selenium toxicity under aerobic conditions in prokaryotic cells, as suggested by several studies (Turner et al., 1998; Bebien et al., 2002; Kessi and Hanselmann, 2004). Selenite is the only known compound that induces both iron and manganese superoxide dismutases (SodB and SodA, respectively) in E. coli. The effect of the oxyanion on the proteomic response of the microorganism strengthens the hypothesis that selenium toxicity involves several molecular circuits and is not directed to a specific and single target. In general, it has been assumed that the toxicity of tellurite is a consequence of the strong oxidizing properties. For example, one recognizes that the redox chemistry of the reaction of tellurite with nitrate reductase is unbalanced, as nitrate reductase
674
Francesca Borsetti, Pier Luigi Martelli, Rita Casadio and Davide Zannoni
would undergo two electron reductions while the reduction of TeO32– to Te(0) requires four electrons. The two electron reduced TeO32– would likely result in the formation of a radical species or radical oxygen ions being formed. Furthermore, the reaction of tellurite with glutathione would sequester glutathione and change the glutathione/ glutaredoxin/ thioredoxin redox balance, and produce H2O2 and O2•–, as discussed above. This latter observation is consistent with the evidence that the induction of the cambialistic Mn/Fe superoxide dismutase of Rba. capsulatus leads to a significant increase in tellurite resistance. Interestingly, it has also been reported that superoxide dismutase activity is increased by the addition of a sub-lethal amount of K2TeO3 (Borsetti et al., 2005). If tellurite is allowed to enter cellular metabolism, then it can find itself incorporated into enzymes as telluromethionine and tellurocysteine, replacing sulfate, sulfite or phosphate in various reactions. Tellurite toxicity in cells of the obligate aerobe and PCB degrader P. pseudoalcaligenes KF707 has been linked to the production of ROS (Tremaroli et al., 2006). This study also indicates that although O2•– generation is clearly linked to tellurite reduction by reduced thiol (RSH) oxidation, the time courses of the two processes are different. The interaction of tellurite with the electron transport chain via the DsbB link to the quinone pool leads to a short circuit in the electron transfer pathways of Rba. capsulatus (Borsetti et al., 2007), and also avoids over-reduction of the quinone pool with a consequent stimulation of light-dependent electron transport under highly reducing conditions. Thus, under unfavorable growth conditions, i.e., phototrophic growth under anaerobiosis, sub-inhibitory amounts of tellurite might exert a positive effect on the redox state of the electron transport components of the facultative phototroph Rba. capsulatus. Overall, the toxicity of tellurite can be considered as a combination of specific targeted thiol chemistry and the resulting production of ROS that poisons the electrochemistry of cells. The difference in toxicity between the oxyanions of the different chalcogens may lie in the rate or ability to repair thiol oxidation and to process RS Ch(II) species (such as GSSe vs GSTe) as well as the rate of ROS production, all of which lead to subsequent damage to respiratory and biosynthetic pathways.
sure is methylation, with the methylated species formed in microorganisms including dimethyl selenide, CH3SeCH3; dimethyl selenenyl sulfide, CH3SeSCH3; dimethyl diselenide, CH3SeSeCH3; dimethyl telluride, CH3TeCH3; dimethyl ditelluride, CH3TeTeCH3 (Chasteen and Bentley, 2003). Methylation of selenium and tellurium in microorganisms has been reviewed extensively by Chasteen and Bentley (2003), and the reader is directed to their publication for an extensive overview. Metalloid methylation in bacteria is a common phenomenon, and examples of methylated products have been reported from the +IV and +VI, as well as 0, redox states of Te and Se. Similarly, some organisms will convert organometalloid compounds such as selenomethionine or telluromethione to methylated derivatives (Chasteen and Bentley, 2003). The biochemical mechanism of this methylation has been explored in only a few organisms. Nonetheless, available work suggests that the methyl group originates from S-adenosylmethionine (SAM), and that methyl cobalamin also contributes to Se methylation (Thompson-Eagle et al., 1989). Overall, the methylation of Se is thought to occur via a variation of the Challenger mechanism. This includes a series of reduction methylation steps, changing the redox state of the Se from +VI to +IV, and finally a dual reduction of dimethylselenone through a Se(III) to Se(II) to dimethylselenide (Challenger, 1945). This mechanism was later modified to account for dimethyl diselenide via a methyl selenide intermediate (Reamer and Zoller, 1980). The majority of studies have explored mixed microbial populations in soils, waters, sediments and effluents from metal contaminated areas as well as in sewage sludge described in early reports (Chau et al., 1976, Cooke and Bruland, 1987). Methylated metals and metalloids are commonly observed in gases released from anaerobic wastewater treatment facilities, presumably due to the microbial activity (Michalke et al., 2000). The review by Chasteen and Bentley (2003) provides a list of organisms that have been associated with the biomethylation of selenium. According to this list, surprisingly, Rba. sphaeroides, Rhodocyclus tenuis and Rsp. rubrum have the ability to use Se(0) and Te(0) as a substrate (Van Fleet-Stalder and Chasteen, 1998). This provides the possibility of biomining of minerals of these metalloids by these species.
4. Methylation
5. Other Metalloids
A common biological response to Se and Te expo-
By definition, metalloids include elements such as
Chapter 33
Metals and Metalloids in Photosynthetic Bacteria
Bo, Al, Si, Ge, As, Sb, Te and Po. As considerable research was focused on bacterial resistance to arsenite, arsenate and antimonite, the microbiology of the latter elements will not be discussed here, and the reader is referred to the following reviews: Rosen (2002a,b), Rosen et al. (1999), Bhattacharejee et al. (2000), Mukhpadhyay et al. (2002), Silver and Phung (2005a,b). It suffices to mention that in E. coli the arsenate resistance determinants are the arsABC encoding an ATP-dependent efflux pump (ArsAB) and an arsenate reductase, which reduces arsenate to arsenite (ArsC). Turner et al (1992a) have observed that this operon also provides tellurite resistance and effectively pumps out tellurite, keeping its accumulation sufficiently low to prevent the ‘cell blackening’ by cytosolic tellurite reduction. Surprisingly, the thiol-dependent arsenate reductase, ArsC, is also required for full resistance of E. coli. In line with this, arsenate reductase activity was also found on the plasmid pI258 from S. aureus. This enzyme can catalyze the reduction of selenate and is inhibited by tellurite (Ji et al., 1994). No work is available on polonium (Po) biochemistry because of its extremely low natural abundance and its high radioactivity. The chemistry of soluble forms of Po are expected to be similar to that of other chalcogens, however, Po toxicity would be greater due to its radioactivity. Although no in-depth investigations are available, studies by Momoshima et al. (2001, 2002) on aquatic samples suggest that microorganisms can generate volatile Po species, most likely through methylation, and they can also reduce the oxyanion forms of Po into elemental Po. This observation is supported by an earlier work indicating that phytoplankton and bacteria accumulate 210 Po (Wildgust et al., 1998). VI. Metal(loid)s Homeostasis in Phototrophs as Revealed by Genome Analysis A. Model Microorganisms for the Study of Metal(loid)s/Bacterial Interactions In this section, genomic sequences generally linked to metal(loid)s resistance and homeostasis in prokaryotes have been examined to survey possible mechanisms of uptake and resistance using a large set of representative phototrophs. Microorganisms taken as ‘model systems’ belong mostly to different genera and species of non-phototrophs (listed in
675
Table 1). Among the species defined as ‘metallophile,’ which are able to grow in environments where the contamination with heavy metals produces extreme conditions (Nies, 2000), the α-proteobacterium R. metallidurans (formerly Alcaligenes eutrophus or Ralstonia eutropha) is best characterized. The strain CH34, isolated in sediments of a decantation basin of a zinc factory in Belgium, is highly resistant to Zn2+, Cd2+, Co2+, Ni2+, Cu2+, CrO42–, Hg2+ and Pb2+, compared to other model microorganisms like E. coli (Mergeay et al., 1978; Mergeay et al., 1985; Nies, 2003). The resistance phenotype is associated with two large plasmids, pMOL28 and pMOL30, but genome sequence analysis of this strain has recently revealed the presence of numerous metal-specific chromosomal loci as well (Mergeay et al., 2003). R. metallidurans CH34 is therefore taken as a ‘prototype organism’ to survey the bacterial response to heavy metals. Comparison of these loci with other prokaryotic genomes has shown that Ralstonia sp. is well adapted to extremely contaminated environments through the use of several systems known as czc, cnr, chr, mer, pbr, and cop. In the closely related strain R. metallidurans 31A additional Ni2+, Cd2+and Co2+ resistant mechanisms (ncc and nre systems) have also been found (Schmidt and Schlegel, 1994; Grass et al., 2001). Another well-characterized Cd2+ resistance efflux system is the plasmid-encoded Cad system of S. aureus (Smith and Novick, 1972) while among the genera with remarkable adaptability to diverse environments, the genus Pseudomonas contains numerous species that could be used as interesting model systems. Indeed, various metal(loid)s-specific resistant mechanisms have been found in P. aeruginosa (Czr system), P. stutzeri (Mer system) and P. syringae (Tmp system) (Barbieri et al., 1989; Cournoyer et al., 1998; Hassan et al., 1999). Notably, E. coli is considered the reference microorganism for copper homeostasis, because the acidic conditions in the digestive tract where this bacterium lives increases the intrinsic toxicity of this metal. Several E. coli strains harbor multiple chromosomal and plasmidencoded resistance systems (e.g., copA, cus, cueO and pco) that allow survival in copper-rich environments (Rensing and Grass, 2003). Metalloids resistance systems such as ars, (harbored by the plasmid R773) and ter, teh, kilAtelAtelB were also analyzed for the first time in E. coli (Hedges and Baumberg, 1973; Turner et al., 1992a, 1994a,b; Taylor, 1999) whereas other genes able to confer tellurite resistance have been isolated in recent years from Geobacillus stearothermophilus and Rba. sphaeroides (isc and
676
Francesca Borsetti, Pier Luigi Martelli, Rita Casadio and Davide Zannoni
telA/cysK genes, respectively) (O’Gara et al., 1997; Rojas and Vásquez, 2005). B. Database Selection To perform our genome analysis an initial reference set of 308 gene sequences was downloaded from the SwissProt Database (release 50.6; Bairoch et al., 2004) These formed an exhaustive sampling of operon families (mer, ars, aso, aox, czc, cad, cnr, ncc, czr, pbr, cop, cus, cue, pco, chr, smt, ter, the, kil, tel, tmp, cysK and iscS) that have been experimentally classified as important in mechanisms for metal resistance in prokaryotes (mostly non-phototrophic bacteria) (Table 1). According to the Gene Ontology (GO) annotations reported in the collected entries, the selected operons comprise genes coding for specific enzymes and for proteins involved in the transport of metals across the inner and the outer membranes, in metal binding, in the cysteine metabolism and in the regulation of gene expression. Some 9% of the selected genes coded for proteins, whose molecular function and subcellular location are still uncharacterized. This data base also included 20 phototrophic bacteria belonging to 10 different genera and 16 different species. The complete genome sequences of the photosynthetic species, Chlorobium chlorochromatii CaD3, Chlorobium tepidum TLS, Erythrobacter litoralis HTCC2594, Rba. sphaeroides 2.4.1, Rhodoferax ferrireducens T118, Rps. palustris BisA53, Rps. palustris BisB18, Rps. palustris BisB5, Rps. palustris CGA009, Rps. palustris HaA2, Rsp. rubrum ATCC11170 and Roseobacter denitrificans Ohc114, were downloaded from the NCBI genome web site (www.ncbi.nlm.nih.gov/Genomes). For the sake of completeness, the genome sequences of eight other species, Chlorobium ferrooxidans DSM13031, Chlorobium limicola DSM245, Chloroflexus aurantiacus J-10-fl, Chloroflexus aggregans DSM 9485, Dinoroseobacter shibae DFL 12; Roseiflexus sp. RS-1, Roseiflexus castenholzii DSM13941 and Rubrivivax gelatinosus PM1, were also downloaded from the same web site although these genomes are presently available as draft assemblies. 1. The ‘Bait and Prey’ Virtual Strategy To screen our data base for genes involved in metalresistance we have followed a two step procedure. Namely, we first clustered the selected prokaryotic genes of known function into similarity sets. Then,
we adopted these clusters as baits for screening our genome data collection and to find, by sequence similarity, genes involved in metal resistance (our preys). The procedure used is described in detail below. a. Clustering of Prokaryotic Genes in Similarity Sets Initially, we searched for similarities among the experimentally validated 308 genes of known function. This was done using the BLAST search program and adopting a similarity threshold equal to 30% (McGinnis and Madden, 2004). Sequence clusters were built with the transitive closure procedure. The sets are listed in Table 2 as a function of the known activity, namely: Enzyme activity (E), Transport activity (T), Metal binding (B), Cysteine metabolism (C), Regulatory activity (R) and Uncharacterized proteins (U). In one case (P-type ATPase), the resulting cluster contained proteins with very different functions. This cluster was further separated into seven sub-families on the basis of a phylogenetic tree constructed using the UPGMA algorithm (Unweighted Pair-Group Method using Arithmetic averages) and the PHYLIP package (available at: http://evolution.genetics.washington.edu/phylip.html). In Table 2, these functional sub-families are highlighted with a star. The procedure described above gave rise to 62 different data sets with each cluster having different numbers of functional proteins, i.e., Enzymes (8 E), Transport proteins (22 T), Metal binding proteins (8 B), proteins for the Cysteine metabolism (2 C), Regulatory proteins (7 R) and Uncharacterized proteins (15 U) (Table 2). It is noteworthy that the adopted procedure assigned each sequence to only one data set. Interestingly, several sets contain proteins derived from only one gene family, indicating that some of the known mechanisms for metal resistance are highly specific. In other cases, proteins related to different gene families clustered together, suggesting a possible evolutionary relationship among proteins with similar overall function but different metal specificity. This is particularly true for proteins involved in transport and regulatory mechanisms. In some cases, proteins of one gene family were split into two different sets, suggesting that despite the same evolutionary origin, they have probably diverged while specializing their functions. This is, for example, the case of the arsC gene family, encoding arsenate reductases, which is split in two different similarity clusters, possibly reflecting their different
Chapter 33
Metals and Metalloids in Photosynthetic Bacteria
677
Table 2. Clustering of the genes in the reference data set Enzyme activity Cluster No of seq. E1 24
Genes arsC
Localization -
Metals As
Functions Arsenate reductase (thioredoxin)
PDB^ yes
E2
10
arsC
-
As
Arsenate reductase (glutaredoxin)
yes
E3
2
aoxA
-
As
Arsenate reductase
yes
E4
2
aoxB
-
As
Arsenate reductase
yes
E5
25
tpm
Cytoplasm
Te, Se
Thiopurine S-methyltranserase
yes
E6
5
cueO
Periplasm
Cu
Oxidoreductase
yes
E7
10
merB
-
Hg
Alkylmercury lyase
yes
E8 *
12
merA
-
Hg
Mercury (II) reductase
No
Genes cnrA, cusA, czcA, czrA, nccA
Localization Membrane
Metals Cd, Co, Cu, Zn
Functions Antiporter (RND-A)
PDB^ No
Transport activity Cluster No of seq. T1 8
T2
5
czcB, czrB, nccB
Membrane
Cd, Co, Ni, Zn
Metal ion transport (RND-B)
No
T3
3
cusB
Membrane
Cu
Antiporter (RND-B)
No
T4
4
cnrC, czcC, nccC
Outer membrane
Cd, Co, Ni, Zn
Transport (RND-C)
No
T5
3
cusC
Outer membrane
Cu
Transport (RND-C)
No
T6
2
czcN, nccN
Membrane
Cd, Co, Ni, Zn
Transport (RND-N)
No
T7
2
czcD
Membrane
Cd, Co, Zn
Transport (CDF)
No
T8 *
9
copA, copB
Membrane
Cu, Hg
P-type ATPase
Yes
T9 *
10
cadA, pbrA
Cd, Co, Cu, Pb, Zn
P-type ATPase
Yes
T10 *
6
copA
Membrane
Cu
P-type ATPase
Yes
T11*
5
copP
Membrane
Cu
Copper exportin ATPase
Yes
T12
3
copD, pcoD
Membrane
Cu
Transport
No
T13
16
arsB
Membrane
As
Arsenite transporter
No
T14
1
arsB
Membrane
As
Arsenite transporter
No
T15
6
arsA
Membrane
As
Arsenite transporting ATPase
Yes
T16
1
merT
Membrane
Hg
Mercuric transport protein
No
T17
9
merT
Membrane
Hg
Mercuric transport protein
No
T18
1
merT
Membrane
Hg
Mercuric transport protein
No
T19
1
merC
Membrane
Hg
Mercury transport
No
T20
1
chrA
Membrane
Cr
Chromate transport protein
No
T21
1
chrA
Membrane
Cr
Chromate transport protein
No
T22
1
pbrI
Membrane
Pb
Permease
No
Francesca Borsetti, Pier Luigi Martelli, Rita Casadio and Davide Zannoni
678 Table 2. Continued Metal binding proteins Cluster B1*
No of seq. 6
Genes copA, copB, pcoA, pcoB
Localization Outer membrane/ Periplasm
Metals Cu
Functions -
PDB^ No
B2*
10
merP
Periplasm
Cu, Hg
Transport protein, periplasmic element
No
B3
3
cusF
Periplasm
Cu
-
No
B4
1
pcoE
Periplasm
Cu
-
No
B5
1
czcI
Periplasm
Cd, Co, Zn
-
No
B6
2
cnrR, nccX
Periplasm
Cd, Co, Ni -
No
B7
2
copC, pcoC
Periplasm
Cu
-
Yes
B8
1
smtA
-
Cd, Cu, Hg, Zn
Metallothionein
Yes
Genes cysK
Localization -
Metals -
Functions Cysteine synthase
PDB^ Yes
C2 1 Regulation activity Cluster No of seq. R1 14
iscS
-
-
Cysteine desulfurase
Yes
Genes arsR, cadC, copS, czcS, merR, pcoS, smtB
Localization Metals Membrane Cd, Co, Cu, Hg, Zn
Functions Two component sensor
PDB^ Yes
R2
26
cueR, merD, merR, pbrR
Cytoplasm
Cu, Hg
Transcription factor
Yes
R3
9
cadC, copR, cusR, czcR, pcoR
Membrane
Cd, Co, Cu, Zn
Signal transduction
Yes
R4
6
copS, cusS, czcS, pcoS
Membrane
Cd, Co, Cu, Zn
Two component sensor
Yes
R5
3
arsD
-
As
-
No
R6
2
cnrH, nccH
Cytoplasm
Cd, Co, Ni Transcription factor
Yes
Cu
-
Yes
Cysteine metabolism Cluster No of seq. C1 3
R7 1 Uncharacterized proteins Cluster No of seq. U1 2
copY
-
Genes tehB
Localization Metals Cytoplasm Te
Functions -
PDB^ No
U2
6
terD, terE, terZ, terX
-
Te
-
No
U3
2
terC
Membrane
Te
-
No
U4
1
terA
-
Te
-
No
U5
1
terB
-
Te
-
No
U6
2
tehA
Membrane
Te
-
No
U7
1
klaA
-
Te
-
No
U8
1
klaB
-
Te
-
No
U9
2
telA
-
Te
-
No
U10
2
merC
-
Hg
-
No
U11
1
merE
-
Hg
-
No
Chapter 33
Metals and Metalloids in Photosynthetic Bacteria
679
Table 2. Continued U12
2
cadI
-
Cd
-
No
U13
1
chrB
-
Cr
-
No
U14
1
pbrD
-
Pb
-
No
U15 2 cnrY, nccY Membrane Cd, Co, Ni No * Cluster deriving from a subdivision of a giant cluster ^ PDB: Protein Data Bank; this column indicates whether a protein having more than 30% identity with clustered proteins is known at atomic resolution
specificities for the reducing molecule, thioredoxin or glutaredoxin (see section V above.). The column PDB (Protein Data Bank) of Table 2 indicates whether a protein with a defined structure and having more than 30% identity with clustered proteins is available at atomic resolution. Indeed, this structure could be used as a template for 3-D modeling at low resolution of the other proteins in the cluster, e.g. for most of the regulatory proteins. Unfortunately, this is not the case for the majority of transport proteins, which are mainly integral membrane proteins, and for most of the metal binding proteins, which are often small polypeptides that are folded and stabilized only when interacting with the metal ions. b. Genome Screening with Known Baits: Looking for Preys All the proteins in the genome data base were screened by BLAST to find sequences similar to the reference set. Only matches with a sequence similarity higher than 40% with a coverage higher than 50% on both sequences were retained. These constraints were adopted to ensure a high similarity between sequences. In doing so, short false positives and also true negatives likely to match in the case of distantly related sequences were avoided. However the retained sequences were unambiguously similar with a pairwise identity higher than 40%. The number of matches for each genome in the different functional clusters is reported in Table 3. It is worth noting that most of the sequences detected with this procedure are not present in the SwissProt data base; owing to this, the annotations of the corresponding genes are at the moment quite broad. The results show that 171 genes out of 47,443 genes forming the genome set that has been analyzed are likely to be involved in metal resistance. Although in most cases our classification matched the annotation reported in the deposited files, 11 new genes, whose functions were not known previously, have also been identified. Five of these new genes
are involved in tellurium resistance of Rsp. rubrum, two are periplasmic Cu-binding proteins of Rps. palustris BisB5, three are related to arsenite/arsenate oxidation-reduction by the two analyzed Chloroflexus species and the remaining one is a mercuric transport protein in Rubrivivax gelatinosus. In general, it appears that all analyzed genomes have at least one gene coding for an arsenate reductase, and one gene coding for a P-type ATPase linked to metal transport, as likely features characterizing primitive forms of inorganic metabolism (Chapter 27, Zannoni et al.). In other words, phototrophic bacteria not only had to face metal toxicity since the beginning of life, but they also had to ‘learn’ how to use metals such as arsenic as an electron acceptor under anaerobic conditions (Castresana, 2004; Chapter 27, Zannoni et al.). Another major observation emerging from the data shown in Tables 2 and 3 is that various strains of the species Rps. palustris differ greatly in the number of genes for metal resistance. It is however worth noting that in some cases, the genes for metal resistance are encoded by plasmids, which are unfortunately at present only partially annotated. VII. Concluding Remarks The most intriguing puzzle in bacterial resistance to metals and metalloids concerns the observation that the levels of resistance frequently observed are higher than the concentration of inorganics typically experienced by bacteria in the environment. This is particularly true for metalloids such as tellurite and selenite although striking differences between the cellular response against tellurite or selenite exist in several species of phototrophic bacteria. The recent observation that the membrane associated thiol: disulfide oxidoreductase, DsbB, of Rba. capsulatus may allow the transfer of oxidizing equivalents from the metalloid tellurite to membrane embedded quinols opens up new perspectives for both microbial
680
Table 3a. Putative metal resistance genes in complete genomes of phototrophic bacteria: enzymes (E) and transport proteins (T) E1 As
Chlorobium chlorochromatii CaD3 Chlorobium tepidum TLS Chlorobium ferrooxidans DSM13031* Chlorobium limicola DSM245 * Chloroflexus aurantiacus J-10-fl * Chloroflexus aggregans DSM 9485 * Dinoroseobacter shibae DFL 12 * Erythrobacter litoralis HTCC2594 Rhodobacter sphaeroides 2.4.1 Rhodoferax ferrireducens T118 Rhodopseudomonas palustris BisA53 Rhodopseudomonas palustris BisB18 Rhodopseudomonas palustris BisB5 Rhodopseudomonas palustris CGA009 Rhodopseudomonas palustris HaA2 Rhodospirillum rubrum ATCC 11170 Roseiflexus castenholzii DSM13941 * Roseiflexus sp. RS-1 * Roseobacter denitrificans Och 114 Rubrivivax gelatinosus PM1 *
1 1 1 2
*Genome available as draft assembly
E2 As
E3 As
1 1(#) 1(#)
E4 As
1 1
2 1
1 1 1 2 1 1
T2 Cd Co Ni Zn
T3 Cu
T4 Cd Co Ni Zn
T5 Cu
T7 Cd Co Zn
1(#)
1 1 1 1
T1 Cd Co Cu Zn
1
1
1
1
1
1 1 7 3 2
2 2
1 2
2
T8 Cu Hg
2 1 1 1 2 1 3 1 1 4 1 1 2 1 1 1 1 1 1 3
T9 Cd Co Cu Pb Zn
2 2 1 1 2 1
2
T12 Cu
T13 As
T14 As
T15 As
1 1 1
1 1 1
T11 Cu
T17 Hg
T20 Cr
T22 Pb
1 1 4
1 1
1
1
1
1
1 1
1 1
1 1 1
1(#)
1
Francesca Borsetti, Pier Luigi Martelli, Rita Casadio and Davide Zannoni
Cluster Metals
Chapter 33
Cluster
B1
B2
B3
Metals
Cu
Cu Hg
Cu
Chlorobium chlorochromatii CaD3 Chlorobium tepidum TLS Chlorobium ferrooxidans DSM13031* Chlorobium limicola DSM245 * Chloroflexus aurantiacus J-10-fl * Chloroflexus aggregans DSM 9485 * Dinoroseobacter shibae DFL 12 * Erythrobacter litoralis HTCC2594 Rhodobacter sphaeroides 2.4.1 Rhodoferax ferrireducens T118 Rhodopseudomonas palustris BisA53 Rhodopseudomonas palustris BisB18 Rhodopseudomonas palustris BisB5 Rhodopseudomonas palustris CGA009 Rhodopseudomonas palustris HaA2 Rhodospirillum rubrum ATCC 11170 Roseiflexus castenholzii DSM13941 * Roseiflexus sp. RS-1 * Roseobacter denitrificans Och 114 Rubrivivax gelatinosus PM1 *
1 1 1
1(#)
1(#)
1
C1
1 1 2 1 1 1 2 2 2 3 2 2 2 4 1 2 1 1 1
R1
R2
R3
R4
R5
U2
U3
U5
U9
U13
Cd Co Cu Hg Zn 1 1 1 2 1 1 1
Cu Hg
Cd Co Cu Zn
Cd Co Cu Zn
As
Te
Te
Te
Te
Cr
1
1
1 1 3 2 2 3
1
1 1 2 2 1 1
2 2 1
1 4
1 1 6 8 1 1 3 5 6 5 4 4 3 3 13 10 6
1
1 1 1
1(#)
1
2(#)
2 (#)
Metals and Metalloids in Photosynthetic Bacteria
Table 3b. Putative metal resistance genes in complete genomes of phototrophic bacteria: metal binding proteins (B), proteins related to cysteine metabolism (C), regulatory proteins (R) and uncharacterized proteins (U)
1
1
* Genome available as draft assembly; # Sequences previously non annotated
681
682
Francesca Borsetti, Pier Luigi Martelli, Rita Casadio and Davide Zannoni
physiologists and biochemists. Indeed, the possibility that membrane-bound disulfide proteins might act as electron wires between exogenous metalloids and membrane redox complexes raises the question of whether metalloids can be considered solely toxic per se, or they might also be used to support cell growth under unfavorable reducing conditions such as those often experienced by facultative phototrophs. A series of open questions concerns the way metalloids get into the cell as no specific carriers have yet been identified in both phototrophs and non-phototrophs although several non-specific mechanisms are known. Further, both the cytosolic fate and the reduction mechanisms to generate less toxic elemental forms are still uncertain for several toxic inorganics. In conclusion, although knowledge in the research area of metal(loid)s in photosynthetic bacteria is lacking, compared with analogous research in nonphototrophs, important advances have indeed been accomplished, as described in this review. We are therefore confident that our work will contribute to bring the metal(loid) metabolism in photosynthetic bacteria out of the dark age, and furthermore, we hope to have convinced more microbial biochemists to move into this fascinating area of microbiology. Acknowledgments This work was supported by MIUR (PRIN 2005) to DZ. FB was supported by a grant from the University of Bologna (Ph.D. program: 2003-05). RC and PLM were supported by MIUR (FIRB 2003) and NOE BioSapiens EU FP6. References Andersen C, Hughes C and Koronakis V (2001) Protein export and drug efflux through bacterial channel-tunnels. Curr Opin Cell Biol 13: 412–416 Angle JS and Chaney RL (1989) Cadmium resistance screening in nitrilotriacetate-buffered minimal media. Appl Environ Microbiol 55: 2030–2035 Anton A, Große C, Reißman J, Pribyl T and Nies DH (1999) CzcD is a heavy metal ion transporter involved in regulation of heavy metal resistance in Ralstonia sp. strain CH34. J Bacteriol 181: 6876–6881 Anton A, Weltrowski A, Haney CJ, Franke S, Grass G, Rensing C and Nies DH (2004) Characteristics of zinc transport by two bacterial cation diffusion facilitators from Ralstonia metallidurans CH34 and Escherichia coli. J Bacteriol 186: 7499–7507 Avazeri C, Turner RJ, Pommier J, Weiner JH, Giordano G and
Verméglio A (1997) Tellurite and selenate reductase activity of nitrate reductases from Escherichia coli: Correlation with tellurite resistance. Microbiol 143: 1181–1189 Bairoch A, Apweiler R, Wu CH, Barker WC, Boeckmann B, Ferro S, Gasteiger E, Huang H, Lopez R, Magrane M, Martin MJ, Natale DA, O’Donovan C, Redaschi N, Yeh LS (2005) The Universal Protein Resource (UniProt). Nucleic Acids Res 33: D154–D159 Barbieri P, Galassi G and Galli E (1989) Plasmid encoded mercury resistance in a Pseudomonas stutzeri strain that degrades o-xylene. FEMS Microbiol Ecol 62: 375–384 Barceloux DG (1999) Selenium. J Toxicol Clin Toxicol 37: 145–172 Barrnett RJ and Palade GA (1957) Histochemical demonstration of the sites of activity of dehydrogenase systems with the electron microscope. J Biophys Biochem Cytology 3: 577–588 Bebien M, Chauvin JP, Adriano JM, Grosse S and Verméglio A (2001) Effect of selenite on growth and protein synthesis in the phototrophic bacterium Rhodobacter sphaeroides. Appl Env Microbiol 67: 4440–4447 Bebien M, Lagniel G, Garin J, Touati D, Verméglio A and Labarre J (2002) Involvement of superoxide dismutases in the response of Escherichia coli to selenium oxides. J Bacteriol 184: 1556–1564 Beswick PH, Hall GH, Hook AJ, Little K, Mc Brien DC and Lott KA (1976) Copper toxicity: evidence for the conversion of cupric to cuprous copper in vivo under anaerobic conditions. Chem Biol Interact 14: 347–356 Beveridge TJ and Murray GE (1976) Uptake and retention of metals by cell walls of Bacillus subtilis. J Bacteriol 127: 1502–1518 Bhattacharjee H, Zhou T, Li J, Gatti DL, Walmsley AR and Rosen BP (2000). Structure-function relationships in an anion-translocating ATPase. Biochem Soc Tran 28: 520–526 Birringer M, Pilawa S and Flohe L (2002) Trends in selenium biochemistry. Nat Prod Rep 19: 693–718 Bjornstedt M, Kumar S and Holmgren A (1992) Selenodiglutathione is a highly efficient oxidant of reduced thioredoxin and a substrate for mammalian thioredoxin reductase. J Biol Chem 267: 8030–8034 Blackadder ES and Manderson WG (1975) Occupational absorption of tellurium: A report of two cases. Br J Ind Med 32: 59–61 Borghese R, Borsetti F, Foladori P, Ziglio G and Zannoni D (2004) Effects of the metalloid oxyanion tellurite (TeO32–) on growth characteristics of the phototrophic bacterium Rhodobacter capsulatus. Appl Env Microbiol 70: 6595–6602 Borghese R, Marchetti D and Zannoni D (2007) The highly toxic oxyanion tellurite (TeO32–) enters the phototrophic bacterium Rhodobacter capsulatus via an as yet uncharacterized monocarboxylate transport system. Arch Microbil, in press Borremans B, Hobman JL, Provoost A, Brown NL and Van der Lelie D (2001) Cloning and functional analysis of the pbr lead resistance determinant of Ralstonia metallidurans CH34. J Bacteriol 183: 5651–5658 Borsetti F, Borghese R, Francia F, Randi MR, Fedi S and Zannoni D (2003a) Reduction of potassium tellurite to elemental tellurium and its effect on the plasma membrane redox components of the facultative phototroph Rhodobacter capsulatus. Protoplasma 221: 153–161 Borsetti F, Toninello A and Zannoni D (2003b) Tellurite uptake
Chapter 33
Metals and Metalloids in Photosynthetic Bacteria
by cells of the facultative phototroph Rhodobacter capsulatus is a ∆pH-dependent process. FEBS Lett 554: 315–318 Borsetti F, Tremaroli V, Michelacci F, Borghese R, Winterstein C, Daldal F and Zannoni D (2005) Tellurite effects on Rhodobacter capsulatus cell viability and superoxide dismutase activity under oxidative stress conditions. Res Microbiol 156: 807–813 Borsetti F, Francia F, Turner RJ and Zannoni D (2007) The disulfide binding protein DsbB allows the transfer of oxidizing equivalents from the toxic metalloid tellurite (TeO32–) to the plasma membrane electron transport system of Rhodobacter capsulatus. J Bacteriol 189: 851–859 Bröer S, Ji G, Bröer A and Silver S (1993) Arsenic efflux governed by the arsenic resistance determinant of Staphylococcus aureus plasmid pI258. J Bacteriol 175: 3480–3485 Brown NL, Barret SR, Camakaris J, Lee BT and Rouch DA (1995) Molecular genetics and transport analysis of the copper-resistance determinant pco from Escherichia coli plasmid pRJ1004. Mol Microbiol 17: 1153–1166 Brown NL, Misra TK, Winnie JN, Schmidt A, Seiff M and Silver S (1986) The nucleotide sequence of the mercuric resistance operons of plasmid R100 and transposon Tn501: Further evidence for mer genes which enhance the activity of mercuric ion detoxification system. Mol Gen Genet 202: 143–151 Brown TA and Shrift A (1982) Selective assimilation of selenite by E. coli. Can J Microbiol 28: 307–310 Brudvig GW (1987) The tetranuclear manganese complex of photosystem II. J Bioenerg Biomembr 19: 91–104 Bruins MR, Kapil S and Oehme FW (2000) Microbial resistance to metals in the environment. Ecotoxicol Environ Safety 45: 198–207 Burian J, Tu N, Klucar L, Guller L, Lloyd-Jones G, Stuchlik S, Fejdi P, Siekel P and Turna J (1998) In vivo and in vitro cloning and phenotype characterization of tellurite resistance determinant conferred by plasmid pTE53 of a clinical isolate of Escherichia coli. Folia Microbiol 43: 589–599 Canovas D, Cases I and de Lorenzo V (2003) Heavy metal tolerance and metal homeostasis in Pseudomonas putida as revealed by complete genome analysis. Environ Microbiol 5: 1242–1256 Castresana J (2004) Evolution and Phylogenetic analysis of Respiration. In: Zannoni D (ed) Respiration in Archaea and Bacteria. Diversity of Prokaryotic Electron Transport Carriers (Advances in Photosynthesis and Respiration, Vol 15), pp 1–14. Kluwer Academic Publishers, Dordrecht Cervantes C, Ohtake H, Chu L, Misra TK and Silver S (1990) Cloning, nucleotide sequence and expression of the chromate resistance determinant of Pseudomonas aeruginosa plasmid pUM505. J Bacteriol 172: 287–291 Cervantes C, Campos-Garcia J, Devars S, Gutierrez-Corona F, Loza-Tavera H, Torres-Guzman JC and Moreno-Sanchez R (2001) Interactions of chromium with microorganisms and plants. FEMS Microbiol Rev 25: 335–347 Challenger F (1945) Biological methylation. Chemical Rev 36: 325–361 Chang JS, Hong J, Oa O and Bh O (1993) Interaction of mercuric ions with the bacterial growth medium and its effects on enzymatic reduction of mercury. Biotechnol Progr 9: 526–532 Chao Y and Fu D (2004) Kinetic study of the antiport mechanism of an Escherichia coli zinc transporter ZitB. J Biol Chem 279: 12043–12050 Chau YK, Wong PTS, Silverberg BA, Luxon PL and Bengert GA (1976) Methylation of selenium in the aquatic environment.
683
Science 192: 1130–1131 Chasteen TG and Bentley R (2003) Biomethylation of selenium and tellurium: Microorganisms and plants. Chem Rev 103: 1–25 Chen P, Greenberg B, Taghavi S, Romano C, van der Lelie D and He C (2005) An exceptionally selective lead(II)-regulatory protein from Ralstonia metallidurans: Development of a fluorescent lead(II) probe. Angew Chem Int Ed Engl 44: 2715–2719 Claverys JP (2001) A new family of high-affinity ABC manganese and zinc permeases. Res Microbiol 152: 231–243 Combs G F Jr, Garbisu C, Lee BC, Yee A, Carlson DE, Smith NR, Magyarosy AC, Leighton T and Buchanan BB (1996) Bioavailability of selenium accumulated by selenite-reducing bacteria. Biol Trace Elem Res 52: 209–225 Cooke TD and Bruland KW (1987) Aquatic chemistry of selenium: Evidence of biomethylation. Environ Sci Technol 21: 1214–1219 Cournoyer B, Watanabe S and Vivian A (1998) A tellurite resistance genetic determinant from phytopathogenic pseudomonas encodes a thiopurine methyltransferase: Evidence of a widelyconserved family of methyltransferase. Biochem Biophys Acta 1397: 161–168 Croal LR, Gralnick JA, Malasarn D and Newman DK (2004) The genetics of geochemistry. Annu Rev Genet 38: 175–202 Csotonyi J, Stackebrandt E and Yurkov V (2006) Anaerobic respiration on tellurate and other metalloids in bacteria from hydrothermal vent fields in the eastern Pacific Ocean. Appl Environ Microbiol 72: 4950–4956 Dey S and Rosen BP (1995) Mechanisms of drug transport in prokaryotes and eukaryotes. In: Georgopapadakou NH (ed) Drug Transport in Antimicrobial and Anticancer Chemotherapy, pp 103–132. Marcel Dekker, New York Deley RG and Cole SP (2006) Substrate recognition and transport by multidrug resistance protein 1 (ABCC1). FEBS Lett 580: 1103–1111 Diels L, Faelen M, Mergeay M and Nies D (1985) Mercury transposons from plasmids governing multiple resistance to heavy metals in Alcaligenes eutrophus CH34. Arch Int Physiol Biochem 93: 27–28 Diels L, Qinghan D, Van der Lelie D, Baeyens W and Mergeay M (1995) The czc operon of Alcaligenes eutrophus CH34: From resistance mechanism to the removal of heavy metals. J Ind Microbiol 14: 142–153 Di Tomaso G, Fedi S, Carnevali M, Manegatti M, Taddei C and Zannoni D (2002) The membrane-bound respiratory chain of Pseudomonas pseudoalcaligenes KF707 cells grown in the presence or absence of potassium tellurite. Microbiol 148: 1699–708 Donovan TJ and van Netten P (1995) Culture media for the isolation and enumeration of pathogenic Vibrio species in foods and environmental samples. Int J Food Microbiol 26: 77–91 Dopp E, Hartmann LM, Florea AM, Rettenmeier AW and Hirner AV (2004) Environmental distribution, analysis and toxicity of organometal(loid) compounds. Crit Rev Toxicol 34: 301–333 Dowdle PR and Oremland RS (1998) Microbial oxidation of elemental selenium in soil slurries and bacterial cultures. Environ Sci Technol 32: 3749–3755 Dutton PL and Prince RC (1978) Reaction centre driven cytochrome interaction in electron and proton translocation and energy coupling. In: Clayton RK and Sistrom WR (eds) The pho-
684
Francesca Borsetti, Pier Luigi Martelli, Rita Casadio and Davide Zannoni
tosynthetic bacteria, pp 525–570. Plenum Press, New York Dyllick-Brenzinger M, Liu M, Winstone TL, Taylor DE and Turner RJ (2000) The role of cysteine residues in tellurite resistance mediated by the TehAB determinant. Biochem Biophys Res Comm 277: 394–400 Ehrenreich A and Widdel F (1994) Anaerobic oxidation of ferrous iron by purple bacteria, a new type of phototrophic metabolism. Appl Env Microbiol 60: 4517–4526 Ehrlich HL (2002) Geomicrobiology of iron. In: Ehrilch HL (ed) Geomicrobiology, pp 345–429. Dekker, New York Fagan MJ and Saier Jr MH (1994) P-type ATPasee of eukaryotes and bacteria: Sequence comparisons and construction of phylogenetic trees. J Mol Evol 38: 57–99 Fahey RC, Brown WC, Adams WB and Worsham MB (1978) Occurrence of glutathione in bacteria. J Bacteriol 133: 1126–1129 Fan AM (1990) Selenium. Nutritional, toxicological and clinical aspects. West J Med. 153: 160–167 Franke K W and Painter E P (1938) A study of the toxicity and selenium content of seleniferous diets: With statistical consideration. Cereal Chemistry 15: 1–24 Gadd GM (2000) Bioremedial potential of microbial mechanisms of metal mobilization and immobilization. Curr Opin Biotech 11: 271–279 Ganther H E (1968) Selenotrisulfides. Formation by the reaction of thiols and selenous acid. Biochemistry 7: 2898–2905 Gerrard TL, Telford JN and Williams HH (1974) Detection of selenium deposits in Escherichia coli by electron microscopy. J Bacteriol 119: 1057–1060 Giotta L, Agostiano A, Italiano F, Milano F and Trotta M (2006) Heavy metal ion influence on the photosynthetic growth of Rhodobacter sphaeroides. Chemosph 62: 1490–1499 Gladysheva TB, Oden KL and Rosen BP (1994) Properties of the arsenate reductase of plasmid R773. Biochemistry 33: 7288–7293 Grass G, Fan B, Rosen BP, Lemke K, Schlegel HG and Rensing C (2001) NreB from Achromobacter xylosoxidans 31A is a nickel-induced transporter conferring nickel resistance. J Bacteriol 183: 2803–2807 Griffin HG, Foster TJ, Silver S and Misra TK (1987) Cloning and DNA sequence of the mercuric- and organomercurial-resistance determinants of plasmid pDU1358. Proc Natl Acad Sci USA 84: 3112–3116 Grunden AM and Shanmugam KT (1997) Molybdate transport and regulation in bacteria. Arch Microbiol 168: 345–354 Gupta A, Whitton BA, Morby AP, Huckle JW and Robinson NJ (1992) Amplification and rearrangement of a prokaryotic methallothionein locus smt in Synechoccus PCC 6301 selected for tolerance to cadmium. Proc Roy Soc Lond 248: 273–281 Gupta A, Morby AP, Turner JS, Whitton BA and Robinson NJ (1993) Deletion within the metallothionein locus of cadmiumtolerant Synechococcus PCC 6301 involving a highly iterated palindrome (HIP1). Mol Microbiol 7: 189–195 Hallenbeck PC (1992) Mutations affecting nitrogenase switchoff in Rhodobacter capsulatus. Biochim Biophys Acta 1118: 161–168 Hamlett NV, Landale EC, Davis BH and Summers AO (1992) Roles of the Tn21 merT, merP and merC gene products in mercury resistance and mercury binding. J Bacteriol 174: 6377–6385 Haney CJ, Grass G, Franke S and Rensing C (2005) New develop-
ments in the understanding of the cation diffusion facilitator family. J Ind Microbiol Biotechnol 32: 215–226 Harris RM, Webb DC, Howitt SM and Cox GB (2001) Characterization of PitA and PitB from Escherichia coli. J Bacteriol 183: 5008–5014 Harrison JJ, Ceri H, Stremick CA and Turner RJ (2004) Biofilm susceptibility to metal toxicity. Environ Microbiol 6: 1220–1227 Harrison JJ, Ceri H, Badry EA, Roper NJ, Tomlin KL and Turner RJ (2005) Effects of the twin-arginine translocase on the structure and antimicrobial susceptibility of Escherichia coli biofilms. Can J Microbiol 51: 671–683 Hassan M-e-T , van der Lelie D, Springael D, Römling U, Ahmed N and Mergeay M (1999) Identification of a gene cluster, czr, involved in cadmium and zinc resistance in Pseudomonas aeruginosa. Gene 238: 417–425 Hassen A, Saidi N, Cherif M and Boudabous A (1998) Resistance of environmental bacteria to heavy metals. Biores Technol 64: 7–15 Hedges RW and Baumberg S (1973) Resistance to arsenic compounds conferred by a plasmid transmissible between strains of Escherichia coli. J Bacteriol 115: 459–460 Heider J and Böck A (1993) Selenium metabolism in microorganisms. Adv Microbial Physiol 35: 71–109 Heising S and Schink B (1998) Phototrophic oxidation of ferrous iron by a Rhodomicrobium vannielii strain. Microbiol 144: 2263–2269 Higham DP, Sadler PJ and Scawen MD (1984) Cadmium-resistant Pseudomonas putida synthesizes novel cadmium binding proteins. Science 225: 1043–1046 Hollins J G (1969) The metabolism of tellurium in rats. Health Phys 17: 497–505 Horne IM, Pemberton JM and McEwan AG (1998) Manganous ions suppress photosynthesis gene expression in Rhodobacter sphaeroides. Microbiol 144: 2255–2261 Hoyle B and Beveridge TJ (1983) Binding of metallic ions to the outer membrane of Escherichia coli. Appl Environ Microbiol 46: 749–752 Inoue S, Sugawara K and Kusano T (1991) The merR regulatory gene in Thiobacillus ferrooxidans is spaced apart from the mer structural genes. Mol Microbiol 5: 2707–2718 Jasper P and Silver S (1978) Divalent cation transport systems of Rhodopseudomonas capsulata. J Bacteriol 133: 1323–1328 Ji G and Silver S (1992) Regulation and expression of the arsenic resistance operon from Staphylococcus aureus plasmid pI258. J Bacteriol 174: 3684–3694 Ji G, Garberg EA, Armes LG, Chen CM, Fuchs JA and Silver S (1994) Arsenate reductase of Staphylococcus aureus plasmid pI258. Biochemistry 33: 7294–7299 Juhnke S, Peitzsch N, Hübener N, Große C and Nies DH (2002) New genes involved in chromate resistance in Ralstonia metallidurans strain CH34. Arch Microbiol 179: 15–25 Kachur AV, Koch CJ and Biaglow JE (1998) Mechanism of copper catalyzed oxidation of glutathione. Free Radical Res 28: 259–269 Kessi J (2006) Enzymic systems proposed to be involved in the dissimilatory reduction of selenite in the purple non-sulfur bacterium Rhodospirillum rubrum and Rhodobacter capsulatus. Microbiol 152: 731–743 Kessi J and Hanselmann KW (2004) Similarities between the abiotic reduction of selenite with glutathione and the dissimilatory
Chapter 33
Metals and Metalloids in Photosynthetic Bacteria
reaction mediated by Rhodospirillum rubrum and Escherichia coli. J Biol Chem 279: 50662–50669 Kho DH, Yoo SB, Kim JS, Kim EJ and Lee JK (2004) Characterization of Cu- and Zn-containing superoxide dismutase of Rhodobacter sphaeroides. FEMS Microbiol Lett 234: 261–267 Kimura T and Nishioka H (1997) Intracellular generation of superoxide by copper sulphate in Escherichia coli. Mutat Res 389: 237–242 Kitchin KT (2001) Recent advances in carcinogenesis: Modes of action, animal model systems and methylated arsenic metabolites. Toxicol Appl Pharmacol 172: 249–261 Klett A (1900) Zur kenntniss der reducirenden eigenschaften der bakterien. Z Hyg Infektionskr 33: 137–160 Klonowska A, Heulin T and Verméglio A (2005) Selenite and tellurite reduction by Shewanella oneidensis. Appl Environ Microbiol 71: 5607–5609 Kryukov GV and Gladyshev VN (2004) The prokaryotic selenoproteome. EMBO Reports 5: 538–543 Kuhn R and Jerchel D (1941) Uber investseifen. VIII Metteil. Reduktion von tetrazoliumsalzen durk bakterien, garende hefe und keimende samen. Ber Dtsch Chem Ges 74: 949–952 Kumar S, Bjornstedt M and Holmgren A (1992) Selenite is a substrate for calf thymus thioredoxin reductase and thioredoxin elicits a large non-stoichiometric oxidation of NADPH in the presence of oxygen. Eur J Biochem 207: 435–439 Lancaster VL, LoBrutto R, Selvaraj FM and Blankenship RE (2004) A cambialistic superoxide dismutase in the thermophilic photosynthetic bacterium Chloroflexus aurantiacus. J Bacteriol 186: 3408–3414 Liu M, Turner RJ, Winstone TL, Saetre A, Brenzinger MD, Jickling G, Tari LW, Weiner JH and Taylor DE (2000) Escherichia coli TehB requires S-adenosilmethionine as a cofactor to mediate tellurite resistance. J Bacteriol 182: 6509–6513 Lloyd JR (2003) Microbial reduction of metals and radionuclides. FEMS Microbiol Rev 27: 411–425 Lloyd JR, Lovley DR and Macaskie LE (2003) Biotechnological application of metal-reducing microorganisms. Adv Appl Microbiol 53: 85–128 Lloyd-Jones G, Ritchie DA and Strike P (1991) Biochemical and biophysical anaylsis of plasmid pMJ600-encoded tellurite [TeO32–] resistance. FEMS Microbiology Lett 81: 19–24 Lloyd-Jones G, Osborn AM, Ritchie DA, Strike P, Hobman JL, Brown NL and Rouch DA (1994) Accumulation and intracellular fate of tellurite in tellurite-resistant Escherichia coli: A model for the mechanism of resistance. FEMS Microbiol Lett 118: 113–120 Losi ME, Amrhein C and Frankenberg WTJ (1994) Environmental biochemistry of chromium. Rev Environ Contam Toxic 136: 91–121 Lovley DR (1991) Dissimilatory Fe(III) and Mn(IV) reduction. Microbiol Rev 55: 259–287 Mattison PL (1992) Bioremediation of metals: Putting it to work. Cognis, Santa Rosa, CA McEntee JD, Woodrow JR and Quirk AV (1986) Investigation of cadmium resistance in Alcaligenes sp. Appl Environ Microbiol 51: 515–520 McGinnis S, Madden TL (2004) BLAST: At the core of a powerful and diverse set of sequence analysis tools. Nucleic Acids Res 32: W20–W25 Meng YL, Liu Z, Rosen BP (2004) As (III) and Sb (III) uptake by GlpF and efflux by ArsB in Escherichia coli. J Biol Chem
685
279: 18334–18341 Mergeay M (1991) Towards an understanding of the genetics of bacterial metal resistance. Trends Biotechnol 9: 17–24 Mergeay M, Houba C and Gerits J (1978) Extracromosomal inheritance controlling resistance to cadmium, cobalt, copper and zinc ions: Evidence from curing in a Pseudomonas. Arch Int Physiol Biochem 86: 440–442 Mergeay M, Nies D, Schlegel HG, Gerits J, Charles P and Von Gijsegem F (1985) Alcaligens eutrophus CH34 is a facultative chemolithotroph with plasmid-bound resistance to heavy metals. J Bacteriol 162: 328–334 Mergeay M, Monchy S, Vallaeys T, Auquier V, Benotmane A, Bertin P, Taghavi S, Dunn J, van der Lelie D and Wattiez R (2003) Ralstonia metallidurans, a bacterium specifically adapted to toxic metals: Towards a catalogue of metal-responsive genes. FEMS Microbiol Rev 27: 385–410 Michalke K, Wickenheiser EB, Mehring M, Hirner AV and Hensel R (2000) Production of volatile derivatives of metal(loid)s by microflora involved in anaerobic digestion of sewage sludge. App Environ Microbiol 66: 2791–2796 Milne JB, Lorusso L, Lear W and Charlebois R (1994) The bioremediation of Se-containing effluents. Reduction of Se by Escherichia coli with selenocysteine lyase and cysteine desulferase activities. J Biol Chem 272: 22417–22424 Momoshima N, Song L.X, Osaki S and Maeda Y (2001) Formation and emission of volatile polonium compound by microbial activity and polonium methylation with methylcobalamin. Environ Sci Technol 35: 2956–2960 Momoshima N, Song LX, Osaki S and Maeda Y (2002). Biologically induced Po emission from fresh water. J Environ Radioact 63: 187–197 Moore MD and Kaplan S (1992) Identification of intrinsic highlevel resistance to rare-earth oxides and oxyanions in members of the class Proteobacteria: Characterization of tellurite, selenite, and rhodium sesquioxide reduction in Rhodobacter sphaeroides. J Bacteriol 174: 1505–1514 Moore MD and Kaplan S (1994) Members of the family Rhodospirillaceae reduce heavy-metal oxyanions to maintain redox poise during photosynthetic growth. ASM News 60:17–23 Morby AP, Turner JS, Huckle JW and Robinson NJ (1993) SmtB is a metal-dependent repressor of the cyanobacterial metallothionein gene smtA: Identification of a Zn inhibited DNA-protein complex. Nucleic Acids Res 91: 921–925 Mukhopadhyay R, Rosen BP, Phung LT and Silver S (2002) Microbial arsenic: From geocycles to genes and enzymes. FEMS Microbiol Rev 26: 311–325 Murata K, Fukuda Y, Shimosaka M, Watanabe K, Saikusa T and Kimura A (1985) Phenotypic character of the methylglyoxal resistance gene in Saccharomyces cerevisae: expression in Escherichia coli and application to breeding wild-type yeast strains. Appl Environ Microbiol 50: 1200–1207 Murphy RJ and Levy JF (1983) Production of copper oxalate by some copper tolerant fungi. Trans Br Mycol Soc 81: 165–168 Nascimento AMA and Chartone-Souza E (2003) Operon mer: Bacterial resistance to mercury and potential for bioremediation of contaminated environments. Genet Mol Res 2: 92–101 Nealson KH and Saffarini D (1994) Iron and manganese in anaerobic respiration: Environmental significance, physiology and regulation. Annu Rev Microbiol 48: 311–343 Nepple BB, Kessi J and Bachofen R (2000) Chromate reduc-
686
Francesca Borsetti, Pier Luigi Martelli, Rita Casadio and Davide Zannoni
tion by Rhodobacter sphaeroides. J Ind Microbiol Biotech 25: 198–203 Nies DH (1992a) Resistance to cadmium, cobalt, zinc and nickel in microbes. Plasmid 27: 17–28 Nies DH (1992b) CzcR and CzcD, gene products affecting resistance to cobalt, zinc and cadmium (czc system) in Alcaligenes eutrophus. J Bacteriol 174: 8102–8110 Nies DH (1999) Microbial heavy-metal resistance. Appl Microbiol Biotechnol 51: 730–750 Nies DH (2000) Heavy metal-resistant bacteria as extremophiles: Molecular physiology and biotechnological use of Ralstonia sp. CH34. Extremophiles 4: 77–82 Nies DH (2003) Efflux-mediated heavy metal resistance in prokaryotes. FEMS Microbiol Rev 27: 313–339 Nies DH and Silver S (1989a) Metal ion uptake by plasmid-free metal-sensitive Alcaligenes eutrophus strain. J Bacteriol 171: 4073–4075 Nies DH and Silver S (1989b) Plasmid-determined inducible efflux is responsible for resistance to cadmium, zinc and cobalt in Alcaligenes eutrophus. J Bacteriol 171: 896–900 Nies DH and Silver S (1995) Ion efflux systems involved in bacterial metals resistance. J Ind Microbiol 14: 186–199 Nies A, Nies DH and Silver S (1990) Nucleotide sequence and expression of a plasmid-encoded chromate resistance determinant from Alcaligenes eutrophus. J Biol Chem 265: 5648–5653 Novick RP, Murphy RP, Gryczan TJ, Barone E and Edelman I (1979) Penicillase plasmids of Staphylococcus aureus: restriction-deletion maps. Plasmid 2: 109–209 Nucifora G, Chu L, Misra TK and Silver S (1989) Cadmium resistance from Staphylococcus aureus plasmid, pI258 cadA results from a cadmium-efflux ATPase. Proc Natl Acad Sci USA 86: 3544–3548 O’Gara JP, Gomelsky M and Kaplan S (1997) Identification and molecular genetics analysis of multiple loci contributing to high-level tellurite resistance in Rhodobacter sphaeroides 2.4.1. Appl Environ Microbiol 63: 4713–4720 Odermatt A, Suter H, Krapf R and Solioz M (1993) Primary structure of two P-type ATPase involved in copper homeostasis in Enterococcus hirae. J Biol Chem 268: 12775–1777 Oh J-II and Kaplan S (2004) Oxydases as redox sensors in pigment synthesis. In: Zannoni D (ed) Respiration in Archaea and Bacteria. Diversity of Prokaryotic Respiratory Systems (Advances in Photosynthesis and Respiration, Vol 16), pp 287–309, Kluwer Academic Publishers, Dordrecht Ohtake H, Fujii E and Toda K (1990) Bacterial reduction of hexavalent chromium: Kinetic aspects of chromate reduction by Enterobacter cloacae HO1. Biocatalysis 4: 227–235 Oremland RS, Steinberg NA, Maest AS, Miller LG and Hollibaugh JT (1990) Measurements of in situ rates of selenate removal by dissimilatory bacterial reduction in sediments. Environ Sci Tech 24: 1157–1164 Oremland RS, Steinberg NA, Presser TS and Miller LG (1991) In situ bacterial selenate reduction in the agricultural drainage systems of western Nevada. Appl Environ Microbiol 57: 615–617 Osborn AM, Kenneth DB, Strike P and Ritchie DA (1997) Distribution, diversity and evolution of the bacterial mercury resistance (mer) operon. FEMS Microbiol Rev 19: 239–262 Painter EP (1941) The chemistry and toxicity of selenium compounds with special reference to the selenium problem. Chem Rev 28: 179–213
Park MH, Wong BB and Lusk JE (1976) Mutants in three genes affecting transport of magnesium in Escherichia coli. J Bacteriol 126: 1096–1103 Paulsen IT and Saier Jr MR (1997) A novel family of ubiquitous heavy metal ion transport proteins. J Membr Biol 156: 99–103 Pierru B, Grosse S, Pignol D and Sabaty M (2006) Genetic and biochemical evidence for the involvement of a molybdenumdependent enzymes in one of the selenite reduction pathways of Rhodobacter sphaeroides f. sp. denitrificans IL106. Appl Env Microbiol 72: 3147–3153 Poole K, Krebes K, McNally C and Neshat S (1993) Multiple antibiotic resistance in Pseudomonas aeruginosa: Evidence for involvement of an efflux operon. J Bacteriol 175: 7363–7372 Qian H, Sahlman L, Eriksson PO, Hambraeus C, Edlund U and Sethson I (1998) NMR solution structure of the oxidized form of MerP, a mercuric ion binding protein involved in bacterial mercuric ion resistance. Biochemistry 37: 9316–9322 Rabenstein D L and Tan K-S (1988) 77Se NMR studies of bis(alkylthio)selenides of biological thiols. Magn Res Chem 26: 1079–1085 Rathgeber C, Yurkova N, Stackebrandt E, Beatty JT and Yurkov V (2002). Isolation of tellurite- and selenite-resistant bacteria from hydrothermal vents of the Juan de Fuca Ridge in the Pacific Ocean. Appl Environm Microbiol 68: 4613–4622 Reamer DC and Zoller WH (1980) Selenium biomethylation products from soil and sewage sludge. Science 208: 500–502 Reniero D, Galli E and Barbieri P (1995) Cloning and comparison of mercury and organomercurial-resistance determinants from a Pseudomonas stutzeri plasmid. Gene 166: 77–82 Rensing C and Grass G (2003) Escherichia coli mechanisms of copper homeostasis in a changing environment. FEMS Microbiol Rev 27: 385–410 Rojas DM and Vásquez CC (2005) Sensitivity to potassium tellurite of Escherichia coli cells deficient in CSD, CsdB and IscS cysteine desulfurases. Res Microbiol 156: 465–471 Rosen BP (1996) Bacterial resistance to heavy metals and metalloids. J Biol Inorg Chem 1: 273–277 Rosen BP (2002a) Transport and detoxification systems for transition metals, heavy metals and metalloids in eukaryotic and prokaryotic microbes. Comp Biochem Physiol A Mol Integr Physiol 133: 689–693 Rosen BP (2002b) Biochemistry of arsenic detoxification. FEBS Lett 529: 86–92 Rosen BP and Borbolla MG (1984) A plasmid-encoded arsenite pump produces arsenite resistance in Escherichia coli. Biochem Biophys Res Comm 124: 760–765 Rosen BP, Weigel U, Karkaria C and Gangola P (1988) Molecular characterization of an anion pump. The arsA gene product is an arsenite (antimonite)-stimulated ATPase. J Biol Chem 263: 3067–3070 Rosen BP, Bhattacharjee H, Zhou T and Walmsley AR (1999) Mechanism of the ArsA ATPase. Biochim Biophys Acta 1461: 207–215 Rouch DA, Lee BTD and Morby AP (1995) Understanding cellular responses to toxic agents: A model for mechanism choice in bacterial metal resistance. J Ind Microbiol 14: 132–141 Sabaty M, Avazeri C, Pignol D and Verméglio A (2001) Characterization of the reduction of selenate and tellurite by nitrate reductases. Appl Env Microbiol 67: 5122–5126 Sahlmann L, Wong W and Powlowski J (1997) A mercuric ion
Chapter 33
Metals and Metalloids in Photosynthetic Bacteria
uptake role for the integral inner membrane protein MerC, involved in bacterial mercuric ion resistance. J Biol Chem 272: 29518–29526 Saltikov CW and Newman DK (2003) Genetic identification of a respiratory arsenate reductase. Proc Natl Acad Sci USA 100: 10983–10988 Saltikov CW, Cifuentes A, Venkateswaran K and Newman DK (2003) The ars detoxification system is advantageous but not required for As(V) respiration by the genetically tractable Shewanella species strain ANA-3. Appl Environ Microbiol 69: 2800–2809 Sanders OI, Rensing C, Kuroda M, Mitra B and Rosen BP (1997) Antimonite is accumulated by the glycerol facilitator GlpF in Escherichia coli. J Bacteriol 179: 3365–3367 Sandrin TR and Maier RM (2003) Impact of metals on the biodegradation of organic pollutants. Environ Health Perspect 111: 1093–1101 Schiering N, Kabsch W, Moore MJ, Distefano MD, Walsh CT and Pai EF (1991) Structure of the detoxification catalyst mercuric ion reductase from Bacillus sp. strain RC607. Nature 352: 168–172 Schmidt T and Schlegel HG (1994) Combined nickel-cobaltcadmium resistance encoded by the ncc locus of Alcaligenes xylosoxidans 31A. J Bacteriol 176: 7045–7054 Schreiner O and Sullivan M X (1911) Reduction by roots. Bot Gaz 51: 121–130 Scott JA and Palmer SJ (1990) Sites of cadmium uptake in bacteria used for biosorption. Appl Environ Microbiol 33: 221–225 Seki H, Suzuki A and Mitsueda S (1998) Biosorption of heavy metal ions on Rhodobacter sphaeroides and Alcaligenes eutrophus H16. J Coll Inter Sci 197: 185–190 Seki H and Suzuki A (2002) Adsorption of heavy metals to floctype biosorbents. J Coll Inter Sci 249: 295–300 Shamberger RJ (1985) The genotoxicity of selenium. Mutat Res 154: 29–48 Shi J, Vlamis-Gardikas A, Aslund F, Holmgren A and Rosen BP (1999) Reactivity of glutaredoxin 1,2 and 3 from Escherichia coli shows that glutaredoxin 2 is the primary hydrogen donor to ArsC-catalyzed arsenate reduction. J Biol Chem 274: 36039–36042 Shrift A (1969) Aspects of selenium metabolism in higher plants. Annu Rev Plant Physiol 20: 475–494 Sidique T, Zhang Y, Oleke B C and Frankenberger W T (2006) Characterization of sediment bacteria involved in selenium reduction. Bioresource Technology 97: 1041–1049 Silver S and Keach D (1982) Energy-dependent arsenate efflux: The mechanism of plasmid-mediated resistance. Proc Natl Acad Sci USA 79: 6114–6118 Silver S and Walderhaug M (1992) Gene regulation of plasmidand chromosomal-determined inorganic transport in bacteria. Microbiol Rev 56: 195–228 Silver S and Phung LT (1996) Bacterial heavy metal resistance: new surprises. Ann Rev Microbiol 50: 753–789 Silver S and Phung LT (2005a) A bacterial view of the periodic table: Genes and protein for toxic inorganic ions. J Ind Microbiol Biotechnol 32: 587–605 Silver S and Phung LT (2005b) Genes and enzymes involved in bacterial oxidation and reduction of inorganic arsenic. Appl Environ Microbiol 71: 599–608 Silver S, Budd K, Leahy KM, Shaw WV, Hammond D, Novick RP, Willsky GR, Malamy MH and Rosenberg H (1981) Induc-
687
ible plasmid-determined resistance to arsenate, arsenite and antimony (III) in Escherichia coli and Staphylococcus aureus. J Bacteriol 146: 983–996 Silver S, Nucifora G and Phung LT (1993) Human Menkes Xchromosome disease and the staphylococcal cadmium resistance ATPase: A remarkable similarity in protein sequences. Mol Microbiol 10: 7–12 Smejan A, Wilkinson KJ and Rossier C (2003) Cd bioaccumulation by a freshwater bacterium, Rhodospirillum rubrum. Environ Sci Technol 37: 701–706 Smith K and Novick RP (1972) Genetic studies on plamid-linked cadmium resistance in Staphylococcus aureus. J Bacteriol 112: 761–772 Snavely MD, Florer JB, Miller CG and Maguire ME (1989) Magnesium transport in Salmonella typhimurium: 28Mg2+ transport by CorA, MgtA and MgtB systems. J Bacteriol 171: 4761–4766 Solioz M and Stoyanov JV (2003) Copper homeostasis in Enterococcus hirae. FEMS Microbiol Rev 27: 183–195 Solioz M and Vulpe C (1996) CPx-type ATPase: a class of Ptype ATPases that pump heavy metals. Trends Biochem Sci 21: 237–241 Springer SE and Huber RE (1973) Sulfate and selenate uptake and transport in wild and in two selenate-tolerant strains of Escherichia coli K12. Arch Biochem Biophys 156: 595–603 Stadtman TC (1974) Selenium biochemistry. Science 183: 915–922 Steinberg NA and Oremland RS (1990) Dissimilatory selenate reduction potential in a diversity of sediments types. Appl Environ Microbiol 56: 3550–3557 Stohs SJ and Bagchi D (1995) Oxidative mechanisms in the toxicity of metal ions. Free Radic Biol Med 18: 321–336 Stolz JF and Oremland RS (1999) Bacterial respiration of arsenic and selenium. FEMS Microbiol Rev 23: 615–627 Stolz JF, Basu P, Oremland RS (2002) Microbial transformation of elements: the case of arsenic and selenium. Int Microbiol 5: 201–207 Stolz JF, Basu P, Santini JM and Oremland RS (2006) Arsenic and selenium in microbial metabolism. Ann Rev Microbiol 60: 107–130 Straub KL, Benz M and Schink B (2001) Iron metabolism in anoxic environments at near neutral pH. FEMS Microbiol Ecol 34: 181–186 Switzer Blum J, Bindi AB, Buzzelli J, Stolz JF and Oremland RS (1998) Bacillus arsenoselenatis sp.nov., and Bacillus selenitireducens sp. nov.: Two haloalkaliphiles from Mono Lake, California, which respire oxyanions of selenium and arsenic. Arch Microbiol 171: 19–30 Tabares LC, Bittel C, Carrillo N, Bortolotti A and Cortez N (2003) The single superoxide dismutase of Rhodobacter capsulatus is a cambialistic, manganese-containing enzyme. J Bact 185: 3223–3227 Takai K, Hirayama H, Sakihama Y, Inagaki F, Yamato Y and Horikoshi K (2002) Isolation and metabolic characteristics of previously uncultured members of the order Aquificales in a subsurface gold mine. Appl Environ Microbiol 68: 3046–3054 Tantaleàn JC, Araya MA, Saavedra CP, Fuentes DE, Perez JM, Calderon IL, Youderian P and Vasquez CC (2003) The Geobacillus stearothermophilus V iscS Gene, encoding cysteine desulfurase, confers resistance to potassium tellurite in Escherichia coli K-12. J Bacteriol 185: 5831–5737
688
Francesca Borsetti, Pier Luigi Martelli, Rita Casadio and Davide Zannoni
Taylor A (1996) Biochemistry of tellurium. Biol Trace Elem Res 55: 231–239 Taylor DE (1999) Bacterial tellurite resistance. Trends Microbiol 7:111–115 Tetaz TJ and Luke RK (1983) Plasmid-controlled resistance to copper in Escherichia coli. J Bacteriol 154: 1263–1268 Thompson-Eagle ET, Frankenberger WT and Karlson U (1989) Volatilization of selenium by Alternaria alternata. Appl Environ Microbiol 55: 1406–1413 Tomas JM and Kay WW (1986) Tellurite susceptibility and nonplasmid mediated resistance in Escherichia coli. Antimicrob Agents Chemother 30: 127–131 Toptchieva A, Sisson G, Bryden LJ, Taylor DE and Hoffman PS (2003) An inducible tellurite resistance operon in Proteus mirabilis. Microbiol 149: 1285–1295 Tremaroli V, Fedi S and Zannoni D (2007) Evidence for a tellurite dependent generation of reactive oxygen species and absence of a tellurite mediated adaptive response to oxidative stress in cells of Pseudomonas pseudoalcaligenes KF707. Arch Microbiol 187: 127–135 Trevors JT, Stratton GW and Gadd GM (1986) Cadmium transport, resistance and toxicity in bacteria, algae and fungi. Can J Microbiol 32: 447–464 Trutko SM, Akimenko VK, Suzina NE, Anisimova LA, Shlyapnikov MG, Baskunov BP, Duda VI and Boronin AM (2000) Involvement of the respiratory chain of gram-negative bacteria in the reduction of tellurite. Arch Microbiol 173: 178–186 Tsai KJ and Linet AL (1993) Formation of a phosphorylated enzyme intermediate by the cadA Cd2+- ATPase. Arch Biochem Biophys 305: 267–270 Turner RJ, Hou Y, Weiner JH and Taylor DE (1992a) The arsenical ATPase efflux pump mediates tellurite resistance. J Bacteriol 174: 3092–3094 Turner RJ, Weiner JH and Taylor DE (1992b) Use of diethyldithiocarbamate for quantitative determination of tellurite uptake by bacteria. Anal Biochem 204: 292–295 Turner JS, Morby AP, Whitton BA, Gupta A and Robinson NJ (1993) Construction of Zn2+/Cd2+ hypersensitive cyanobacterial mutants lacking a functional metallothionein locus. J Biol Chem 268: 4494–4498 Turner RJ, Weiner JH and Taylor DE (1994a) In vivo complementation and site-specific mutagenesis of the tellurite resistance determinant kilAtelAB from IncP alpha plasmid RK2Ter. Microbiol 140: 1319–1326 Turner RJ, Weiner JH and Taylor DE (1994b) Characterization of the growth inhibition phenotype of the kilAtelAB operon from IncP alpha plasmid RK2Ter. Biochem Cell Biol 72: 333–342 Turner RJ, Weiner JH and Taylor DE (1995) Neither reduced uptake nor increased efflux is encoded by tellurite resistance determinants expressed in Escherichia coli. Can J Microbiol 41: 92–98 Turner RJ, Taylor DE and Weiner JH (1997) Expression of Escherichia coli TehA gives resistance to antiseptics and disinfectants similar to that conferred by multidrug resistance efflux pumps. Antimicrob Agents Chemotherap 41: 440–444 Turner RJ, Weiner JH and Taylor DE (1998) Selenium metabolism in Escherichia coli. BioMetals 11: 223–227 Turner RJ, Aharonowitz Y, Weiner JH and Taylor DE (2001) Glutathione is a target in bacterial tellurite toxicity and is protected by tellurite resistance determinants in Escherichia coli. Can J Microbiol 47: 33–40
Tynecka Z, Gos Z and Zajac J (1981) Energy-dependent efflux of cadmium coded by a plasmid resistance determinant in Staphylococcus aureus. J Bacteriol 147: 313–319 Van Fleet-Stalder V, Gürleyük H, Bachofen R and Chasteen T G (1997) Effects of growth conditions on production of methyl selenides in cultures of Rhodobacter sphaeroides. Ind Microbiol Biotechnol 19: 98–103 Van Fleet-Stalder V and Chasteen TG (1998) Using fluorine-induced chemiluminescence to detect organo-metalloids in the headspace of phototrophic bacterial cultures amended with selenium and tellurium. J Photochem Photobiol 43: 193–203 Van Fleet-Stalder V, Chasteen TG, Pickering IJ, George GN and Prince RC (2000) Fate of selenate and selenite metabolized by Rhodobacter sphaeroides. Appl Env Microbiol 66: 4849–4853 Van Iterson W and Leene W (1964a) A cytochemical localization of reductive sites in a Gram-negative bacterium: Tellurite reduction in Proteus vulgaris. J Cell Biol 20: 377–387 Van Iterson W and Leene W (1964b) A cytochemical localization of reductive sites in a Gram-positive bacterium: Tellurite reduction by Bacillus subtilis. J Cell Biol 20: 362–375 Van Veen HW (1997) Phosphate transporter in prokaryotes: Molecules, mediators and mechanisms. Antonie van Leeuwenhoek 72: 299–315 Vargas M, Kashefi K, Blunt-Harris EL and Lovley DR (1998) Microbiological evidence for Fe(III) reduction on early Earth. Nature 395: 65–67 Vinceti M, Wei E T, Malagoli C, Bergomi M and Vivoli G (2001) Adverse health effects of selenium in humans. Rev Environ Health 16: 233–251 Watanabe M, Kawahara K, Sasaki K and Noparatnaraporn N (2003) Biosorption of cadmium ions using a photosynthetic bacterium, Rhodobacter sphaeroides S and a marine photosynthetic bacterium, Rhodovulum sp. and their biosorption kinetics. J Biosci Bioeng 95: 374–378 Watt RK and Ludden PW (1999) Ni2+ transport and accumulation in Rhodospirillum rubrum. J Bacteriol 181: 4554–4560 White C, Sayer JA and Gadd GM (1997) Microbial solubilization and immobilization of toxic metals: key biogeochemical processes for treatment of contamination. FEMS Microbiol Rev 20: 503–516 Whittaker JW (2000) Manganese superoxide dismutase. Met Ions Biol Syst 37: 587–611 Widdel F, Schnell S, Heising S, Ehrenreich A, Assmus B and Schink B (1993) Ferrous iron oxidation by anoxygenic phototrophic bacteria. Nature 362: 834–836 Wilson JR, Leang C, Morby AP, Hobman JL and Brown NL (2000) MerF is a mercury transport protein: Different structures but a common mechanism for mercuric ion transporters? FEBS Lett 472: 78–82 Wildgust MA, McDonald P and White KN (1998) Temporal changes of 210Po in temperate coastal waters. Sci Total Environ 214: 1–10 Xiong AM and Jayaswal RK (1998) Molecular characterization of a chromosomal determinant conferring resistance to zinc and cobalt ions in Staphylococcus aureus. J Bacteriol 180: 4024–4029 Yarema MC and Curry SC (2005) Acute tellurium toxicity from ingestion of metal oxidizing solutions. Pediatrics 116: 319–321 Yoch DC (1979) Manganese, an essential trace element for N2
Chapter 33
Metals and Metalloids in Photosynthetic Bacteria
fixation by Rhodospirillum rubrum and Rhodopseudomonas capsulata: Role in nitrogenase regulation. J Bacteriol 140: 987–995 Yurkov V, Jappè J and Verméglio A (1996) Tellurite resistance and reduction by obligately aerobic photosynthetic bacteria. Appl Env Microbiol 62: 4195–4198 Zadvorny OA, Zorin NA and Gogotov IN (2006) Transformation of metals and metal ions by hydrogenases from phototrophic bacteria. Arch Microbiol 184: 279–285 Zadik PM, Chapman, PA and Siddons CA (1993) Use of tellurite for the selection of verocytotoxigenic Escherichia coli O157. J Med Microbiol 39: 155–158 Zannoni D, Borsetti F, Harrison JJ and Turner RJ (2007) The
689
bacterial response to the chalcogen metalloids Se and Te. Adv Microbial Physiol 53: 1–71 Zawadzka AM, Crawford RL and Paszczynski AJ (2006) Pyridine-2,6-bis(thiocarboxylic acid) produced by Pseudomonas stutzeri KC reduces and precipitates selenium and tellurium oxyanions. Appl Environ Microbiol 72: 3119–3129 Zhang Y, Fomenko DE and Gladyshev VN (2005) The microbial selenoproteome of the Sargasso Sea. Genome Biology 6: R37 Zhang Y and Frankenberger WT Jr (2005) Removal of selenium from river water by a microbial community enhanced with Enterobacter taylorae in organic carbon coated sand columns. Sci Tot Environ 346: 280–285
Chapter 34 Purple Bacterial Genomics Madhusudan Choudhary1, Chris Mackenzie1, Timothy J. Donohue2 and Samuel Kaplan1* 1
Department of Microbiology and Molecular Genetics, The University of Texas Health Science Center, Houston, Texas 77030, U.S.A.; 2Department of Bacteriology, University of WisconsinMadison, Madison, Wisconsin 53076, U.S.A.
Summary ............................................................................................................................................................... 691 I. Introduction..................................................................................................................................................... 692 II. Genome Architecture and Characteristics...................................................................................................... 692 A. Genome Comparisons .................................................................................................................... 692 B. Evolutionary Relationships and the Origin of Mitochondria ............................................................. 694 C. Continual Evolution of Purple Bacteria ............................................................................................ 694 III. Gene Homologs and Metabolic Versatility ..................................................................................................... 696 A. Flagellum Biosynthesis .................................................................................................................... 696 B. CO2 Utilization and Photosynthesis ................................................................................................. 699 C. Energy Production ........................................................................................................................... 699 D. Tetrapyrrole Biosynthesis ................................................................................................................ 700 E. Sigma Factors ................................................................................................................................. 700 F. Molecular Chaperones .................................................................................................................... 701 IV. Variation in Transcriptional Regulation and Adaptation to Changing Environments ...................................... 701 V. Transposons and Genomic Rearrangements ................................................................................................ 702 VI. Circadian Clock and Gas Vesicle Proteins ..................................................................................................... 702 VII. Inorganic Compounds as Reducing Power .................................................................................................... 703 VIII. Genomic Insights into the Photosynthetic Lifestyle ........................................................................................ 703 Acknowledgments ................................................................................................................................................. 704 References ............................................................................................................................................................ 704
Summary Genomes of several purple bacteria have recently been sequenced and many of these genomes have been fully annotated and are now available on either NCBI or/and other publicly accessible databases. This chapter gives a comparative analysis of three representative genomes, Rhodobacter (Rba.) sphaeroides 2.4.1,Rhodospirillum (Rsp.) rubrum ATCC 11170, and Rhodopseudomonas (Rps.) palustris CGA009. The results reveal that these three genomes possess some remarkable similarities, although these species differ in their genome architectures and the numbers and sizes of plasmids. The genomes of these three species encode two or multiple homologs (orthologs or paralogs) of many protein-coding genes representing a wide variety of metabolic pathways, which substantiate the enormous amount of metabolic versatilities shown by organisms belonging to the α-Proteobacteria. Paralogs are genes related by duplication within a genome, and therefore evolve new functions, even if these are related to the original one. In addition to abundant gene paralogs revealed by their genomes, the *Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 691–706. © 2009 Springer Science + Business Media B.V.
692
Madhusudan Choudhary, Chris Mackenzie, Timothy J. Donohue and Samuel Kaplan
above strains of these species displayed different numbers of sigma factors, transcriptional regulators (activators or repressors) and corresponding DNA binding motifs in their respective genomes. The diversity in these regulatory components indicates the use of different regulatory strategies for the genome-wide transcriptional regulation in these organisms. The different numbers of insertion sequences and transposases found among strains of these three species suggests that transposon-induced genomic variants may play a major role in strain differentiation within species. I. Introduction Several studies involving morphology, physiology, and genetics of purple bacteria have been conducted over many decades, and these studies have led to the demonstration that the primary metabolism of these organisms is a photosynthetic process requiring light and a low availability of oxygen. However the biochemical reaction of CO2 reduction is carried out with reductant derived not from water as in green plants but from a variety of organic substances, as well as H2 or hydrogen sulfide (van Niel 1944; also see Chapter 1, Madigan and Jung; Chapter 28, Romangnoli and Tabita; Chapter 39, Vignais). Also, members of this group can fix atmospheric nitrogen (Chapter 37, Masepohl and Kranz), and they can grow fermentatively. Three representative species of purple photosynthetic bacteria, Rba. sphaeroides 2.4.1, Rsp. rubrum ATCC 11170, and Rps. palustris CGA009, which belong to the α-subgroup of Proteobacteria (Woese et al., 1984), are chosen for genome comparison as their genomes have been sequenced and the genomic sequence data are publicly available. The sequence of the complete genome of Rps. palustris CGA009 is available under GenBank/EMBL/DDBI accession numbers BX571963 (chromosome) and BX571964 (plasmid). The sequence of the Rba. sphaeroides 2.4.1 genome is available under NCBI accession numbers NC_007488 to NC_07494. The complete genome sequence of Rsp. rubrum ATCC11170 is under NCBI accession numbers NC_007643 (chromosome) and NC_007641 (plasmid). These representative species along with other members of the α-Proteobacteria are widely distributed in nature and considered the most metabolically versatile Abbreviations: ALA – 5-aminolevulinic acid; ALAS – 5 -aminolevulinic acid synthase; CDS – Coding sequence; CI – Large chromosome of Rba. sphaeroides; CII – Small chromosome of Rba. sphaeroides; IS – Insertion sequence; LH – Light harvesting complex; P. – Pseudomonas; PGC – Photosynthesis gene cluster; Rba. – Rhodobacter; RC – Reaction center ; Rps. – Rhodopseudomonas; Rsp. – Rhodospirillum
group of organisms (Gest 1972). Recent biochemical and genetic studies of these bacteria have shown induced levels of expression of genes encoding the photosynthetic apparatus, and enzymes involved in photosynthesis or anaerobic respiration, under a reduced oxygen tension (Zeiltra-Ryalls et al., 1998; Grammel et al., 2003; Roh et al., 2004; Braatsch et al., 2006; Mackenzie et al., 2007; Chapter 35, Bauer et al.). Although multiple strains of Rba. sphaeroides (Choudhary et al., 2007) and Rps. palustris (http:// img.jgi.doe.gov) have been recently sequenced, unless otherwise stated our genome comparison of these three species is based on a single representative strain selected from each of the three species (http://img. jgi.doe.gov): Rba. sphaeroides 2.4.1; Rsp. rubrum ATCC 11170; and Rps. palustris CGA009. II. Genome Architecture and Characteristics The genome characteristics of Rba. sphaeroides, Rsp. rubrum, and Rps. palustris are described in Table 1, and the original data of these three species are available on the publicly available website (http://img. jgi.doe.gov). A. Genome Comparisons The genome sizes of Rba. sphaeroides, Rsp. rubrum, and Rps. palustris are ~4.60, 4.40, and 5.46 Mb, respectively. The genomes of Rba. sphaeroides and Rsp. rubrum are of similar sizes, but the genome of Rps. palustris is 1.2 Mb larger than the genomes of the other two species. Besides the different genome sizes, the genome architecture of the three species also appears to be different. The genome of Rba. sphaeroides is comprised of two circular chromosomes, CI (large chromosome) and CII (small chromosome), and five plasmids (Suwanto and Kaplan, 1989b). The sizes of CI and CII of Rba. sphaeroides are 3,188,631 and 943,022 bp, respectively. In contrast the genomes of Rsp. rubrum and Rps. palustris consist of only one
Chapter 34
Purple Bacterial Genomics
693
Table 1. Comparison of Genome Characteristics Genome characteristics Total number of nucleotides
Rba. sphaeroides 2.4.1 4,603,060
Rsp. rubrum ATCC 11170 4,406,557
Rps. palustris CGA009 5,467,640
% DNA coding
88.28
88.93
87.90
% GC content
68.79
65.38
65.03
DNA scaffolds
7 (2 chromosomes) CI: 3,188,631 bp and CII: 943,022 bp)
2 (1 chromosome)
2 (1 chromosome)
Number of genes
4367
3917
4895
Protein coding genes
4304 (98.56%)
3850 (98.29%)
4838 (98.84%)
RNA genes
63
67
57
rRNA genes
9
12
8
5S rRNA
3
4
2
16S rRNA
3
4
2
23S rRNA
3
4
2
tRNA
54
55
49
Genes with function prediction
3038 (69.57%)
2861 (73.04%)
3406 (69.58%)
Genes in ortholog clusters
4115 (94.23%)
3627 (92.60%)
4723 (96.49%)
Genes in paralog clusters 2136 (50.33%) 1972 (50.34%) 2712 (55.40%) Source of data: http://img.jgi.doe.gov, Rba. sphaeroides (version 1.6), and Rps. palustris and Rsp. rubrum (version 2.1).
chromosome and a plasmid. Several strains of Rsp. rubrum possess a very similar size ~55 kb plasmid with very little sequence divergence, but these are not identical and the plasmid is required for photosynthesis (Kuhl et al., 1983). Thus, similar plasmids seem to be disseminated throughout strains of Rsp. rubrum. Similarly, the genome of Rps. palustris contains one chromosome and a single plasmid of ~8.4 kb (Larimer et al., 2004). Besides the size variation of plasmids (8–15 kb) among strains of Rps. palustris, the plasmid replicates in closely related species of Bradyrhizobium, but not in Rba. sphaeroides and Rhizobium species. Thus, the origin of this plasmid has been thought to be from one of the strains of Bradyrhizobium (Giraud et al., 2000). Despite its large genome size, the genome of Rps. palustris contains a single chromosome, which also suggests that genome size is not the sole factor for genome complexity, including the existence of multiple chromosomes in bacteria. Genomes of these three species exhibit remarkable similarities in gross genome structure. The total number of predicted genes in Rba. sphaeroides, Rsp. rubrum, and Rps. palustris are 4367, 3917, and 4897, respectively, which reflects their relative genome sizes. Of the total annotated genes, ~98% represent
open reading frames, which potentially encode protein functions in all three species. Of these protein coding genes, ~70% can be assigned predicted functions, and >90% of these genes are classified in ortholog clusters in each of the three species. Orthologs are genes that evolve from a common ancestral gene by speciation. Since orthologs conserve the same function through evolution, identification of orthologs is valuable for prediction of gene function in newly sequenced genomes. Also, the coding capability of each of the three genomes is approximately 88%. The genomes of different bacterial species in varied habitats reveal a narrow variation of their G+C compositions, which indicates that primarily, global environmental factors as opposed to different lifestyles of the microorganisms influence the G+C composition of a microbial community (Foerstner et al., 2005). The variation in G+C composition of different species is either maintained by selective pressure or caused by mutational bias. The observed G+C differences have a direct impact on the amino acid composition of proteins in the organisms living in their respective environments (Bharanidharan et al., 2004). The G+C composition of the genomes of Rsp. rubrum and Rps. palustris is ~65%, which is ~3% lower than the ~68% G+C content of the genome of
694
Madhusudan Choudhary, Chris Mackenzie, Timothy J. Donohue and Samuel Kaplan
Rba. sphaeroides. The differences in G+C content exist across species, however there is a tendency of large genomes to be G+C-rich and small genomes to be A+T-rich (Glass et al., 2000; Moran, 2002; Rocha and Danchin, 2002). Bacteria with large genomes are found in more complex and diverse environments, and therefore G+C content may possibly be reflective of their niche complexity (Rocha and Danchin, 2002; Moran, 2002). Since the nucleotide composition of the three genomes discussed in this chapter is G+Crich, it may correlate with a similar distribution of these species in diverse environments. A higher G+C composition of Rba. sphaeroides is possibly related to its genome architecture, as the genome of Rba. sphaeroides possesses two chromosomes, but this issue remains unsettled. B. Evolutionary Relationships and the Origin of Mitochondria Despite the gene level similarities, genome data also suggest significant evolutionary differences among these three bacteria. Based on 16S rDNA sequences, Rba. sphaeroides is part of a distinct clade within the α-3 sub-group of Proteobacteria. The group of α-Proteobacteria also contains organisms that associate with eukaryotes, like Rhizobacteria (essential for N2 fixation in legumes), Agrobacteria (plant pathogens), and Rickettsia (intracellular animal pathogens) (Woese, 1987). The ability of α-Proteobacteria to associate with eukaryotes and their diverse metabolic activities led to the proposal that they are evolutionary ancestors of mitochondria (Woese, 1987). The current thought is that the species belonging to the order Rickettsiales of α-Proteobacteria are the closest relatives of mitochondria (Andersson et al., 1998; Gray et al., 1999; Emelyanov, 2001). More than 150 nucleus-encoded mitochondrial proteins of Saccharomyces cerevisiae share significant sequence homology with Rickettsia prowazekii proteins, and also several gene clusters in the mitochondrial genome are reminiscent of those found in Rickettsia, including a similar repertoire of proteins involved in ATP production, and transport (Andersson et al., 1998). However, the relationship of mitochondria to the Ricketsiales is limited to some conserved indels (signature protein sequences), and the closest relationship of mitochondria was seen for Rsp. rubrum rather than Ricketsiales (Esser et al., 2004). Thus, additional data are required to resolve the above conflicting observations.
C. Continual Evolution of Purple Bacteria Several hypotheses have been proposed to explain the complex patterns of sequence relationships observed in microbial genomes. The continual ‘horizontal transfer’ hypothesis suggests that gene acquisitions are ongoing processes in microorganisms (Jain et al., 2002), whereas the early ‘massive transfer’ hypothesis proposes that massive exchanges occurred early in prokaryotic evolution, long before the diversification of modern bacterial species (Woese, 1998). Indeed, the presence of photosynthetic and non-photosynthetic species among α-Proteobacteria has been used to propose a continual evolution of traits acquiring by horizontal gene transfer in this group (Woese, 1987). Although the Proteobacteria are considered as the earliest lineage among the photosynthetic prokaryotes (Xiong et al., 2000), the evidence suggests that the Cyanobacteria constitute one of the earliest prokaryotic photosynthetic lineages that existed ~2.5 billion years ago. Thus, photosynthetic purple bacteria could have independently evolved or acquired photosynthesis gene clusters from one of the members of Cyanobacteria. Therefore, it is possible that non-photosynthetic purple bacteria may have evolved by losing the photosynthesis gene clusters from their ancestor lineages. Indeed, the notion of continual evolution among α-Proteobacteria is reinforced by genome-wide comparisons of Rba. sphaeroides, Rsp. rubrum and Rps. palustris. Although more than 40 bacterial species, which belong to diverse groups of bacteria, are known to have multiple chromosome-like replicons, most of the species containing multiple chromosomes have been found within the α-subgroup of Proteobacteria (Mackenzie et al., 2004). Among 25 isolates examined, the presence of two chromosomes is found in all these strains of Rba. sphaeroides (Nereng and Kaplan, 1999). The presence of two chromosomes in Rba. sphaeroides was originally proposed from pulse-field gel electrophoresis, but it has been confirmed by either optical mapping or genome sequence analysis of several isolates of this species (Choudhary et al., 2007; T. Donohue unpublished). In addition, by comparing genome sequence-derived restriction maps or optical maps of Rba. sphaeroides (Zhou et al., 2003), Rsp. rubrum (Reslewic et al., 2005) and Rps. palustris, no large regions of genome similarity except the photosynthesis gene cluster (PGC) are found between these three α-Proteobacteria. The total length of conserved regions among Rba. sphaeroi-
Chapter 34
Purple Bacterial Genomics
des, Rsp. rubrum, and Rps. palustris is ~330 kb of DNA that span over 107 common collinear blocks (Choudhary and Kaplan, unpublished). The majority of common DNA blocks consist of an average ~3 kb of DNA length. Also, the overall nucleotide similarity over all common collinear blocks was only ~25% among these three species. This is unlike the finding of significant regions of conservation among many Rba. sphaeroides isolates (Choudhary et al., 2007), and it is probably not surprising given the various activities and habitats described for these and other photosynthetic α-Proteobacteria (Woese 1987; also see Chapters 1, Madigan and Jung; Chapter 3, Yurkov and Csotonyi). A comparison of the coding sequences (CDSs) of the Rsp. rubrum and Rps. palustris genomes when used to query an Rba. sphaeroides protein database indicates that each has a similar number of genes that are orthologous to Rba. sphaeroides genes (Fig. 1). Each has approximately twice the percentage of
695 polypeptides with high quality matches (low P-values) to Rba. sphaeroides as compared to the E. coli outlier. The genomes of both Rsp. rubrum and Rps. palustris also encode relatively few polypeptides with high P-values (poor quality matches) whereas a larger proportion of the E. coli CDSs fall into this class. This is perhaps not surprising, since the three phototrophic bacteria are α-proteobacteria and E. coli is a γ-Proteobacterium. Interestingly, the Rsp. rubrum and Rps. palustris genomes have only approximately half of the high quality CDS matches (~2% of the genome having P-values <10–200), as does Rba. sphaeroides to members of the marine Roseobacter clade (Moran et al., 2004) or Paracoccus denitrificans (http://genome.ornl.gov) (~5% of the genome having P-values <10–200). This suggests that although Rsp. rubrum, Rps. palustris, and Rba. sphaeroides are similar because of their photosynthetic properties, they are in fact quite distantly related when the marine Roseobacters or the non-photosynthetic Paracoccus
Fig. 1. The CDSs from the genomes of Rps. palustris CGA009 and Rsp. rubrum ATCC 11170 and E. coli (for comparison) were run in a WU-BLAST 2.0 (W. Gish, personal communication) search using BLASTP mode against an Rba. sphaeroides 2.4.1 CDS database. The output of each BLASTP search was extracted to give the top hit found in the Rba. sphaeroides genome. The top hits were then combined into a list, then sorted on the basis of their P-value. The percentage of genes in each genome falling within the following P-value ranges; <10–200, 10–200–150, 10–150–100, 10–100–75, 10–75–50, 10–50–35, 10–35–20, 10–20–10, 1010–4, 10–4–0.05, 0.05–0.03 and >0.3 was then determined (x-axis). To give an indication of the conversion of P-value to amino acid identity/similarity, consider the following two examples; a polypeptide of 275 amino acid (aa) residues having a good match, 73% identity (85% similar) to a polypeptide in the Rba. sphaeroides database had a P-value of 5.7e 110, whereas a borderline match was a polypeptide of 219 aa residues with 29% identity (44% similar) having a P-value of 5.4e–18. An imaginary cutoff for relevant matches lies approximately in the 10–20–10 range.
696
Madhusudan Choudhary, Chris Mackenzie, Timothy J. Donohue and Samuel Kaplan
denitrificans are added into the analysis (Mackenzie et al., 2007).
which descend from speciation events. Of the sample of genes listed in Table 2, many homologs exist as paralogs in the genomes of all three species, Rba. sphaeroides, Rsp. rubrum, and Rps. palustris.
III. Gene Homologs and Metabolic Versatility
A. Flagellum Biosynthesis
A gene related to another gene within or between genomes evolves by descent from a common ancestral DNA sequence and is defined as a homolog. Homologs are classified into two families, orthologs and paralogs (Fitch, 2000). Orthologs represent homologous genes in different species that evolved from a common ancestral gene by the speciation event, which leads to the evolution of two different lineages. Most orthologs retain the same physiological function across species in the course of evolution. In contrast, gene paralogs are related by a gene duplication event followed by sequence divergence of these duplicated genes within a genome, and therefore paralogs evolve new functions. Gene paralogs are abundantly found in bacterial genomes, and thought to descend from sequence duplications (Gevers et al., 2004). Genome analysis of Rba. sphaeroides, Rps. palustris and Rsp. rubrum reveals ~50% of the genes in paralog clusters as shown in Table 1. Although the numbers of rRNA and tRNA genes in all three genomes are similar, the genomes of Rba. sphaeroides and Rsp. rubrum contain a higher number of these genes than Rps. palustris as shown in Table 1. The genomes of Rps. palustris, Rba. sphaeroides, and Rsp. rubrum encode 2, 3, and 4 copies of 16S and 23S rRNA genes, respectively. Rba. sphaeroides and Rsp. rubrum contain 54 and 55 tRNA genes, respectively, which are higher than the 49 tRNA genes identified in the Rps. palustris genome. Genome analyses of Rba. sphaeroides revealed numerous homologous genes representing many metabolic pathways, such as flagellum biosynthesis, photosynthesis, carbon metabolism, and chemosensory pathways (Mackenzie et al., 2001; Choudhary et al., 2004). The majority of gene homologs found in the Rba. sphaeroides genome displayed less similarity with each other than to orthologs from species closely related to Rba. sphaeroides. Thus, the majority of the gene homologs in Rba. sphaeroides, for example rdxA and rdxB, and hemA and hemT, are thought to be very old and possibly occurred prior to the evolution of the Rba. sphaeroides lineage (Choudhary et al., 2004), and therefore these gene homologs are orthologs,
In Rba. sphaeroides, there are two sets of flagellar biosynthesis genes on CI, albeit the second set is incomplete. The first set of complete flagellar genes (RSP0032-RSP0084) is responsible for the synthesis and rotation of the subequatorial flagellum while the functions of the second set of incomplete flagellar genes (RSP1302-RSP1330) are not known. In contrast, the genome of Rsp. rubrum encodes a single operon containing genes for flagellum formation on its chromosome. However, the genome of Rps. palustris encodes a single copy of the genes required for flagellum formation, except flgC, flgE, and flgG, which are encoded in gene paralogs. The Rba. sphaeroides flagellar genes in the first operon are expressed in all growth conditions examined, such as aerobic, semiaerobic, and photosynthetic (J. Roh, unpublished results). Although microarray expression of the second set of Rba. sphaeroides genes was undetected in any of the growth conditions tested (J. Roh, unpublished), and therefore the functions of these genes are unknown, the finding of two copies of most of the structural genes for flagellum formation is very significant. The cluster of the second set of genes in Rba. sphaeroides appears to be horizontally acquired, and these genes were shown to coexist with the endogenous (first) set of flagellar genes (Poggio et al., 2007). It was speculated that the second set of genes for flagellum biosynthesis in Rba. sphaeroides could be required for surface translocation using lateral flagella during biofilm production or another hypothetical type of alternative life style of this organism, as has been shown in other organisms (Capdevila et al., 2004; Kanbe et al., 2007). Bacterial cell motility in other organisms is very important in colonization since non-motile mutants of Pseudomonas (P.) fluorescens are severely impaired in root colonization (Capdevila et al., 2004). It has been noticed that the predominance of flagella variants with enhanced surface motility reach the distal part of the rhizosphere, which is not easily traveled by the wild type strains of P. fluorescens (Sanchez-Contreras et al., 2002). It remains to be seen if the second set of flagellar genes of Rba. sphaeroides has strong homology with the flagellar genes of organisms such
Chapter 34
Purple Bacterial Genomics
697
Table 2. List of gene homologs and number of genes in functional classes Metabolic pathways/Genes
Number of genes encoded in genomes of Rba. sphaeroides Rsp. rubrum
Rps. palustris
Flagellum biosynthesis motA flgA flgC flgE flgG flgH flgI flgL fliE fliF fliG fliH fliM fliQ fliR flhB
2 2 1 2 1 2 2 2 2 2 2 2 2 2 2 2
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
1 1 2 2 2 1 1 1 1 1 1 1 1 1 1 1
Chemotaxis cheA cheB cheY cheW cheR
3 2 10 4 3
3 3 21 4 4
3 2 3 4 4
Photosynthesis pucA pucB pucC
2 2 1
2 2 2
6 8 1
CO2 fixation cbbA cbbF cbbP cbbL cbbG cbbT dxs
2 2 2 2 3 2 2
1 1 1 1 1 2 2
1 1 1 2 1 2 1
Energy production (Extra ATPase subunit) nuoA nuoB nuoC nuoD nuoE nuoF nuoG nuoH nuoI nuoJ nuoK nuoL nuoM nuoN atp (b subunit)
2 2 2 2 2 2 2 2 1 2 1 1 2 1 2
1 2 2 2 1 1 1 2 1 1 2 1 2 2 2
1 2 2 2 2 2 2 2 2 2 2 2 2 2 2
698
Madhusudan Choudhary, Chris Mackenzie, Timothy J. Donohue and Samuel Kaplan Table 2. Continued Metabolic pathways/Genes
Number of genes encoded in genomes of Rba. sphaeroides Rsp. rubrum
Rps. palustris
DNA Replication and partitioning parA repA parB repB dnaE dnaN
4 3 7 2 2 1
1 3 1 1 1 3
3 1 3 1 2 2
Amino acid biosynthesis trpB glnA
2 5
2 2
2 4
Lipid metabolism Ech (Enoyl-CoA hydratase)
6
6
23
Heat shock proteins groES groEL
1 3
2 2
2 2
Tetrapyrrole biosynthesis hemA/hemT hemN
2 2
2 1
2 1
Sigma factors rpoN rpoH
4 2
1 1
1 1
Transcription regulators DoeR family LysR family LuxR family MarR family ArsR family AsnC family Crp family GntR family MerR family TetR family IclR Family AraC Family Helix-turn-helix, Fis type Helix-turn-helix, CopG family Histidine kinases Response regulators Sigma factors
4 23 13 5 6 10 8 27 6 9 2 10 1 2 29 51 17
6 26 6 10 6 10 5 12 5 18 2 18 16 10 54 47 17
1 27 11 20 9 5 13 13 3 39 6 23 8 2 64 70 19
Others Transposases 34 26 16 Dehydrogenases 195 111 194 Source of data: (http://img.jgi.doe.gov; Choudhary et al., 2004; Larimer et al., 2004).
as P. fluorescens, where the functions of these genes have been implicated for root colonization. Flagellar motility and behavioral responses are discussed in this book elsewhere (Chapter 32, Armitage). All three of Rba. sphaeroides, Rsp. rubrum and
Rps. palustris possess a number of gene paralogs in chemosensory pathways. There are 21 copies of cheY (chemotaxis-specific response regulator) in Rsp. rubrum, which is twice the number of cheY paralogs (10 copies) present in the Rba. sphaeroides genome.
Chapter 34
Purple Bacterial Genomics
However, the genome of Rps. palustris encodes only 3 copies of cheY. Thus, Rsp. rubrum and Rba. sphaeroides display a high level of cheY diversity compared to Rps. palustris, which may allow the former two species to exploit more diverse environments. However, the number of methyl accepting chemotaxis genes, such as cheA, cheB, and cheR moderately vary (2 to 4 copies) in the three species, as shown in Table 2. Thus, methyl accepting chemotaxis genes, being the best indicators of the ability of a bacterium to exploit diverse environments, do not indicate great difference in these three species. B. CO2 Utilization and Photosynthesis The purple photosynthetic bacteria serve as model organisms for the study of autotrophy and its relationship to photosynthesis. This group of organisms can grow in different growth conditions that use the Calvin cycle for CO2 assimilation. CO2 fixation and carbon metabolism are described elsewhere in this book (Chapter 28, Romangnoli and Tabita). Genes encoding enzymes and regulatory proteins of the Calvin cycle have been identified from a variety of bacteria, but the organization and the regulation of these genes are known in detail in only some organisms, including Rba. sphaeroides. Gene paralogs in the pathway of Calvin cycle are more abundant in Rba. sphaeroides than in Rsp. rubrum and Rps. palustris. For example, there are two homologs of cbbA, cbbF, cbbP, cbbL, cbbT, and dxs; and three homologs of cbbG in Rba. sphaeroides as shown in Table 1. However, Rsp. rubrum possesses two homologs of cbbT and dxs, while the genome of Rps. palustris contains paralogs of cbbT and cbbL, but on the other hand Rsp. rubrum does not have paralog of cbbL. Photosynthesis is the essential characteristic of this group of organisms. Most of the genes involved in photosynthesis are located in a photosynthesis gene cluster (PGC) that contains reaction center (RC), light-harvesting complexes, and bacteriochlorophyll and carotenoid biosynthesis genes. The pucBA structural genes encode the apoproteins of the light-harvesting complex, LH2 (B800-850) βand α-polypeptides. Paralogs of pucBA are located about 1.36 Mb apart on CI of Rba. sphaeroides. The puc2B-encoded polypeptide is almost identical to the puc1B-encoded polypeptide. However, the N-terminal 48 amino acid residues of puc2A and puc1A exhibit a lower level of sequence conservation, and it is suspected that the puc2A gene is a pseudogene or/and
699 may be in the process of acquiring a new function (Zeng et al., 2003). The expressions of these two puc operons in Rba. sphaeroides are both oxygen and light dependent. In contrast, the genome of Rsp. rubrum lacks pucBA (Bérard et al., 1986; Hessner et al., 1991), while the genome of Rps. palustris encodes four complete sets and one incomplete set of LH2 genes (Larimer et al., 2004). Two of the four complete sets of pucBA genes are located adjacent to the bacteriophytochrome genes that may regulate LH2 complex gene expression. However, only one pucBA operon is located in the PGC of Rba. sphaeroides. The high number of LH2 (pucBA) genes appear to encode different types of LH2 complexes in Rps. palustris, which help to adapt Rps. palustris in harvesting different wavelengths of light (Hartigan et al., 2002; Tharia et al., 1999; Chapter 40, Evans et al.). Thus, Rps. palustris appears to be more versatile for photosynthetic growth. The pufBA genes, which encode LH1 polypeptides, are highly conserved in Rba. sphaeroides, Rps. palustris and Rsp. rubrum. In addition, the genes encoding the RC L, M, and H polypeptides of Rps. palustris are highly conserved and are 45 to 60% similar at the amino acid sequence level to the homologs of Rba. sphaeroides, but these genes are more closely related to Bradyrhizobium ORS278 (Giraud et al., 2000), and therefore these three genes (pufL, pufM, and pufH) may have been acquired by horizontal transfer from Bradyrhizobium, or vice versa. C. Energy Production The NADH: ubiquinone oxidoreductase (complex I), an enzyme of the photosynthetic as well as aerobic and anaerobic respiratory chains (Hirst, 2005), consists of 13 subunits, NuoA through NuoN. The genomes of Rba. sphaeroides, Rsp. rubrum, and Rps. palustris encode two nuo operons, one of which contains an incomplete set of nuo genes. Many of these nuo genes were identified as paralogs in these species as shown in Table 2, however their functions are not yet known. Expression of the nuo genes in P. fluorescens is increased under low oxygen tension, and their levels of expression vary during different growth phases (Camacho Carvajal et al., 2002). Mutation in the nuo gene(s) in P. fluorescens has been shown to impair the root-colonization ability in tomato rhizosphere, due to the low oxygen availability (Camacho Carvajal et al., 2002). A better understanding of the regulation of the paralogs of nuo genes could possibly indicate
700
Madhusudan Choudhary, Chris Mackenzie, Timothy J. Donohue and Samuel Kaplan
whether Rba. sphaeroides, Rsp. rubrum or Rps. palustris is involved in a symbiotic association involving the gene products of the nuo genes. Enzymes in the F0F1 family of proton-translocating ATPases and ATP synthases are essential in all bacterial growth conditions. The general structure of ATP synthase is highly conserved (Borghese et al., 1998a,b), and it was found that there are ten genes contained in two operons, atpHAGDC and atpFXEBI, in Rba. capsulatus and Rba. sphaeroides. This type of gene organization seems to be unique to other members of Rhodospirillaceae. All three representatives of the purple photosynthetic bacteria examined in this chapter possess a total of these 10 genes, of which eight genes encode essential subunits of the ATP synthase. In addition to these essential subunits, atpX and atpI encode for b´ and I subunits, respectively. While the function of the atpI gene remains unknown, a duplicated and divergent copy of the b subunit gene (b´ subunit) was identified in all three species and its function was linked to proton transport. In Cyanobacteria, it has been shown that when an additional b subunit gene is present, a b2 heterodimer (bb´) is formed (Dunn et al., 2001), that could lead to different rates of proton transport (Turina et al., 2006). Since the duplication of a divergent b subunit gene was found only in the photosynthetic prokaryotes and the plant chloroplast (Borghese et al., 1998b), the role of gene duplication has been implicated in an alternative life style or photosynthesis, which may possibly require different levels of protons per ATP hydrolyzed in order to adapt to a specific growth condition (Cross and Taiz, 1990). Rba. sphaeroides possesses gene homologs, rdxB and rdxA (Neidle and Kaplan, 1992), encoding membrane bound ferredoxin-like proteins that are homologs of the Rhizobium meliloti fixG. RdxB is involved in photosynthesis gene expression under aerobic growth, carotenoid biosynthesis, and the expression of the cbb3 cytochrome oxidase. These findings suggest that rdxB along with other genes in this cluster (rdxBHIS) on CI are part of the same signal transduction pathway, and might control directly or indirectly the redox state of the cell. In contrast, the other copy (rdxA) on CII is expressed in a number of growth conditions (Pappas et al., 2004), but no specific function has been determined. A homolog of rdxBHIS of Rba. sphaeroides is also present in the genome of Rps. palustris, but no such homolog is found in Rsp. rubrum. Therefore, rdxBHIS genes are not essential effectors of photosynthesis, but they
may possibly modulate photosynthesis gene expression differentially in the environments where both Rba. sphaeroides and Rps. palustris are naturally established. D. Tetrapyrrole Biosynthesis The enzyme 5-aminolevulinic acid synthase (ALAS) catalyzes the first and rate-limiting step in a branched pathway for tetrapyrrole biosynthesis, which is discussed thoroughly elsewhere in this book (Chapter 39, Zeilstra-Ryalls). There are two ALAS genes, hemA and hemT in the Rba. sphaeroides genome, localized on CI and CII, respectively (Neidle and Kaplan, 1993a; 1993b). Both Rsp. rubrum and Rps. palustris encode homologs of the gene for ALAS on their chromosomes. It has been demonstrated that both hemA and hemT contribute to the overall cellular levels of ALA in Rba. sphaeroides, however the mRNA from hemT was not detected (Zeilstra-Ryalls and Kaplan, 1995). In addition to the difference in gene regulation and possibly the biochemical properties of the two isozymes, their cellular localizations are thought to be different; one is localized in the cytoplasm and other appears to be in the membrane bound fraction (Fanica-Gaignier and Clement-Metral 1973; Bolt et al., 1999). It remains to be examined whether paralog copies of ALAS genes also follow the same pattern of expression in Rsp. rubrum and Rps. palustris. E. Sigma Factors Prokaryotic transcription initiation requires different types of sigma (σ) factors, which are activated in response to different environmental conditions. The genomes of Rsp. rubrum and Rps. palustris contain a single σ54 homolog encoded by the rpoN gene. However, the Rba. sphaeroides genome contains four σ54 homologs encoded by rpoNI, rpoNIII and rpoNIV , located on CI, and rpoNII located on CII. rpoNI is located within the nif gene cluster containing genes for nitrogen fixation (Suwanto and Kaplan, 1989a). Deletion of rpoNI resulted in reduced diazotrophic growth of Rba. sphaeroides (Meijer and Tabita, 1992). In contrast, rpoNII, is not involved in either diazotrophic growth or in the synthesis of nitrogenase activity (Smith and Kaplan, unpublished). The rpoNIII gene lies ~100 bp downstream of fliC and might be involved in flagellar biosynthesis and/or cell motility (Mackenzie et al., 2001). The physical proximity of
Chapter 34
Purple Bacterial Genomics
rpoNIII and fliC suggests their involvement in flagellum formation as in other bacterial species. The negative control of flagellum formation in Pseudomonas species was seen to be mediated by either structural proteins of the flagella or regulatory proteins such as RpoN, RpoF (FliA), or/and FleQ (Kieboom et al., 1998). In Pseudomonas aeruginosa, both copies of the rpoN gene, which encode the alternative sigma factor, σ54, as well as fliC, encoding the flagellar subunit protein flagellin, were found to be essential for flagellum formation and motility (Garrett et al., 1999). There is no obvious operon near the Rba. sphaeroides rpoNII or rpoNIV that could provide a clue as to their possible function(s). Similarly, two rpoH genes encoding σ32 homologs are present in the Rba. sphaeroides genome, but found only in single copy in both Rsp. rubrum and Rps. palustris. Each of these two alternative sigma factors (RpoHI and RpoHII) may contribute to the heat-shock response since cells lacking only one of these proteins are able to grow at elevated temperatures (Karls et al., 1998; Green and Donohue, 2006). However, RpoHII expression appears to be dependent on the extracytoplasmic function of sigma factor, σE (Anthony et al., 2005), and it was suggested that the role of RpoHII is tightly linked to the singlet oxygen stress response. In contrast, cells lacking RpoHI are more sensitive than wild type cells to oxyanions (Karls et al., 1998), suggesting a role for this protein in another potential stress pathway. F. Molecular Chaperones Molecular chaperones are commonly called heatshock proteins, which help protein folding. Chaperones exist in many different families and they are expressed under conditions of high stress. The GroEL/GroES complex (HSP60), well-characterized in E. coli, is a large (~1 MDa) chaperone in bacteria. The genome of Rba. sphaeroides contains two groESL operons, one of which encodes groES and the other groEL genes. Transcription of the groESL1 genes was observed to be low under photoautotrophic growth conditions. However, no transcript was detected for the groESL2 operon under growth conditions examined so far (Smith and Kaplan, unpublished). Seemingly, Rsp. rubrum and Rps. palustris also possess one each of groES and groEL, but their functions are not yet determined.
701 IV. Variation in Transcriptional Regulation and Adaptation to Changing Environments Transcriptional regulation serves the primary mechanisms of adaptation in prokaryotes. Bacterial cells sense environmental signals using sensor kinases, and typically transmit these signals by phosphorylation of response regulators, which may activate or repress transcription of specific genes. Rba. sphaeroides, Rsp. rubrum, and Rps. palustris possess 80, 101, and 134 signal transduction genes (encoding histidine kinases and response regulators), respectively. The greater number of signal transduction genes in Rps. palustris may possibly provide this species the ability to control many different physiological processes involving transcriptional regulators. The networks of transcriptional controls are discussed elsewhere in this book (Chapter 35, Bauer et al., Chapter 36, Klug and Masuda), therefore only a brief overview of diverse families of regulatory proteins found in these species is provided here. The families of transcription regulatory proteins (Martinez-Bueno et al., 2004), such as AsnC, AraC, Crp, GntR, LysR, LuxR, TetR and the number of paralogs encoded in the Rba. sphaeroides, Rsp. rubrum, and Rps. palustris genomes are listed in Table 2. Many of these proteins are predicted to contain multiple domain motifs (Pareja et al., 2006). Since LysR regulatory proteins act as either activator or repressor, which are commonly found in bacteria and archaeae, it is not surprising that all three species possess a comparable number of LysR in their genomes. LysR regulatory proteins induce gene expression involved in multiple pathways, including pathogenicity, and biofilm production in other species (Kovacikova et al., 2005). However, their specific functions have not been yet determined in these three species. The Rba. sphaeroides genome encodes 27 GntRlike regulatory repressors, which are also found in diverse groups of bacterial species. In contrast, Rps. palustris and Rsp. rubrum encode only half the number of GntR proteins as compared Rba. sphaeroides. GntR regulatory proteins have been shown to sense cellular signals, and stimulate antibiotic production and carbon utilization (Hillerich and Westpheling, 2006). Thus, the possession of a variety of GntR regulators in these species may possibly play an important regulatory role in secondary metabolite production and carbon metabolism. In comparison to Rba. sphaeroides and Rsp. rubrum, the genome of Rps. palustris encodes more
702
Madhusudan Choudhary, Chris Mackenzie, Timothy J. Donohue and Samuel Kaplan
copies of MarR regulatory proteins. This class of regulator often negatively regulates the expression of antibiotic resistance genes. In Rps. palustris, the BadR protein, a member of the MarR family, positively regulates expression of genes for anaerobic benzoate degradation (Egland and Harwood, 1999). In addition, the genome of Rps. palustris possesses a much higher number of TetR and AraC regulatory proteins. TetR regulators are involved in transcriptional control of multidrug efflux pumps, antibiotic production, pathogenicity, and responses to osmotic stress (Ramos et al., 2005). Activators belonging to AraC family are involved in quorum sensing, signaling, and the production of virulence factors (Ramos, 2002). In addition to the high copies of LysR type transcriptional regulators, Rsp. rubrum also possesses moderately high copies of TetR, AraC, MarR, and AsnC families of transcriptional regulators. Although the comparative analysis of gene sequences has been helpful in the prediction of the structure-function of proteins, the regulation of these genes depends on their upstream regions with different regulatory motifs. For example, DNA sequences predicted to bind helix-turn-helix motifs are more abundantly found in Rsp. rubrum than the same sequences found in the genomes of the other two species (M. Choudhary, unpublished). Therefore, the number of regulatory proteins found within each of these organisms and the corresponding variation in number of DNA binding motifs located in the respective genome, may possibly play an important role in different strategies for global transcriptional regulation in each of the three species. V. Transposons and Genomic Rearrangements Insertion sequences (IS) consist of a large group of bacterial mobile elements, and are known to comprise up to 2 to 5% of genomic DNA in most bacterial genomes (Mahillion et al., 1999; Chandler and Mahillion, 2002). IS elements encode transposases required for transposition that helps the bacterial genome to acquire accessory functions and antibiotic resistance genes. Rba. sphaeroides and Rsp. rubrum contain 34 and 26 transposase genes, respectively. However, Rps. palustris contains only seven transposase genes indicating a minor role of transposon-induced genomic variation in this species. In contrast to strain Rps. palustris CGA009, other strains of Rps. palustris
(http://img.jgi.doe.gov) possess equivalent number of transposases as found in either Rba. sphaeroides 2.4.1 or Rsp. rubrum ATCC 11170. On the contrary to Rba. sphaeroides 2.4.1, another strain of Rba. sphaeroides (ATCC 17029) contains a reduced number of transposases in its genome (http://img. jgi.doe.gov). Members of the IS66 family are found in all three of the genomes considered here, and are widely distributed in Gram-negative bacteria, such as in Agrobacterium, Rhizobium, Escherichia, Pseudomonas, and Vibrio (Han et al., 2001). Although Rba. sphaeroides, Rsp. rubrum, and Rps. palustris harbor similar families of transposons in their genomes, the numbers vary in different strains within species, and therefore transposon-generated functions possibly play an important role in strain differentiation or intra-species variation. This hypothesis has recently been supported by a comparative genome analysis of the three strains of Rba. sphaeroides, 2.4.1, ATCC 17029, and ACC 17025, which revealed a rapid nucleotide divergence between orthologs of CII-specific DNA sequences (Choudhary et al., 2007). The DNA divergence could be the result of rearranged DNA sequences mediated by insertion sequences, gene duplications or/and newly acquired genetic elements on CII. The rapid evolution of CII suggests that accumulation of genetic variants in CII may play a major role in strain differentiation in Rba. sphaeroides. VI. Circadian Clock and Gas Vesicle Proteins The existence of circadian clock genes, kaiB and kaiC, has been reported among the purple bacteria (Mackenzie et al., 2001; Larimer et al., 2004). In Rba. sphaeroides, both genes are located on a plasmid while in Rsp. rubrum and Rps. palustris, these two circadian clock genes are located on the chromosome. Although circadian clock genes have previously been identified in many cyanobacterial species (Johnson and Golden, 1999), the function of these genes in the anoxygenic photosynthetic bacteria has yet to be determined. These species generate enough energy by photophosphorylation during daylight hours, but an inadequate amount of sunlight is available at night. Therefore, these organisms may use other organic compounds as an energy source when sunlight is absent or limiting. Thus, circadian regulation of redox reactions such as nitrogen fixa-
Chapter 34
Purple Bacterial Genomics
tion or other physiological processes could require these time-keeping genes. For example, gas vesicle proteins are encoded on CII of Rba. sphaeroides. Surprisingly, gas vesicle proteins are not encoded in the genome of either Rsp. rubrum or Rps. palustris. We speculate that gas vesicles may possibly play an important role in the flotation of Rba. sphaeroides at low light intensities or sinking of the cells under intense light, in conjunction with day-night cycles. It remains to be seen whether the expression of gas vesicle genes and circadian clock genes are linked in Rba. sphaeroides. If so, Rba. sphaeroides and Rps. palustris may use different strategies to adjust to different light intensities. VII. Inorganic Compounds as Reducing Power In addition to carbon dioxide and nitrogen assimilation, many inorganic compounds such as thiosulfate, hydrogen gas, as well as organic compounds such as formate, serve as reductants for autotrophic growth. The genomes of Rps. palustris, Rba. sphaeroides, and Rsp. rubrum encode a cluster of 20 genes for the synthesis of a nickel-containing uptake hydrogenase (Larimer et al., 2004), which may provide reducing power for autotrophic growth. Genes homologous to the sox genes encode cytochrome c oxidoreductases and are found in many sulfur-oxidizing organisms (Friedrich et al., 2001). Genomes of Rba. sphaeroides, Rps. palustris, and Rsp. rubrum encode sox genes, which may allow autotrophic growth by oxidation of reduced sulfur-compounds. In addition, the genomes of Rba. sphaeroides, Rps. palustris, and Rsp. rubrum encode carbon monoxide dehydrogenases and a formate dehydrogenase. This is not a surprising observation since these three species exist in diverse ecological niches where carbon monoxide and formate may be produced by either usual metabolism or/and photooxidation of atmospheric hydrocarbons (Conrad, 1988). Thus, it is possible that all three of these organisms may be able to oxidize CO and reduced sulfur compounds. VIII. Genomic Insights into the Photosynthetic Lifestyle These genome sequences have also provided new per-
703 spectives on the photosynthetic lifestyle of bacteria. As discussed in section III.B on CO2 utilization and photosynthesis, Rba. sphaeroides and several other purple photosynthetic bacteria contain a ~65 kb region of the genome, which is known as the photosynthesis gene cluster (PGC) that houses many of the genes required for photosynthetic growth. Genes within this PGC encode enzymes for bacteriochlorophyll and carotenoid biosynthesis, the structural genes for pigment-binding proteins of the light-harvesting and RC complexes, electron carriers like cytochrome c2, and conserved proteins of unknown function (Choudhary and Kaplan, 2000). This is a striking example of a large region with high conservation among the genomes of Rba. sphaeroides, Rsp. rubrum and Rps. palustris. Given what is known about the control of Rba. sphaeroides photosynthetic membrane assembly, it is not surprising that expression of the genes in this region respond to signals that regulate the synthesis of the photosynthetic apparatus (Roh et al., 2004). Genome-wide expression studies have also provided insight into new gene products or metabolic pathways, which are required for photosynthetic growth in Rba. sphaeroides (Tavano et al., 2005) and possibly other bacteria. Given the differences in gene expression that have been noted among other photosynthetic bacteria, it will be interesting to see if these patterns are conserved across genus and species lines. Another genome-based approach to identify what is needed for bacterial photosynthesis has been based on obtaining protein blueprints of whole cells or subcellular fractions that contain the photosynthetic apparatus. For example, by determining the protein abundance and composition of either whole cells or purified cell fractions (cytoplasm, cytoplasmic membrane, periplasm, outer membrane), including the purified photosynthetic membrane (chromatophores) ~2000 proteins have been identified (Fejes et al., 2003; Callister et al., 2006a,b; VerBerkmoes et al., 2006). When analyzing the protein composition of Rba. sphaeroides chromatophores of known purity, ~100 proteins have been identified that were present or enriched in this fraction, including the light-harvesting, reaction center and bioenergetic enzymes (cytochrome bc1 complex, F1F0 ATPase) that were known to be present in these photosynthetic membranes (Zeng et al., 2007). However, many chromatophore proteins are either of unknown function or are homologs of proteins that were not found in the photosynthetic apparatus. In contrast, many of the PGC proteins that are required for pigment biosynthesis were not found
704
Madhusudan Choudhary, Chris Mackenzie, Timothy J. Donohue and Samuel Kaplan
in the purified chromatophores, suggesting that these pigments are synthesized elsewhere and subsequently inserted into intact pigment-protein complexes before or during insertion into the specialized domain of the inner membrane. Indeed, most of the chromatophore proteins are encoded by genes that lie outside the PGC (Fejes et al., 2003; Zeng et al., 2007). This illustrates the types of new information that can be derived from genome-enabled analysis of solar energy utilization in these and other bacteria. Acknowledgments This work was supported by: NIH grant GM1559039 to SK; NIH grant GM075273 and DOE grant DE-FG02-05ER15653 to TJD; and the Department of Energy (DOE) grant DOE ER63232-10182200007203 to SK and TJD. References Andersson SG, Zomorodipour A, Andersson JO, Sicheritz-Ponten T, Alsmark UC, Podowski RM, Naslund AK, Eriksson AS, Winkler HH and Kurland CG (1998) The genome sequence of Rickettsia prowazekii and the origin of mitochondria. Nature 396:133–140 Anthony JR, Warczak K and Donohue TJ (2005) A transcriptional response to singlet oxygen, a toxic byproduct of photosynthesis. Proc Natl Acad Sci USA 102: 6502–6507 Bérard J, Bélanger G, Corriveau P and Gingras G (1986) Molecular cloning and sequence of the B880 holochrome gene from Rhodospirillum rubrum. J Biol Chem 262: 82–87 Bharanidharan DG, Bhargavi R, Uthanumallian K and Gautam N (2004) Correlation between nucleotide frequencies and amino acid composition in 115 bacterial species. Biochem Biophys Res Commun 315: 1097–1103 Bolt EL, Kryszak L, Zeilstra-Ryalls J, Shoolingin-Jordon PM and Warren MJ (1999) Characterization of the Rhodobacter sphaeroides 5-aminolaevulinic acid synthase isozynes, HemA and HemT, isolated from Escherichia coli. Eur J Biochem 265: 290–299 Borghese R, Crimi M, Fava L and Melandri BA (1998a) The ATP synthase atpHAGDC (F1) operon from Rhodobacter capsulatus. J Bacteriol 180: 416–421 Borghese R, Turina P, Lambertini L and Melandri BR (1998b) The atpIBEXF operon coding for the F0 sector of the ATP synthase from the purple nonsulfur photosynthetic bacterium Rhodobacter capsulatus. Arch Microbiol 170: 385–388 Braatsch S, Bernstein JR, Lessner F, Morgan J, Liao JC, Harwood CS and Beatty JT (2006) Rhodopseudomonas palustris CGA009 has two functional ppsR genes, each of which encodes a repressor of photosynthesis gene expression. Biochemistry 45:14441–14451 Callister SJ, Dominguez MA, Nicora CD, Zeng X, Tavano CL,
Kaplan S, Donohue TJ, Smith RD and Lipton MS (2006a) Application of the accurate mass and time tag approach to the proteome analysis of sub-cellular fractions obtained from Rhodobacter sphaeroides 2.4.1. Aerobic and photosynthetic cell cultures. J Proteome Res 5: 1940–1947 Callister SJ, Nicora CD, Zeng X, Roh JH, Dominguez MA, Tavano CL, Monroe ME, Kaplan S, Donohue TJ, Smith RD and Lipton MS (2006b) Comparison of aerobic and photosynthetic Rhodobacter sphaeroides 2.4.1 proteomes. J Microbiol Methods 67: 424–436 Camacho Carvajal MM, Wijfjes AH, Mulders IH, Lugtenberg BJ and Bloemberg GV (2002) Characterization of NADH dehydrogenases of Pseudomonas fluorescens WCS365 and their role in competitive root colonization. Mol Plant Microbe Interact 15: 662–671 Capdevila S, Martinez-Granero FM, Sanchez-Contreras M, Rivilla R and Martin M (2004) Analysis of Pseudomonas fluorescens F113 genes implicated in flagellar filament synthesis and their role in competitive root colonization. Microbiology 150: 3889–3897 Chandler M and Mahillon J (2002) Insertion sequences revisited. In: Craig NL, Craigie R, Gellert M and Lambowitz AM (ed) Mobile DNA II, pp 305–366. ASM Press, Washington, DC Choudhary M and Kaplan S (2000) DNA sequence analysis of the photosynthesis region of Rhodobacter sphaeroides 2.4.1. Nucl Acids Res 28: 862–867 Choudhary M, Fu YX, Mackenzie C and Kaplan S (2004) DNA sequence duplication in R. sphaeroides 2.4.1: Evidence of an ancient partnership between chromosome I and II. J Bacteriol 186: 2019–2027 Choudhary M, Zanhua X, Fu YX and Kaplan S (2007) Genome analysis of three strains of Rhodobacter sphaeroides: Evidence of rapid evolution of chromosome II. J Bacteriol 189: 1914–1921 Conrad R (1988) Biogeochemistry and ecophysiology of atmospheric CO and H2. Adv Microb Ecol 10: 231–283 Cross R L and Taiz L (1990) Gene duplication as a means for altering H+/ATP ratio during the evolution of F0F1 ATPases and synthases. FEBS Lett 259: 227–229. Dunn SD, Kellner E and Lill H (2001) Specific heterodimer formation by the cytoplasmic domains of the b and b´ subunits of cyanobacterial ATP synthase. Biochemistry 40: 187–192 Egland PG and Harwood CS (1999) BadR, a new marR family member, regulates anaerobic benzoate degradation by Rhodopseudomonas palustris in concert with AadR and Fnr family members. J Bacteriol 181: 2102–2109 Emelyanov VV (2001) Evolutionary relationship of Rickettsiae and mitochondria. FEBS Lett 501:11–18 Esser C, Ahmadinejad N, Wiegand C, Rotte C, Sebastiani F, Gelius-Dietrich G, Henze K, Kretschmann E, Richly E, Leister D, Bryant D, Steel MA, Lockhart PJ, Penny D and Martin W (2004) A genome phylogeny for mitochondria among alphaproteobacteria and a predominantly eubacterial ancestry of yeast nuclear genes. Mol Biol Evol 21:1643–1660 Fanica-Gaignier M and Clement-Metral J (1973) Cellular compartmentation of two species of 5-aminolevulinic acid synthetase in a facultative photohetrotrophic bacterium (Rps. sphaeroides Y). Biochem Biophys Res Comm 55: 610–615 Fejes AP, Yi EC, Goodlett DR and Beatty JT (2003) Shotgun proteomic analysis of a chromatophore-enriched preparation from the purple phototrophic bacterium Rhodopseudomonas
Chapter 34
Purple Bacterial Genomics
palustris. Photosynth Res 78: 195–203 Fitch WM (2000) Homology: A personal view on some of the problems. Trends Genet 16: 227–231 Friedrich CG, Rother D, Bardischewsky F, Quentmeier A and Fischer J (2001) Oxidation of reduced inorganic sulfur compounds by bacteria: Emergence of a common mechanism? Appl Environ Microbiol 67: 2873–2882 Foerstner, KU, von Mering C, Hooper, SD and Bork P (2005) Environments shape the nucleotide composition of genomes. EMBO Reports 6: 1208–1213 Garrett ES, Perlegas D and Wozniak DJ (1999) Negative control of flagellum synthesis in Pseudomonas aeruginosa is modulated by the alternative sigma factor AlgT (AlgU). J Bacteriol 181: 7401–7404 Gest H (1972) Energy conservation and generation of reducing power in bacterial photosynthesis. Adv Microb Physiol 7: 243–282 Gevers D, Vandepoele K, Simillion C and Deeper YV (2004) Gene duplication and biased functional retention of paralogs in bacterial genomes. Trends Microbiol 12: 148–154 Giraud E, Hannibal L, Fardoux J, Verméglio A and Dreyfus B (2000) Effects of Braydirhizobium photosynthesis on stem nodulation of Aeschynomene sensitive. Proc Natl Acad Sci USA 97: 14795–14800 Glass JI, Lefkowitz, EJ, Glass JS, Heiner CR, Chen EY and Cassell GH (2000) The complete sequence of the mucosal pathogen Ureaplasma urealyticum. Nature 407; 757–762 Grammel H, Gilles ET and Ghosh R (2003) Microaerophilic cooperation of reductive and oxidative pathways allows maximal photosynthetic membrane biosynthesis in Rhodospirillum rubrum. Appl Envir Microbiol 69: 6577–6586 Gray MW, Burger G and Lang BF (1999) Mitochondrial evolution. Science 283:1476–1481 Green HA and Donohue TJ (2006) Activity of Rhodobacter sphaeroides RpoHII, a second member of the heat shock sigma factor family. J Bacteriol 188: 5712–5721 Han C, Shiga Y, Tobe T, Sasakawa C and Ohtsubo E (2001) Structural and functional characterization of IS679 and IS66-family elements. J Bacteriol 183: 4296–4304 Hartigan N, Tharia HA, Sweeney F, Lawless AM and Papiz MZ (2002) The 7.5-Å electron density and spectroscopic properties of a novel low-light B800 LH2 from Rhodopseudomonas palustris. Biophys J 82: 963–977 Hessner MJ, Wejksnora PJ and Collins ML (1991) Construction, characterization, and complementation of Rhodospirillum rubrum puf region mutants. J Bacteriol 173:5712–5722 Hillerich B and Westpheling J (2006) A new GntR family transcriptional regulator in Streptomyces coelicolor is required for morphogenesis and antibiotic production and controls transcription of an ABC transporter in response to carbon source. J Bacteriol 188: 7477–7487 Hirst J (2005) Energy transduction by respiratory complex I-an evaluation of current knowledge. Biochem Soc Trans 33: 525–529 Jain R, Rivera MC, Moore JE and Lake JA (2002) Horizontal gene transfer in microbial genome evolution. Theor Popul Biol 61:489–495 Johnson CH and Golden SS (1999) Circadian programs in cyanobacteria: Adaptiveness and mechanism. Annu Rev Microbiol 53: 389–409 Kanbe M, Yagasaki J, Zehner S, Gottfert M and Aizawa S (2007)
705 Characterization of two sets of subpolar flagella in Bradyrhizobium japonicum. J Bacteriol 189: 1083–1089 Karls RK, Brooks J, Rossmeissl P, Luedke J and Donohue TJ (1998) Metabolic roles of a Rhodobacter sphaeroides member of the σ32 family. J Bacteriol 180: 10–19 Kieboom J, Dennis JJ, de Bont AMJ and Zylstra GJ (1998) Identification and molecular characterization of an efflux pump involved in Pseudomonas putida S12 solvent tolerance. J Biol Chem 273: 85–91 Kovacikova G, Lin W and Skorupski K (2005) Dual regulation of genes involved in acetoin biosynthesis and motility/biofilm formation by the virulence activator AphA and the acetateresponsive LysR-type regulator AlsR in Vibrio cholerae. Mol Microbiol 57: 420–433 Kuhl SA, David WN and Yoch DC (1983) Characterization of Rhodospirillum rubrum plasmid: Loss of photosynthetic growth in plasmidless strains. J Bacteriol 156: 737–742 Larimer FW, Chain P, Hauser L, Lamerdin J, Malfatti S, Do L, Land ML, Pelletier DA, Beatty JT, Lang AS, Tabita FR, Gibson JL, Hanson TE, Bobst C, Torresy Torres JL, Peres C, Harrison FH, Gibson J and Harwood CS (2004) Complete genome sequence of the metabolically versatile photosynthetic bacterium Rhodopseudomonas palutris. Nature Biotechnol 22: 55–61 Mackenzie C, Choudhary M, Larimer FW, Predki PF, Stilwagen S, Armitage JP, Barber RD, Donohue TJ, Hosler JP, Newman JE, Shapleigh JP, Sockett RE, Zeilstra-Ryalls J and Kaplan S (2001) The home stretch, a first analysis of the nearly completed genome of Rhodobacter sphaeroides 2.4.1. Photosynth Res 70: 19–41 Mackenzie C, Kaplan S and Choudhary, M (2004) Multiple chromosomes. In: Miller RV and Day MJ (ed) Microbial Evolution, pp 82–101. ASM Press, Washington, DC Mackenzie C, Eraso JM, Choudhary M, Roh JH, Zeng X, Brucella P, Puskás A and Kaplan S (2007) Post-genomic adventures with Rhodobacter sphaeroides. Annu Rev Microbiol 61: 283–307 Mahillion J, Leonard C and Chandler M (1999) IS elements as constituents of bacterial genomes. Res Microbiol Mol Biol Rev 150: 675–687 Martinez-Bueno M, Molina-Henares AJ, Praeja E, Ramos JL and Tobes R (2004) BactTregulators: A database of transcriptional regulators in bacteria and archaea. Bioinformatics 20: 2787–2791 Meijer WG and Tabita R (1992) Isolation and characterization of the nifUSVW-rpoN gene cluster from Rhodobacter sphaeroides. J Bacteriol 174: 3855–3866 Moran NA (2002) Microbial minimalism: Genome reduction in bacterial pathogens. Cell 108: 583–586 Moran MA, Buchan A, Gonzalez JM, Heidelberg JF, Whitman WB, Kiene RP, Henriksen JR, King GM, Belas R, Fuqua C, Brinkac L, Lewis M, Johri S, Weaver B, Pai G, Eisen JA, Rahe E, Sheldon WM, Ye W, Miller TR, Carlton J, Rasko DA, Paulsen IT, Ren Q, Daugherty SC, Deboy RT, Dodson RJ, Durkin AS, Madupu R, Nelson WC, Sullivan SA, Rosovitz MJ, Haft DH, Selengut J and Ward N (2004) Genome sequence of Silicibacter pomeroyi reveals adaptations to the marine environment. Nature 432: 910–913 Neidle E and Kaplan S (1992) Rhodobacter sphaeroides rdxA, a homolog of Rhizobium meliloti fixG, encodes a membrane protein which may bind cytoplasmic (4Fe-4S) clusters. J Bacteriol 74: 6444–6454 Neidle E and Kaplan S (1993a) Expression of Rhodobacter
706
Madhusudan Choudhary, Chris Mackenzie, Timothy J. Donohue and Samuel Kaplan
sphaeroides hemA and hemT genes encoding two aminolevulinic acid synthase isozymes. J Bacteriol 175: 2292–2303 Neidle E and Kaplan S (1993b) 5-aminolevulinic acid availability and control of spectral complex formation in HemA and HemT mutants of Rhodobacter sphaeroides. J Bacteriol 175: 2304–2313 Nereng K and Kaplan S (1999) Genomic complexity among strains of the facultative phototrophic bacterium, Rhodobacter sphaeroides. J Bacteriol 181: 1684–1688 Pappas CT, Sram J, Moskvin OV, Ivanov PS, Mackenzie RC, Choudhary M, Land ML, Larimer FW, Kaplan S and Gomelsky M (2004) Construction and validation of the Rhodobacter sphaeroides 2.4.1 DNA microarray: Transcriptome flexibility at diverse growth modes. J Bacteriol 186: 4748–4758. Pareja E, Pareja-Tobes P, Manrique M, Pareja-Tobes E, Bonal J and Tobes R (2006) Extra Train: A database of extragenic regions and transcriptional information in prokaryotic organisms. BMC Microbiol 6: 29–38 Poggio S, Abreu-Goodger C, Fabela S, Osorio A, Dreyfus G, Vinuesa P and Camarena L. (2007) A complete set of flagellar genes acquired by horizontal transfer coexists with the endogenous flagellar system in Rhodobacter sphaeroides. J Bacteriol 189: 3208–3216 Ramos JL (2002) AraC-XylS database: a family of positive transcriptional regulators in bacteria. Nucl Acid Res 30: 318–321 Ramos JL, Martinez-Bueno M, Molina-Henares AJ, Teran W, Watanabe K, Zhang X, Gallegoes MT, Brennan R and Tobes R (2005) The TetR family of transcriptional repressors. Microbial Mol Biol Rev 69: 326–356 Reslewic S, Zhou S, Place M, Zhang Y, Briska A, Goldstein S, Churas C, Runnheim R, Forrest D, Lim A, Lapidus A, Han CS, Roberts GP and Schwartz DC (2005) Whole-genome shotgun optical mapping of Rhodospirillum rubrum. Appl and Environ Microbiol 71: 5511–5522 Rocha EP and Danchin A (2002) Base composition bias might result from competition for metabolic resources. Trend Genet 18: 291–294 Roh JH, Smith WE and Kaplan S (2004) Effects of oxygen and light intensity on transcriptome expression in Rhodobacter sphaeroides 2.4.1. J Biol Chem 279: 9146–9155 Sanchez-Contreras M, Martin M, Villacieros M, O’Gara F, Bonilla I and Rivilla R (2002) Phenotypic selection and phase variation occur during alfalfa root colonization by Pseudomonas fluorescens F113. J Bacteriol 184: 1587–1596 Suwanto A and Kaplan S (1989a) Physical and genetic mapping of the Rhodobacter sphaeroides 2.4.1 genome: Genome size, fragment identification, and gene localization. J Bacteriol 171: 5840–5849 Suwanto A and Kaplan S (1989b) Physical and genetic mapping of the Rhodobacter sphaeroides 2.4.1 genome: Presence of two unique circular chromosomes. J Bacteriol 171: 5850–5859
Tavano CL, Podevels A and Donohue TJ (2005) Gene products required to recycle reducing power produced under photosynthetic conditions. J Bacteriol 187: 5249–5258 Tharia HA, Nightingale TD, Papiz MZ and AM Lawless AM (1999) Characterisation of hydrophobic peptides by RP-HPLC from different spectral forms of LH2 isolated from Rhodopseudomonas palustris. Photosynth Res 61: 157–167 Turina P, Rebecchi A, D’Alessandro M, Anefors S and Melandri BA (2006) Modulation of proton pumping efficiency in bacterial ATP synthases. Biochim Biophys Acta 1757: 320–325 Van Niel CB (1944) The culture, general physiology, morphology, and classification of the non-sulfur purple and brown bacteria. Bacteriol Rev 8: 1–118 VerBerkmoes NC, Shah MB, Lankford PK, Pelletier DA, Strader MB, Tabb DL, McDonald WH, Barton JW, Hurst GB, Hauser L, Davison BH, Beatty JT, Harwood CS, Tabita FR, Hettich RL and Larimer FW (2006) Determination and comparison of the baseline proteomes of the versatile microbe Rhodopseudomonas palustris under its major metabolic states. J Proteome Res 5: 287–298 Woese, CR (1987) Bacterial evolution. Microbiol Rev 51: 221–271 Woese C (1998) The universal ancestor. Proc Natl Acad Sci U S A.95: 6854–6859 Woese CR, Stachebrandt E, Weisburg WG, Paster BJ, Madigan MT, Fowler CRM, Hahn CM, Blanz P, Gupta R, Nealson KH and Fox GE (1984) The phylogeny of the purple bacteria: the α-subdivision. Syst Appl Microbiol 5: 315–326 Xiong J, Fischer WM, Inoue K, Nakahara M and Bauer CE (2000) Molecular evidence for the early evolution of photosynthesis. Science 289:1724–1730 Zeilstra-Ryalls J and Kaplan S (1995) Regulation of 5-aminolevulinic acid synthase in Rhodobacter sphaeroides 2.4.1: The genetic basis of mutant H-5 auxotrophy. J Bacteriol 177: 2760–2768 Zeilstra-Ryalls J, Gomelsky M, Eraso JM, Yeliseev A, O’Gara J and Kaplan S (1998) Control of photosystem formation in Rhodobacter sphaeroides. J Bacteriol 180: 2801–1809 Zeng X, Choudhary M and Kaplan S (2003) A second and unusual pucBA operon of Rhodobacter sphaeroides 2.4.1: Genetics and function of the encoded polypeptides. J Bacteriol 185: 6171–6184 Zeng X, Roh JH, Callister SJ, Tavano CL, Donohue TJ, Lipton ML and Kaplan S (2007) Proteomic characterization of the Rhodobacter sphaeroides photosynthetic membrane. J Bacteriol 189: 7464–7474 Zhou SE, Kile AK, Severin J, Forrest D, Runnheim R, Churas C, Anantharaman TS, Hickman JW, Mackenzie C, Choudhary M, Donohue TJ, Kaplan S and Schwartz DC (2003) Whole-genome shotgun optical mapping of Rhodobacter sphaeroides 2.4.1 and its use for whole-genome shotgun sequence assembly. Genome Res 13: 2142–2151
Chapter 35 Regulation of Gene Expression in Response to Oxygen Tension Carl E. Bauer1*, Aaron Setterdahl1, Jiang Wu1 and Brigitte R. Robinson2 1
Departments of Biology and 2Chemistry, Indiana University, Bloomington, IN 47405-7170, U.S.A.
Summary ............................................................................................................................................................... 707 I. Introduction..................................................................................................................................................... 707 II. RegB/RegA Two-Component Signal Transduction System ........................................................................... 708 A. The Sensor Kinase RegB ................................................................................................................. 712 1. Kinase Activity ......................................................................................................................... 713 2. Phosphatase Activity ............................................................................................................... 713 3. Redox Sensing ........................................................................................................................ 713 B. The Response Regulator RegA ....................................................................................................... 714 1. Effect of Phosphorylation ........................................................................................................ 715 2. DNA Binding Sites................................................................................................................... 715 C. Global Significance .......................................................................................................................... 716 III. Aerobic repression by CrtJ ............................................................................................................................. 716 A. Redox and Light Regulation ............................................................................................................. 717 B. Phylogenetic Analysis ..................................................................................................................... 719 IV. Regulation by Fnr ........................................................................................................................................... 721 Acknowledgments ................................................................................................................................................. 722 References ............................................................................................................................................................ 722
Summary Purple photosynthetic bacteria control numerous energy-generating and energy-utilizing processes in response to alterations in cellular redox, which is affected by environmental oxygen tension. The list of redox-regulated events includes synthesis of the pigmented and cytochrome components of the photosystem, enzymes for fixation of carbon and nitrogen, the synthesis of several terminal respiratory electron transport complexes, and synthesis of the energy-generating hydrogenase complex. Regulating synthesis of these components involves several well-characterized transcription factors including the sensor kinase RegB and its cognate response regulator RegA. Other redox-responding regulators include CrtJ and Fnr. Mechanisms of redox sensing by these transcription factors are discussed. I. Introduction Many metabolic processes such as photosynthesis, respiration, carbon and nitrogen fixation are highly regulated in response to alterations in environmental
oxygen tension (redox). Some processes, such as photosynthesis, appear to be predominately regulated by oxygen tension (Cohen-Bazire et al., 1957), while other processes, such as nitrogen and carbon fixation, are regulated by the availability of nitrogen or carbon,
*Author for correspondence, email: [email protected] C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 707–725. © 2009 Springer Science + Business Media B.V.
708
Carl E. Bauer, Aaron Setterdahl, Jiang Wu and Brigitte R. Robinson
respectively, as well as by alteration in cellular redox (Madigan, 1995; Tabita, 1995; Bowman et al., 1999). To a large extent, redox control of many different processes occurs by controlling gene expression either directly in response to the presence or absence of molecular oxygen or in response to changes in the overall redox state of the cell. This chapter is focused on both oxygen- and redox-responding master regulators of gene expression in purple bacteria. Studies on redox-regulation of gene expression in purple photosynthetic bacteria initially focused on photosynthesis because a change in pigment biosynthesis provided a simple visible screen for mutants that either failed to synthesize a photosystem under anaerobic conditions (Sganga and Bauer 1992; Mosley et al., 1994) or aberrantly synthesized photopigments aerobically (Penfold and Pemberton, 1994; Du et al., 1998). These studies resulted in the identification of several conserved transcription factors that are responsible for controlling synthesis of the photosystem in Rhodobacter (Rba.) capsulatus and Rba. sphaeroides. The first identified redox regulators were the RegB/RegA two-component transduction system from Rba. capsulatus (Sganga and Bauer 1992; Mosley et al., 1994). This twocomponent system induces the synthesis of nearly all of the components of the bacterial photosystem comprised of the light-harvesting, reaction center and cytochrome apoproteins (Elsen et al., 2004), as well as bacteriochlorophyll, carotenoids and heme (Smart et al., 2004; Willet et al., 2007). In addition to RegB-RegA, there is a redox-responding aerobic repressor of photopigment, lightharvesting and cytochrome biosynthesis called CrtJ (designated as PpsR in other species) that has also been extensively characterized. Like RegA/RegB, CrtJ controls the synthesis of individual components of the photosystem such as the tetrapyrroles bacteriochlorophyll (Penfold and Pemberton, 1994; Ponnampalam et al., 1995) and heme (Smart et al., 2004), as well as apoproteins for cytochromes (Swem and Bauer, 2002) and the light-harvesting complex (Ponnampalam et al., 1995). Finally, a redox-responding homolog of the Escherichia (E.) coli transcriptional regulator FNR (fumarate-nitrate reduction) is also involved in controlling photosynthetic and respiratory growth in Rba. capsulatus (Swem and Bauer, Abbreviations: B. – Bradyrhizobium; bp – base pair(s); E. – Escherichia; NMR – nuclear magnetic resonance; P. – Pseudomonas; PAS – Per-ARNT-Sim ; Rba. – Rhodobacter; Rps. – Rhodopseudomonas
2002; Smart et al., 2004) and in Rba. sphaeroides (Zeilstra-Ryalls and Kaplan, 1995; Mouncey and Kaplan, 1997; Zeilstra-Ryalls and Kaplan, 1998). This chapter is focused on biochemical analysis of redox-responding transcription factors that have been characterized in purple photosynthetic bacteria. II. RegB/RegA Two-Component Signal Transduction System RegA and RegB were initially identified in genetic screens designed to isolate mutants defective in oxygen control of photosystem synthesis in Rba. capsulatus (Sganga and Bauer, 1992; Mosley et al., 1994). Mutants defective in anaerobic synthesis of the photosystem were isolated that exhibited significantly reduced expression of the puh, puf and puc operons that encode apoproteins for the light-harvesting 1, light-harvesting 2 and reaction center complexes (Sganga and Bauer, 1992; Mosley et al., 1994). Sequence analysis demonstrated that these strains contained mutations in two linked genes, regB and regA, with RegB exhibiting homology to histidine protein kinase (McCleary and Stocks, 1994; Mosley et al., 1994), and RegA exhibiting homology to a DNA-binding response regulator (Parkinson and Kofoid, 1992; Sganga and Bauer, 1992; McCleary and Stocks, 1994). Subsequent to the discovery of RegB and RegA from Rba. capsulatus, homologous two-component systems were found and genetically characterized in many other species such as: RegB/ RegA homologs in Rba. sphaeroides (called PrrB/ PrrA) (Eraso and Kaplan, 1994, 1995; Phillips-Jones and Hunter, 1994); RegS/RegR from Bradyrhizobium (B.) japonicum (Bauer et al., 1998); ActS/ActR from Sinorhizobium meliloti (Tiwari et al., 1996); RoxS/ RoxR from Pseudomonas (P.) aeruginosa (Comolli and Donohue, 2002); and RegB/RegA from Rhodovulum sulfidophilum and Roseobacter denitrificans (Masuda et al., 1999). Genome sequence studies have also identified RegA and RegB homologs in many other photosynthetic as well as non-photosynthetic α- and γ-proteobacterial species (>90 species) with a partial list of these homologs present in Table 1. Genetic shuttling studies demonstrated that RegB and RegA homologs from different species are interchangeable in vitro and in vivo. This indicates that phosphotransfer can be observed between different RegB and RegA homologs (Comolli and Donohue, 2002; Emmerich et al., 2000), and that some RegA
Organism
RegB H-Box
YES Q9L906 YES Q3J6C1 YES A0VNJ9 YES A3S9L0 YES A3VGE1 YES Q5LLQ5 YES O82866 YES O82869 YES Q2CJX1 YES A3KAZ2 YESA3W499 YES Q8KYV6 YES A3V2H4 YES Q28JY5 YES A1B5R9 YES A1H0J9 YES Q7X352 YES A0P3M9 YES Q98C40 YES Q11B16 YES Q8G321 YES Q2YP02 YES Q26N86 YES Q8YER2 YES A5P1X6 YES O86124 YES Q1YF90 YES Q3SWG3 YES Q92TA1 YES Q6NCA0 YES Q2KE47 YES Q52912 YES Q1QRL7 YES Q9ABH9 YES Q1MNA6 YES Q0FYD3 YES Q0BWS2 YES Q8UJ81
58% 57% 55% 55% 54% 54% 55% 53% 53% 53% 53% 52% 51% 47% 39% 39% 39% 38% 38% 37% 37% 38% 37% 37% 36% 35% 35% 36% 36% 36% 36% 34% 38% 36% 36% 36% 36%
YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES
Q-binding site YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES
Redox-active cystein YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES
RegA
RegA Identity Acid-Box
YES P42508 YES Q53228 YES A0VNJ7 YES A3S9K8 YES A3VGE3 YES Q1GCP6 YES O82868 YES Q9ZNM4 YES Q2CJX3 YES A3KAZ4 YES Q0FV45 YES Q8KYV8 YES A3V2H5 YES Q28JX7 YES Q3PEU7 YES A1H0J8 YES Q7X351 YES A0P3M8 YES Q98C39 YES Q11B17 YES Q8G319 YES Q57FN7 YES Q26N85 YES Q8YER6 YES A5P1X5 YES Q89VZ0 YES Q1YF91 YES Q3SWG2 YES Q52913 YES Q6NCA1 YES Q2KE48 YES Q0MH93 YES Q1QRL6 YES Q9ABI0 YES Q1MNA7 YES Q0FYD2 YES Q0BWS1 YES Q8UJ82
83% 82% 79% 80% 82% 83% 81% 84% 82% 82% 78% 83% 81% 77% 70% 66% 71% 69% 70% 67% 67% 69% 67% YES 69% 68% 70% 70% 70% 69% 70% 70% 68% 69% 68% 69% 68%
YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES
DNA-Binding domain YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES
709
Identity
Redox Control of Gene Expression
Rhodobacter capsulatus Rhodobacter sphaeroides Dinoroseobacter shibae Sulfitobacter sp. Rhodobacterales bacterium Silicibacter pomeroyi Rhodovulum sulfidophilum Roseobacter denitrificans Oceanicola granulosus Sagittula stellata Roseovarius sp. uncultured proteobacterium Loktanella vestfoldensis Jannaschia sp. Paracoccus denitrificans Parvibaculum lavamentivorans Uncultured Acidobacteria bacterium Stappia aggregata Rhizobium loti Mesorhizobium sp. Brucella suis Brucella abortus Xanthobacter sp. Brucella melitensis Methylobacterium sp. Bradyrhizobium japonicum Aurantimonas sp. Nitrobacter winogradskyi Rhizobium meliloti Rhodopseudomonas palustris Rhizobium etli Sinorhizobium medicae Nitrobacter hamburgensis Caulobacter crescentus Rhizobium leguminosarum Fulvimarina pelagi Hyphomonas neptunium Agrobacterium tumefaciens
RegB
Chapter 35
Table 1. RegB and RegA homologs identified by similarity based search
710
Table 1. Continued Organism Identity
RegB H-Box
YES Q0AKP8 YES A3VQ48 YES A3UI09 YES Q1CZ19 YES Q2YD54 YES A4A0N7 YES Q4ZNL1 YES Q1QWH9 YES Q3KI43 YES Q4IY97 YES Q1GZ69 YES Q2II98 YES Q2XJG8 YES Q2BQZ7 YES Q82V00 YES Q9HVS7 YES Q4FNH0 YES Q08V74 YES Q1V1K9 YES Q1IEE3 YES Q1VI62 YES Q47FP6 YES Q1GT06 YES Q1N6Y2 YES Q0EZG0 NO2 NO2 NO2 NO2 NO2 NO2 NO2 NO2 NO2 NO2 NO2 NO2 NO2
36% 36% 36% 31% 27% 27% 30% 26% 29% 28% 27% 27% 28% 26% 25% 28% 28% 28% 29% 29% 28% 27% 26% 24% 24%
YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES
Q-binding site YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES NPF>NPL YES YES YES
Redox-active cystein YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES
RegA YES Q0AKP9 YES A3VQ47 YES A3UI10 NO1 YES Q2YD55 NO1 YES Q87WJ3 YES Q1QWI0 YES Q3KI42 YES Q4IY98 YES Q1GZ68 YES Q2II97 YES Q88PG2 YES Q2BQZ8 YES Q820M1 YES Q9HVS8 YES Q4FP64 YES Q08V73 YES Q1V0V8 YES Q1IEE2 YES Q1VJU0 YES Q47FP7 YES Q1GT07 NO1 NO1 YES A4BPC8 YES Q3N6K1 YES Q3SFG7 YES Q476X9 YES A6G990 YES Q1LS55 YES Q0A850 YES Q21GP2 YES Q3IBV7 YES Q3NW53 YES Q44X66 YES Q7NZM5 YES Q4BNG8
RegA Identity Acid-Box 67% 68% 65%
YES YES YES
DNA-Binding domain YES YES YES
50%
YES
YES
52% 47% 52% 50% 51% 47% 50% 49% 50% 50% 67% 51% 67% 50% 65% 44% 45%
YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES
YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES
50% 49% 49% 48% 48% 46% 45% 39% 42% 42% 43% 47% 43%
YES YES YES YES YES YES YES YES YES YES YES YES YES
YES YES YES YES YES YES YES YES YES YES YES YES YES
Carl E. Bauer, Aaron Setterdahl, Jiang Wu and Brigitte R. Robinson
Maricaulis maris Parvularcula bermudensis Oceanicaulis alexandrii Myxococcus xanthus Nitrosospira multiformis Blastopirellula marina Pseudomonas syringae Chromohalobacter salexigens Pseudomonas fluorescens Azotobacter vinelandii Methylobacillus flagellatus Anaeromyxobacter dehalogenans Pseudomonas putida Oceanospirillum sp Nitrosomonas europaea Pseudomonas aeruginosa Pelagibacter ubique Stigmatella aurantiaca Candidatus Pelagibacter ubique Pseudomonas entomophila Psychroflexus torquis Dechloromonas aromatica Sphingopyxis alaskensis Sphingomonas sp. Mariprofundus ferrooxydans Nitrococcus mobilis Nitrosomonas eutropha Thiobacillus denitrificans Alcaligenes eutrophus Plesiocystis pacifica Ralstonia metallidurans Alkalilimnicola ehrlichei Saccharophagus degradans Pseudoalteromonas haloplanktis Shewanella frigidimarina Burkholderia cenocepacia Chromobacterium violaceum Burkholderia vietnamiensis
RegB
Organism RegB
Identity
Q-binding site
Redox-active cystein
RegA
RegA Identity Acid-Box
YES Q8Y3E0 YES Q3JXA5 YES Q2T278 YES Q62F01 YES Q3FK41 YES Q47UR4 YES Q602T5 YES Q5QWI7 YES Q3QI67 YES Q3P3L2 YES Q2ZVU1 YES Q8E9U1 YES Q3Q7V3 YES Q2SM70 YES Q36SD4 YES Q7UHV2 YES Q7VUY1 YES Q7WMV8 YES Q7WBD8 YES Q2KWQ0
45% 43% 43% 51% 43% 42% 46% 38% 42% 37% 38% 38% 37% 42% 38% 40% 42% 42% 42% 42%
YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES
DNA-Binding domain YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES
Redox Control of Gene Expression
Ralstonia solanacearum NO2 Burkholderia pseudomallei NO2 Burkholderia thailandensis NO2 Burkholderia mallei NO2 Burkholderia ambifaria NO2 Colwellia psychrerythraea NO2 Methylococcus capsulatus NO2 Idiomarina loihiensis NO2 Shewanella amazonensis NO2 Shewanella denitrificans NO2 Shewanella putrefaciens NO2 Shewanella oneidensis NO2 Shewanella baltica NO2 Hahella chejuensis NO2 Marinobacter aquaeolei NO2 Rhodopirellula baltica NO2 Bordetella pertussis NO2 Bordetella bronchiseptica NO2 Bordetella parapertussis NO2 Bordetella avium NO2 1. No homologs to RegA were found in Blast2 complete database 2. No homologs to RegB were found in Blast2 complete database
RegB H-Box
Chapter 35
Table 1. Continued
711
712
Carl E. Bauer, Aaron Setterdahl, Jiang Wu and Brigitte R. Robinson
homologs can bind to promoters and regulate gene transcription in other species (Emmerich et al., 2000; Comolli and Donohue, 2002). It is now well established that RegB and RegA constitute a highly conserved global regulatory system that provides an overlying layer of redox-control on a variety of energy-generating and energy-utilizing biological processes in many diverse species of bacteria (Fig. 1A) (Elsen et al., 2004). Furthermore, recent genome array experiments indicate that nearly 20% of Rba. sphaeroides genes are part of the RegB-RegA regulon (Kaplan et al., 2005). Predictive hierarchical clustering analysis of the Rba. sphaeroides genome, for putative RegA binding sites, also indicates the presence of a large number of putative RegA binding sites well beyond that of photosynthesis and respiratory genes (Mao et al., 2005). These results
indicate that RegB-RegA provide redox control of many cellular processes. A. The Sensor Kinase RegB The Rba. capsulatus regB gene encodes a 50.1 kDa histidine protein kinase composed of 460 amino acids. The N-terminal region comprises a transmembrane domain containing six hydrophobic membrane-spanning regions which is followed by a C-terminal cytoplasmic ‘transmitter’ domain (Tiwari et al., 1996; Ouchane and Kaplan, 1999; Chen et al., 2000). A recent study identified the ubiquinone pool as a redox signal for RegB with a highly conserved quinone binding site (Table 1) found to be located in the transmembrane domain thereby indicating that this region plays a role in redox-sensing (Swem et
Fig. 1. A) Biological processes regulated by the RegB/RegA system through a direct interaction with the ubiquinone pool. B) Oxidized redox sensors of RegB include direct interaction with oxidized ubiquinones (UQ) that inhibits kinase activity, the formation of a disulfide bond (S-S) that stimulates tetramerization, and formation of a putative cysteine sulfonate (SOH).
Chapter 35
Redox Control of Gene Expression
al., 2006). The transmembrane domain is followed by a cytosolic domain that contains an H-box site of autophosphorylation (His225), and the N, G1, F, and G2 boxes that define the nucleotide binding cleft (McCleary and Stocks, 1994; Mosley et al., 1994; Ouchane and Kaplan, 1999). The cytosolic domain contains a conserved redox-active cysteine located in a conserved ‘redox box’ just downstream of the Hbox (Fig. 1B; Table 1) (Swem et al., 2003). This Cys is capable of regulating the activity of RegB through forming an intermolecular disulfide bond in response to the redox state (Swem et al., 2003). 1. Kinase Activity Initial kinase assays demonstrated that a His-tagged cytosolic domain of Rba. capsulatus RegB was capable of autophosphorylation in vitro as well as phosphotransfer to its cognate response regulator, RegA (Inoue et al., 1995; Bird et al., 1999). As the kinetics were not affected by ATP concentration, it was suggested that the rate-limiting step of RegB autophosphorylation was phosphotransfer from ATP to the histidine residue, rather than binding of ATP (Bird et al., 1999). Similar results have been obtained with cytosolic versions of RegB homologs from Rba. sphaeroides and from B. japonicum (Emmerich et al., 1999; Comolli et al., 2002). It was shown that phosphorylated full-length RegB exhibits decreased stability of phosphate compared to the truncated version of RegB (half-life of about 34 min versus 5.5 to 6 h), pointing to a role of the transmembrane domain in regulating the phosphorylation state of RegB (Masuda et al., 1999; Potter et al., 2002). The first demonstration of phosphotransfer from Rba. capsulatus RegB~P to RegA was reported by Inoue et al., (1995). Phosphotransfer studies showed that the transfer of phosphate is rapid (<1 min) from the cytosolic domain of RegB to RegA in vitro (Bird et al., 1999). Since autophosphorylation of RegB is slow and phosphotransfer from RegB~P to RegA is very rapid, it appears that autophosphorylation of RegB is a rate-limiting step in controlling the phosphorylation level of RegA (Bird et al., 1999). No back-transfer of phosphate from RegA~P to RegB has been observed (Comolli and Donohue, 2002; Potter et al., 2002). 2. Phosphatase Activity Histidine kinases can modulate the level of phos-
713 phorylation of a cognate phosphorylated response regulator not only through phosphorylation, but also by exerting phosphatase activity on the receiver’s Asp~P. Dephosphorylation of RegA~P in vitro was shown to be dependent on the amount of unphosphorylated RegB, which indicates that RegB can dephosphorylate RegA (Bird et al., 1999). Results from the B. japonicum RegB homolog (RegS) (Emmerich et al., 1999), and both full-length version and truncated forms of RegB from Rba. sphaeroides (Potter et al., 2002), showed that the presence of RegB resulted in >16-fold reduction in the stability of the phosphate on RegA~P (Comolli and Donohue, 2002). Because both truncated and full-length RegB exhibit the same phosphatase activity, it is assumed that modulation of phosphatase activity does not require the N-terminal region of RegB. However, further studies are required to determine whether phosphatase activity is redox-regulated. 3. Redox Sensing In vivo studies indicated that expression of many RegB-RegA-regulated photosynthesis genes were inhibited by growth under aerobic conditions (Bauer et al., 1988; Sganga and Bauer, 1992; Mosely et al., 1994). It was therefore presumed that the kinase activity of RegB was directly inhibited by oxygen (Mosely et al., 1994). However, this possibility was subsequently excluded because Rba. capsulatus is fully capable of inducing pigment biosynthesis under chemoautotrophic growth conditions in the presence of oxygen, hydrogen and carbon dioxide (Madigan and Gest, 1979). Another signal proposed to regulate RegB was the redox state of the respiratory electron transport chain (O’Gara and Kaplan, 1997; Eraso and Kaplan, 2000; Oh and Kaplan, 2000; Roh and Kaplan, 2000). This conclusion was based on the observation that mutations of cytochrome cbb3 oxidase in Rba. sphaeroides and Rba. capsulatus lead to elevated aerobic expression of RegB-RegA-regulated genes (Eraso and Kaplan, 1994; Buggy and Bauer, 1995). It was suggested that cytochrome cbb3 oxidase generates an ‘inhibitory signal’ that represses the RegB/RegA two-component system. Recently, Swem et al., (2006) demonstrated that the redox state of the ubiquinone pool, which is known to be affected by respiration and photosynthesis, is a direct signal controlling RegB autophosphorylation. In this study, autophosphorylation activity of full-length RegB
714
Carl E. Bauer, Aaron Setterdahl, Jiang Wu and Brigitte R. Robinson
was significantly inhibited in vitro by the presence of oxidized, but not reduced, ubiquinone. A highly conserved ubiquinone-binding site was identified in a short periplasmic loop between transmembrane helices three and four with the use of 14C-azidoquinone photo-affinity cross-linking (Swem et al., 2006). There is a heptapeptide sequence of GGXXNPF that is 100% conserved among all known RegB homologs in the same region (Table 1) (Swem et al., 2006). It has been proposed that oxidized ubiquinone binds to this heptapeptide through π-π interactions between its para-hydroxybenzoate ring and the aromatic side group of the conserved phenylalanine (Phe112), as well as hydrogen bond interaction between ubiquinone and the conserved asparagine (Asp111) (Swem et al., 2006). The binding of oxidized ubiquinone presumably results in allosteric modification of RegB that leads to inhibition of autophosphorylation. When the ubiquinone pool is shifted to a protonated form under anaerobic conditions, a hydrogen bond between asparagine and ubiquinone could be disrupted triggering a structural change that allows autophosphorylation of RegB (Fig. 1B). Subsequent in vivo mutational study on Phe112 found elevated aerobic synthesis of the photosystem, confirming that this amino acid is involved in the sensing of the redox state of the ubiquinone pool and regulating RegB kinase activity (Swem et al., 2006). There is a large ubiquinone pool in membranes from purple photosynthetic bacteria that provides electron carriers for photosynthesis and respiration (Bolton, 1978). The redox state of the ubiquinone pool varies in response to changes in oxygen tension, being predominantly oxidized under aerobic conditions and predominantly reduced under anaerobic conditions (Parson, 1978). Ubiquinone is a good redox signal given that the redox state of ubiquinones reflects changes in the redox state of cells in general. Ubiquinone as a redox signal for controlling RegB activity also correlates well with the observation that mutations in cytochrome cbb3 oxidase lead to elevated RegB activity (Buggy and Bauer, 1995). In this case, a mutation in a terminal respiratory electron acceptor such as a cytochrome cbb3 oxidase would result in a more reduced ubiquinone pool, which would subsequently lead to an elevation of RegB kinase activity. In addition to the ubiquinone-binding site, there is a fully conserved cysteine (Cys 265) that is involved in redox-sensing (Table 1) (Swem et al., 2003). This redox-active cysteine is located in a ‘redox-box’
that is harbored in a cytosolic dimerization interface downstream of the H-box (Fig. 2 and Table 1). In vitro analysis using truncated RegB without the transmembrane domain indicates that an intermolecular disulfide bond forms between RegB dimers under oxidizing conditions, converting active dimers into inactive tetramers (Fig. 1B) (Swem et al., 2003). In vitro disulfide bond formation was shown to require the presence of a divalent metal ion which may help to fold RegB into a functional structure (Swem et al., 2003). The involvement of an intermolecular disulfide bond in the control of RegB activity is also supported by an increase in vitro of full-length RegB phosphorylation in the presence of DTT (Potter et al., 2002). Furthermore, Western blot analysis has confirmed that the RegB intermolecular disulfide bond can form under aerobic growth conditions in vivo (Swem et al., 2003). However, Western blot analysis also showed that <20% of wild type RegB forms a disulfide bond in vivo when cells are shifted from anaerobic to aerobic growth conditions. The remainder of the RegB dimers have been proposed to form other oxidized derivatives of Cys 265 such as sulfonic acid (Cys-S-OH), to regulate kinase activity (Swem et al., 2003). Several mutations have been constructed to probe the roles of ubiquinone and Cys265 redox signals in regulating RegB activity. A Cys265 to Ala mutation (C265A) in full-length RegB led to attenuated, but not absent redox control by coenzyme Q1 in vitro and reduced redox control in vivo (Swem et al., 2006; Swem et al., 2003). Similar results were observed with mutations in the ubiquinone binding domain, which show elevated aerobic expression but still harbor a low level of redox control (Swem et al., 2006). These results indicate that the ubiquinone pool is a redox signal independent of the redox state of Cys265 (Swem et al., 2006).Given that the ubiquinone binding site is located in the transmembrane domain, and Cys265 is located in the cytosolic domain, they are not likely to directly interact. So, it seems that ubiquinone-binding and Cys265 function independently and that they both contribute to redox control of RegB autophosphorylation activity. B. The Response Regulator RegA RegA is a 20.4 kDa response regulator comprised of 184 amino acid residues, that consists of a receiver domain and a DNA binding domain that are linked by a four proline hinge (Elsen et al., 2004). The N-
Chapter 35
Redox Control of Gene Expression
terminal receiver domain is similar to that of other two-component response regulators, containing an aspartate residue (Asp63) where phosphorylation occurs and a highly conserved ‘acid pocket’ consisting of two aspartate residues. The C-terminal DNA binding domain is made up of a 3 α-helices (α-6, α-7, and α-8). The α-7 and α-8 helices comprise a wing-turn-helix DNA binding motif (Sganga and Bauer, 1992; Du et al., 1998; Laguri et al., 2003; Elsen et al., 2004). This binding domain was found to be 100% conserved in RegA homologs from a variety of α-proteobacterial species (Elsen et al., 2004). 1. Effect of Phosphorylation For response regulators to bind DNA, there typically needs to be an interruption of intramolecular forces between the receiver and DNA-binding domains to allow access of DNA binding helices to the DNA surface. The structural change that mediates this opening typically occurs upon phosphorylation of a conserved aspartate (Du et al., 1998; Laguri et al., 2003). A recent crystal structure of a RegA homolog from Mycobacterium tuberculosis provides evidence that domain interface is indeed destabilized upon phosphorylation, yielding a more extended confirmation (Nowak et al., 2006). The affinity of Rba. capsulatus RegA for the puc promoter, and the effect of phosphorylation on DNA binding, has been determined (Bird et al., 1999). The DNA binding affinity of wild type RegA increases 16-fold upon phosphorylation. Similar phosphorylation-induced increases in binding affinities have been reported for B. japonicum RegR and for P. aeruginosa RoxR (Emmerich et al., 1999; Comolli and Donohue, 2002). There are exceptions to the need of phosphorylation for promoting DNA-binding activity of RegA. One exception is the isolation of a hyperactive mutant of RegA called RegA* that contains an Ala95 to Ser mutation, which is near the hinge region in the receiver domain. This mutation allows RegA to promote photosynthesis gene expression even in the absence of RegB, indicating that this substitution increases protein dynamics in favor of an open pseudo-phosphorylated state even under conditions of nonphosphorylation (Du et al., 1998). Nonphosphorylated RegA* has a similar DNA binding affinity as that of phosphorylated wild type RegA. RegA* can also be phosphorylated resulting in an additional six-fold increase in binding activity beyond that of
715 phosphorylated wild type RegA. Furthermore, the phosphate bound to RegA*~P was shown to be significantly more stable with a much longer half-life than that observed with wild type RegA~P (Bird et al., 1999). Both phosphorylated and nonphosphorylated RegA have been observed to play a role in regulating gene expression. For example, while phosphorylated RegA is an anaerobic repressor of cytochrome cbb3 oxidase, nonphosphorylated RegA also functions as an aerobic activator of cytochrome cbb3 oxidase (Swem et al., 2001; Swem and Bauer, 2002a). Additionally, both phosphorylated and nonphosphorylated RegA participate in activation and repression of ubiquinol oxidase expression as well as expression of the cbb and hupSLC operons (Qian and Tabita, 1996; Elsen et al., 2000; Swem and Bauer, 2002a). It remains unclear how phosphorylated and nonphosphorylated forms of RegA are capable of selectively binding and repressing to different promoters. Mutational studies have been performed on the phosphorylation site of RegA from Rba. capsulatus, as well as RegA homologs from B. japonicum and Rba. sphaeroides. An Asp65 to Lys mutant (D63A) of RegA from Rba. capsulatus retained the ability to bind DNA despite being unable to be phosphorylated (Hemschemeier et al., 2000). Comolli and Donohue (2002) also demonstrated that both the nonphosphorylated and phosphorylated versions of wild-type Rba. sphaeroides PrrA activated transcription in vitro, but that the D63A mutant could not. This suggests that Asp63 is also involved in transcriptional activation in some manner, perhaps involving interaction with RNA polymerase. This conclusion must be tempered by the conflicting observation that a Asp65 to Asn mutant of RegR from B. japonicum was incapable of DNA-binding to the fixR-nifA promoter (Emmerich et al., 1999). 2. DNA Binding Sites DNA recognition sequences bound by RegA have been identified by DNase I footprinting and oligonucleotide retention assays. RegA binding sites have been defined at numerous Rba. capsulatus promoters including the puf, puc, nifA2, hupSLC, regB, senC-regA-hvrA, petABC, cycA, cycY, cydAB, ccoNOPQ, cbbI, cbbII, cheOp2, bchE and crtI-crtA (Du et al., 1999; Elsen et al., 2000; Vichivanives et al., 2000; Swem et al., 2001). Alignment of RegA binding sites from Rba. capsulatus yields a consensus
716
Carl E. Bauer, Aaron Setterdahl, Jiang Wu and Brigitte R. Robinson
sequence of 5´-G(C/T)G(G/C)(G/C)(G/A)NN(T/ A)(T/A)NNC(G/A)C-3´ (Swem et al., 2001). The alignment reveals a partially conserved GCG…CGC palindromic binding sequence separated by three to nine bp. A similar inverted repeat is also seen in the B. japonicum RegR consensus sequence that was identified by oligonucleotide affinity trapping. (5´-GNC(A/G)C(A/G)TTNNGNCGC-3´) (Emmerich et al., 2000). Two RegA molecules bind each half of the palindromic sequence in gel shift assays indicating that binding is highly cooperative (Du et al., 1998; Laguri et al., 2003). The variable distances between the GCG…CGC palindrome suggests that RegA proteins bound to palindrome half-sites may interact differently, depending on spacing. The ATrich region in the middle of the consensus sequence is presumably flexible, and may facilitate interaction of RegA molecules with each other or with the promoter DNA sequence. It was proposed that RegA may recognize the shape of the DNA rather than just the sequence (Swem et al., 2001; Elsen et al., 2004). This suggestion is supported by nuclear magnetic resonance (NMR) studies of RegA bound to DNA, which show that RegA binds to kinks formed by pyrimidine-purine steps approximately one helix turn apart (Laguri et al., 2003). This may provide a structural recognition element to aid RegA in finding correct DNA sequences that vary slightly in composition. The NMR experiments (Laguri et al., 2003) also confirmed that RegA binds to each half of the palindromic sequence in a symmetric fashion. The structure revealed that the DNA binding domain consists of the three predicted helices α-6, α-7, and α-8. The α-8 helix is the DNA recognition helix and lies within the major groove interacting with the base pairs of the palindrome half site. The α-6 helix interacts non-specifically with the phosphate backbone of DNA to increase binding affinity. These NMR studies also allowed a refinement of the consensus sequence to 5´YGCGTCRxTATAxGNCGC-3´ (where x is a variable number of nucleotides). The full-length NMR structure of Rba. sphaeroides RegA reveals that the interaction between the receiver and DNA-binding domains only involves the α-6 helix (Laguri et al., 2003). Therefore, the inactive, nonphosphorylated state does not directly block the entire DNA recognition helix, but rather, just the helix that is involved in stabilizing DNA-binding. Mutational analysis of the C-terminal DNA-binding domain identified several residues critical for both binding and transcription. The critical residues correlate well with the proposed
recognition site from the NMR analysis (Jones et al., 2005). Additionally, Laguri et al. (2006) demonstrated through NMR diffusion times, gel filtration, and analytical ultracentrifugation that dimerization of PrrA occurs when a BeF3– generated phosphate analog of PrrA is constructed. This may indicate that phosphorylation induces dimerization and may be important for correct alignment and binding. C. Global Significance As more RegB/RegA homologs and Reg-regulated metabolic processes are discovered, it is becoming evident that RegB/RegA act as a global regulatory system that provides an overlying layer of redox regulation on many important biological processes. Unraveling the redox control mechanism of RegB will provide insight into how a large number of α and γ proteobacteria bacteria balance synthesis of many different redox-responding cellular processes. Analysis of redox control by RegB/RegA also can have practical application in medicine and agriculture. This regulon is present in a number of pathogenic bacteria including P. aeruginosa, Brucella melitensis, Brucella suis, and several species with agricultural importance, such as Agrobacterium tumefaciens, Rhizobium leguminosarum, Sinorhizobium meliloti, and B. japonicum. III. Aerobic repression by CrtJ Present in the photosynthesis gene cluster of several purple photosynthetic bacteria is a redox-responding transcription factor called CrtJ or PpsR, depending on the species. Studies have indicated that CrtJ/PpsR is an aerobic repressor of bacteriochlorophyll (bch), heme (hem) and carotenoid (crt) biosynthesis genes in Rba. sphaeroides and Rba. capsulatus (Penfold and Pemberton, 1994; Gomelsky and Kaplan, 1995a; Ponnampalam and Bauer, 1997; Elsen et al., 1998; Cho et al., 2004; Smart et al., 2004; Kovacs et al., 2005; Moskvin et al., 2005). CrtJ also aerobically represses synthesis of light-harvesting 2 polypeptides (PucBA) that bind bacteriochlorophyll, and carotenoids in these species (Gomelsky and Kaplan, 1995a; Ponnampalam et al., 1995; Moskvin et al., 2005). As discussed above, anaerobic induction of bacteriochlorophyll, carotenoid and light-harvesting genes also requires phosphorylated RegA. So together, CrtJ and RegA regulate synthesis of the photosystem by coordinating aerobic repression, and anaerobic activation of
Chapter 35
Redox Control of Gene Expression
photosystem genes, respectively. A good example of coordinate regulation by CrtJ and RegA occurs in the Rba. capsulatus puc operon. DNase footprint analysis demonstrated that CrtJ and RegA~P have binding sites that overlap near the -35 promoter recognition sequence (Bowman et al., 1999). Oxidized CrtJ was shown to effectively out-compete RegA~P for binding to this promoter region, based on alteration of footprint protection patterns during titration experiments. Oxidized CrtJ also repressed RegA~Pmediated in vitro transcription of puc expression (Bowman et al., 1999). Rba. capsulatus CrtJ and RegA also coordinately regulate synthesis of the respiratory terminal electron acceptors, quinol oxidase (cydAB) and cytochrome cbb3 oxidase (ccoNOPQ) (Swem and Bauer, 2002). Quinol oxidase has a high affinity for oxygen and is maximally expressed under low oxygen tension. This is contrasted by cytochrome cbb3 oxidase that has a low affinity for oxygen, but a high turnover rate and is maximally expressed under high oxygen conditions. Together CrtJ and RegA regulate expression of these oxidases by a coordinating activation and repression. For example, under conditions of high oxygen tension, expression of cytochrome cbb3 oxidase is activated by dephosphorylated RegA while expression of quinol oxidase is repressed by CrtJ. Under low oxygen tension, phosphorylated RegA represses expression of cytochrome cbb3 oxidase and also stimulates expression of quinol oxidase (Swem and Bauer, 2002). The mechanism by which phosphorylated and nonphosphorylated RegA functions as both an activator and repressor of these respiratory promoters remains to be addressed. Protein domain analysis indicates that there are two putative Per-ARNT-Sim motifs called PAS domains in the central region of CrtJ/PpsR, followed by a highly conserved helix-turn-helix DNA binding domain at the C-terminal end (Gomelsky et al., 2000). Gel retardation analysis indicates that PpsR is present in solution as a tetramer (Gomelsky et al., 2000; Masuda and Bauer, 2002) suggesting that at least one of the PAS domains may be involved in tetramerization of CrtJ, however ligand binding to one or both PAS sites cannot be ruled out. In all cases where it has been examined, CrtJ cooperatively binds to two copies of the palandromic sequence TGT-N12-ACA (Ponnampalam and Bauer, 1997; Elsen et al., 1998; Ponnampalam et al., 1998). The palandromic sequence is found either eight base pairs apart, or at sites that are distantly separated. For example, the Rba. capsulatus bchC promoter region
717 has a CrtJ recognition palindrome that spans the –35 promoter region, and a second CrtJ palindrome located 8 bp away that spans the –10 promoter region (Ponnampalam et al., 1998). Binding to these two palindromes is cooperative, so if the 8 bp space between the two palindromes in the bchC promoter region is altered by the addition or deletion of just a few bp, then CrtJ is unable to bind to either palindrome effectively (Ponnampalam et al., 1998). CrtJ also cooperatively binds to palindrome pairs at other promoters, but these palindromes are separated by more than 100–150 base pairs (Elsen et al., 1998). An example of this type of binding occurs in the intergenic region between crtA and crtI which contains two promoters, one that is responsible for driving expression of the crtA-bchI-bchD operon, and a second divergent promoter >100 bp away that is responsible for expression of the crtI-crtB operon (Elsen et al., 1998). The promoter for the crtA-bchI-bchD transcript has a single CrtJ binding site that spans the –10 promoter sequence, while the crtI-crtB promoter also has a single CrtJ recognition sequence that spans the -35 recognition sequence. Cooperative binding of CrtJ to these two palindromes coordinately represses expression of both the crtA-bchI-bchD and crtI-crtB operons (Elsen et al., 1998). This affects synthesis of both bacteriochlorophyll and carotenoids since BchI and BchD are subunits of Mg-chelatase, which is the first committed enzyme of the bacteriochlorophyll branch of the tetrapyrrole biosynthetic pathway (Bollivar et al., 1994). In addition, crtI and crtB code for phytoene dehydrogenase and phytoene synthase, respectively, which are enzymes for the first two committed steps of carotenoid biosynthesis (Armstrong et al., 1990). A similar example of distantly removed CrtJ binding sites occurs in the Rba. capsulatus puc operon that has one CrtJ binding site overlapping the –35 promoter recognition sequence and a second site located 240 bp upstream (Elsen et al., 1998). Like that of the crtA and crtI promoters, binding of CrtJ to the two distant binding sites in the puc promoter region occurs cooperatively. Presumably, binding of CrtJ to distant sites involves looping of the DNA so that tetrameric CrtJ can bind cooperatively to both of the recognition palindromes (Elsen et al., 1998). A. Redox and Light Regulation Redox regulation of CrtJ/PpsR repression in Rba. capsulatus and Rba. sphaeroides has been shown to involve the oxidation and reduction of an intramolecular disulfide bond (Fig. 2) (Masuda et al.,
718
Carl E. Bauer, Aaron Setterdahl, Jiang Wu and Brigitte R. Robinson
2002). In the case of Rba. capsulatus CrtJ, the redox potential (Em) of the disulfide bond is –180 mV at pH 7.0, which means that CrtJ would be predominately reduced in the cytosol, that has a redox state of –220 mV (Masuda et al., 2002). It is therefore assumed that disulfide bond formation may be directly stimulated by molecular oxygen in this species (Masuda et al., 2002). Once oxygen is depleted, the redox poise of the cytosol would effectively reduce the disulfide in CrtJ to promote derepression of photosystem gene expression. In this model, CrtJ would be a direct sensor of the presence of molecular oxygen. Interestingly, the
disulfide bond in Rba. sphaeroides PpsR has an Em= –320 mV at pH 7.0, which would make the cysteines predominately oxidized in the cytosol. There is also a redox- and light-responding antirepressor called AppA in Rba. sphaeroides, which is required to keep the cysteines in PpsR reduced in the absence of oxygen (Fig. 2) (Kim et al., 2006). Rba. sphaeroides strains that are disrupted in AppA thus constitutively have PpsR in its active oxidized state resulting in constitutive repression by PpsR (Gomelsky and Kaplan, 1995b; Masuda and Bauer, 2002). The redox potential of the disulfide in AppA
Fig. 2. The role of AppA in the regulation of PpsR repressor activity. As discussed in the text, AppA has two activities: one is to reduce a disulfide bond in PpsR which partially disrupts DNA binding activity of the repressor; the second activity of AppA is to convert the PpsR tetramer into an AppA-(PpsR)2 complex that is incapable of binding DNA. This latter reaction is inhibited by blue light excitation of a flavin in AppA.
Chapter 35
Redox Control of Gene Expression
has been determined to be from –315 to –325 mV, which is isopotential with the disulfide in PpsR (–320 mV), thereby allowing efficient electron transfer from AppA to PpsR (Kim et al., 2006). AppA has recently been shown to bind heme although the function of heme in redox sensing, or in disulfide bond formation, is not yet clear (Han et al., 2007; Moskvin et al., 2007). AppA is also a blue light photoreceptor that contains a photoactive flavin with a conserved BLUF domain (Gomelsky and Kaplan, 1998; Masuda and Bauer, 2002, 2005; Gomelsky and Klug, 2002). The flavin undergoes a photocycle (for details of the BLUF photocycle see Chapter 36, Klug and Masuda) upon light excitation where a conformational change occurs in AppA, disrupting the interaction between AppA and PpsR (Fig. 2). Under dark conditions, AppA is able to convert PpsR from an active tetramer into an inactive AppA(PpsR)2 complex (Masuda and Bauer, 2002). Under high intensity blue light AppA releases PpsR, which then is able to repress the expression of the photosystem. Even though the function of CrtJ/PpsR as a repressor of photosystem synthesis is well supported in Rba. capsulatus and in Rba. sphaeroides, the function of CrtJ/PpsR in other species has some significant differences. For example, PpsR from Rubriviax gelatinosus is capable of activating light-harvesting 2 (puc) gene expression while also repressing bacteriochlorophyll and carotenoid gene expression (Steunou et al., 2004). In Bradyrhizhobium there are two PpsR genes ppsR1 and ppsR2 (Giraud et al., 2002; Jaubert et al., 2004; Elsen et al., 2005). PpsR1 has only one Cys and yet is still able to respond to redox through the formation of an intermolecular disulfide bond (Jaubert et al., 2004). PpsR1 functions as an activator of photosystem gene expression, and contrary to its homologs, its DNAbinding affinity is higher under reducing conditions, thus promoting expression of the photosystem under low oxygen levels (Jaubert et al., 2004). PpsR2 is active as a repressor that binds to the same target DNA sequence as PpsR1. However PpsR2 does not respond to changes in redox and instead responds to light via interaction with a light-responding antirepressor not unlike the light-dependent interaction of AppA with PpsR in Rba. sphaeroides. A significant difference is that the antirepressor of PpsR2 in Bradyrhizhobium is not a member of the blue light-absorbing BLUF family of flavin photoreceptors like AppA. Instead, the Bradyrhizhobium antirepressor of PpsR2 is a red-light absorbing bacteriophytochrome photoreceptor that
719 uses a bilin as a photoreceptor (Jaubert et al., 2004). The mode of interaction between the phytochrome antirepressor and PpsR2 is unclear. Rhodopseudomonas (Rps.) palustris also has two PpsR transcriptional regulators; however, in contrast to Bradyrhizhobium, both function as transcriptional repressors in response to oxygen (Braatsch et al., 2006). PpsR1 contains three cysteines, two of which would potentially be capable of forming intramolecular disulfide bonds in response to oxygen like that shown for Rba. capsulatus and Rba. sphaeroides, while PpsR2 contains only one cysteine. Like Bradyrhizhobium, Rps. palustris contains a bacteriophytochrome, BphP1, which upon illumination at 750 nm triggers derepression of the photosynthetic genes by one of its two PpsR repressor proteins, PpsR2 (Braatsch et al., 2007; see Chapter 40, Evans et al. for a review). B. Phylogenetic Analysis Phylogenetic analysis of the available CrtJ/PpsR protein sequences reveal three distinct clusters (Fig. 3). Cluster III contains Rba. sphaeroides PpsR and Rba. capsulatus CrtJ. All members of this group have the two well conserved cysteines, (Cys-420 and Cys-249 as defined in Rba. capsulatus CrtJ) that have been shown to be involved in disulfide bond formation (Masuda and Bauer, 2002; Masuda et al., 2002; Kim et al., 2006;). The two closely related Cluster III CrtJ/PpsR proteins from Rba. capsulatus and Rba. sphaeroides have been shown to form disulfides in vitro with disulfide reduction potentials determined (Kim et al., 2006; Masuda et al., 2002). These CrtJ/ PpsR proteins are most likely able to respond to redox via formation of a intramolecular disulfide bond. Clusters I and II CrtJ/PpsR proteins are clearly different in that they do not contain the same set of conserved cysteines. Specifically, the cysteine that is located between the internal PAS domains (equivalent to Cys249 in Rba. capsulatus) are replaced by hydrophobic residues such as leucine or valine. Furthermore, only a subset of CrtJ/PpsR proteins in Cluster I (Rhodospirillum centenum, Rubrivivax gelatinosus, Thiocapsa roeopersicina, Rhodospirillum rubrum, Bradyrhizobium sp. ORS278, Bradyrhizobium sp. BTAi1, Erythrobacter sp. NAP1, Gamma proteobacterium KT 71, Uncultured proteobacterium AAL76373, Uncultured proteobacterium AAM48620) contains the redox active cysteine near the helix-turn-helix (equivalent to Cys420 in Rba.
720
Carl E. Bauer, Aaron Setterdahl, Jiang Wu and Brigitte R. Robinson
capsulatus). Presumably this subset of PpsR/CrtJ proteins could respond to redox through a intermolecular disulfide interaction between Cys420 as has been documented to occur for Bradyrhizobium (Jaubert et al., 2004). The remaining proteins in Cluster I (Rps. palustris BisB5, Rps. palustris HaA2, Rps. palustris CGA009, Rps. palustris BisB18, Rps. palustris BisA53,) have a serine at the Cys420 equivalent position suggesting that these proteins may be redox-active. Interestingly, Rps. palustris BisB5, Rps. palustris HaA2, Rps. palustris CGA009 all have a conserved cysteine at position 377 which is 52 amino
acids before Cys420 (relative to the amino terminus), as well as a cysteine at position 240 which is 16 amino acids before redox-active Cys249 in Rba. capsulatus/ Rba. sphaeroides CrtJ/PpsR. No mutational studies have been reported for these additional Cys in these other species so it is unclear if they are involved in controlling DNA-binding activity. Cluster II CrtJ/PpsR proteins (Fig. 3) also are lacking the corresponding Cys at position 249 and all but one (Thiocapsa roseopersicina PpsR2) are lacking Cys at position 420. These proteins would most likely be responding to another mechanism
Fig. 3. Phylogenetic analysis of CrtJ/PpsR peptide sequences. Sequences were aligned using Neighbor-Joining method with 1000 replicates. For Cluster I peptide sequences, only half contain the C-terminal cysteine required for redox activity and do not contain the middle cysteine. Cluster II proteins do not contain either cysteine. Cluster III proteins contain both cysteines. Horizontal branch lengths represent relative evolutionary distances with the scale bar corresponding to 10 amino acid substitutions per 100. Bootstrap values (%) are indicated at each node except basal nodes.
Chapter 35
Redox Control of Gene Expression
such as in response to light as is the case of PpsR2 from Bradyrhizobium, and PpsR2 from Rps. palustris CGA009, which is regulated by a light responding phytochrome (Giraud et al., 2002, 2004; Braatsch et al., 2007). The presence of PAS domains also raises the possibility that proteins lacking Cys may bind a photoactive or redox active ligand though no evidence has been published as of yet for ligand binding to these domains. IV. Regulation by Fnr FnrL in Rba. sphaeroides and Rba. capsulatus is a homolog of the E. coli anaerobic regulatory protein, Fnr (Zeilstra-Ryalls and Kaplan, 1995). In E. coli, Fnr regulates fumarate and nitrate reduction as well as a host of other cellular processes such as respiration (reviewed in Bauer et al., 1999). FnrL is a member of the Crp family of transcription factors that bind to a palindrome target sequence as a dimer. The amino terminal region of FnrL contains a ferredoxin-like cysteine cluster (Cys-X3-Cys-X2-Cys-X5-Cys). For E. coli Fnr, it is known that an Fe-S center is formed with three of these clustered Cys, as well as with a fourth conserved Cys located ~100 amino acid residues away. Exposure of the FeS cluster to oxygen leads to disassembly of the cluster and subsequent loss of DNA binding activity of Fnr (Fig. 4). Rba. capsulatus and Rba. sphaeroides FnrL have not been isolated and biochemically characterized, but they do contain the same conserved Cys, so it is likely that they also form an oxygen sensitive Fe-S cluster not unlike that of E. coli Fnr. Fnr and FnrL also contain a highly
721 conserved helix-turn-helix motif near the carboxyl terminus, so it is also likely that they bind to a similar DNA sequence of TTGAT-N4-ATCAA (ZeilstraRyalls and Kaplan, 1998). In Rba. sphaeroides this sequence is found upstream of several tetrapyrrole and nitrogen regulatory genes such as fnrL, hemA (ALA synthase), hemN (coproporphyrinogen III oxidase), hemZ (coproporphyrinogen III oxidase), bchE (Mg2+-protoporphyrin monomethyl esterase), ccoNOQP (cbb3 cytochrome oxidase), rdxBHIS/ccoGHIS (membrane-localized redox complex), ctaD (subunit of the aa3 cytochrome c terminal oxidase), ctaABC (subunit of the aa3 cytochrome c terminal oxidase), cycP (cytochrome c´), dorS (sensor protein of the dor operon), pucBAC (α and β structural polypeptides of LHII), nnrR (activator of nir and nor), nirK (nitrite reductase structural gene), and norCBQD (nitric oxide reductase operon)(Zeilstra-Ryalls et al., 1997; Mao et al., 2005). However, this sequence is found in only a few promoter regions in Rba. capsulatus such as hemZ, porphobilinogen synthase (hemB), fnrL, and ccoNOQP (Zeilstra-Ryalls et al., 1997; Swem and Bauer, 2002b; Smart et al., 2004; Choi et al., 2005; Chapter 39, Zeilstra-Ryalls). A fnrL mutant of Rba. sphaeroides cannot grow photosynthetically (Mouncey and Kaplan, 1998). For cultures grown in 2% oxygen, the expression of ccoN is increased greatly in an fnrL knockout, yet under 30% oxygen growth conditions the expression is much reduced (Mouncey and Kaplan, 1998). FnrL has been shown to activate the expression of cbb3 cytochrome oxidase and ALA synthase under anaerobic conditions in Rba. sphaeroides (Mouncey and Kaplan, 1998; Zeilstra-Ryalls and Kaplan, 1995;
Fig. 4. Model of oxygen inactivation of FnrL activity. E. coli Fnr has a Fe-S center that is disrupted by molecular oxygen. Based on results from the Fnr homolog in E. coli, we propose that oxygen-dependent disruption of the Fe-S center in FnrL from Rba. capsulatus and Rba. sphaeroides leads to inhibition of DNA-binding activity.
722
Carl E. Bauer, Aaron Setterdahl, Jiang Wu and Brigitte R. Robinson
Zeilstra-Ryalls and Kaplan, 1998). A fnrL disruption in Rba. capsulatus is somewhat different in that loss of FnrL does not significantly affect the ability of these cells to grow photosynthetically (Swem and Bauer, 2002). Expression of ubiquinol oxidase in a Rba. capsulatus fnrL knockout was significantly increased when grown semiaerobically (Swem and Bauer, 2002). Expression studies of cbb3 oxidase in an fnrL knockout strain showed a decrease in expression under anaerobic and semi-aerobic conditions (Swem and Bauer, 2002). In Rubrivivax gelatinosus, a null mutant of fnrL was unable to grow under anaerobic conditions; however photosynthetic complexes were produced under high oxygen levels. FnrL was shown to regulate the expression of hemN (oxygen dependent coproporphyrin III dehydrogenase), and bchE (Mg2+-protoporphyrin monomethyl cyclase) in response to oxygen tension (Ouchane et al., 2007). Clearly much more research needs to be undertaken on this important regulator, to better understand its role in controlling the synthesis of the photosystem and respiration in response to alterations in oxygen tension. Acknowledgments Research from the CE Bauer laboratory is supported by funding from the National Institutes of Health (GM 40941). We thank Marina Resendes de Sousa Antonio for generating the CrtJ/PpsR phylogenetic tree in Fig. 3. References Armstrong GA, Schmidt A, Sandmann G and Hearst JE (1990) Genetic and biochemical characterization of carotenoid biosynthesis mutants of Rhodobacter capsulatus. J Biol Chem 265:8329–8338 Bauer CE, Young DA and Marrs BL (1988) Analysis of the Rhodobacter capsulatus puf operon: Location of the oxygen regulated promoter region and the identification of an additional puf encoded gene. J Biol Chem 263: 4820–4827 Bauer CE, Elsen S and Bird TH (1999) Mechanisms for redox control of gene expression. Ann Rev Microbiol 53: 495–523 Bauer E, Kaspar T, Fischer HM and Hennecke H (1998) Expression of the fixR-nifA operon in Bradyrhizobium japonicum depends on a new response regulator, RegR. J Bacteriol 180: 3853–3863 Bird TH, Du S and Bauer CE (1999) Autophosphorylation, phosphotransfer, and DNA-binding properties of the RegB/RegA two-component regulatory system in Rhodobacter capsulatus. J Biol Chem 274: 16343–16348
Bollivar DW, Suzuki JY, Beatty JT, Dobrowlski JD and Bauer CE (1994) Directed mutational analysis of bacteriochlorophyll a biosynthesis in Rhodobacter capsulatus. J Mol Biol 237: 622–640 Bolton JR (1978) Primary electron acceptor. In: Clayton RK and Sistrom WR (ed) The Photosynthetic Bacteria, pp 419–429. Plenum Press, New York Bowman WC, Du S, Bauer CE and Kranz RG (1999) In vitro activation and repression of photosynthesis gene transcription in Rhodobacter capsulatus. Mol Microbiol 33: 429–437 Bratsch S, Gomelsky M, Kuphal S and Klug G (2002) A single flavoprotein, AppA, integrates both redox and light signals in Rhodobacter sphaeroides. Mol Microbiol 45: 827–836 Braatsch S, Bernstein JR, Lessner F, Morgan J, Liao JC, Harwood CS and Beatty JT (2006) Rhodopseudomonas palustris CGA009 has two functional ppsR genes, each of which encodes a repressor of photosynthesis gene expression. Biochemistry 45:14441–14451 Braatsch S, Johnson JA, Noll K and Beatty JT (2007) The O2responsive repressor PpsR2 but not PpsR1 transduces a light signal sensed by the BphP1 phytochrome in Rhodopseudomonas palustris CGA009. FEMS Microbiol Lett 272: 60–64 Buggy J and Bauer CE (1995). Cloning and characterization of senC, a gene involved in both aerobic respiration and photosynthesis gene expression in Rhodobacter capsulatus. J Bacteriol 177, 6958–6965 Chen W, Jager A and Klug G (2000) Correction of the DNA sequence of the regB gene of Rhodobacter capsulatus with implications for the membrane topology of the sensor kinase RegB. J Bacteriol 182: 818–820 Cho SH, Youn SH, Lee SR, Yim HS and Kang SO (2004) Redox property and regulation of PpsR, a transcriptional repressor of photosystem gene expression in Rhodobacter sphaeroides. Microbio 150: 697–706 Cohen-Bazire G, Sistrom WR and Stanier RY (1957) Kinetic studies of pigment synthesis by nonsulfur purple bacteria. J Cellular Comp Physiol 49: 25–68 Comolli JC and Donohue TJ (2002) Pseudomonas aeruginosa RoxR, a response regulator related to Rhodobacter sphaeroides PrrA, activates expression of the cyanide-insensitive terminal oxidase Mol Micro 45: 755–768 Dong C, Elsen S, Swem LR and Bauer CE (2002) AerR, a second aerobic repressor of photosynthesis gene expression in Rhodobacter capsulatus. J Bacteriol 184: 2805–2814 Du S, Bird TH and Bauer CE (1998) DNA binding characteristics of RegA. A constitutively active anaerobic activator of photosynthesis gene expression in Rhodobacter capsulatus. J Biol Chem 273:18509–18513 Du S, Kouadio J-L K and Bauer CE (1999) Regulated expression of a highly conserved regulatory gene cluster is necessary for controlling photosynthesis gene expression in response to anaerobiosis in Rhodobacter capsulatus. J Bacteriol 181: 4334–4341 Elsen S, Ponnampalam SN and Bauer CE (1998) CrtJ bound to distant binding sites interacts cooperatively to aerobically repress photopigment biosynthesis and light harvesting II gene expression in Rhodobacter capsulatus. J Biol Chem 273: 30762–30769 Elsen S, Dischert W, Colbeau A and Bauer CE (2000) Expression of uptake hydrogenase and molybdenum nitrogenase in Rhodobacter capsulatus is coregulated by the RegB-RegA
Chapter 35
Redox Control of Gene Expression
two-component regulatory system. J Bact 182: 2831–2837 Elsen S, Swem LR, Swem DL and Bauer CE (2004) RegB/RegA, a highly conserved redox-responding global two-component regulatory system. Microbio Molec Biol Rev 68: 263–279 Elsen S, Jaubert M, Pignol D and Giraud E (2005) PpsR: A multifaceted regulator of photosynthesis gene expression in purple bacteria. Mol Microbiol 57: 17–26 Emmerich R, Panglungtshang K, Strehler P, Hennecke H and Fischer H-M (1999) Phosphorylation, dephosphorylation and DNA-binding of the Bradyrhizobium japonicum RegSR two-component regulatory proteins. Eur J Biochem 263: 455–463 Emmerich R, Strehler P, Hennecke H and Fischer H-M (2000a) An imperfect inverted repeat is critical for DNA binding of the response regulator RegR of Bradyrhizobium japonicum. Nuc Acids Res 28: 4166–4171 Emmerich R, Hennecke H and Fischer H-M (2000b) Evidence for a functional similarity between the two-component regulatory systems RegSR, ActSR and RegBA (PrrBA) in α-proteobacteria. Arch Microbiol 174:307–313 Eraso JM and Kaplan S (1994) prrA, a putative response regulator involved in oxygen regulation of photosynthesis gene expression in Rhodobacter sphaeroides. J Bacteriol 176: 32–43 Eraso JM and Kaplan S (1995) Oxygen-insensitive synthesis of the photosynthetic membranes of Rhodobacter sphaeroides: A mutant histidine kinase. J Bacteriol 177: 2695–2706 Eraso JM and Kaplan S (2000) From redox flow to gene regulation: Role of the PrrC protein of Rhodobacter sphaeroides 2.4.1. Biochem 39: 2052–2062 Giraud E, Fardoux J, Fourrier N, Hannibal L, Genty B, Bouyer P, Dreyfus B and Verméglio A (2002) Bacteriophytochrome controls photosystem synthesis in anoxygenic bacteria. Nature 417: 202–205 Giraud E, Zappa S, Jaubert M, Hannibal L, Fardoux J, Adriano JM, Bouyer P, Genty B, Pignol D and Verméglio A (2004) Bacteriophytochrome and regulation of the synthesis of the photosynthetic apparatus in Rhodopseudomonas palustris: Pitfalls of using laboratory strains. Photochem Photobiol Sci 3: 587–591 Gomelsky M and Kaplan S (1995a) Genetic evidence that PpsR from Rhodobacter sphaeroides 2.4.1 functions as a repressor of puc and bchF expression. J Bacteriol 177: 1634–1637 Gomelsky M and Kaplan S (1995b). appA, a novel gene encoding a trans-acting factor involved in the regulation of photosynthesis gene expression in Rhodobacter sphaeroides. J Bacteriol 177: 4609–4618 Gomelsky M and Klug G (2002) BLUF: A novel FAD-binding domain involved in sensory transduction in microorganisms. Trends Biochem. Sci. 27: 497–500 Gomelsky M, Horne IM, Lee HJ, Pemberton JM, McEwan AG, Kaplan S (2000) Domain structure, oligomeric state, and mutational analysis of PpsR, the Rhodobacter sphaeroides repressor of photosystem gene expression. J Bacteriol 182: 2253–2261 Han Y, Meyer MH, Keusgen M, Klug G (2007) A haem cofactor is required for redox and light signalling by the AppA protein of Rhodobacter sphaeroides. Mol Microbiol 64:1090–1104 Hemschemeier SK, Ebel U, Jager A, Balzer A, Kirndorfer M and Klug G (2000) In vivo and in vitro analysis of RegA response regulator mutants of Rhodobacter capsulatus. J Mol Microbiol Biotech 2: 291–300
723 Inoue K, Mosley C, Kouadio J-L and Bauer C (1995) Isolation and in vitro phosphorylation of sensory transduction components controlling anaerobic induction of light harvesting and reaction center gene expression in R. capsulatus. Biochemistry 34: 391–396 Jaubert M, Zappa S, Fardoux J, Adriano J M, Hannibal L, Elsen S, Lavergne J, Verméglio A, Giraud E and Pignol D (2004) Light and redox control of photosynthesis gene expression in Bradyrhizobium: Dual roles of two PpsR. J Biol Chem 279: 44407–44416 Jones DF, Stenzel RA and Donohue TJ (2005) Mutational analysis of the C-terminal domain of the Rhodobacter sphaeroides response regulator PrrA. Microbiol 151: 4103–4110 Joshi HM and Tabita FR (1996) A global two component signal transduction system that integrates the control of photosynthesis, carbon dioxide assimilation, and nitrogen fixation. Proc Natl Acad Sci USA 93: 14515–14520 Kaplan S, Eraso J and Roh JH (2005) Interacting regulatory networks in the facultative photosynthetic bacterium, Rhodobacter sphaeroides 2.4.1. Biochem Soc Trans 33: 51–55 Kim, SK, Mason, JT, Knaff, DB, Bauer, CE and Setterdahl, AT (2006) Redox properties of the Rhodobacter sphaeroides transcriptional regulatory proteins PpsR and AppA. Photosynth Res 89: 89–98 Kovacs AT, Rakhely G and Kovacs KL (2005) The PpsR regulator family. Res Microbiol 156: 619–625 Laguri C, Phillips-Jones MK and Williamson MP (2003) Solution structure and DNA binding of the effector domain from the global regulator PrrA (RegA) from Rhodobacter sphaeroides: Insights into DNA binding specificity. Nucl Acids Res 31: 6778–6787 Laguri C, Stenzel RA, Donohue TJ, Phillips-Jones MK and Williamson MP (2006) Activation of the global gene regulator PrrA (RegA) from Rhodobacter sphaeroides. Biochem 45: 7872–7881 Laratta WP, Choi PS, Tosques IE and Shapleigh JP (2002) Involvement of the PrrB/PrrA two-component system in nitrite respiration in Rhodobacter sphaeroides 2.4.1: Evidence for transcriptional regulation. J Bact 184: 3521–3529 Madigan MT (1995) The microbiology of nitrogen fixation by anoxygenic photosynthetic bacteria, In: Blankenship RE, Madigan MT and Bauer CE (ed) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 915–928. Kluwer Academic Publishing, Dordrecht Madigan MT and Gest H (1979) Growth of the photosynthetic bacterium Rhodopseudomonas capsulata chemoautotrophically in darkness with H2 as the energy source. J Bacteriol 137: 524–530 Mao L, Mackenzie C, Roh JH, Eraso JM, Kaplan S and Resat H (2005) Combining microarray and genomic data to predict DNA binding motifs. Microbiology 151: 3197–3213 Masuda S and Bauer CE (2002) AppA is a blue light photoreceptor that antirepresses photosynthesis gene expression in Rhodobacter sphaeroides. Cell 110: 613–623 Masuda S and Bauer CE (2005) The antirepressor AppA uses the novel flavin-binding BLUF domain as blue-light-absorbing photoreceptor to control photosystem synthesis. In: Briggs W and Spudich J (eds) Handbook of Photosensory Receptors, pp 433–446. Wiley-VCH publishing, Weinheim Masuda S, Matsumoto Y, Nagashima KVP, Shimada K, Inoue K, Bauer CE and Matsuura K (1999) Structural and functional
724
Carl E. Bauer, Aaron Setterdahl, Jiang Wu and Brigitte R. Robinson
analyses of photosynthetic regulatory genes regA and regB from Rhodovulum sulfidophilum, Roseobacter denitrificans, and Rhodobacter capsulatus. J Bacteriol 181: 4205–4215 Masuda S, Dong C, Swem D, Setterdahl AT, Knaff DB and Bauer CE (2002) Repression of photosynthesis gene expression by formation of a disulfide bond in CrtJ. Proc Natl Acad Sci USA 99: 7078–7083 McCleary WR and Stocks JB (1994) Acetyl phosphate and the activation of two-component response regulators. J Biol Chem 269: 31567–31572 Moskvin OV, Gomelsky L and Gomelsky M (2005) Transcriptome analysis of the Rhodobacter sphaeroides PpsR regulon: PpsR as a master regulator of photosystem development J Bacteriol 187: 2148–2156 Moskvin OV, Kaplan S, Gilles-Gonzalez MA and Gomelsky M (2007) Novel heme-based oxygen sensor with a revealing evolutionary history. J Biol Chem 282: 28740–28748 Mosley CS, Suzuki JY and Bauer CE (1994) Identification and molecular genetic characterization of a sensor kinase responsible for coordinately regulating light harvesting and reaction center gene expression in response to anaerobiosis. J Bacteriol 176: 7566–7573 Mouncey N J and Kaplan S (1998) Oxygen regulation of the ccoN gene encoding a component of the cbb3 oxidase in Rhodobacter sphaeroides 2.4.1: Involvement of the FnrL protein. J Bacteriol 180: 2228–2231 Nowak E, Panjikar S, Konarev P, Svergun DI and Tucker PA (2006) The structural basis of signal transduction for the response regulator PrrA from Mycobacterium tuberculosis. J Biol Chem 281: 9659–9666 O’Gara JP and Kaplan S (1997) Evidence for the role of redox carriers in photosynthesis gene expression and carotenoid biosynthesis in Rhodobacter sphaeroides. 2.4.1. J Bacteriol 179: 1951–1961 Oh JI and Kaplan S (2000) Redox signaling: Globalization of gene expression EMBO J 19: 4237–4247 Ouchane S and Kaplan S (1999) Topological analysis of the membrane-localized redox-responsive sensor kinase PrrB from Rhodobacter sphaeroides 2.4.1. J Biol Chem 274: 17290–17296 Ouchane S, Picaud M, Therizols P, Reiss-Husson F and Astier C (2007) Global Regulation of Photosynthesis and Respiration by FnrL: The first two targets in the tetrapyrrole pathway. J Biol Chem 282: 7690–7699 Parkinson JS and Kofoid EC (1992) Communication modules in bacterial signaling proteins. Ann Rev Genet 26: 71–112 Parson W (1978) Quinones as secondary electron acceptor, In: RK Clayton and Sistrom WR (ed) The Photosynthetic Bacteria, pp 455–469. Plenum Press, New York Penfold RJ and Pemberton JM (1994) Sequencing, chromosomal inactivation, and functional expression in Escherichia coli of ppsR, a gene which represses carotenoid and bacteriochlorophyll synthesis in Rhodobacter sphaeroides. J Bacteriol 176: 2869–2876 Pfenning N (1978) General physiology and ecology of photosynthetic bacteria, In: RK Clayton and Sistrom WR (ed) The Photosynthetic Bacteria, pp 1–18. Plenum Press, New York Phillips-Jones MK and Hunter CN (1994) Cloning and nucleotide sequencing of RegA, a putative response regulator gene of Rhodobacter sphaeroides. FEMS Microbiol Lett 116: 269–275 Ponnampalam SN and Bauer CE (1997) DNA binding charac-
teristics of CrtJ A redox-responding repressor of bacteriochlorophyll, carotenoid, and light harvesting-II gene expression in Rhodobacter capsulatus. J Biol Chem 272: 18391–18396 Ponnampalam SN, Buggy JJ, Bauer CE (1995) Characterization of an aerobic repressor that coordinately regulates bacteriochlorophyll, carotenoid, and light harvesting-II expression in Rhodobacter capsulatus. J Bacteriol 177: 2990–2997 Ponnampalam SN, Elsen S and Bauer CE (1998) Aerobic repression of the Rhodobacter capsulatus bchC promoter involves cooperative interactions between CrtJ bound to neighboring palindromes. J Biol Chem 273: 30757–30761 Potter CA, Ward A, Laguri C, Williamson MP, Henderson PJ and Phillips-Jones MK (2002) Expression, purification and characterization of full-length histidine protein kinase RegB from Rhodobacter sphaeroides. J Mol Biol 320: 201–213 Qian, Y and Tabita FR (1996) A global signal transduction system regulates aerobic and anaerobic CO2 fixation in Rhodobacter sphaeroides. J Bact 178: 12–18 Roh JH and Kaplan S (2000) Genetic and phenotypic analyses of the rdx locus of Rhodobacter sphaeroides 2.4.1. J Bacteriol 182: 3475–3481 Sganga MW and Bauer CE (1992) Regulatory factors controlling photosynthetic reaction center and light-harvesting gene expression in Rhodobacter capsulatus. Cell 68: 945–954 Smart JL, Willett JW and Bauer CE (2004) Regulation of hem gene expression in Rhodobacter capsulatus by redox and photosystem regulators RegA, CrtJ, FnrL, and AerR. J Mol Biol 342: 1171–1186 Steunou AS, Astier C and Ouchane S (2004) Regulation of photosynthesis genes in Rubrivivax gelatinosus: Transcription factor PpsR is involved in both negative and positive control. J Bacteriol 186: 3133–3142 Swem DL and Bauer CE (2002) Coordination of ubiquinol oxidase and cytochrome cbb3 oxidase expression by multiple regulators in Rhodobacter capsulatus. J Bacteriol 184: 2815–2820 Swem LR, Elsen S, Bird TH, Swem DL, Koch HG, Myllykallio H, Daldal F and Bauer CE (2001) The RegB/RegA two-component regulatory system controls synthesis of photosynthesis and respiratory electron transfer components in Rhodobacter capsulatus. J Mol Biol 309:121–138 Swem LR, Kraft BJ, Swem DL, Setterdahl AT, Masuda S, Knaff DB, Zaleski JM and Bauer CE (2003) Signal transduction by the global regulator RegB is mediated by a redox-active cysteine. EMBO J 22: 4699–4708 Swem LR, Gong X, Yu CA and Bauer CE (2006) Identification of a ubiquinone-binding site that affects autophosphorylation of the sensor kinase RegB. J Biol Chem 281: 6768–6775 Tabita FR (1995) The biochemistry and metabolic regulation of carbon metabolism and CO2 fixation in purple bacteria. In: Blankenship RE, Madigan MT and Bauer CE (ed) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 885–914. Kluwer Academic Publishing, Dordrecht Tiwari RP, Reeve WG, Dilworth MJ and Glenn AR (1996) Acid tolerance in Rhizobium meliloti strain WSM419 involves a two-component sensor-regulator system Microbiol 142: 1693–1704 Vichivanives P, Bird TH, Bauer CE and Tabita FR (2000) Multiple regulators and their interactions in vivo and in vitro with the cbb regulons of Rhodobacter capsulatus. J Mol Biol 300: 1079–1099
Chapter 35
Redox Control of Gene Expression
Willett J, Smart JL and Bauer CE (2007) RegA controls synthesis of bacteriochlorophyll and carotenoid photopigments in Rhodobacter capsulatus. J Bacteriol, In Press Zeilstra-Ryalls JH, Gabbert K, Mouncey NJ, Kaplan S and Kranz RG (1997) Analysis of the fnrL gene and its function in Rhodobacter capsulatus. J Bacteriol 179: 7264–7273
725 Zeilstra-Ryalls JH and Kaplan S (1995) Aerobic and anaerobic regulation in Rhodobacter sphaeroides 2.4.1: The role of the fnrL gene. J Bacteriol 177: 6422–6431 Zeilstra-Ryalls JH and Kaplan S (1998) Role of the fnrL gene in photosystem gene expression and photosynthetic growth of Rhodobacter sphaeroides 2.4.1. J Bacteriol 180: 1496–1503
Chapter 36 Regulation of Genes by Light Gabriele Klug* Institute for Microbiology and Molecular Biology, University of Giessen, Heinrich-Buff-Ring 26-32, D-35392 Giessen, Germany
Shinji Masuda* Graduate School of Bioscience and Biotechnology, Tokyo Institute of Technology, Yokohama 226-8501, Japan
Summary ............................................................................................................................................................... 727 I. Introduction..................................................................................................................................................... 728 II. Photoreceptors in Purple Photosynthetic Bacteria ......................................................................................... 728 A. Bacteriophytochromes...................................................................................................................... 728 B. Sensory Rhodopsins ........................................................................................................................ 730 C. Phototropin-Related Proteins ........................................................................................................... 731 D. BLUF Domain Proteins .................................................................................................................... 732 E. Cryptochromes ................................................................................................................................. 734 F. Photoactive Yellow Protein ............................................................................................................... 735 III. Light-Dependent Responses That Do Not Depend on Photoreceptors ......................................................... 735 A. Photosynthetic Electron Transport ................................................................................................... 735 B. Singlet Oxygen ................................................................................................................................. 736 IV. Concluding Remarks ...................................................................................................................................... 737 Acknowledgments ................................................................................................................................................. 737 References ............................................................................................................................................................ 737
Summary Our current understanding of light-dependent regulation of gene expression in purple bacteria is summarized. Most of the regulatory systems utilize photoreceptor proteins that transmit a light-dependent signal to different downstream components to control a wide variety of physiological responses. The photoreceptors identified so far are (bacterio)phytochrome, sensory rhodopsin, phototropin-related proteins, BLUF domain proteins, cryptochrome, and photoactive yellow protein. They use different chromophores such as bilin, retinal, flavin or p-coumaric acid that absorb different wavelengths of light. Based on structural and spectroscopic studies, photochemical reaction mechanisms are beginning to be revealed, which show how the photoreceptors translate a light signal into protein structural changes. On the other hand, downstream factors as well as their signaling pathways are still largely unknown. Purple bacteria also respond to the light environment independently of the photoreceptors. Recent biochemical and genetic data have established that the responses involve photosynthetic and respiratory electron transport chains as well as reactive oxygen species such as singlet oxygen. Regulatory mechanisms of the photoreceptor-independent light responses are also discussed. *Authors for correspondence, email: [email protected]; [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 727–741. © 2009 Springer Science + Business Media B.V.
728 I. Introduction Light is beneficial for photosynthetic organisms as an energy source. Nevertheless, light of high energy can also be harmful due to the generation of reactive oxygen species that damage cellular components. Therefore it is not surprising that photosynthetic bacteria have evolved a variety of responses to light. Some of these are behavioral responses that allow the cells to approach or escape from light, others lead to an adaptation of the cellular composition and allow light to be used as an energy source or to counteract photooxidative stress. Light responses were first observed for plants. When molecular genetic approaches were applied to model organisms, different types of photoreceptors could be identified. It was anticipated that such photoreceptors evolved in eukaryotes and that simple prokaryotic organisms would lack such signaling pathways, responding instead to changes in photosynthetic electron flow or to the presence of reactive oxygen species. With the discovery of the Photoactive Yellow Protein (PYP) in purple bacteria (Meyer, 1985) and bacteriophytochromes in cyanobacteria (Kehoe and Grossman, 1996; Yeh et al., 1997), this view had to be revised. Indeed bacteria contain a variety of photoreceptors, some of them with similarity to plant photoreceptors. Our understanding of signaling by these proteins is still very limited. Notably, all types of photo-responsive proteins discovered to date are conserved in the genus of purple bacteria (proteobacteria). We will give an overview of the photoreceptors found in purple photosynthetic bacteria and summarize our current knowledge of their biological functions, the molecular mechanisms of light signaling and photochemistry. We also include those light responses that turned out to be independent of photoreceptors and discuss the underlying regulatory mechanisms.
Abbreviations: BLUF – sensors of Blue Light Using FAD; BV – biliverdin IXα; E. – Euglena; FMN – flavin mononucleotide; FTIR – Fourier transform infrared; GAF – cyclic GMP, adenylyl cyclase, FhlA; LOV – light oxygen voltage; NMR – nuclear magnetic resonance; PAS – Per Arnt Sim; PYP – Photoactive Yellow Protein; Rba. – Rhodobacter; Rcs. – Rhodocista; Rps. – Rhodopseudomonas; Rsp. – Rhodospirillum; Rvi. – Rubrivivax; SRII – sensory rhodopsin II; Tch. – Thermochromatium
Gabriele Klug and Shinji Masuda II. Photoreceptors in Purple Photosynthetic Bacteria A. Bacteriophytochromes In plants, phytochromes together with cryptochromes and phototropins are the main photomorphogenesis regulators. The discovery of bacterial phytochromes (called bacteriophytochromes) within the last decade has considerably advanced our knowledge of phytochrome structure and function (Kyndt et al., 2004; Rockwell et al., 2006; Chapter 40, Evans et al.). The bacteriophytochromes were first identified in cyanobacteria: RcaE, involved in complementary chromatic adaptation in Fremyella diplosiphon (Kehoe and Grossman, 1996) and Cph1 of Synechocystis sp. PCC6803, a light regulated histidine kinase (Hughes et al., 1997; Yeh et al., 1997). Later on, genes encoding bacteriophytochrome were also identified in the genomes of other cyanobacteria, filamentous fungi, several purple bacteria, and nonphotosynthetic eubacteria such as Deinococcus radiodurans and Pseudomonas aeruginosa. Currently a function can be assigned to only some of these phytochromes (Davis et al., 1999; Jiang et al., 1999; Wu and Lagarias, 2000; Giraud et al., 2002; 2005; Catlett et al., 2003; Evans et al., 2005). All (bacterio)phytochromes seem to carry a bilin chromophore that is bound to a cyclic GMP, adenylyl cyclase, FhlA (GAF) domain, which is combined to different protein domains in the individual phytochromes (Fig. 1). The bacterial phytochromes show different types of domain composition. Most bacteriophytochromes carry a Per Arnt Sim (PAS) domain at the N-terminus, while the cyanobacterial Cph2 lacks this PAS domain but contains another GAF domain. The regular output domain of many prokaryotic phytochromes is related to the output domain of two-component histidine kinases. Some of the bacteriophytochromes from the purple bacteria Bradyrhizobium sp. and Rhodopseudomonas (Rps.) palustris, however, have a C-terminal domain without similarity to known domains (Giraud et al., 2002), and those from Rhodobacter (Rba.) sphaeroides and Thermochromatium (Tch.) tepidum have GGDEF/EAL type output domains typical for diguanylate cyclases/ phosphodiesterases (Kyndt et al., 2005; Tarutina et al., 2006). A special form of bacteriophytochrome is observed in Ppr and PpcI of Rhodocista (Rcs.) centenaria (formerly called Rhodospirillum centenum)
Chapter 36
Regulation of Genes by Light
Fig. 1. Domain composition of the plant phytochromes of Arabidopsis thaliana, the cyanobacterial phytochrome Cph1 and bacteriophytochromes from purple bacteria.
and Tch. tepidum, respectively (Jiang et al., 1999; Kyndt et al., 2004). Besides a histidine kinase and/or GGDEF/EAL output domain at the C-terminus, these proteins have a PYP domain N-terminally fused to the GAF domain. As indicated in the following chapter, PYP is a blue-light sensing domain that binds a p-coumaric acid as a chromophore. In contrast to the wild type, a ppr deletion mutant of Rcs. centenaria shows low, blue light independent expression of the gene for chalcone synthase (Jiang et al., 1999), suggesting that these bacteriophytochromes could sense not only red/far-red light, but also blue light. Surprisingly, genome sequencing revealed that Rps. palustris harbors six phytochrome-like genes (Larimer et al., 2004). Four of the bacteriophytochrome genes (encoding RpBphP1-4) are located close to photosynthesis genes. Bacteriophytochromes were shown to control photosystem synthesis at low aeration in the purple bacteria Rps. palustris and Bradyrhizobium ORS278 (Giraud et al., 2002). In both species the corresponding bacteriophytochrome genes are localized just downstream of the ppsR gene. PpsR is a transcriptional repressor that is involved in the redox regulation of photosynthesis genes in several purple phototrophic bacteria (Elsen et al., 2005). Bacteriophytochrome genes identified in Rba. sphaeroides and/or Rcs. centenaria did not influence photosystem synthesis (Giraud et al., 2002). While red light (740nm) stimulates the formation of photosynthetic complexes in Rps. palustris and Bradyrhizobium ORS278, this is not the case for the other species. The two BphP1 bacteriophytochromes from Rps. palustris and Bradyrhizobium do not
729 contain histidine kinase output domains. Instead the GAF domain is fused to an S-box domain that corresponds to a strongly conserved region in PAS domains. It was therefore suggested that these bacteriophytochromes transmit signals to downstream regulators by protein-protein interaction (Giraud et al., 2002). The same study indeed revealed that the transcriptional repressor PpsR is part of the lightdependent signal chain, as also shown by Braatsch et al. (2007) (Fig. 2). How the Pr form of bacteriophytochrome antagonizes repression of genes by PpsR is not known to date. RpBphP2 and RpBphP3 of Rps. palustris were further analyzed (Evans et al., 2005; Giraud et al., 2005). They are organized in tandem downstream of the pucBAd genes encoding the pigment binding proteins of the LH4 complex. Both of these phytochromes contain a histidine kinase output domain and autophosphorylate in their dark adapted Pr form. The phosphate is then transferred to a common response regulator, Rpa3017 (Giraud et al., 2005). Analysis of mutant strains suggests that RpBphP2 and RpBphP3 operate in tandem to control the synthesis of LH4 complexes dependent on light quality (Giraud et al., 2005). Like plant phytochromes, bacteriophytochromes in purple bacteria contain bilin chromophores that enable photoconversion between Pr and Pfr forms with light-induced cis-trans isomerization of the C15 double bond between rings C and D of the open tetrapyrrole chromophore (Rockwell et al., 2006). However, they have distinct properties that are not shown in plant and cyanobacterial phytochromes. One is that they use biliverdin IXα (BV) as a bilin chromophore (Bhoo et al., 2001; Wagner et al., 2005), unlike plant and cyanobacterial phytochromes that use phytochromobilin or phycocyanobilin (Lagarias and Rapoport, 1980; Park et al., 2000; Lamparter et al., 2001). Both bilins in plant and cyanobacteria are covalently attached at C31 to a conserved Cys residue in the GAF domain; on the other hand, the BV chromophore is attached at C32 to a conserved Cys residue at the N-terminus that is localized outside of the GAF domain (Lamparter et al., 2004; Wagner et al., 2005). Because there is one additional double bond in the π-electron system of BV versus phytochromobilin or phycocyanobilin, the absorption maxima of BVbound bacteriophytochrome in Pr and Pfr forms (~690 and ~750 nm) are slightly red-shifted compared to plant and cyanobacterial phytochromes. Given that purple bacteria use longer wavelengths of light for
730
Gabriele Klug and Shinji Masuda
Fig. 2. Schematic overview of photoreceptors and photoreceptor homologues in purple bacteria, of downstream signaling factors and of light responses.
photosynthesis compared to plant and cyanobacteria, the different chromophore composition may reflect the requested light quality composition incident on these organisms. Recently, unique photochemical properties have been observed in two bacteriophytochromes from purple bacteria. One is Rps. palustris RpBphP3 that converts the Pr form to a form absorbing at shorter wavelength around 645 nm, designated as Pnr form for near-red absorbing form (Giraud et al., 2005). Given that RpBphP3 regulates the synthesis of LH4 complexes as described above, this type of regulation may be crucial for complementary chromatic adaptation of this bacterium. The other observation is that the Pr form of Rcs. centenaria Ppr is reversibly photobleached rather than showing a spectral shift upon even white-light illumination (Kyndt et al., 2004; Chung et al., 2007). This suggests that Ppr senses light intensity rather than light quality. In both proteins, the molecular basis of the unusual spectral properties during the photocycle has not been elucidated. B. Sensory Rhodopsins Rhodopsins known as visual pigments and photoreceptors in halophilic archaea have been intensively investigated with regard to their structure and photocycle (Spudich et al., 2000). They share identical topologies characterized by seven transmembrane helices that form a pocket in which retinal is covalently bound as a chromophore. By genomic analysis of naturally occurring marine phytoplankton the first eubacterial rhodopsin was discovered in an uncultivated purple bacterium classified into the gamma subclass (Béjà et al., 2000). When functionally expressed in E. coli this protein exhibited a photocycle characteristic of archaeal proton-pumping rhodopsins
and functioned as a light-driven proton pump (Beja et al., 2000). The first eubacterial sensory rhodopsin, a green light-activated photoreceptor, was found in Anabaena sp. PCC7120 (Jung et al., 2003). The same study suggested that this rhodopsin signals through a soluble cytoplasmic protein. More recently a gene encoding an archaeal sensory rhodopsin II (SRII) like protein was identified in the genome of the purple bacterium Tch. tepidum (Kyndt et al., 2004). The sequenced genome of other purple bacteria, Rba. sphaeroides, Rba. capsulatus, Rsp. rubrum and Rps. palustris, do not contain genes for opsin family members. The SRII gene of Tch. tepidum is localized next to a gene cluster for chemotaxis related proteins and HtrII, a transducer for SRII family proteins presumably involved in negative phototaxis of this bacterium (Kyndt et al., 2004) (Fig. 2). The presence of sensory rhodopsins in all domains of life suggests an early evolutionary origin (Braatsch and Klug, 2004a). Crystal structures of archaeal rhodopsins and sensory rhodopsins indicated that these proteins share a nearly identical positioning of seven transmembrane helices with a similar binding-pocket of the all-trans retinal chromophore that is bound to the conserved Lys residue in helix G as a protonated Schiff base. The chromophore-binding pocket is comprised of residues from each of the seven helices, and some of them are substituted in order to tune the wavelengths of light absorbed by the chromophore, photocycle kinetics, and functional characteristics. The photocycle reaction of rhodopsins is initiated by a light-induced 13-cis isomerization of the bound retinal, which is followed by several intermediates (K, L, M, N, and O intermediates) characterized by different absorption maxima. During the photocycle reaction, a proton is transferred through the retinal Schiff base and three carboxylate residues (Asp and/or Glu). Most sensory
Chapter 36
Regulation of Genes by Light
rhodopsins lack one of the carboxylate residues used for the Schiff base proton donor, and do not pump protons when complexed with their cognate transducers. This causes a longer life-time of the M intermediate that is suggested as the signaling state in sensory rhodopsins. The crystal structure of the SRII-HtrII complex from the archaeon Natronobacterium pharaonis indicates that the transmembrane segments of dimeric HtrII are composed of four helices comprised of TM1 and TM2 for one monomer (Gordeliy et al., 2002). TM2 of each HtrII monomer is contacted by helices G and F of SRII by extensive van der Waals interactions with a few hydrogen bonds between Tyr199 of SRII and Asn74 of HtrII, indicating a role of TM2 as a signal-propagating helix. Based on the structure as well as on spectroscopic data, the authors proposed a model of signal transduction mechanism in the SRII-HtrII complex in which light-dependent conformational changes in helix F of SRII are crucial for triggering a rotation of the cytoplasmic end of TM2 in HtrII, that sends a signal for activating downstream domain components. However, in the case of SRII of the purple bacterium Tch. tepidum, the conserved Tyr residue making a hydrogen bond with HtrII (Tyr199 for archaeal SRII) is not conserved and is changed to Ile, suggesting that the interaction mechanism is somewhat different. To date, neither SRII nor HtrII of Tch. tepidum has been characterized further. C. Phototropin-Related Proteins Phototropins represent the main photoreceptors for phototropism in plants and undergo a blue light-dependent phosphorylation (Briggs et al., 2001). Plant phototropins possess two N-terminal photoactive light oxygen voltage (LOV) domains, and a C-terminal serine/threonine kinase domain. Each LOV domain non-covalently binds one molecule of flavin mononucleotide (FMN), and works as a blue-light sensing domain (Salomon et al., 2000). The VIVID protein of Neurospora, a photoreceptor involved in the entrainment of the circadian clock (Schwerdtfeger and Linden, 2003; Elvin et al., 2005), consists only of the LOV domain, suggesting that it signals by protein-protein interaction to an output domain. LOV domains containing the photoactive flavin consensus sequence are found in many bacteria (Crosson et al., 2003; Losi, 2004). Bacterial phototropin-like proteins combine a single LOV domain to a variety of output domains: histidine kinase domains, response regulator domains, EAL and GGDEF domains, and
731 STAS domains (sigma factor antagonist). The Rba. sphaeroides genome encodes a LOV domain without additional output domain similar to that of the Neurospora VIVID protein. Knockout of the corresponding gene influences the blue light-dependent expression of few proteins, yet to be identified (Hendrischk and Klug, unpublished data). To date nothing is known about the signaling pathway initiated by the Rba. sphaeroides LOV-containing protein. The variety of domain compositions of LOV-containing proteins indicates that the LOV domain is a light-sensing module for different output domains. How does the LOV domain convert blue-light signals into protein structural change for controlling regulatory domains? It has been established that light-absorption in the FMN chromophore results in formation of a flavin triplet, followed by transient formation of a covalent bond with a conserved Cys residue at C4(a) position of the flavin ring. The exact reaction pathways for the flavin-cysteinyl adduct formation following the activated triplet state have not been completely elucidated (Salomon et al., 2000; Swartz et al., 2001; Kennis et al., 2003; Kottke et al., 2003; Schleicher et al., 2004; Sato et al., 2005; Chapter 41, Hendriks et al.). Crystal structures of phototropin LOV domains in the dark and light states indicated that there is only a slight structural change limited to the chromophore-binding pocket (Crosson and Moffat, 2001, 2002). However, Fourier transform infrared (FTIR) and nuclear magnetic resonance (NMR) studies have detected global structural changes in the protein moiety upon light irradiation (Swartz et al., 2002; Harper et al., 2003; Iwata et al., 2003). Importantly, structural changes detected in the NMR spectra are coupled to the Cterminal Jα helix contacted on the solvent-exposed surface of the β-sheet in the core LOV domain, and the interaction is broken by light excitation (Harper et al., 2004). Adduct formation in the FMN-binding site triggers conformational changes of the β-sheet, perhaps through ~180° rotation of the conserved Gln residue (Nozaki et al., 2004), causing partial unfolding of the Jα helix that leads to kinase activation for phototropin. The rotation of the Gln residue is also suggested as a key step in the activation processes of the other flavin-based photoreceptor BLUF. Given that the hydrogen bond network in the flavin-binding pocket and the overall protein topology of LOV and BLUF domains resemble each other (Anderson et al., 2005), it is not surprising that the two photoreceptors have similar photo-activation mechanisms.
732 Bacterial LOV-containing proteins were photochemically characterized for the YtvA protein of Bacillus subtilis comprising a STAS output domain (Losi et al., 2002) and for PpSB2-LOV from Pseudomonas putida with an alpha helical C-terminal segment (Krauss et al., 2005), in each case before the biological functions of these proteins was known. A role of YtvA in the sigma B-dependent general stress response of Bacillus subtilis was recently revealed (Ávila-Pérez et al., 2006; Gaidenko et al., 2006). No phototropins from anoxygenic phototrophic bacteria have been characterized to date. It can be expected that further investigations of these proteins will lead to new insights into light-dependent signaling in bacteria. D. BLUF Domain Proteins Shimada et al. (1992) showed the repression by blue light of photosynthesis genes of Rba. sphaeroides grown at intermediate oxygen tension. It took ten more years to identify the AppA protein as photoreceptor of this response (Braatsch et al., 2002; Masuda and Bauer, 2002). This blue light repression does not occur at very low oxygen tension, and during anaerobic growth blue light even stimulates photosynthesis gene expression (Braatsch et al., 2002). AppA was originally identified as a protein involved in the redox control of photosynthesis genes (Gomelsky and Kaplan, 1995a). A genetic approach revealed the PpsR repressor of photosynthesis genes (Gomelsky and Kaplan, 1995b), which acts as an antagonist of AppA (Gomelsky and Kaplan, 1997). Light perception by AppA involves a flavin chromophore (Braatsch et al., 2002), which is bound to a new type of protein domain named BLUF for ‘sensors of Blue Light Using FAD’ (Gomelsky and Klug, 2002). The BLUF domain fused to adenylyl cyclase domains is also involved in a blue light response of the unicellular alga Euglena (E.) gracilis (Iseki et al., 2002). Interestingly, one of the BLUF domains of the PACα subunit from E. gracilis could restore the blue light-dependent gene expression in the Rba. sphaeroides appA– mutant, when fused to the C-terminal domain of AppA (Han et al., 2004). Furthermore, the BLUF domain could also signal to the C-terminal part of AppA, when the two domains were not covalently linked (Han et al., 2004). This demonstrated the modular character of the BLUF domain that can signal to different output domains. The C-terminal part of AppA, which shows no similarities to known protein domains, is
Gabriele Klug and Shinji Masuda sufficient to mediate redox control of photosynthesis genes if the BLUF domain is deleted (Gomelsky and Kaplan, 1998), implying that this domain interacts with the repressor PpsR (Han et al., 2004). This suggests that the BLUF domain of AppA interacts with the C-terminal domain competitively with PpsR in a light-dependent manner (Fig. 2). As reported in a previous chapter (Chapter 40, Evans et al.) the interaction of bacteriophytochrome to PpsR controls light-dependent expression of photosynthesis genes in Rps. palustris (Girauld et al., 2002) (Fig. 2). Apparently, expression of photosynthesis genes can be controlled by different light qualities due to interaction of PpsR with different photoreceptors. PpsR harbors two PAS domains and binds to a DNA consensus sequence (TGT-N12-ACA) in the proximity of promoters and represses photosynthesis genes in Rhodobacter species (Penfold and Pemberton, 1994; Gomelsky and Kaplan, 1995b; Ponnampalam et al., 1995). Some purple bacteria contain more than one PpsR protein and different effects on gene expression (repressing and activating, redox-dependent or not) have been observed (Elsen et al., 2005). Although most photosynthesis genes of the Rba. sphaeroides PpsR regulon contain PpsR binding sites (Moskvin et al., 2005), this is not true for the puf operon that encodes proteins required for the formation of lightharvesting I and reaction center complexes, although this operon is clearly under control of PpsR. The mechanisms by which PpsR affects puf transcription are not known to date. The DNA binding activity of PpsR of Rba. sphaeroides (and CrtJ, its counterpart in Rba. capsulatus) has been proposed to be regulated through oxidation and reduction of conserved Cys residues (Ponnampalam and Bauer, 1997; Masuda and Bauer, 2002; Masuda et al., 2002); however, the regulated modification of Cys is not involved in the light-dependent regulation of PpsR activity (Masuda and Bauer, 2002; Cho et al., 2004). Biochemical analysis showed that AppA releases PpsR from the DNA only in dark or low-light conditions, and that this effect results from the light-induced dissociation of PpsR from an active tetramer form to an inactive dimer form (Masuda and Bauer, 2002). Initial analysis using recombinant His-tagged proteins indicated a light-dependent complex formation of monomeric AppA with dimeric PpsR for the anti-repression. Further investigations still need to fully unravel how the illumination of the BLUF domain or redox state affect AppA-PpsR interaction. Several BLUF-domain containing proteins have
Chapter 36
Regulation of Genes by Light
been genetically characterized, which include the E. gracilis PAC protein involved in blue-light induced photo-avoidance response (Iseki et al., 2002), Synechocystis Slr1694 involved in the phototaxis response (Masuda and Ono, 2004; Okajima et al., 2005), and the E. coli YcgF (Blrp) protein involved in the light-dependent cell motility response (Masuda and Takamiya, unpublished). These results suggest that BLUF proteins are conserved in many microorganisms and are involved in the avoidance of photodamage from strong light and/or UV light. In these BLUF proteins, a variety of regulatory domains are fused to the BLUF domain, indicating that BLUF is a light ‘input’ domain that modulates the activities of different regulatory ‘output’ domains. In this sense, the BLUF domain is functionally similar to the LOV domain. Some bacterial proteins mainly consist of the BLUF domain without any output domain and are believed to signal to other proteins by protein-protein interaction (Gomelsky and Klug, 2002; Han et al., 2004). AppA is the only BLUF protein whose downstream component has been revealed, the transcriptional repressor PpsR. The photocycle reaction of BLUF proteins was first discovered in AppA. Upon blue-light irradiation, AppA showed a unique photocycle, which is characterized by a long-lived (approximately 30 min) 10 nm red-shift of a UV-visible absorption spectrum (Masuda and Bauer, 2002). The spectral changes are different from those of other flavin-based photoreceptors such as the phototropin LOV domain and/or cryptochromes. Thus, the molecular mechanisms of the photochemical reaction of BLUF must be unique to BLUF, and they are still not fully understood. A fluence-response curve generated from a series of light-minus-dark difference spectra shows that the photocycle is induced at a low-light intensity of 5 µmol m–2 s–1 with saturation of the cycle occurring at a light intensity of 400 µmol m–2 s–1. These values correlate with physiological light-intensities needed for repression of photosystem synthesis. Furthermore, light-excited AppA needs more than 20 min to recover its antirepressor activity when shifted to dark conditions (Masuda and Bauer, 2002). These results suggest that the light-induced spectral red-shift corresponds to a signaling state which converts AppA from a functional antirepressor in the dark form into an inactive state in the light form. Later studies indicated that other BLUF photoreceptors, such as Slr1694, PAC, Tll0078, YcgF and BlrB show similar spectral red shift upon light irradiation (Masuda et
733 al., 2004; Rajagopal et al., 2004; Fukushima et al., 2005; Ito et al., 2005; Jung et al., 2005), suggesting that this reaction is conserved among all BLUF photoreceptors. Ultrafast absorption spectroscopy revealed that the signaling state showing 10 nm red-shift is formed directly from the singlet-excited flavin within 10–8 sec after light absorption (Gauden et al., 2005). The quantum yield of the signaling state formation in AppA is calculated to be ~25%, and a compatible value was reported for other BLUF proteins (Fukushima et al., 2005; Gauden et al., 2005, 2006; Zirak et al., 2005; Chapter 41, Hendriks et al.). Interestingly, the dark decay kinetics in the photocycle of each BLUF protein are very different with half lives of ~5 sec to ~10 min. The different decay kinetics may reflect the physiological light conditions where these organisms exist in nature. Recently, crystal structures of BLUF domains of AppA, Tll0078 and BlrB were determined (Anderson et al., 2005; Jung et al., 2005; Kita et al., 2005). The backbone structures of these BLUF domains are very similar, consisting of a five-stranded mixed β-sheet with two α-helices running parallel with it. A flavin molecule is sandwiched between the α-helices. The dimethyl benzene ring of the flavin is packed by hydrophobic side-chains and the pyrimidine ring of flavin is hydrogen bonded with specific amino acids including conserved Gln63, Tyr21 and Asn45 for AppA. FTIR and Raman spectroscopic analyses of several BLUF proteins have indicated that the photocycle reaction of BLUF is accompanied by small structural changes of the flavin chromophore with alternations of chromophore-apoprotein interactions, including changes in the hydrogen bond network around the flavin-binding pocket (Kraft et al., 2003; Laan et al., 2003; Masuda et al., 2004, 2005a; Unno, et al., 2005). Specifically, light-induced creation of a new hydrogen bond is proposed between an amino-acid side chain and the C4=O carboxyl oxygen of the flavin. From the spectroscopic data as well as the crystal structures, a possible model was proposed for the BLUF photocycle in which blue-light absorption by flavin results in a rotation of a conserved Gln (Gln63 for AppA) side chain by ~180° to form a new hydrogen bond with the flavin C4=O group (Anderson et al., 2005). The hydrogen bond to C4=O may result in the 10nm red-shift in the bound flavin in the visible region of the spectrum (Hasegawa et al., 2005; Unno et al., 2005). It was proposed that excited flavin transiently accepts an electron and a proton from the conserved Tyr residue (corresponds to Tyr21 in AppA) resulting
734 in the formation of anionic and neutral flavin radical species (Dragnea et al., 2005; Gauden et al., 2006). This reaction may induce a rearrangement of the hydrogen-bond network around the flavin-binding pocket including Gln63, Tyr21 and Trp104 for AppA (Unno et al., 2006). The alternation of the hydrogenbond network is critical for the β-sheet structural changes detected by FTIR and NMR spectroscopy (Masuda et al., 2005b; Grinstead et al., 2006), which may be critical for transferring the light signal to the next component such as PpsR for AppA. Jung et al. (2005) also presented an alternative model in which the excitation of flavin stimulates changes in the hydrogen bond network around C2=O involving a proton transfer from the conserved Arg residue to the carboxyl oxygen. This influences output modules such as the adenine moiety of the flavin, which is located on the opposite side of the β-sheet. Hence, the two models are completely inconsistent with each other, and further photochemical and spectral analyses are needed in order to clarify the exact mechanisms of the BLUF photocycle. E. Cryptochromes Although the exact photocycle could not be solved up to now, it is widely accepted that cryptochromes are blue light photoreceptors in plants (Lin, 2002) and circadian clock components in animals (Panda et al., 2002). Cryptochromes and photolyases belong to the same family of proteins. Photolyases catalyze blue/UV light-dependent repair of DNA damage resulting from exposure to high-energy short-wavelength (< 350 nm) UV light (UV-B and UV-C). This process, based upon light-induced reversible electron transfer, is called photoreactivation. Cryptochromes do not show the activity of photoreactivation. Proteins of the cryptochrome/photolyase family contain two non-covalently bound chromophore cofactors. Reduced FAD is the catalytic chromophore during photoreactivation by photolyases, and in addition it functions in blue light perception in cryptochromes. The second cofactor is either metenyltetrahydrofolate or 8-hydroxy-7,8,-dimethyl-5-deazariboflavin and serves as photoantenna (Sancar, 2004). So far, no functional bacterial cryptochrome has been identified. The function of bacterial proteins of the cryptochrome/photolyase family that are not involved in DNA repair is mostly not known. The sll1629 gene product of Synechocystis sp. PCC 6803 was designated as a cryptochrome since it has no function in
Gabriele Klug and Shinji Masuda photoreactivation (Hitomi et al., 2000). Brudler et al. (2003) demonstrated that the Sll1629 gene product can bind DNA. Transcription profiling using DNA microarrays revealed two open reading frames of unknown function that showed significantly enhanced expression in the mutant when compared to the wild type (Brudler et al., 2003). Thus, Sll1629 appears to affect expression of only a minor subset of genes. However, a light-dependent function of the gene product was not experimentally proven. In plants light-regulated protein degradation appears to be central to cryptochrome signaling (Chen et al., 2004). Animal and plant cryptochromes interact with the E3 ubiquitin ligase COP1. COP1 degrades several transcription factors in the dark; light prevents this degradation. Cryptochrome-COP1 interaction occurs in both the light and the dark. Most likely the light-driven conformational modification of cryptochrome induces a structural modification of COP1. Cryptochromes also interact with a number of other proteins including phytochromes (Chen et al., 2004), but the functional implications of these interactions are still unclear. The photoactivation mechanism of cryptochromes is largely unknown. Given that photolyases use light-driven electron transfer from the reduced flavin chromophore FADH– to the substrate for catalysis, similar processes may be involved in the photoactivation processes of cryptochrome. Indeed, it was reported for Arabidopsis cry1 that electron transfer could be involved in cryptochrome signaling (Giovani et al., 2003). Purified cry1 from baculovirus-transfected insect cells contains fully oxidized FAD, and upon photoexcitation a semireduced radical FADH• is formed concomitantly with a neutral Trp radical. This suggests that the FAD chromophore is photoreduced in cryptochromes as in photolyases involving Trp (and/or Tyr) radicals. However, physiological electron donors and acceptors for the flavin radical species have not been identified, although these may mediate cryptochrome light signaling (Zeugner et al., 2005). Genes encoding proteins of the cryptochrome/photolyase family are also found in the genomes of purple photosynthetic bacteria. Rba. capsulatus encodes a single protein of this family that was shown to function as photolyase (Braatsch and Klug, 2004b). Its relative, Rba. sphaeroides, encodes three proteins of this family. Proteins with high similarity to such putative cryptochromes are also present in Rubrivivax (Rvi.) gelatinosus and Roseobacter denitrificans. It is
Chapter 36
Regulation of Genes by Light
likely that such proteins in the purple photosynthetic bacteria are involved in blue light-dependent gene expression as suggested for cyanobacteria. However, this hypothesis still awaits verification. F. Photoactive Yellow Protein The PYP proteins resemble the PAS domain and are present in several purple photosynthetic bacteria. PYP was first described for the halophilic purple bacterium Ectothiorhodospira halophila, which swims away from blue light (Meyer, 1985) and evidence was presented that it functions as blue light photoreceptor for this response (Sprenger et al., 1993). The chromophore of PYP, later also designated xanthopsin (Kort et al., 1996), is p-coumaric acid. PYP is also part of the phytochrome like Ppr protein of Rc. centenaria. Since a separate chapter of this book deals with PYP (Chapter 41, Hendriks et al.), it is not discussed further here. III. Light-Dependent Responses that Do Not Depend on Photoreceptors A. Photosynthetic Electron Transport For many years it was anticipated that photosynthetic electron transport is the major factor in controlling bacterial responses to light. Despite the discovery of several bacterial photoreceptors for the perception of specific light qualities, some responses indeed depend on photosynthetic electron transport and are activated by light qualities that are absorbed by bacteriochlorophyll. This is the case for some phototactic responses but also for photosynthesis gene expression in Rhodobacter species. Early studies demonstrated that the action spectrum for phototaxis in Rsp. rubrum is closely related to that for photosynthesis (Clayton, 1963). Mutants in components of the photosynthetic complexes of Rsp. rubrum showed impaired phototaxis (Weaver, 1971). Photosynthetic electron transport and pigment content of Rba. sphaeroides were also shown to affect phototactic responses (Armitage et al., 1985; Grishanin et al., 1997; Romagnoli and Armitage, 1999). Unlike E. coli, Rba. sphaeroides harbors several chemosensory operons. Mutants in the first chemosensory operon (cheOp1) show wild type photoresponses, while mutants in cheA2 of the second operon show inverse photoresponse (Romangnoli and Armitage, 1999;
735 Chapter 32, Armitage). Likewise, other components of the second chemosensory operon and all components of the third operon, CheOp3, except CheB2 and Slp (putative chromosomal segregation protein), are required for phototaxis (Porter et al., 2002). Interestingly, the operon comprising the cheA2, cheW2 and cheW3 genes is under control of the response regulator PrrA (Armitage and Hellingwerf, 2003). As discussed below the two-component system PrrB/PrrA in Rba. sphaeroides (called RegB/RegA in Rba. capsulatus) which affects expression of photosynthesis genes depends on signals transmitted by photosynthetic and respiratory electron transport (Elsen et al., 2004; Happ et al., 2005). While light-dependent synthesis of photosynthetic complexes was assigned to electron transport-dependent signals in the past (Oh and Kaplan, 2000), several photoreceptors have now been identified and shown to regulate photosynthesis gene expression in different species. The AppA/PpsR system was found to be responsible for the blue light-dependent repression of photosynthesis genes at intermediate oxygen tension in Rba. sphaeroides (Braatsch et al., 2002; Masuda and Bauer, 2002). However, at very low oxygen tension or under anaerobic conditions blue light stimulates photosynthesis gene expression without the involvement of the AppA/PpsR system (Braatsch et al., 2002; Happ et al., 2005). A similar observation was made for Rba. capsulatus that lacks the AppA protein (Happ et al., 2005). This response turned out not to be specific for blue light but to be mediated also by other wavelengths absorbed by bacteriochlorophyll in the photosynthetic apparatus. Mutant analyses revealed that induction of photosynthesis genes by light under anaerobic conditions requires photosynthetic electron transport, components of the respiratory chain, and the PrrB/PrrA(RegB/RegA) two component system. PrrB(RegB) is a sensor kinase that is autophosphorylated in response to redox changes in the photosynthetic and respiratory electron transport chains. Mechanisms for sensing electron flow by PrrB have not been fully elucidated and two different models have emerged, based on several genetic and biochemical data. One is that PrrB kinase activity is inhibited by the extent of electron flow through the cbb3 oxidase (Kaplan et al., 2005), and the other model is that oxidized quinone directly inhibits the PrrB kinase activity (Swem et al., 2006). The influence of cbb3 on PrrB activity and electron flow through cbb3 even under anaerobic conditions had been demonstrated previously in Rba. sphaeroi-
736 des (Oh and Kaplan, 2002; Oh et al., 2004), and the redox status of the quinone pool is also known to be drastically changed depending on growth conditions under different oxygen tensions and light intensities (Takamiya and Takamiya, 1969). Thus, both electron flow through the cbb3 oxidase and the redox status of the quinone pool could control PrrB activity, although available evidence is insufficient to establish the correct mechanism. A mutant lacking PrrB showed light-dependent repression of photosynthesis genes, even under anaerobic conditions, revealing that the repressing effect of AppA/PpsR is present under these conditions but overwritten by PrrA dependent activation (Happ et al., 2005). These results supported the close interplay of the two regulatory systems in controlling expression of photosynthesis genes that guarantees adequate responses to changes in environmental conditions. Both systems are redox responsive and both systems can transmit light signals, in the case of AppA with involvement of a photoreceptor, in the case of PrrB/PrrA by sensing electron transport. Buggy et al. (1994) found a trans-acting factor HvrA from Rba. capsulatus that appeared to be responsible for light-mediated regulation of several photosynthesis genes, and the corresponding factor, Spb, was also identified as putative light regulator in Rba. sphaeroides by Shimada et al. (1996). HvrA was first identified as a low-light activator for puf and puh operons under anaerobic conditions (Buggy et al., 1994), but it also works as a repressor for several photosynthesis genes under aerobic conditions (Masuda and Bauer, unpublished). Spb was identified as a high-light repressor for photosynthesis genes under anaerobic conditions (Nishimura et al., 1998). The two proteins show 53% amino-acid identity with a putative helix-turn-helix DNA-binding motif, suggesting that these proteins are functionally similar. Later studies have indicated that HvrA/Spb belongs to a family of H-NS (histone-like nucleoid structuring) proteins conserved in many bacteria. This protein family, which mostly binds to curved DNA with no consensus sequences, negatively and/or positively influences the activities of many promoters on the genome through changing local DNA topology (Dorman and Deighan, 2003). These results suggest that HvrA/Spb regulates the expression of not only photosynthesis genes but also many other genes through changing chromosome structure (Masuda and Bauer 2004). The expression of HvrA is under the control
Gabriele Klug and Shinji Masuda of the PrrB/PrrA (RegB/RegA) two-component system (Du et al., 1999), suggesting that light- and oxygen-mediated changes of redox status of electron transport chains are critical for the HvrA/Spb-dependent regulation of photosynthesis genes. Rcs. centenaria shows a unique response to light. Cells grown on an agar surface differentiate into multi-flagellated swarm cells and allow colonies to move at speeds higher than 75 mm per hour dependent on light intensity and light direction (Ragatz et al., 1994, 1995). Rcs. centenaria colonies are attracted by infra-red light of wavelengths around 800–850 nm, whereas they are repelled by light at 590 and 475–550 nm. This response was shown to depend on photosynthetic electron transport (Jiang et al., 1998). Interestingly, the wavelengths that cause both attraction and repulsion are absorbed by bacteriochlorophyll. It is not known to date which downstream signaling events are responsible for the different responses. The phytochrome-PYP fusion protein Ppr which was later identified in this species is not required for the light-dependent colony movement (Jiang et al., 1999). B. Singlet Oxygen Some light responses have been shown to be independent of photoreceptors even in bacteria that lack photosynthetic electron transport. For example, E. coli shows a photophobic response at high fluence rates that is also observed at lower fluence rates in the presence of photosensitizers (Macnab and Koshland, 1974). Photosensitizers like methylene blue are known to generate singlet oxygen, a highly reactive molecule that is also naturally generated by the simultaneous presence of light, oxygen and chlorophylls or other tetrapyrroles. The accumulation of reactive oxygen species in the presence of light is called photooxidative stress. Although most organisms, especially photosynthetic organisms, have to defend against photooxidative stress, our knowledge of the underlying mechanisms is limited. A transcriptome analysis for Rba. sphaeroides identified genes that respond to blue light (Braatsch et al., 2004). Many of these genes were repressed by blue light and belong to the AppA/PpsR regulon. The expression of other genes however was increased and based on the promoter sequences it was predicted that some of these genes belong to the sigma-E regulon (Braatsch et al., 2004). Sigma-E is a member of
Chapter 36
Regulation of Genes by Light
the extra-cytoplasmatic function (ECF) family of sigma factors. ChrR belongs to a new family of zinc containing anti-sigma factors (Newman et al., 2001) and inhibits the ability of sigma-E to form a stable complex with core RNA polymerase by the formation of a sigma-E:ChrR complex (Anthony et al., 2004). Anthony et al. (2005) demonstrated that singlet oxygen increased sigma-E activity, but the mechanisms of activation await identification. Carotenoids are known as potent quenchers of singlet oxygen and Rba. sphaeroides and Rvi. gelatinosus carotenoid mutants are sensitive even to low light intensities (Anthony et al., 2005; Glaeser and Klug, 2005). In carotenoid-less mutants of Rvi. gelatinosus photooxidative stress stimulates illegitimate recombination and mutability (Ouchane et al., 1997). Despite their protective function, carotenoids do not accumulate in Rba. sphaeroides under photooxidative stress and expression of carotenoid synthesis genes is not increased (Glaeser and Klug, 2005). IV. Concluding Remarks Although our knowledge of light sensing by purple photosynthetic bacteria has increased remarkably over the last few years, our understanding of the molecular basis of sensing and signaling is still at the very beginning. Genome sequencing has revealed the presence of photoreceptor homologs in several species, but the function of most of these proteins still awaits elucidation. For known photoreceptors the downstream signaling pathway is not known or the interaction to the downstream partners needs to be investigated (Fig. 2). Despite this, it has become clear that purple photosynthetic bacteria have evolved different strategies and complex regulatory networks in order to respond adequately to changes in light quantity or quality in their environment. Acknowledgments Research on light regulation of photosynthesis genes of the authors was supported by Deutsche Forschungsgemeinschaft and Fonds der Chemischen Industrie (G.K) and Sumitomo Foundation, Foundation for Opto-Science & Technology and Grant-in-Aid from MEXT of Japan (S.M.), respectively.
737 References Anderson S, Dragnea V, Masuda S, Ybe J, Moffat K and Bauer C (2005) Structure of a novel photoreceptor, the BLUF domain of AppA from Rhodobacter sphaeroides. Biochemistry 44: 7998–8005 Anthony JR, Newman JD and Donohue TJ (2004) Interactions between the Rhodobacter sphaeroides ECF sigma factor, σ E, and its anti-sigma factor, ChrR. J Mol Biol 341: 345–360 Anthony JR, Warczak KL and Donohue TJ (2005) A transcriptional response to singlet oxygen, a toxic byproduct of photosynthesis. Proc Natl Acad Sci USA 102: 6502–6507 Armitage JP and Hellingwerf KJ (2003) Light-induced behavioural responses (Phototaxis) in prokaryotes. Photosynth Res 76: 145–155 Armitage JP, Ingham C and Evans MC (1985) Role of proton motive force in phototactic and aerotactic responses of Rhodopseudomonas sphaeroides. J Bacteriol 161: 967–972 Ávila-Pérez M, Hellingwerf KJ and Kort R (2006) Blue light activates the sigma-E dependent stress response of Bacillus subtilis via YtvA. J Bacteriol 188: 6411–6414 Béjà O, Aravind L, Koonin EV, Suzuki MT, Hadd A, Nguyen LP, Jovanovich SB, Gates CM, Feldman RA, Spudich JL, Spudich EN and DeLong EF (2000) Bacterial rhodopsin: Evidence for a new type of phototrophy in the sea. Science 289: 1902–1906 Bhoo S, Davis SJ, Walker J, Karniol B and Vierstra RD (2001) Bacteriophytochromes are photochromic histidine kinases using a biliverdin chromophore. Nature 414: 776–779 Braatsch S and Klug G (2004a) Blue light perception in bacteria. Photosynth Res 79: 45–57 Braatsch S and Klug G (2004b) ORF90, a gene required for photoreactivation in Rhodobacter capsulatus SB1003 encodes a cyclobutane pyrimidine dimer photolyase. Photosynth Res 79: 167–177 Braatsch S, Gomelsky M, Kuphal S and Klug G (2002) A single flavoprotein, AppA, can integrate both redox and light signals in Rhodobacter sphaeroides. Mol Microbiol 45: 827–836 Braatsch S, Moskvin O, Klug G and Gomelsky M (2004) Responses of the Rhodobacter sphaeroides transcriptome to blue light under semiaerobic conditions. J Bacteriol 186: 7726–7735 Braatsch S, Johnson JA, Noll K and Beatty JT (2007) The O2responsive repressor PpsR2 but not PpsR1 transduces a light signal sensed by the BphP1 phytochrome in Rhodopseudomonas palustris CGA009. FEMS Microbiol Lett 272: 60–64 Briggs WR, Christie JM and Samomon M (2001) Phototropins: A new family of flavin-binding blue light receptors in plants. Antioxid Redox Signal 3: 775–788 Brudler R, Hitomi K, Daiyasu H, Toh H, Kucho KI, Ishiura M, Kanehisa M, Roberts VA, Todo T, Tainer JA and Getzoff ED (2003) Identification of a new cryptochrome class: Structure, function, and evolution. Mol Cell 11: 59–67 Buggy JJ, Sganga MW and Bauer CE (1994) Characterization of a light-responding trans-activator responsible for differentially controlling reaction center and light-harvesting-I gene expression in Rhodobacter capsulatus. J Bacteriol 176: 6936–6943 Catlett NL, Yoder OC and Turgeon BG (2003) Whole-genome analysis of two-component signal transduction genes in fungal pathogens. Eukayotic Cell 2: 1151–1161
738 Chen M, Chory J and Frankhauser C (2004) Light signal transduction in higher plants. Ann Rev Genet 38: 87–117 Cho SH, Youn SH, Lee SR, Yim HS and Kang SO (2004) Redox property and regulation of PpsR, a transcriptional repressor of photosystem gene expression in Rhodobacter sphaeroides. Microbiology 150: 697–706 Chung Y-H, Masuda S and Bauer CE (2007) Purification and reconstitution of PYP-phytochrome (Ppr) with biliverdin and 4-hydroxycinnamic acid. Methods Enzymol 422: 184–189 Clayton RK (1963) An automatic energy switch in photosynthesis. J Theor Biol 5:497–499 Crosson S and Moffat K (2001) Structure of a flavin-binding plant photoreceptor domain: Insight into light-mediated signal transduction. Proc Natl Acad Sci USA 98: 2995–3000 Crosson S and Moffat K (2002) Photoexcited structure of a plant photoreceptor domain reveals a light-driven molecular switch. Plant Cell 14: 1067–1075 Crosson S, Rajagopal S and Moffat K (2003) The LOV domain family: Photoresponsive signaling modules coupled to diverse output domains. Biochemistry 42: 2–10 Davis SJ, Vener AV and Vierstra RD (1999) Bacteriophytochromes: Phytochrome-like photoreceptors from nonphotosynthetic eubacteria. Nature 286: 2517–2520 Dorman CJ and Deighan P (2003) Regulation of gene expression by histone-like proteins in bacteria. Curr Opin Gen Dev 13: 179–184 Dragnea V, Waegele M, Balascuta S, Bauer C and Dragnea B (2005) Time-resolved spectroscopic studies of the AppA bluelight receptor BLUF domain from Rhodobacter sphaeroides. Biochemistry. 44: 15978–85 Du S, Kouadio JLK and Bauer CE (1999) Regulated expression of a highly conserved regulatory gene cluster is necessary for controlling photosynthesis gene expression in response to anaerobiosis in Rhodobacter capsulatus. J Bacteriol 181: 4334–4341 Elsen S, Swem LR, Swem DL and Bauer CE (2004) RegB/RegA, a highly conserved redox-responding global two-component regulatory system. Microbiol Mol Biol Rev 68: 263–279 Elsen S, Jaubert M, Pignol D and Giraud E (2005) PpsR: A multifaceted regulator of photosynthesis gene expression in purple bacteria. Mol Microbiol 57: 17–26 Elvin M, Loros JJ, Dunlap JC and Heintzen C (2005) The PAS/LOV protein VIVID supports a rapidly dampened daytime oscillator that facilitates entrainment of the Neurospora circadian clock. Genes Dev 19: 2593–2605 Evans K, Fordham-Skelton AP, Mistry H, Reynolds CD, Lawless AM and Papiz MZ (2005) A bacteriophytochrome regulates the synthesis of LH4 complexes in Rhodopseudomonas palustris. Photosynth Res 85: 169–180 Fukushima Y, Okajima K, Shibata Y, Ikeuchi M and Itoh S (2005) Primary intermediate in the photocycle of a blue-light sensory BLUF FAD-protein, Tll0078, of Thermosynechococcus elongatus BP-1. Biochemistry 44: 5149–5158 Gaidenko TA, Kim T-J, Weigel AL, Brody MS and Price CW (2006) The blue light receptor YtvA acts in the environmental stress signaling pathway of Bacillus subtilis. J Bacteriol. 188: 6387–6395 Gauden M, Yeremenko S, Laan W, van Stokkum IHM, Ihalainen JA, van Grondelle R, Hellingwerf KJ and Kennis JTM (2005) Photocycle of the flavin-binding photoreceptor AppA, a bacterial transcriptional antirepressor of photosynthesis genes.
Gabriele Klug and Shinji Masuda Biochemistry 44: 3653–3662 Gauden M, van Stokkum IHM, Key JM, Luhrs DC, van Grondelle R, Hegemann P and Kennis JTM (2006) Hydrogen bond switching via radical pair mechanism in a flavin-binding photoreceptor. Proc Natl Acad Sci 103:10895–10900 Giovani B, Byrdin M, Ahmad M and Brettel K (2003) Lightinduced electron transfer in a cryptochrome blue-light photoreceptor. Nat Struct Biol 10: 489–490. Giraud E, Fardoux J, Fourrier N, Hannibal L, Genty B, Bouyer P, Dreyfus B and Verméglio A (2002) Bacteriophytochrome controls photosystem synthesis in anoxygenic bacteria. Nature 417: 202–205 Giraud E, Zappa S, Vuillet L, Adriano JM, Hannibal L, Fardoux J, Berthomieu C, Bouyer P, Pignol D and Verméglio A (2005) A new type of bacteriophytochrome acts in tandem with a classical bacteriophytochrome to control the antennae synthesis in Rhodopseudomonas palustris. J Biol Chem 280:32389–32397 Glaeser J and Klug G (2005) Photo-oxidative stress in Rhodobacter sphaeroides: Protective role of carotenoids and expression of selected genes. Microbiology 151: 1927–1938 Gomelsky M and Kaplan S (1998) AppA, a redox regulator of photosystem formation in Rhodobacter sphaeroides 2.4.1, is a flavoprotein. Identification of a novel FAD binding domain. J. Biol. Chem. 273:35319–35325 Gomelsky M and Kaplan S (1995a) AppA, a novel gene encoding a trans-acting factor involved in the regulation of photosynthesis gene expression in Rhodobacter sphaeroides 2.4.1. J Bacteriol 177: 4609–4618 Gomelsky M and Kaplan S (1995b) Genetic evidence that PpsR from Rhodobacter sphaeroides 2.4.1 functions as a repressor of puc and bchF expression. J Bacteriol 177: 1634–1637 Gomelsky M and Kaplan S (1997) Molecular genetic analysis suggesting interaction between AppA and PpsR in regulation of photosynthesis gene expression in Rhodobacter sphaeroides 2.4.1. J Bacteriol 179: 128–134 Gomelsky M and Klug G (2002) BLUF: A novel FAD-binding domain involved in sensory transduction in bacteria. Trends Biochem Sci 27: 497–500 Gordeliy VI, Labahn J, Moukhametzianov R, Efremov R, Granzin J, Schlesinger R, Buldt G, Savopol T, Scheidig AJ, Klare JP and Engelhard M (2002) Molecular basis of transmembrane signaling by sensory rhodopsis II-transducer complex. Nature 419: 484–487 Grinstead JS, Hsu SD, Laan W, Bonvin AMJJ, Hellingwerf KJ, Boelens R and Kaptein R (2006) The solution structure of the AppA BLUF domain: Insight into the Mechanism of lightinduced signaling. Chem Bio Chem 7: 187–193 Grishanin RN, Gauden DE and Armitage JP (1997) Photoresponses in Rhodobacter sphaeroides: Role of photosynthetic electron transport. J. Bacteriol. 179: 24–30 Han Y, Braatsch S, Osterloh L and Klug G (2004) A eukaryotic BLUF domain mediates light-dependent gene expression in the purple bacterium Rhodobacter sphaeroides 2.4.1. Proc Natl Acad Sci USA 101: 12306–12311 Happ H, Braatsch S, Broschek V, Osterloh V and Klug G (2005) A blue light dependent expression of photosynthesis genes under anaerobic conditions in Rhodobacter capsulatus and Rhodobacter sphaeroides is mediated by photosynthetic electron transport. Mol Microbiol 58: 903–914 Harper SM, Neil LC and Gardner KH (2003) Structural basis of a phototropin light switch. Science 301: 1541–1544
Chapter 36
Regulation of Genes by Light
Harper SM, Christie JM and Gardner KH (2004) Disruption of the LOV-Ja Helix interaction activates phototropin kinase activity. Biochemistry 43: 16184–16192 Hasegawa K, Masuda S and Ono T-A (2005) Spectroscopic analysis of the dark relaxation process of a photocycle in a sensor of blue light using FAD (BLUF) protein Slr1694 of the cyanobacterium Synechocystis sp. PCC6803. Plant Cell Physiol 46: 136–146 Hitomi K, Okamoto K, Daiyasu H, Miyashita H, Iwai S, Toh H, Ishiura M and Todo T (2000) Bacterial cryptochrome and photolyase: Characterization of two photolyase-like genes of Synechocystis sp. PCC6803. Nucleic Acids Res. 28:2353–2362 Hughes J, Lamparter T, Mittmann F, Hartmann E, Gärtner W, Wilde A and Börner T (1997) A prokaryotic phytochrome. Nature 386: 663 Iseki M, Matsunaga S, Murakami A, Ohno K, Shiga K, Yoshida K, Sugai M, Takahashi T, Hori T and Watanabe M (2002) A blue-light-activated adenylyl cyclase mediated photoavoidance in Euglena gracilis. Nature 415: 1047–1051 Ito S, Murakami A, Sato K, Nishina Y, Shiga K, Takahashi T, Higuchi S, Iseki M and Watanabe M (2005) Photocycle features of heterologously expressed and assembled eukaryotic flavinbinding BLUF domains of photoactivated adenylyl cyclase (PAC), a blue-light receptor in Euglena gracilis. Photochem Photobiol Sci 4: 762–769 Iwata T, Nozaki D, Tokutomi S, Kagawa T, Wada M and Kandori H (2003) Light-induced structural changes in the LOV2 domain of Adiantum phytochrome 3 studied by low-temperature FTIR and UV-visible spectroscopy. Biochemistry 42: 8183–8191 Jiang ZY, Rushing BG, Bai Y, Gest H and Bauer CE (1998) Isolation of Rhodospirillum centeneum mutants defective in phototactic colony motility by transposon mutagenesis. J Bacteriol 180: 1248–1255 Jiang ZY, Swem LR, Rushing BG, Devanathan S, Tollin G and Bauer CE (1999) Bacterial photoreceptor with similarity to photoactive yellow protein and plant phytochrome. Science 285: 406–409 Jung KH, Trivedi VD and Spudich JL (2003) Demonstration of a sensory rhodopsin in eubacteria. Mol. Microbiol. 47: 1513–1522 Jung A, Domratcheva T, Tarutina M, Wu Q, Ko WH, Shoeman RL, Gomelsky M, Gardner KH and Schlichting I (2005) Abstract structure of a bacterial BLUF photoreceptor: Insights into blue light-mediated signal transduction. Proc Natl Acad Sci USA 102: 12350–12355 Kaplan S, Eraso J and Roh JH (2005) Interacting regulatory networks in the facultative photosynthetic bacterium, Rhodobacter sphaeroides 2.4.1. Biochem Soc Trans 33: 51–55 Kehoe DM and Grossman AR (1996) Similarity of a chromatic adaptation sensor to phytochrome and ethylene receptors. Science 273: 1409–1412 Kennis JTM, Crosson S, Gauden M, Stokkum IHM, Moffat K and van Grondelle R (2003) Primary reactions of the LOV2 domain of phototropin, a plant blue-light photoreceptor. Biochemistry 42: 3385–3392 Kita A, Okajima K, Morimoto Y, Ikeuchi M and Miki K (2005) Structural of a cyanobacterial BLUF protein, Tll0078, containing a novel FAD-binding blue-light sensor domain. J Mol. Biol.349: 1–9 Kort R, Hoff WD, van West M, Kroon AR, Hoffer SM, Vlieg KH, Crielaard W, van Beeumen JJ and Hellingwerf KJ (1996) The
739 xanthopsins: A new family of eubacterial blue-light photoreceptors. EMBO J 15: 3209–3218 Kottke T, Heberle J, Hehn D, Dick B and Hegemann P (2003) Phot-LOV1: Photocycle of a blue-light receptor domain from the green Alga Chlamydomonas reinhardtii. Biophys J 84: 1192–1201 Krauss U, Losi A, Gärtner W, Jaeger KF and Eggert T (2005) Initial characterization of a blue-light sensing, phototropin-related protein from Pseudomonas putida: A paradigm for an extended LOC construct. Phys Chem Chem Phys 7: 2804–2811 Kraft BJ, Masuda S, Kikuchi J, Dragnea V, Tollin G, Zaleski JM and Bauer CE (2003) Spectroscopic and mutational analysis of the blue-light photoreceptor AppA: A novel photocycle involving flavin stacking with an aromatic amino acid. Biochemistry 42: 6726–6734 Kyndt JA, Meyer TE and Cusanovich MA (2004) Photoactive yellow protein, bacteriophytochrome, and sensory rhodopsin in purple phototrophic bacteria. Photochem Photobiol Sci 3: 519–530 Laan W, van der Horst MA, van Stokkum I.H and Hellingwerf KJ (2003) Initial characterization of the primary photochemistry of AppA, a blue-light-using flavin adenine dinucleotide-domain containing transcriptional antirepressor protein from Rhodobacter sphaeroides: A key role for reversible intramolecular proton transfer from the flavin adenine dinucleotide chromophore to a conserved tyrosine? Photochem Photobiol. 78: 290–297 Lagarias JC and Rapoport H (1980) Chromopeptides from phytochrome. The structure and linkage of the Pr form of the phytochrome chromophore. J Am Chem Soc 102: 4821–1828 Lamparter T, Esteban B and Hughes J (2001) Phytochrome Cph1 from the cyanobacterium Synechocystis PCC6803-Purification, assembly, and quaternary structure. Eur J Biochem 268: 4720–4730 Lamparter T, Carrascal M, Michel N, Martinez E, Rottwinkel G and Abian J (2004) The biliverdin chromophore binds covalently to a conserved cysteine residue in the N-terminus of Agrobacterium phytochrome Agp1. Biochemistry 43: 3659–3669 Larimer FW, Chain P, Hauser L, Lamerdin J, Malfatti S, Do L, Land ML, Pelletier DA, Beatty JT, Lang AS, Tabita FR, Gibson JL, Hanson TE, Bobst C, Torres JL, Peres C, Harrison FH, Gibson J and Harwood CS (2004) Complete genome sequence of the metabolically versatile photosynthetic bacterium Rhodopseudomonas palustris. Nat Biotechnol 22: 55–61 Lin C (2002) Blue light receptors and signal transduction. Plant Cell 14: 207–225 Losi A (2004) The bacterial counterparts of plant phototropins. Photochem Photobiol Sci 3: 566–574 Losi A, Polverini E, Quest B and Gärtner W (2002) First evidence for phototropin-related blue-light receptors in prokaryotes. Biophys J 82: 2627–2634 Macnab R and Koshland Jr DE (1974) Bacterial motility and chemotaxis: Light-induced tumbling response and visualization of individual flagella. J Mol Biol 84: 399–406 Masuda S and Bauer CE (2002) AppA is a blue light photoreceptor that antirepresses photosynthesis gene expression in Rhodobacter sphaeroides. Cell 110: 613–623 Masuda S and Bauer CE (2004) Null mutation of HvrA compensates for loss of an essential relA/spoT-like gene in Rhodobacter capsulatus. J Bacteriol 186: 235–239 Masuda S and Ono T-A (2004) Biochemical characterization
740 of the major adenylyl cyclase, Cya1, in the cyanobacterium Synechocystis sp. PCC 6803. FEBS Lett. 577: 255–258 Masuda S, Dong C, Swem DL, Setterdahl AT, Knaff DB and Bauer CE (2002) Repression of photosynthesis gene expression by formation of a disulfide bond in CrtJ. Proc Natl Acad Sci USA 99: 7078–7083 Masuda S, Hasegawa K, Ishii A and Ono T-A (2004) Lightinduced structural changes in a putative blue-light receptor with a novel FAD binding fold sensor of blue-light using FAD (BLUF); Slr1694 of Synechocystis sp. PCC6803. Biochemistry 43: 5304–5313 Masuda S, Hasegawa K and Ono T-A (2005a) Light-induced structural changes of apoprotein and chromophore in the sensor of blue light using FAD (BLUF) domain of AppA for a signaling state. Biochemistry 44: 1215–1224 Masuda S, Hasegawa K and Ono T-A (2005b) Tryptophan at position 104 is involved in transforming light signal into changes of β-sheet structure for the signaling state in the BLUF domain of AppA. Plant Cell Physiol. 46:1894–1901 Meyer TE (1985) Isolation and characterization of soluble cytochromes, ferredoxins and other chromophoric proteins from the halophilic phototrophic bacterium Ectothiorhodospira halophila. Biochim Biophys Acta 806: 175–183 Moskvin OV, Gomelsky L and Gomelsky M (2005) Transcriptome analysis of the Rhodobacter sphaeroides PpsR regulon: PpsR as a master regulator of photosystem development. J Bacteriol 187: 2148–2156 Newman JD, Anthony JR and Donohue TJ (2001) The importance of zinc-binding to the function of Rhodobacter sphaeroides ChrR as an anti-sigma factor. J Mol Biol 313: 485–499 Nishimura K, Shimada H, Hatanaka S, Mizoguchi H, Ohta H, Masuda T and Takamiya K (1998) Growth, pigmentation, and expression of the puf and puc operons in a light-respondingrespressor (SPB)-disrupted Rhodobacter sphaeroides. Plant Cell Physiol 39: 411–417 Nozaki D, Iwata T, Ishikawa T, Todo T, Tokutomi S and Kandori H (2004) Role of Gln1029 in the photoactivation processes of the LOV2 domain in Adiantum phytochrome3. Biochemistry 43: 8373–8379 Oh JI and Kaplan S (2000) Redox signaling: Globalization of gene expression. EMBO J 19:4237–4247 Oh JI and Kaplan S (2002) Oxygen adaptation. The role of the CcoQ subunit of the cbb3 cytochrome c oxidase of Rhodobacter sphaeroides 2.4.1. J Biol Chem 277: 16220–16228 Oh JI, Ko I-J and Kaplan S (2004) Reconstitution of the Rhodobacter sphaeroides cbb3-PrrBA signal transduction pathway in vitro. Biochemistry 43: 7915–7923 Okajima K, Yoshihara S, Fukushima Y, Geng X, Katayama M, Higashi S, Watanabe M, Sato S, Tabata S, Shibata Y, Itoh S and Ikeuchi M (2005) Biochemical and functional characterization of BLUF-type flavin-binding proteins of two species of cyanobacteria. J Biochem (Tokyo) 137: 741–750 Ouchane S, Picaud M, Vernotte C and Astier C (1997) Photooxidative stress stimulates illegitimate recombination and mutability in carotenoid-less mutants of Rubrivivax gelatinosus. EMBO J 16: 4777–4787 Panda S, Hogenesch JB and Kay SA (2002) Circadian rhythms from flies to humans. Nature 417: 329–335 Park CM, Kim JI, Yang SS, Kang JG, Kang JH, Shim JY, Chung YH, Park YM and Song PS (2000) A second photochromic bacteriophytochrome from Synechocystis sp PCC6803: spe-
Gabriele Klug and Shinji Masuda cial analysis and down-regulation by light. Biochemistry 39: 10840–10847 Penfold RJ and Pemberton JM (1994) Sequencing, chromosomal inactivation, and functional expression in E. coli of ppsR, a gene which represses carotenoid and bacteriochlorophyll synthesis in Rhodobacter sphaeroides. J Bacteriol 176: 2869–2876 Ponnampalam SN and Bauer CE (1997) DNA binding characteristics of CrtJ. J Biol Chem 272: 18391–18396 Ponnampalam SN, Buggy JJ and Bauer CE (1995) Characterization of an aerobic repressor that coordinately regulates bacteriochlorophyll, carotenoid, and light harvesting-II expression in Rhodobacter capsulatus. J Bacteriol 177: 2990–2997 Porter SL, Warren AV, Martin AC and Armitage JP (2002) The third chemotaxis locus of Rhodobacter sphaeroides is essential for chemotaxis. Mol Microbiol 46:1081–1094 Ragatz L, Jiang Z-Y, Bauer CE and Gest H. (1994) Phototactic purple bacteria. Nature 370, 104 Ragatz L, Jiang ZY, Bauer CE and Gest H (1995) Macroscopic phototactic behaviour of the purple photosynthetic bacterium Rhodospirillum centeneum. Arch Microbiol 163: 1–6 Rajagopal S, Key JM, Purcell EB, Boerema DJ and Moffat K (2004) Purification and initial characterization of a putative blue light-regulated phosphodiesterase from Escherichia coli. Photochem Photobiol 80: 542–547 Rockwell NC, Su Y-S and Lagarias JC (2006) Phytochrome structure and signaling mechanisms. Annu Rev Plant Biol 57: 837–858 Romagnoli S and Armitage JP (1999) Roles of chemosensory pathways in transient changes in swimming speed of Rhodobacter sphaeroides induced by changes in photosynthetic electron transport. J Bacteriol 181: 34–39 Salomon M, Christie JM, Knieb E, Lempert U and Briggs WR (2000) Photochemical and mutational analysis of the FMN binding domains of the plant blue light receptor phototropin. Biochemistry 39: 9401–9410 Sancar A. (2004) Photolyase and cryptochrome blue-light photoreceptors. Adv Protein Chem 69:73–100 Sato Y, Iwata T, Tokutomi S and Kandori H (2005) Reactive cystein is protonated in the triplet excited state of the LOV2 domain in Adiantum phytochrome3. J Am Chem Soc 127: 1088–1089 Schleicher E, Kowalczyk RM, Kay CWM, Hegemann P, Bacher A, Fischer M, Bittl R, Richter G and Weber S (2004) On the reaction mechanism of adduct formation in LOV domains of the plant blue-light receptor phototropin. J Am Chem Soc 126: 11067–11076 Schwerdtfeger C and Linden H (2003) VIVID is a flavoprotein and serves as a fungal blue light photoreceptor for photoadaptation. EMBO J 22: 4846–4855 Shimada H, Iba K and Takamiya K (1992) Blue-light irradiation reduces the expression of puf and puc operons of Rhodobacter sphaeroides under semi-aerobic conditions. Plant Cell Physiol 33: 471–475 Shimada H, Wada T, Handa H, Ohta H, Mizoguchi H, Nishimura K, Masuda T, Shioi Y and Takamiya K (1996) A transcription factor with a leucine-zipper motif involved in light-dependent inhibition of expression of the puf operon in the photosynthetic bacterium Rhodobacter sphaeroides. Plant Cell Physiol 37: 515–522 Sprenger WW, Hoff WD, Armitage JP and Hellingwerf KJ (1993) The eubacterium Ectothiorhodospira halophila is negatively phototactic, with a wavelength dependence that fits the absorp-
Chapter 36
Regulation of Genes by Light
tion spectrum of the photoactive yellow protein. J Bacteriol 175: 3096–3104 Spudich JL, Yang CS, Jung KH and Spudich EN (2000) Retinylidene proteins: Structures and functions from archaea to humans. Annu Rev Cell Dev Biol 16: 365–392 Swartz TE, Corchnoy SB, Christie JM, Lewis JW, Szundi I, Briggs WR and Bogomolni RA (2001) The photocycle of a flavinbinding domain of the blue light photoreceptor phototropin. J. Biol. Chem. 276: 36493–36500. Swartz TE, Wenzel PJ, Corchnoy SB, Briggs WR and Bogomolni RA (2002) Vibration spectroscopy reveals light-induced chromophore and protein structural changes in the LOV2 domain of the plant blue-light receptor phototropin 1. Biochemistry 41: 7183–7189 Swem LR, Gong X, Yu CA and Bauer CE (2006) Identification of a ubiquinone-binding site that affects autophosphorylation of the sensor kinase RegB. J. Biol. Chem. 281: 6768–6775 Takamiya K and Takamiya A (1969) Light-induced reactions of ubiquinone in photosynthetic bacterium, Chromatium D. III. Oxidation-reduction state of ubiquinone in intact cells of Chromatium D. Plant Cell Physiol. 10: 363–368 Tarutina M, Ryjenkov DA and Gomelsky M. (2006) An unorthodox bacteriophytochrome from Rhodobacter sphaeroides involved in turnover of the second messenger c-di-GMP. J Biol Chem 281: 34751–34758 Unno M, Sano R, Masuda S, Ono T-A and Yamauchi S (2005) Light-induced structural changes in the active site of the BLUF
741 domain in AppA by Raman spectroscopy. J Phys Chem B 109: 12620–12626 Unno M, Masuda S, Ono T-A and Yamauchi S (2006) Orientation of a key glutamine residue in the BLUF domain from AppA revealed by mutagenesis, spectroscopy, and quantum chemical calculations. J Am Chem Soc 128: 5638–5639 Wagner JR, Brunzelle JS, Forest KT and Vierstra RD (2005) A light-sensing knot revealed by the structure of the chromophore binding domain of phytochrome. Nature 438: 325–331 Weaver P (1971) Temperature-sensitive mutations of the photosynthetic apparatus of Rhodospirillum rubrum. Proc Natl Acad Sci USA 68: 136–138 Wu SH and Lagarias JC (2000) Defining the bilin lyase domain: Lessons from the extended phytochrome superfamily. Biochemistry 39: 13487–13495 Yeh K-C, Wu SH, Murphy JT and Lagarias JC (1997) A cyanobacterial phytochrome two-component light sensory system. Science 277: 1505–1508 Zeugner A, Byrdin M, Bouly J, Bakrim N, Giovani B, Brettel K and Ahmad M (2005) Light-induced electron transfer in Arabidopsis cryptochrome-1 correlates with in vivo function. J Biol Chem 280: 19437–19440 Zirak P, Penzkofer A, Schiereis T, Hegemann P, Jung A and Schlichting I (2005) Absorption and fluorescence spectroscopic characterization of BLUF domain of AppA from Rhodobacter sphaeroides. Chem Phys 315: 142–154
Chapter 37 Regulation of Hydrogenase Gene Expression Paulette M. Vignais* CEA Grenoble, Laboratoire de Biochimie et Biophysique des Systèmes Intégrés, UMR CEA/CNRS/UJF n° 5092, Institut de Recherches en Sciences et Technologies pour le Vivant (iRSTV), 17 rue des Martyrs, 38054 Grenoble cedex 9, France
Summary ............................................................................................................................................................... 743 I. Introduction..................................................................................................................................................... 744 II. Regulation of Hydrogenase Gene Expression: Signaling and Transcription Control ..................................... 744 A. Two Phylogenetically Distinct Classes of Hydrogenases ................................................................ 744 1. Presence of Various Types of [NiFe]-hydrogenases in Photosynthetic Bacteria .................... 745 2. Presence of a [FeFe]-hydrogenase in Rhodopseudomonas palustris and in Dehalococcoides ethenogenes ............................................................................................. 746 B. Signaling and Transcription Control ................................................................................................. 747 1. Response to H2 — A H2-specific Regulatory Cascade............................................................. 747 a. The Two-component Regulatory System, HupT and HupR ......................................... 748 b. The H2-sensing Regulatory [NiFe]-hydrogenase, HupUV ............................................. 750 c. Role of the Global Regulator IHF................................................................................... 752 2. Response to CO...................................................................................................................... 752 3. O2 Regulation .......................................................................................................................... 752 4. Redox regulation ..................................................................................................................... 753 C. Conclusions and Perspectives ......................................................................................................... 754 References ............................................................................................................................................................ 755
Summary Photosynthetic bacteria contain various types of [NiFe]-hydrogenases. The most widespread are respiratory uptake hydrogenases that are usually co-synthesized with nitrogenase. Hydrogenase synthesis responds to environmental signals such as H2, O2, CO. In Rhodobacter species and Rhodopseudomonas palustris, a regulatory [NiFe]-hydrogenase, HupUV, is the direct H2 sensor of the H2 regulatory pathway. It transfers the signal to a two-component regulatory system, which comprises the soluble histidine kinase HupT/HoxJ and the response regulator HupR/HoxA. The transcription factor HupR/HoxA activates the transcription of hydrogenase genes in its nonphosphorylated form. In Rhodospirillum rubrum, the transcriptional regulator CooA, which senses CO and the redox state, activates the synthesis of a CO-tolerant hydrogenase under anaerobic conditions. In Thiocapsa roseopersicina, FnrT activates anaerobic induction of the Hyn hydrogenase. In Rhodobacter species and in Rhodopseudomonas palustris, hydrogenase synthesis is also under the negative control of the redoxresponding global two-component regulatory system, RegB/PrrB/RegS-RegA/Prra/RegR.
*Email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 743–757. © 2009 Springer Science + Business Media B.V.
744 I. Introduction The ability to metabolize hydrogen is widespread among prokaryotes. This ability is carried out by enzymes called hydrogenases that catalyze both the oxidation and production of hydrogen (Cammack et al., 2001). The majority of hydrogenases belong to one of the two phylogenetically distinct classes of metalloenzymes, the [NiFe]-hydrogenases and the [FeFe]-hydrogenases. [NiFe]-hydrogenases are widely distributed among prokaryotes, while [FeFe]hydrogenases are primarily found in the clostridia, Thermotogales and Desulfovibrionaceae (Vignais et al., 2001). As the capacity to take up or to evolve H2 is usually a facultative trait, hydrogenases are predominantly formed only when H2, the substrate, is available. The mechanisms that regulate hydrogenase gene expression have been particularly well studied in Escherichia (E.) coli, which contains four hydrogenase systems (Sawers et al., 2004), in the purple nonsulfur bacterium Rhodobacter (Rba.) capsulatus, and in the β-proteobacterium Ralstonia (R.) eutropha (formerly Alcaligenes eutrophus) and newly-named Cupriavidus necator (Vandamme and Coenye, 2004; Vaneechoutte et al., 2004), which both contain a H2-specific regulatory system (see below). In N2 fixers, for example cyanobacteria (Tamagnini et al., 2002), Rhizobium leguminosarum, Bradyrhizobium (B.) japonicum and Rba. capsulatus, transcription of the hydrogenase genes is co-regulated with nitrogen fixation genes (nif and fix) and is controlled by global activators such as NifA and RegB-RegA (Friedrich et al., 2001; Dubbs and Tabita, 2004; Elsen et al., 2004; Vignais and Colbeau, 2004). Complete microbial genome sequences have revealed the presence of multiple hydrogenase isoenzymes in metabolically versatile bacteria. These isoenzymes are differently regulated depending on the growth conditions (aerobiosis/anaerobiosis) and the mode of anaerobic life (fermentation/anaerobic respiration). This review is focused on the regulation of hydrogenase gene expression in photosynthetic bacteria. Abbreviations: B. – Bradyrhizobium; E. – Escherichia; IHF – integration host factor; nt – nucleotide R. – Ralstonia; Rba. – Rhodobacter; Rps. – Rhodopseudomonas; Rsp. – Rhodospirillum; Tca. – Thiocapsa
Paulette M. Vignais II. Regulation of Hydrogenase Gene Expression: Signaling and Transcription Control A. Two Phylogenetically Distinct Classes of Hydrogenases Hydrogenases are iron-sulfur proteins with two metal atoms at their active site, either a Ni and an Fe atom (in [NiFe]-hydrogenases) or two Fe atoms (in [FeFe]hydrogenases). An exception is the iron-sulfur-free hydrogenase, Hmd, found in methanogenic archaea; it contains an Fe atom at the active site but no iron-sulfur (Fe-S) cluster). The [NiFe]- and [FeFe]-hydrogenases are phylogenetically distinct classes of enzymes (Vignais et al., 2001). The [NiFe]-hydrogenases are the most numerous and best studied hydrogenases. They are found in the domains of Bacteria and Archaea; 360 sequences in Bacteria and 79 in Archaea have been identified mainly by genome sequencing (Vignais and Billoud, 2007). The core enzyme consists of an αβ heterodimer with the large subunit (α-subunit) of ca. 60 kDa hosting the bimetallic NiFe active site and the small subunit (β-subunit) of ca. 30 kDa, the Fe-S clusters. The bimetallic NiFe center is deeply buried in the large subunit; it is coordinated to the protein by four cysteines. Three non-protein ligands, comprising one CO and two CN–, are bound to the Fe atom. The hydrogenase assembly and maturation processes have been reviewed recently (Casalot and Rousset, 2001; Vignais and Colbeau, 2004; Sawers et al., 2004; Böck et al., 2006). At the active site, H2 is split to yield two protons and two electrons. The protons are released via a proton pathway extending from the active site to the surface of the large subunit, and the electrons are conducted by the Fe-S clusters to the binding site of the physiological electron acceptor (or donor) of hydrogenase on the surface of the small subunit (Volbeda et al., 1995; Higuchi et al., 1997). Hydrophobic channels link the active site to the surface of the molecule and facilitate gas access to the active site. Full sequence alignments of the small and large subunits have shown that the two subunits of [NiFe]hydrogenases evolved conjointly. This analysis led to a classification of [NiFe]-hydrogenases into four groups which are consistent with the functions of the enzymes (Vignais et al., 2001; Vignais and Colbeau, 2004; Vignais and Billoud, 2007) (Fig. 1).
Chapter 37
Regulation of Hydrogenase Genes
745
Fig. 1. Schematic representation of the phylogenetic tree of [NiFe]-hydrogenases based on the comparison of the complete sequences of the small and the large subunits (the same tree was obtained with each type of subunit), originally established by Vignais et al. (2001). The numbers of enzymes identified in 2006 in each domain, Bacteria and Archaea, are indicated.
1. Presence of Various Types of [NiFe]hydrogenases in Photosynthetic Bacteria Representatives in each of the four groups of [NiFe]hydrogenases are found in photosynthetic bacteria, namely uptake hydrogenases (group 1), cytoplasmic H2 sensors (group 2), bidirectional heteromultimeric cytoplasmic [NiFe]-hydrogenases (group 3) and H2 evolving, energy-conserving, membrane-associated hydrogenases (group 4) (Fig. 1). The uptake hydrogenases (group 1) are membranebound respiratory enzymes. They link the oxidation of H2 to the reduction of anaerobic electron acceptors, such as NO3–, SO42–, fumarate, DMSO or TMAO (anaerobic respiration) or to O2 (aerobic respiration) with recovery of energy in the form of a protonmotive force. They are connected to the quinone pool of the respiratory chain in the membrane by a third subunit, a di-heme cytochrome b, which, together with the hydrophobic C-terminus of the small subunit, anchors the hydrogenase dimer to the membrane. The uptake hydrogenase can also provide electrons to the Complex I (NADH: ubiquinone oxidoreductase) for CO2 reduction, and to a membrane complex called Rnf (for Rhodobacter-specific nitrogen fixation) involved in electron transfer to nitrogenase (Schmehl
et al., 1993; Jouanneau et al., 1998; reviewed in Vignais et al., 2004). A respiratory hydrogenase is usually present in the N2 fixers; it has been found in Rba. capsulatus, Rba. sphaeroides, Rubrivivax gelatinosa (Rhodocyclus gelatinosus), Rhodopseudomonas (Rps.) palustris, Allochromation vinosum (Chromatium vinosum) and Thiocapsa (Tca.) roseopersicina. In Proteobacteria, the genes encoding H2 uptake hydrogenases are clustered and have been named alphabetically in accordance with their order in the cluster. However, the structural genes encoding the small and large subunits are often named S and L, respectively. Besides the structural genes, the hydrogenase gene cluster (i.e., hup) contains several additional genes, the products of which are involved in the maturation of the heterodimeric enzyme, the insertion of Ni, Fe, and the CO and CN ligands to Fe (hyp genes), and the regulation of hydrogenase gene transcription. The organization of the hydrogenase genes at the hup locus of Rba. capsulatus is shown in Fig. 2. The cytoplasmic H2 sensors (group 2) are regulatory hydrogenases. Their role is to detect the presence of H2 in the environment and to trigger a cascade of cellular reactions controlling the synthesis of respiratory hydrogenases (see below). This type of hydrog-
746
Paulette M. Vignais
Fig. 2. The hydrogenase gene cluster of Rhodobacter capsulatus at the hup locus. Genes involved in the regulation of the biosynthesis of the HupSL uptake hydrogenase, located outside the hup locus are also shown. HK: histidine kinase; RR: response regulator.Transcription start sites are indicated by horizontal arrows.
enase is present in Rba. capsulatus, Rba. sphaeroides, Rps. palustris and Tca. roseopersicina. The bidirectional heteromultimeric cytoplasmic [NiFe]-hydrogenases (group 3) are able to bind soluble cofactors such as NAD or NADP. The genes of hydrogenases assigned to the subgroup 3b have been identified in the genome of phototrophic green sulfur bacteria of the genus Chlorobium and Pelodictyon. The NAD(P)-dependent [NiFe]hydrogenases (subgroup 3d), typically found in cyanobacteria, are composed of two moieties: the heterodimer hydrogenase moiety, encoded by the hoxY and hoxH genes, and the diaphorase moiety, encoded by the hoxU, hoxF and hoxE genes, homologous to some subunits of Complex I of mitochondrial and bacterial respiratory chains that contains NAD(P), FMN and Fe-S binding sites. Physiologically, they function in both direction, either to regenerate reduced nucleotides needed for CO2 fixation, or to dispose of low-potential electrons generated at the onset of illumination (Appel et al., 2000; Cournac et al., 2004). The presence of a heteropentameric NAD+-reducing hydrogenase has been reported in the phototrophic purple sulfur bacterium Tca. roseopercisina (Rákhely et al., 2004) (for recent reviews see Appel and Schulz, 1998; Tamagnini et al. 2002; Vignais and Colbeau, 2004; Vignais and Billoud, 2007). The energy-conserving H2 evolving membraneassociated hydrogenases (group 4) are multimeric enzymes with six or more subunits. They comprise transmembrane subunits homologous to the Complex I subunits involved in proton pumping and energy coupling (reviewed in Vignais, 2007). They appear to be able to couple anaerobic oxidation of C1 organic compounds of low potential, such as formate or car-
bon monoxide, with reduction of protons to H2. These hydrogenases are associated with formate dehydrogenases or CO dehydrogenases. The prototype of these enzymes is the hydrogenase 3, which is part of the formate hydrogen lyase complex in E. coli, although the majority of hydrogenases assigned to group 4 have been found in Archaea (reviewed in Hedderich, 2004). In purple bacteria, the CooLH hydrogenase of Rhodospirillum (Rsp.) rubrum, which is associated with the CO dehydrogenase (Fox et al., 1996), has been best studied (see below). Genome sequencing has revealed the presence of this type of hydrogenase in Rps. palustris as well (Larimer et al., 2004). 2. Presence of a [FeFe]-hydrogenase in Rhodopseudomonas palustris and in Dehalococcoides ethenogenes The [FeFe]-hydrogenases are found in anaerobic prokaryotes, such as clostridia and sulfate reducers (Atta and Meyer, 2000 and refereces therein; for reviews, see Vignais et al., 2001; Vignais and Colbeau 2004; Vignais and Billoud, 2007; Meyer, 2007). They are the only type of hydrogenase to have been found in eukaryotes, and are located exclusively in organelles such as chloroplasts or hydrogenosomes (reviewed in Vignais et al., 2001; Horner et al., 2002; Vignais and Colbeau, 2004). Based on the genome sequence, an [FeFe]-hydrogenase (probably acquired by horizontal gene transfer) would also be present in Rps. palustris (Larimer et al., 2004) and in Dehalococcoides ethenogenes strain 195, a green nonsulfur bacterium related with the Chloroflexi (Seshadri et al., 2005).
Chapter 37
Regulation of Hydrogenase Genes
747
Fig. 3. Expression of the hydrogenase genes is controlled at the transcriptional level. The promoter of the Rhodobacter capsulatus structural hydrogenase operon, hupSLC, was fused to the reporter gene lacZ, which encodes the β-galactosidase enzyme (A). The hupS:: lacZ fusion, carried on a plasmid, was introduced in the wild type strain B10. Cells were grown anaerobically in malate/glutamate (MG) medium (N2-fixing, H2-evolving conditions) or in malate/ammonia (MN) medium in the absence or presence of O2, and where indicated, 10 % H2. Hydrogenase gene expression was correlated with β-galactosidase activity (B), indicating that hupS::lacZ expression was under the same control as the chromosomal hupS gene (Elsen et al., 1996).
B. Signaling and Transcription Control Control of hydrogenase synthesis represents a means to respond quickly and efficiently to changes in the environment, in particular to new energy demands. It is exerted at the transcriptional level; this was demonstrated by the use of transcriptional reporter systems measuring levels of promoter activity of structural hydrogenase operons (for recent reviews, see Friedrich et al., 2001; Vignais and Colbeau, 2004). Fig. 3 shows that expression of a hupS::lacZ fusion, carried on a plasmid, paralleled the activity of hydrogenase expressed from the chromosome of a wild type Rba. capsulatus strain B10. The use of reporter systems has been instrumental in identifying regulatory mutants and assigning genes responding to environmental signals. Transcriptional control of hydrogenase involves usually one or several twocomponent regulatory systems that might act either positively or negatively. In response to a specific signal, the first component, which is a sensor histidine kinase, autophosphorylates at a conserved histidine residue, and then the cognate response regulator is transphosphorylated at a conserved aspartate residue (Hoch and Silhavy, 1995). Hydrogenase synthesis responds to several type
of signals: (1) molecular hydrogen, which is the substrate, activates hydrogenase gene expression in the photosynthetic bacteria Rba. capsulatus, Rba. sphaeroides and Rps. palustris; (2) molecular oxygen, which is an inhibitory signal for most of hydrogenases; (3) nickel ions, which are required for enzyme function; and (4) metabolites such as formate, carbon monoxide, nitrate or sulfur, which act as electron donors or acceptors. 1. Response to H2 — A H2-specific Regulatory Cascade In N2 fixers, uptake hydrogenase is induced concomitantly with nitrogenase. Hydrogenase transcription is co-regulated with nitrogenase transcription by global regulators (see below). In addition, in some Proteobacteria, a H2-specific signal transduction system controls the transcription of uptake hydrogenase genes (termed hupSLC in the photosynthetic bacteria Rba. capsulatus, Rba. sphaeroides, Rps. palustris and Tca. roseopersicina). The H2-specific regulatory system comprises a hydrogen-sensing regulatory hydrogenase (HupUV/HoxBC) and a twocomponent signal transduction system (HupT/HoxJ and HupR/HoxA); it has been studied mainly in Rba.
748
Paulette M. Vignais
Fig. 4. Comparison of the amino acid sequences of the central domain of response regulators of the NtrC subfamily. The underlined motifs, Walker A (GESGTGKE), site of σ54 interaction [ESELFGxxxGAFTGA] and Walker B motif (GThhhDEh) where h represents hydrophobic amino acids, are derived from sequence comparisons of the activation domains of 41 σ54-dependent activators (Osuna et al., 1997). The arrow shows that the glycine residue at position 258 in the Rhodobacter capsulatus HupR protein of the wild-type strain B10 is replaced by an aspartic residue in the mutant RCC8 (Richaud et al., 1991). Note that the GAFTGA sequence (in bold) and the Walker motifs (in italics) are not well conserved in Rhodobacter capsulatus and Rhodobacter sphaeroides HupR. Rhodobacter capsulatus (Rc) HupR (Richaud et al.,, 1991), Rhodobacter sphaeroides (Rs) HupR (A. Colbeau, unpublished, Accession AAF19999; Xu and Wu 2001), Ralstonia eutropha (Re) HoxA (Eberz and Friedrich, 1991), Rhodopseudomonas palustris (Rp) HoxA (Accession RPA0979; Rey et al., 2006), Thiocapsa roseopersicina (Tr) HupR (A. Colbeau, unpublished, Accession L22980; Kovács et al., 2005a), Klebsiella pneumoniae (Kp) NtrC (Drummond et al., 1986) and Rhodobacter capsulatus (Rc) NtrC (Jones and Haselkorn, 1989) (adapted from Dischert et al., 1999).
capsulatus (Elsen et al., 1996, 2003; Toussaint et al., 1997; Dischert et al., 1999; Vignais et al., 2005), and in the Gram-negative aerobic bacterium Ralstonia (R.) eutropha (Lenz and Friedrich, 1998; Lenz et al., 2002; Buhrke et al., 2004; Friedrich et al., 2005). a. The Two-component Regulatory System, HupT and HupR In Rba. capsulatus, Rba. sphaeroides and Tca. roseopersicina, the response regulator of the two-component regulatory system is termed HupR (HoxA in Rps. palustris), and the protein histidine kinase HupT (HoxJ in Rps. palustris). The Rba. capsulatus hupR gene was identified and isolated by complementation of a mutant (RCC8), which was unable to grow autotrophically and which had a low hydrogenase activity. The deduced amino acid sequence of mutated HupR compared to the wild-type revealed a single mutation, a glycine residue changed into an aspartic residue (Richaud et al., 1991) (Fig. 4). The HupR protein shows the typical three-domain organization of the response regulators from the NtrC subfamily (Fig. 5). The N-terminal receiver domain (~ 123 residues) contains the D54 conserved residue, which is the site of phosphorylation. The central domain (~ 240 residues) belongs to the AAA+ (ATPases associated with various cellular activities) superfamily of ATPases that is responsible for ATP hydrolysis and transcriptional activation (Neuwald et
Fig. 5. The two-component regulatory system HupT and HupR of Thiocapsa roseopersicina and Rhodobacter capsulatus. HupR from Thiocapsa roseopersicina has the typical structure of σ54-dependent activators (Fig. 4), and expression of the uptake hydrogenase in Thiocapsa roseopersicina is RpoN-dependent (Kovács et al., 2005a). In Rhodobacter capsulatus HupR some critical sequences are not well conserved in the central domain (Fig. 4), hence HupR is not active as an ATPase (Davies et al., 2006) and hupSL transcription is σ70-dependent in this species. Rhodobacter capsulatus HupR activates hupSL transcription in a non phosphorylated state, unlike other transcription factors of the NtrC-type. In the absence of H2, HupR is phosphorylated by HupT~P and it does not activate transcription (Dischert et al., 1999).
al., 1999; Ogura and Wilkinson, 2001). The C-terminal domain (~ 90 residues) with a helix-turn-helix motif is the DNA-binding domain that recognizes the enhancer (Osuna et al., 1997). An unusual feature of Rhodobacter HupR is that the putative σ54 binding motifs (ESELFGH and GAFTGA) of NtrC in enteric
Chapter 37
Regulation of Hydrogenase Genes
bacteria, in Tca. roseopersicina HupR, in R. eutropha and in Rps. palustris HoxA, are not well conserved (Fig. 4). Consistent with this fact, Rba. capsulatus HupR functions with σ70-RNA polymerase (Dischert et al., 1999) as is the case for NtrC of Rba. capsulatus (Bowman and Kranz, 1998), which lacks the sequence corresponding to the site of σ54 interaction (Fig. 4). On the other hand, in Tca. roseopersicina (Kovacs et al., 2005a) and Rps. palustris (Rey et al., 2006), uptake hydrogenase gene transcription requires a σ54-RNA polymerase holoenzyme. Another unusual feature of Rhodobacter HupR is its lack of ATPase activity. Unlike the HupR from Tca. roseopersicina and HoxA from Rps. palustris, the Walker A (G(E/D)(S/T)G(S/T)GK(E/D) and B (G(S/T)(I/L)FLDE) motifs (Walker et al., 1982) that participate in ATP binding and hydrolysis are not well conserved in the HupR from Rba. capsulatus and Rba. sphaeroides (Fig. 4). Usually, the Walker B motif in the AAA+ ATPases contains the sequence DExx with two conserved carboxylate side chains projecting into the active site. The carboxylate of the conserved Asp residue is involved in the Mg2+ coordination sphere while the conserved Glu residue is believed to be the catalytic base (Ogura and Wilkinson, 2001). These two critical residues are lacking in Rba. capsulatus and Rba. sphaeroides HupR (Fig. 4), and in agreement with this, no ATPase activity has been detected in the isolated central domain of Rba. capsulatus HupR (Davies et al., 2006). However, HupR activity might still require ATP binding because the replacement of a Gly residue in the Walker B motif by an Asp residue renders this transcription factor non functional (Richaud et al., 1991; Toussaint et al., 1997). Activators of σ54-RNA polymerase holoenzyme stimulate transcription by binding to enhancer sites that are at large distance from promoter elements (Xu and Hoover, 2001). Rba. capsulatus HupR has kept some of the features of the enhancer binding proteins: it binds to a palindromic sequence (5´TTG-N5-CAA) centered at 157 nt upstream from the transcription initiation site, and requires the global regulator integration host factor (IHF) to contact the RNA polymerase by DNA looping (Toussaint et al., 1997). However, in contrast to other NtrCtype activators, Rba. capsulatus HupR activates the transcription of hupSL in its non-phosphorylated form and becomes inactive upon phosphorylation. This was demonstrated by replacing the putative phosphorylation site (D54) of the HupR protein with various amino acids or by deleting it using site-di-
749 rected mutagenesis. Strains expressing mutated hupR genes showed high hydrogenase activities even in the absence of H2, indicating that hupSL transcription is activated by binding of unphosphorylated HupR (Dischert et al., 1999). Similarly, hydrogenase gene expression is stimulated by the unphosphorylated form of HoxA, the homolog of HupR in R. eutropha (Lenz and Friedrich, 1998), and in B. japonicum (Van Soom et al., 1999). HupR crystallizes as a dimer in its unphosphorylated state, and forms an elongated V-shape structure with extended arms (Davies et al., 2006). A sketch of the dimer derived from the work of Davies et al. (2006) is shown in Fig. 6. The Rba. capsulatus hupR gene is located within the hyp genes and is transcribed constitutively (Dischert et al., 1999) together with the hyp genes from a promoter located in the hypA gene (Colbeau, A., unpublished) (Fig. 2). The level of the HupR protein is constant while in R. eutropha, under hydrogenase derepressing conditions, hoxA can be transcribed from the strong promoter of the membrane-bound hydrogenase structural genes (PMBH), leading to enhanced synthesis of the H2-responding cascade proteins (HoxABCJ) (Schwartz et al., 1999). The Rba. capsulatus hupT gene, which encodes the cognate soluble histidine kinase, is expressed independently from hupR. It belongs to the hupTUV operon, which is expressed from a promoter localized just upstream from hupT (Fig. 2), and is regulated
Fig. 6. Schematic representation of the envelope of the non phosphorylated HupR dimer crystallized in two-dimensions. The 3-D structure of the negatively stained HupR was determined from tilted electron microscope images, and fit to the 3-D structure of the central domain from NtrC1 Aquifex aeolicus protein (Lee et al., 2003) (adapted from Davies et al., 2006).
750 by the availability of organic carbon. Under autotrophic growth conditions, the hupTUV operon is very poorly expressed and hydrogenase biosynthesis is derepressed (Elsen et al., 1996). The HupT histidine kinase autophosphorylates in the presence of [γ-32P]ATP (Elsen et al., 1997). When HupT~P is mixed with HupR in vitro, HupR becomes phosphorylated on the aspartate residue (D54) conserved in response regulators. Unlike the hupR mutants, hupT mutants exhibit high hydrogenase activities even in the absence of H2, indicating that phosphorylation prevents activation of hydrogenase gene expression by HupR, as illustrated in Fig. 7 (Elsen et al., 1993, 1996). b. The H2-sensing Regulatory [NiFe]hydrogenase, HupUV Detection of the H2 signal involves a regulatory hydrogenase encoded by the hupUV genes in Rba. capsulatus, Rba. sphaeroides, Rps. palustris and Tca. roseopersicina. The HupUV protein is the H2-sensor for the HupT/HupR (HoxJ/HoxA) regulatory system, which controls the transcription of hydrogenase genes in response to H2. An interesting feature of the H2-sensing hydrogenase, which is different from the majority of hydrogenases, is that it is still active in the presence of oxygen (Vignais et al., 2002). In this regulatory hydrogenase, the main gas channel leading to the NiFe active site is considered to be too narrow, due to the presence of amino acids bulkier than in the usual [NiFe]-hydrogenases, restricting molecular oxygen from reaching the active site and inactivating it (Volbeda et al., 2002). This hypothesis has been confirmed by site-directed mutagenesis of the HupUV hydrogenase of Rba. capsulatus (Duché et al., 2005) and of the homologous regulatory HoxBC hydrogenase of R. eutropha (Buhrke et al., 2005). Indeed, replacement of two bulky amino acids by smaller ones enlarged the gas channel leading to the active site, and yielded mutant derivatives that became sensitive to oxygen. Thus, inaccessibility of O2 to the active site of the regulatory hydrogenases permits them to remain functional in the presence of molecular oxygen. It is not yet clear how the H2 signal detected by the HupUV hydrogenase is transmitted to HupT. The histidine kinase HupT contains a PAS (PER, ARNT, SIM) domain in its N-terminal portion (Fig. 5). PAS domains are involved in sensing various factors, such as light, or redox changes, or in protein-protein interactions (Taylor and Zhulin, 1999). PAS-deleted HupT
Paulette M. Vignais protein variants migrated as dimers and tetramers in native gels (Elsen et al., 2003), in agreement with the localization of dimerization determinants within the H region of the kinase transmitter domain (Robinson et al., 2000). Thus, the PAS domain might contribute to the dimerization and tetramerization of HupT, and increase the stability of its oligomeric forms. On native gels, the HupUV protein appears as two enzymatically active bands with molecular masses of about 80 and 170 kDa, indicating that it exists in dimeric HupUV as well as tetrameric (HupUV)2 forms that are in equilibrium (Elsen et al., 2003). The interaction of the H2-sensor HupUV with HupT has been studied in vitro using wild-type and various mutant forms of HupT protein (the PAScontaining N-terminal domain, the PAS-deleted HupT, and the H217N HupT, which is mutated in the conserved phosphorylatable His217 and inactive in signal transduction) overproduced in E. coli (Elsen et al., 2003). To analyze the formation of complexes between HupUV and HupT, the two isolated proteins were first incubated together at 30 °C and then loaded onto a native gel. The HupUV protein was detected either by its hydrogenase activity (reduction of benzyl viologen in the presence of H2) or by antibodies against the His6 tag of HupU (revealed by chemiluminescence) to estimate the amount of each form. Both the heterodimer HupUV and the tetramer (HupUV)2 had hydrogenase activity. In the presence of HupT, a new active band of approx. 300 kDa appeared, suggesting complex formation of the dimeric forms (HupUV)2-(HupT)2. Addition of H2 to the regulatory hydrogenase before or during the incubation with HupT rendered the complex unstable. Similarly, the isolated N-terminal domain of HupT, but not the truncated HupT that lacked its PAS domain, formed a stable complex with HupUV. Thus, the N-terminal PAS domain of HupT is absolutely required for interaction with the regulatory hydrogenase (probably through the small subunit HupU) (Elsen et al., 2003). In Rba. capsulatus, it is not yet elucidated whether transmission of the H2 signal involves merely a change in the interaction between the regulatory hydrogenase (HupUV) and the PAS-containing domain of the histidine kinase (HupT) upon H2 reduction, or whether the PAS domain is also involved in the binding of a cofactor reducible by the regulatory hydrogenase in the presence of H2, as has been proposed in the case of R. eutropha (Bernhard et al., 2001; Lenz et al., 2002; Buhrke et al., 2004). A similar hupTUV operon and a response regulator
Chapter 37
Regulation of Hydrogenase Genes
751
Fig. 7. Schematic representation of the relationships between the H2 sensor HupUV, the histidine kinase HupT and the response regulator HupR in respect to hydrogenase biosynthesis in Rhodobacter capsulatus. A) Hydrogenase activities in wild type and regulatory mutants reflect hydrogenase biosynthesis. Cells were grown anaerobically in malate/glutamate (MG) medium (N2-fixing, H2-evolving conditions) or in malate/ammonia (MN) medium in the absence or presence of O2 and, where indicated in 10% H2. In the absence of HupR, hydrogenase biosynthesis is minimal. Hydrogenase activity is expressed in micromoles of methylene blue (MB) reduced/h per mg of protein. B) In the absence of H2, phosphotransfer between HupT and HupR is favored, and there is no stimulation of hydrogenase gene transcription (upward arrows). In the presence of H2, detected by the regulatory HupUV hydrogenase, HupR remains in the active non-phosphorylated state and induces transcription. Hydrogenase biosynthesis is derepressed in the hupT and the hup(UV)∆ mutants (Elsen et al., 1996; Toussaint et al., 1997; Vignais et al., 2005).
HupR are also present in the purple sulfur bacterium Tca. roseopersicina (Kovács et al., 2005a). However, although the genes of the entire H2 sensing and regulation system are present, expression of the uptake HupSL hydrogenase is unaffected by the presence or absence of H2. This difference might result from the fact that the interaction between the regulatory hydrogenase HupUV and the histidine kinase HupT does not change in the presence of H2, as is the case in Rba. capsulatus and R. eutropha. While the H2 transduction cascade appears to be
specific for the control of uptake hydrogenase synthesis in Rba. capsulatus, there appears to be a HoxA regulon in Rps. palustris. Indeed, in addition to the hydrogenase genes, the HupUV-HoxJA system of Rps. palustris has been shown by microarray experiments to activate expression of a dicarboxylic acid transporter, a formate transporter, and a glutamine synthetase (Rey et al., 2006). In summary, three Rhodobacter proteins, HupUV, HupT and HupR participate in the signal transduction system specific for H2 as follows: the H2 signal (i) is
Paulette M. Vignais
752 detected by the HupUV hydrogenase and transmitted to HupT; (ii) is transduced by phosphotransfer between the two components of the regulatory system HupT/HupR; (iii) is integrated at the promoter of the structural genes of hydrogenase by HupR. Based on the results presented in Fig. 7 it is inferred that, in the absence of H2, the histidine kinase HupT and the regulatory hydrogenase HupUV form a complex in which the HupT protein (maintained in its correct conformation) may have increased autokinase activity. Thus, as the phosphorylated state of HupT and its surface interacting with HupR is being stabilized, phosphotransfer to HupR is favored. The transcription factor in the HupR~P state does not activate transcription. In the presence of H2 HupR remains in its nonphosphorylated state, and activates transcription. Whether this is due to the inhibited kinase activity of HupT or to its increased phosphatase activity on HupR~P remains to be determined. Clearly, as HupR remains active in the absence of HupUV or HupT, the association-dissociation of the HupUV-HupT complex appears to be critical in the response to H2. A similar system functions in B. japonicum, R. eutropha and Rps. palustris. A notable difference is that, in the absence of the H2 sensor, there is no synthesis of the membrane-bound hydrogenase in B. japonicum, R. eutropha and Rps. palustris (Lenz et al., 2002; Friedrich et al., 2005; Burgdorf et al., 2005; Rey et al., 2006), while hydrogenase synthesis is derepressed in Rba. capsulatus (Fig. 7) (Elsen et al., 1996; Toussaint et al., 1997; Vignais et al., 2005). This difference is probably related to the phosphorylation status of the histidine kinase (HoxJ/HupT), which apparently remains phosphorylated in the absence of the regulatory hydrogenase in the former cases and not in the case of Rba. capsulatus. c. Role of the Global Regulator IHF The involvement of IHF in hydrogenase transcription was discovered from the study of a Hup-minus mutant with a mutated himA gene, which encodes the α-subunit of IHF. Rba. capsulatus IHF (αβ = 21 kDa) was isolated and shown to bind to the promoter of the hydrogenase structural genes at the binding site 5´-TCACACACCATTG, centered at –87 nt from the transcription start site (Toussaint et al., 1997). The mutated protein had a point mutation (Arg8 was replaced with a Cys in the α-subunit) that affected the DNA-IHF interaction (Toussaint et al., 1994) so that in the IHF mutant hydrogenase synthesis was reduced
to 25 per cent of that of the wild-type strain (B10), but was still responsive to H2 (Toussaint et al., 1997). As shown in Fig. 8, the role of IHF is to strongly bend DNA to allow interaction of the transcription activator HupR with RNA polymerase. 2. Response to CO In some microorganisms, biosynthesis of hydrogenase is regulated by various metabolites, which are electron donors or acceptors. Regulation of hydrogenase biosynthesis in E. coli by formate through the transcription factor FhlA has been very well studied. FhlA belongs to the NtrC sub-family of regulators. It is not activated by phosphorylation, for its N-terminal domain lacks the phosphorylatable aspartate residue, but is activated by binding of formate, which is the effector molecule (for recent reviews see Leonhartsberger et al., 2002; Sawers et al., 2004; Sawers, 2005; Böck et al., 2006). Carbon monoxide can support anaerobic growth of Rsp. rubrum. CO-dependent growth relies on a CO oxidation system, encoded by coo genes organized in two CO-regulated transcriptional units. The coo regulon comprises CooS, an O2-sensitive CO dehydrogenase and CooLH, a CO-tolerant hydrogenase. Expression of the coo genes depends upon the activity of the CooA (CO-oxidation activator) protein, a member of the CRP/Fnr (cyclic AMP receptor protein/fumarate nitrate reduction) transcriptional regulator family. CooA is a homodimer in which each monomer contains a b-type heme and senses CO under anaerobic conditions (Shelver et al., 1997). Actually, CooA senses both the redox state of the cell and CO, for only the reduced form of the heme Fe (reduced at about –300 mV, Nakajima et al., 2001) can bind CO. CO binding stabilizes a conformation of the dimeric protein that allows sequence specific DNA binding, and transcription is activated through contacts between CooA and RNA polymerase. The crystal structure of Fe(II)CooA has been solved (Lanzilotta et al., 2000). Based on mutational and spectroscopic studies, models invoking movement of the heme group during CO-specific activation of CooA have been proposed (for recent reviews see Aono, 2003; Roberts et al., 2001, 2004, 2005). 3. O2 Regulation Optimal synthesis of hydrogenases usually requires strict anaerobiosis or microaerobiosis and the synthe-
Chapter 37
Regulation of Hydrogenase Genes
HupR Reg A RNA polymerase
IHF
70
hup SLC Fig. 8. Transcription control at the promoter of the uptake hydrogenase hupSLC genes. The non phosphorylated transcription factor HupR binds to the TTG-N5-CAA palindromic sequence centered at –157 nucleotides upstream from the transcription start site. HupR belongs to the class of enhancer-binding proteins that activates σ70 RNA polymerase instead of σ54 RNA polymerase (Dischert et al., 1999) as is the case for the NtrC protein of Rhodobacter capsulatus (Bowman and Kranz, 1998). The global regulator IHF is required to bend DNA and to allow interaction between HupR and the RNA polymerase (Toussaint et al., 1997). The response regulator RegA binds to two sites, and represses transcription by preventing binding of either the RNA polymerase and/or of the IHF to the hupSLC promoter (Elsen et al., 2000) (Adapted and reproduced with permission from Vignais et al., 2005 © The Biochemical Society)
sis of most hydrogenases is negatively regulated by molecular oxygen. The sensing of low O2 concentration involves global regulatory proteins homologous to the E. coli Fnr protein. The E. coli anaerobic regulator Fnr is a cytoplasmic O2-responsive regulator with a sensory and a regulatory DNA-binding domain. Fnr activates the transcription of genes involved in anaerobic respiratory pathways while it represses the expression of genes involved in aerobic energy generation (Iuchi and Lin, 1993). The protein binds as a dimer to an Fnr consensus sequence of dyad symmetry, TTGAT-N4-ATCAA. Fnr activity depends on the presence of a [4Fe-4S]2+ cluster that, in the presence of O2, is converted rapidly to a more O2-stable [2Fe-2S]2+ cluster (Kiley and Beinert, 1998). Indeed, this O2-lability of the [4Fe-4S]2+ cluster makes Fnr an O2 sensor (Beinert and Kiley, 1999; Bates et al., 2000; Kiley and Beinert, 2003). During anaerobiosis the Fnr protein binds and activates the hyp operon in E. coli and thus indirectly affects hydrogenase synthesis. In Rhizobia, Fnr homologs that regulate hydrogenase synthesis are either Fnr-like (such as FixK1 in B. japonicum or FnrN in Rhizobium leguminosarum) or FixK-like (such as FixK2 in B. japonicum). FixK-like proteins lack the N-terminal region of Fnr for binding of the [4Fe-4S] cluster. The main difference between Fnr-like and FixK-like regulators is therefore at the
753 level of redox control. The FixK-like proteins, which lack the redox sensitive cysteines, are activated by an associated O2-sensitive two-component system, FixLJ. In B. japonicum, O2 signal transduction is organized along two regulatory cascades involving the activators FixK2 and NifA (Sciotti et al., 2003). Transcription of hydrogenase structural genes is activated by DNA binding of FixK2, the expression of which depends on the FixL/FixJ two-component system (Nellen-Anthamatten et al., 1998). The transmembrane sensor FixL has a heme-binding domain that belongs to the PAS superfamily and directly senses O2. Association of O2 with FixL induces a conformational change that activates the C-terminal kinase domain, then triggers autophosphorylation from ATP and phosphoryl transfer to FixJ. Similarly, in Rhizobium leguminosarum nodules, hydrogenase transcription is co-regulated with that of nitrogenase, and controlled by NifA and FnrN in response to low O2. In this case, NifA directly activates hydrogenase expression by binding to an upstream activating sequence (UAS) of the promoter region of the hupSL genes. This promoter also binds the global regulator IHF, and is transcribed by a σ54-RNA polymerase (Brito et al., 1997; see Vignais and Colbeau, 2004 for additional references). Recently, the gene for an Fnr homolog, FnrT, has been identified in the genome of Tca. roseopersicina, and an fnrT knockout mutant constructed (Kovács et al., 2005b). Tca. roseopersicina BBS contains a heat-stable membrane-associated hydrogenase, encoded by the hyn operon. In the fnrT mutant anaerobic induction of the hynS expression was abolished, suggesting that FnrT is an activator of the hynS promoter. Complementary experiments showed that the Tca. roseopersicina hynS promoter could be activated in E. coli in an Fnr-dependent manner. In vitro experiments showed that E. coli Fnr could bind to two sites in the hyn regulatory region, and activate transcription initiation (to different extents) when bound at the hynS promoter at the two target sites (Kovács et al., 2005b). 4. Redox regulation Redox regulation was first studied in E. coli. In E. coli, the synthesis of hydrogenase-1 and -2 depends on the global two-component regulatory system ArcB/ArcA (Iuchi and Lin, 1993). ArcB is a transmembrane sensor protein, the histidine kinase activity of which is stimulated in the absence of O2.
754 Under anaerobic conditions, the ArcB sensor kinase autophosphorylates and then transphosphorylates the global transcriptional regulator ArcA. Phosphorylation of ArcA induces multimerization, a prerequisite for DNA binding (Jeon et al., 2001). ArcA~P is the active form which represses target genes of aerobic metabolism, and activates genes of anaerobic metabolism. Quinones are redox signals for the Arc system. Oxidized forms of quinone electron carriers act as direct negative signals and inhibit autophosphorylation of ArcB during aerobiosis, thus providing a link between the respiratory chain and gene expression (Georgellis et al., 2001; Malpica et al., 2004). By oxidizing H2 and generating low potential electrons that can feed energy-consuming processes such as carbon dioxide and dinitrogen fixation, the hydrogenase participates to cellular redox metabolism. A global two-component signal transduction system, called RegB/RegA in Rba. capsulatus and PrrB/PrrA in Rba. sphaeroides (Chapter 35, Bauer et al.; Chapter 36, Klug and Masuda), is implicated in the redox control of the above-mentioned processes (Swem et al., 2001; for recent reviews see Dubbs and Tabita, 2004; Elsen et al., 2004). In vivo, under aerobic conditions, the cbb3 cytochrome oxidase generates an inhibitory signal, preventing accumulation of the activated PrrA (Oh et al., 2004). A redox-responsible cysteine (Cys265) located in the dimerization interface of the sensor kinase RegB appears to be involved in redox sensing and in the regulation of RegB autophosphorylation (Swem et al., 2003). It has been shown recently that a periplasmic loop between the transmembrane helices 3 and 4 of RegB contains a ubiquinone-binding site. This domain was suggested to be responsible for sensing the redox state of the ubiquinone pool and for controlling autophosphorylation of RegB (Swem et al., 2006). The response regulator RegA of Rba. capsulatus was shown by footprinting experiments to bind to the hupS promoter at two sites, a high affinity site located between the binding sites for the global regulator IHF and for RNA polymerase, and a low affinity site overlapping the binding site IHF (Elsen et al., 2000) (Fig. 8). RegB- and RegA-defective mutants had 3 to 5 times more activity than the wild type, either in the presence or absence of H2. These data indicate that, in Rba. capsulatus, the RegB-RegA couple exerts a negative control on hydrogenase synthesis, and that global regulation by RegB-RegA is superimposed on H2-regulation (Elsen et al., 2000).
Paulette M. Vignais In Rps. palustris, the homologous RegS-RegR two-component regulatory system also represses hydrogenase gene expression. But, in contrast to Rhodobacter, RegS-RegR does not play a pivotal role in global gene regulation in Rps. palustris (Rey et al., 2006). Homologs of the highly conserved RegB and RegA proteins have been found in a wide number of photosynthetic and nonphotosynthetic bacteria, with evidence suggesting that RegB-RegA plays a fundamental role in the transcription of redox-regulated genes in many bacterial species (Elsen et al. 2004). The use of a gene chip with a PrrA mutant strain has revealed that approx. 20% of genes/orfs (i.e., more than 800) present in Rba. sphaeroides underwent significant changes in expression profiles in the absence of PrrA relative to the wild-type (Kaplan et al., 2005), indicating that PrrA has a more universal role extending beyond the regulation of redox processes. C. Conclusions and Perspectives Genome sequencing has revealed the wide distribution of hydrogenases in the prokaryotes, and the presence of multiple hydrogenases in Archaea and in Bacteria, including photosynthetic prokaryotes (Fig. 1). Phylogenetic analyses have led to the identification of two phylogenetically distinct groups, the [NiFe]- and the [FeFe]-hydrogenases, and to various subgroups that form the basis of a coherent system of classification. Studies of regulation have mainly concerned the [NiFe]-hydrogenases of Proteobacteria, in particular their regulation in response to H2, O2 and CO. Hydrogen utilization has recently been shown to be also under redox control and to belong to the Reg regulon. Post-genomic analysis (transcriptome, proteome, metabolome) has and will be essential to elucidating the metabolic roles of hydrogenases and the regulation of their biosynthesis and activity. Although this chapter is focused on H2 uptake, some hydrogenases function physiologically as H2 producers. The current interest in H2 as an alternative to fossil fuels has led to an increased interest in these enzymes. Studies of H2 metabolism and regulation will be important in engineering microorganisms at the cellular level in order to maximize H2 production.
Chapter 37
Regulation of Hydrogenase Genes
References Aono S (2003) Biochemical and biophysical properties of the CO-sensing transcriptional activator CooA. Acc Chem Res 36: 825–831 Appel J and Schulz R (1998) Hydrogen metabolism in organisms with oxygenic photosynthesis: Hydrogenases as important regulatory devices for a proper redox poising? J Photochem Photobiol B-Biol 47: 1–11 Appel J, Phunpruch S, Steinmuller K and Schulz R (2000) The bidirectional hydrogenase of Synechocystis sp. PCC 6803 works as an electron valve during photosynthesis. Arch Microbiol 173: 333–338 Atta M and Meyer J (2000) Characterization of the gene encoding the [Fe]-hydrogenase from Megasphaera elsdenii. Biochim Biophys Acta 1476: 368–371 Bates DM, Popescu CV, Khoroshilova N, Vogt K, Beinert H, Munck E and Kiley PJ (2000) Substitution of leucine 28 with histidine in the Escherichia coli transcription factor FNR results in increased stability of the [4Fe-4S](2+) cluster to oxygen. J Biol Chem 275: 6234–6240 Beinert H and Kiley PJ (1999) Fe-S proteins in sensing and regulatory functions. Curr Opin Chem Biol 3: 152–157 Bernhard M, Buhrke T, Bleijlevens B, De Lacey AL, Fernandez VM, Albracht SP, Friedrich B (2001) The H2 sensor of Ralstonia eutropha. Biochemical characteristics, spectroscopic properties, and its interaction with a histidine protein kinase. J Biol Chem 276:15592–15597 Böck A, King PW, Blokesch M, Posewitz MC (2006) Maturation of hydrogenases. Adv Microb Physiol 51: 1–71 Bowman WC and Kranz RG (1998) A bacterial ATP-dependent enhancer binding protein that activates the housekeeping RNA polymerase. Gene Dev 12: 1884–1893 Brito B, Martinez M, Fernandez D, Rey L, Cabrera E, Palacios JM, Imperial J and Ruiz-Argueso T (1997) Hydrogenase genes from Rhizobium leguminosarum bv. viciae are controlled by the nitrogen fixation regulatory protein NifA. Proc Natl Acad Sci U S A 94: 6019–6024 Buhrke T, Lenz O, Porthun A and Friedrich B (2004) The H2sensing complex of Ralstonia eutropha: Interaction between a regulatory [NiFe] hydrogenase and a histidine protein kinase. Mol Microbiol 51: 1677–1689 Buhrke T, Lenz O, Krauss N and Friedrich B (2005) Oxygen tolerance of the H2-sensing [NiFe] hydrogenase from Ralstonia eutropha H16 is based on limited access of oxygen to the active site. J Biol Chem 280: 23791–23796 Burgdorf T, Lenz O, Buhrke T, van der Linden E, Jones AK, Albracht SP, Friedrich B (2005) [NiFe]-hydrogenases of Ralstonia eutropha H16: Modular enzymes for oxygen-tolerant biological hydrogen oxidation. J Mol Microbiol Biotechnol 10: 181–196 Cammack R, Frey M and Robson R (2001) Hydrogen as a Fuel. Learning from Nature. Taylor and Francis, London and New York Casalot L, Rousset M (2001) Maturation of the [NiFe] hydrogenases. Trends Microbiol 9: 228–237 Cournac L, Guedeney G, Peltier G and Vignais PM (2004) Sustained photoevolution of molecular hydrogen in a mutant of Synechocystis sp. strain PCC 6803 deficient in the type I NADPH-
755 dehydrogenase complex. J Bacteriol 186: 1737–1746 Davies KM, Skamnaki V, Johnson LN and Vénien-Bryan C (2006) Structural and functional studies of the response regulator HupR. J Mol Biol 359: 276–288 Dischert W, Vignais PM and Colbeau A (1999) The synthesis of Rhodobacter capsulatus HupSL hydrogenase is regulated by the two-component HupT/HupR system. Mol Microbiol 34: 995–1006 Drummond MH, Whitty P and Wootton JC (1986) Sequence and domain relationship of ntrC and nifA from Klebsiella pneumoniae: homologies to other regulatory proteins. EMBO J. 5: 441–447 Dubbs JM and Tabita FR (2004) Regulators of nonsulfur purple phototrophic bacteria and the interactive control of CO2 assimilation, nitrogen fixation, hydrogen metabolism and energy generation. FEMS Microbiol Rev 28: 353–376 Duché O, Elsen S, Cournac L and Colbeau A (2005) Enlarging the gas access channel to the active site renders the regulatory hydrogenase HupUV of Rhodobacter capsulatus O2 sensitive without affecting its transductory activity. FEBS J 272: 3899–3908 Eberz G and Friedrich B (1991) Three trans-acting functions control hydrogenase expression in Alcaligenes eutrophus. J. Bacteriol. 173: 1845–1854 Elsen S, Richaud P, Colbeau A and Vignais PM (1993) Sequence analysis and interposon mutagenesis of the hupT gene, which encodes a sensor protein involved in repression of hydrogenase synthesis in Rhodobacter capsulatus. J Bacteriol 175: 7404–7412 Elsen S, Colbeau A, Chabert J and Vignais PM (1996) The hupTUV operon is involved in negative control of hydrogenase synthesis in Rhodobacter capsulatus. J Bacteriol 178: 5174–5181 Elsen S, Colbeau A and Vignais PM (1997) Purification and in vitro phosphorylation of HupT, a regulatory protein controlling hydrogenase gene expression in Rhodobacter capsulatus. J Bacteriol 179: 968–971 Elsen S, Dischert W, Colbeau A and Bauer CE (2000) Expression of uptake hydrogenase and molybdenum nitrogenase in Rhodobacter capsulatus is coregulated by the RegB-RegA twocomponent regulatory system. J Bacteriol 182: 2831–2837 Elsen S, Duché O and Colbeau A (2003) Interaction between the H2 sensor HupUV and the histidine kinase HupT controls HupSL hydrogenase synthesis in Rhodobacter capsulatus. J Bacteriol 185: 7111–7119 Elsen S, Swem LR, Swem DL and Bauer CE (2004) RegB/RegA, a highly conserved redox-responding global two-component regulatory system. Microbiol Mol Biol Rev 68: 263–279 Fox JD, Kerby RL, Roberts GP and Ludden PW (1996) Characterization of the CO-induced, CO-tolerant hydrogenase from Rhodospirillum rubrum and the gene encoding the large subunit of the enzyme. J Bacteriol 178: 1515–1524 Friedrich B, Vignais PM, Lenz O and Colbeau A (2001) Regulation of hydrogenase gene expression. In: Cammack R, Frey M and Robson R (eds) Hydrogen as a fuel. Learning from Nature, pp 33–56. Taylor and Francis, London and New York Friedrich B, Buhrke T, Burgdorf T and Lenz O (2005) A hydrogen-sensing multiprotein complex controls aerobic hydrogen metabolism in Ralstonia eutropha. Biochem Soc Trans 33: 97–101
756 Georgellis D, Kwon O and Lin EC (2001) Quinones as the redox signal for the Arc two-component system of bacteria. Science 292: 2314–2316 Hedderich R (2004) Energy-converting [NiFe] hydrogenases from archaea and extremophiles: ancestors of complex I. J Bioenerg Biomembr 36: 65–75 Higuchi Y, Yagi T and Yasuoka N (1997) Unusual ligand structure in Ni-Fe active center and an additional Mg site in hydrogenase revealed by high resolution X-ray structure analysis. Structure 5: 1671–1680 Hoch JA and Silhavy TJ (1995) Two-component signal transduction. American Society for Microbiology, Washington DC. Horner DS, Heil B, Happe T and Embley TM (2002) Iron hydrogenases–ancient enzymes in modern eukaryotes. Trends Biochem Sci 27: 148–153 Iuchi S and Lin EC (1993) Adaptation of Escherichia coli to redox environments by gene expression. Mol Microbiol 9: 9–15 Jeon Y, Lee YS, Han JS, Kim JB and Hwang DS (2001) Multimerization of phosphorylated and non-phosphorylated ArcA is necessary for the response regulator function of the Arc two-component signal transduction system. J Biol Chem 276: 40873–40879 Jones R and Haselkorn R (1989) The DNA sequence of the Rhodobacter capsulatus ntrA, ntrB and ntrC analogues required for nitrogen fixation. Mol Gen Genet 217:507–516 Jouanneau Y, Jeong HS, Hugo N, Meyer C and Willison JC (1998) Overexpression in Escherichia coli of the rnf genes from Rhodobacter capsulatus — characterization of two membrane-bound iron-sulfur proteins. Eur J Biochem 251: 54–64 Kaplan S, Eraso J and Roh JH (2005) Interacting regulatory networks in the facultative photosynthetic bacterium, Rhodobacter sphaeroides 2.4.1. Biochem Soc Trans 33: 51–55 Kiley PJ and Beinert H (1998) Oxygen sensing by the global regulator, FNR: The role of the iron-sulfur cluster. FEMS Microbiol Rev 22: 341–352 Kiley PJ and Beinert H (2003) The role of Fe-S proteins in sensing and regulation in bacteria. Curr Opin Microbiol 6: 181–185 Kovács AT, Rákhely G, Balogh J, Maróti G, Cournac L, Carrier P, Meszaros LS, Peltier G and Kovács KL (2005a) Hydrogen independent expression of hupSL genes in Thiocapsa roseopersicina BBS. FEBS J 272: 4807–48016 Kovács AT, Rákhely G, Browning DF, Fulop A, Maróti G, Busby SJ and Kovács KL (2005b) An FNR-type regulator controls the anaerobic expression of hyn hydrogenase in Thiocapsa roseopersicina. J Bacteriol 187: 2618–2627 Lanzilotta WN, Schuller DJ, Thorsteinsson MV, Kerby RL, Roberts GP and Poulos TL (2000) Structure of the CO sensing transcription activator CooA. Nat Struct Biol 7: 876–880 Larimer FW, Chain P, Hauser L, Lamerdin J, Malfatti S, Do L, Land ML, Pelletier DA, Beatty JT, Lang AS, Tabita FR, Gibson JL, Hanson TE, Bobst C, Torres JL, Peres C, Harrison FH, Gibson J, Harwood CS (2004) Complete genome sequence of the metabolically versatile photosynthetic bacterium Rhodopseudomonas palustris. Nat Biotechnol 22: 55–61 Lee SY, De La Torre A, Yan D, Kustu S, Nixon BT and Wemmer DE (2003) Regulation of the transcriptional activator NtrC1: Structural studies of the regulatory and AAA+ ATPase domains. Genes Dev 17: 2552–2563 Lenz O and Friedrich B (1998) A novel multicomponent regulatory system mediates H2 sensing in Alcaligenes eutrophus. Proc Natl Acad Sci USA 95: 12474–12479
Paulette M. Vignais Lenz O, Bernhard M, Buhrke T, Schwartz E and Friedrich B (2002) The hydrogen-sensing apparatus in Ralstonia eutropha. J Mol Microbiol Biotechnol 4: 255–262 Leonhartsberger S, Korsa I and Böck A (2002) The molecular biology of formate metabolism in enterobacteria. J Mol Microbiol Biotechnol 4: 269–276 Malpica R, Franco B, Rodriguez C, Kwon O and Georgellis D (2004) Identification of a quinone-sensitive redox switch in the ArcB sensor kinase. Proc Natl Acad Sci USA 101: 13318–13323 Meyer J (2007) FeFe hydrogenases and their evolution: A genomic perspective. Cellular and Molecular Life Sciences (in press) Nakajima H, Honma Y, Tawara T, Kato T, Park SY, Miyatake H, Shiro Y and Aono S (2001) Redox properties and coordination structure of the heme in the CO-sensing transcriptional activator CooA. J Biol Chem 276: 7055–7061 Nellen-Anthamatten D, Rossi P, Preisig O, Kullik I, Babst M, Fischer HM and Hennecke H (1998) Bradyrhizobium japonicum FixK2, a crucial distributor in the FixLJ-dependent regulatory cascade for control of genes inducible by low oxygen levels. J Bacteriol 180: 5251–5255 Neuwald AF, Aravind L, Spouge JL and Koonin EV (1999) AAA+: A class of chaperone-like ATPases associated with the assembly, operation, and disassembly of protein complexes. Genome Res 9: 27–43 Ogura T and Wilkinson AJ (2001) AAA+ superfamily ATPases: Common structure–diverse function. Genes Cells 6: 575– 597 Oh JI, Ko IJ and Kaplan S (2004) Reconstitution of the Rhodobacter sphaeroides cbb3-PrrBA signal transduction pathway in vitro. Biochemistry 43: 7915–7923 Osuna J, Soberon X and Morett E (1997) A proposed architecture for the central domain of the bacterial enhancer-binding proteins based on secondary structure prediction and fold recognition. Protein Sci 6: 543–555 Rákhely G, Kovács AT, Maróti G, Fodor BD, Csanádi G, Latinovics D and Kovács KL (2004) Cyanobacterial-type, heteropentameric, NAD+-reducing NiFe hydrogenase in the purple sulfur photosynthetic bacterium Thiocapsa roseopersicina. Appl Environ Microbiol 70: 722–728 Rey FE, Oda Y and Harwood CS (2006) Regulation of uptake hydrogenase and effects of hydrogen utilization on gene expression in Rhodopseudomonas palustris. J Bacteriol 188: 6143–6152 Richaud P, Colbeau A, Toussaint B and Vignais PM (1991) Identification and sequence analysis of the hupR1 gene, which encodes a response regulator of the NtrC family required for hydrogenase expression in Rhodobacter capsulatus. J Bacteriol 173: 5928–5932 Roberts GP, Thorsteinsson MV, Kerby RL, Lanzilotta WN and Poulos T (2001) CooA: A heme-containing regulatory protein that serves as a specific sensor of both carbon monoxide and redox state. Prog Nucleic Acid Res Mol Biol 67: 35–63 Roberts GP, Youn H and Kerby RL (2004) CO-sensing mechanisms. Microbiol Mol Biol Rev 68: 453–473 Roberts GP, Kerby RL, Youn H and Conrad M (2005) CooA, a paradigm for gas sensing regulatory proteins. J Inorg Biochem 99: 280–292 Robinson VL, Buckler DR, Stock AM (2000) A tale of two components: A novel kinase and a regulatory switch. Nat Struct Biol 7: 626–633 Sawers RG, Blokesch M and Böck A (2004) Anaerobic formate
Chapter 37
Regulation of Hydrogenase Genes
and hydrogen metabolism. In: Curstiss IR (ed) EcoSal-Escherichia coli and Salmonella: Cellular and Molecular Biology Online (http://www.ecosal.org/ecosal/index.jsp) Chapter 3.5.4. ASM Press, Washington Sawers RG (2005) Formate and its role in hydrogen production in Escherichia coli. Biochem Soc Trans 33: 42–46 Schmehl M, Jahn A, Meyer zu Vilsendorf A, Hennecke S, Masepohl B, Schuppler M, Marxer M, Oelze J and Klipp W (1993) Identification of a new class of nitrogen fixation genes in Rhodobacter capsulatus: A putative membrane complex involved in electron transport to nitrogenase. Mol Gen Genet 241: 602–615 Schwartz E, Buhrke T, Gerischer U and Friedrich B (1999) Positive transcriptional feedback controls hydrogenase expression in Alcaligenes eutrophus H16. J Bacteriol 181: 5684–5692 Sciotti MA, Chanfon A, Hennecke H and Fischer HM (2003) Disparate oxygen responsiveness of two regulatory cascades that control expression of symbiotic genes in Bradyrhizobium japonicum. J Bacteriol 185: 5639–5642 Seshadri R, Adrian L, Fouts DE, Eisen JA, Phillippy AM, Methe BA, Ward NL, Nelson WC, Deboy RT, Khouri HM, Kolonay JF, Dodson RJ, Daugherty SC, Brinkac LM, Sullivan SA, Madupu R, Nelson KE, Kang KH, Impraim M, Tran K, Robinson JM, Forberger HA, Fraser CM, Zinder SH and Heidelberg JF (2005) Genome sequence of the PCE-dechlorinating bacterium Dehalococcoides ethenogenes. Science 307: 105–108 Shelver D, Kerby RL, He Yand Roberts GP (1997) CooA, a CO-sensing transcription factor from Rhodospirillum rubrum, is a CO-binding heme protein. Proc Natl Acad Sci USA 94: 11216–11220 Swem LR, Elsen S, Bird TH, Swem DL, Koch HG, Myllykallio H, Daldal F and Bauer CE (2001) The RegB/RegA two-component regulatory system controls synthesis of photosynthesis and respiratory electron transfer components in Rhodobacter capsulatus. J Mol Biol 309: 121–138 Swem LR, Kraft BJ, Swem DL, Setterdahl AT, Masuda S, Knaff DB, Zaleski JM and Bauer CE (2003) Signal transduction by the global regulator RegB is mediated by a redox-active cysteine. EMBO J 22: 4699–4708 Swem LR, Gong X, Yu CA and Bauer CE (2006) Identification of a ubiquinone-binding site that affects autophosphorylation of the sensor kinase RegB. J Biol Chem 281: 6768–6775 Tamagnini P, Axelsson R, Lindberg P, Oxelfelt F, Wunschiers R and Lindblad P (2002) Hydrogenases and hydrogen metabolism of cyanobacteria. Microbiol Mol Biol Rev 66: 1–20 Taylor BL and Zhulin IB (1999) PAS domains: Internal sensors of oxygen, redox potential, and light. Microbiol Mol Biol Rev 63: 479–506 Toussaint B, David L, de Sury d’Aspremont R, Vignais PM (1994) The IHF proteins of Rhodobacter capsulatus and Pseudomonas aeruginosa. Biochimie 76: 951–957 Toussaint B, de Sury d’Aspremont R, Delic-Attree I, Berchet V, Elsen S, Colbeau A, Dischert W, Lazzaroni Y and Vignais PM (1997) The Rhodobacter capsulatus hupSLC promoter: identification of cis-regulatory elements and of trans-activating factors involved in H2 activation of hupSLC transcription. Mol
757 Microbiol 26: 927–937 Van Soom C, Lerouge I, Vanderleyden J, Ruiz-Argueso T and Palacios JM (1999) Identification and characterization of hupT, a gene involved in negative regulation of hydrogen oxidation in Bradyrhizobium japonicum. J Bacteriol 181: 5085–5089 Vandamme P, Coenye T (2004) Taxonomy of the genus Cupriavidus: A tale of lost and found. Int J Syst Evol Microbiol 54: 2285–2289 Vaneechoutte M, Kampfer P, De Baere T, Falsen E, Verschraegen G (2004) Wautersia gen. nov., a novel genus accommodating the phylogenetic lineage including Ralstonia eutropha and related species, and proposal of Ralstonia [Pseudomonas] syzygii (Roberts et al. 1990) comb. nov. Int J Syst Evol Microbiol 54: 317–327 Vignais PM (2007) Hydrogenases and H+-reduction in primary energy conservation. In: Penefsky H, Schäfer G (eds) Bioenergetics. Structure and Function in Energy-Transducing Systems (Results and Problems in Cell Differentiation). Springer, Berlin, Heidelberg (online: http://dx.doi.org/10.1007/400_ 2006_027) Vignais P and Billoud B (2007) Occurrence, classification, and biological function of hydrogenases: An overview. Chem Rev 107: 4206–4272 Vignais PM and Colbeau A (2004) Molecular biology of microbial hydrogenases. Curr Issues Mol Biol 6: 159–188 Vignais PM, Billoud B and Meyer J (2001) Classification and phylogeny of hydrogenases. FEMS Microbiol Rev 25: 455–501 Vignais PM, Cournac L, Hatchikian EC, Elsen S, Serebryakova L, Zorin N and Dimon B (2002) Continuous monitoring of the activation and activity of [NiFe]-hydrogenases by membrane-inlet mass spectrometry. Intl J Hydrogen Ener 27: 1441–1448 Vignais PM, Willison JC and Colbeau A (2004) H2 respiration. In: Zannoni D (ed) Respiration in Archaea and Bacteria: Diversity of Prokaryotic Respiratory Systems (Advances in Photosynthesis and Respiration, Vol 16), pp 233–260. Springer, Dordrecht Vignais PM, Elsen S and Colbeau A (2005) Transcriptional regulation of the uptake [NiFe]hydrogenase genes in Rhodobacter capsulatus. Biochem Soc Trans 33: 28–32 Volbeda A, Charon MH, Piras C, Hatchikian EC, Frey M and Fontecilla-Camps JC (1995) Crystal structure of the nickel-iron hydrogenase from Desulfovibrio gigas. Nature 373: 580–587 Volbeda A, Montet Y, Vernède X, Hatchikian EC, JC F-C (2002) High-resolution crystallographic analysis of Desulfovibrio fructosovorans [NiFe]hydrogenase. International J Hydrogen Energy 27:1449–1461 Walker JE, Saraste M, Runswick MJ and Gay NJ (1982) Distantly related sequences in the alpha- and beta-subunits of ATP synthase, myosin, kinases and other ATP-requiring enzymes and a common nucleotide binding fold. EMBO J 1: 945–951 Xu DQ and Wu YQ (2001) Isolation and analysis of hupR gene required for the expression of hydrogenase in Rhodobacter sphaeroides. Sheng Wu Hua Xue Yu Sheng Wu Wu Li Xue Bao (Shanghai) 33: 607–614 Xu H and Hoover TR (2001) Transcriptional regulation at a distance in bacteria. Curr Opin Microbiol 4: 138–144
Chapter 38 Regulation of Nitrogen Fixation Bernd Masepohl Lehrstuhl für Biologie der Mikroorganismen, Fakultät für Biologie, Ruhr-Universität Bochum, 44780 Bochum, Germany
Robert G. Kranz* Department of Biology, Washington University, One Brookings Dr. St. Louis, MO 63130, U.S.A.
Summary ............................................................................................................................................................... 760 I. Nitrogen Fixation in Purple Nonsulfur Bacteria .............................................................................................. 760 A. Genes Involved in Nitrogen Fixation ................................................................................................ 761 B. Environmental Factors Regulating Nitrogen Fixation ....................................................................... 761 II. Three Regulatory Levels of Nitrogen Fixation and Molecular Mechanisms Studied in Rhodobacter capsulatus ............................................................................................................................... 762 A. Level 1: Sensing of the Fixed Nitrogen Status by the Ntr System.................................................... 763 1. The Two-Component System NtrB-NtrC................................................................................. 764 2. The NtrC Regulon ................................................................................................................... 765 B. Level 2: The NifA Regulon and the Alternative Sigma Factor RpoN ................................................ 766 1. Transcriptional Control by the NifA Class of Activators........................................................... 766 2. Post-Translational Control of NifA Activity .............................................................................. 766 C. Level 3: Post-Translational Control of Nitrogenase Activity by DraT and DraG ............................... 767 III. Other Factors that Feed into the Nitrogen Regulatory Circuitry ..................................................................... 768 A. Molybdenum Transport and Mo-dependent Regulation by MopA and MopB .................................. 768 B. The Alternative Nitrogenase of Rhodobacter capsulatus ............................................................... 769 C. Role of the Ammonium Transporter AmtB in Regulation of Nitrogen Fixation ................................. 769 D. Linkage of N2-Fixation to Photosynthesis, CO2 Assimilation, and H2 Metabolism ............................ 770 IV. Regulation in Other Purple Photosynthetic Bacteria ...................................................................................... 770 A. Regulation of Nitrogen Fixation in Rhodobacter sphaeroides ........................................................ 770 B. Regulation of Nitrogen Fixation in Rhodopseudomonas palustris .................................................. 771 C. Regulation of Nitrogen Fixation in Rhodospirillum rubrum ............................................................. 771 V. Future Perspectives ....................................................................................................................................... 771 Acknowledgments ................................................................................................................................................. 772 References ............................................................................................................................................................ 772
*Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 759–775. © 2009 Springer Science + Business Media B.V.
760
Bernd Masepohl and Robert G. Kranz
Summary Nitrogen fixation is the process of reducing atmospheric dinitrogen to ammonia, carried out by the enzyme nitrogenase using a source of reductant and energy (ATP). Early studies in the field of nitrogen fixation involved the discovery of nitrogenase and understanding the physiological properties of diazotrophs, including the evidence that most purple photosynthetic bacteria have this capability. More recently, at the forefront has been the understanding of diverse regulatory signals that turn on and off nitrogen fixation genes as well as the structure and mechanisms of nitrogenase. Here is presented the unique and common aspects of regulation of nitrogen fixation in the purple photosynthetic bacteria. The most studied model is Rhodobacter capsulatus, where three levels of control are apparent. The first senses nitrogen and ATP, mediated by the Ntr system. After transcriptional activation by NtrC, the second involves oxygen and more nitrogen sensing via the NifA-type transcriptional activator. The final level of control resides at the activity of nitrogenase, which is modulated by the DraT and DraG system in response to nitrogen and other signals. Some interplay exists between nitrogen fixation circuitry and general regulatory systems for light and redox sensors of purple phototrophs. In summary, the regulatory circuitry is complex; sensors respond to ATP, nitrogen status, oxygen level, light, and the availability of molybdenum and iron, so that nitrogenase is only expressed and active under conditions that make physiological sense. Also discussed is the analysis of other purple phototrophs, the presence of alternative nitrogenases, as well as what the future may hold in this area. For example, light energy is ultimately used to fix nitrogen in these prokaryotes. Given the increased costs and environmental concerns for fertilizers, what is the potential for harnessing these capabilities and what are the possible hurdles, not the least of which is the need to control feedback regulations from all these signals? I. Nitrogen Fixation in Purple Nonsulfur Bacteria The large reservoir of nitrogen gas in our atmosphere is made available for diazotrophic (N2-fixing) bacteria because they can reduce dinitrogen to ammonium. The ability to fix dinitrogen is found only among representatives of two of the primary kingdoms of living organisms, the prokaryotic Archaea and Bacteria. The eight-electron reduction of nitrogen gas to ammonium requires significant energy, as shown here: N2 + 8 e– + 8 H+ + 16–24 ATP → 2 NH3 + 1 H2 + 16–24 ADP + 16–24 Pi Because phototrophs derive energy from sunlight, they are particularly attractive models to investigate this process and how it is controlled. Reduction of N2 to NH3 is catalyzed by an enzyme called nitrogenase, and the genes required for nitrogen fixation are commonly called nif genes. All nitrogenases consist of two dissociable metalloproteins, dinitrogenase (e.g., NifDK) and dinitrogenase reductase (e.g., NifH). The first discovered and most Abbreviations: 2-OG – 2-oxoglutarate; A. – Azotobacter; Ar. – Azorhizobium; E. – Escherichia; K. – Klebsiella; PII – generic term for GlnB, GlnK, and GlnZ; Rba. – Rhodobacter; Rps. – Rhodopseudomonas; Rsp. – Rhodospirillum
commonly occurring nitrogenases have molybdenum in their active center, the iron-molybdenum cofactor (FeMo-co). In addition to Mo-nitrogenase, some species possess homologous alternative nitrogenases, namely a vanadium-containing nitrogenase (V-nitrogenase carrying an iron-vanadium cofactor, FeV-co, encoded by vnf for vanadium-dependent nitrogen fixation) and/or the so-called iron-only nitrogenase (Fe-nitrogenase), which does not contain any heterometal in its cofactor (FeFe-co). Genes specific for the latter nitrogenase are encoded by anf for alternative nitrogen fixation. It should be emphasized that both V- and Fe-nitrogenases are ‘alternative’ nitrogenases in the sense that they are non-Mo enzymes. All three nitrogenases are based on genetically distinct systems, but significant amino acid sequence homology exists among the three forms of nitrogenase, especially among dinitrogenase reductases. The distribution of Mo- and non-Mo nitrogenases in selected purple bacteria is shown in Table 1. Besides these examples, nitrogenase activity has been demonstrated for many other purple nonsulfur bacteria by the acetylene reduction assay (for a review, see Madigan, 1995), suggesting that the ability to reduce N2 is widespread in these genera. All nitrogen-fixing bacteria contain Mo-nitrogenase, and in addition, some species possess one or two non-Mo nitrogenases. No bacterium has yet been
Chapter 38
Regulation of Nitrogen Fixation
761
Table 1. Occurrence of Mo-, V-, and Fe-nitrogenases in purple nonsulfur bacteria Species
Mo-nitrogenase
V-nitrogenase
Fe-nitrogenase
Rhodobacter capsulatus Rhodobacter sphaeroides
+ +
– –
+ –
Rhodopseudomonas palustris Rhodospirillum rubrum
+ +
+ –
+ +
described which contains an alternative nitrogenase but no Mo-nitrogenase. Even closely related species such as Rhodobacter (Rba.) capsulatus and Rba. sphaeroides may differ in their capacity to synthesize alternative nitrogenases. It is worth noting that Rhodopseudomonas (Rps.) palustris is one of just a few species described so far that has the capacity to synthesize all three nitrogenases (Oda et al., 2005). All nitrogenases convert N2 to NH3 with the concomitant obligate production of molecular hydrogen. This side product is a ramification of the active site chemistry, whereby protons are reduced to the dihydrogen. H2 generated by nitrogenase is subsequently oxidized via uptake hydrogenases, which are frequently found in N2-fixing species. Hydrogenases recover the energy or provide electrons to the Rnf membrane complex involved in electron transfer to nitrogenase (see Chapter 37, Vignais). A. Genes Involved in Nitrogen Fixation As many as 50–100 genes involved in nitrogen fixation have been identified in purple bacteria such as Rba. capsulatus or Rps. palustris (for a review, see Masepohl et al., 2004; Oda et al., 2005). These genes may be assigned to eight groups with respect to the function of their products in N2-fixation, namely (i) Mo-nitrogenase structural genes (nifHDK), (ii) V-nitrogenase structural genes (vnfHDK), (iii) Fenitrogenase structural genes (anfHDK), (iv) electron transport genes, (v) cofactor biosynthesis genes, (vi) molybdenum transport genes, (vii) N-regulated genes of unknown function, and (viii) regulatory genes, the topic of this chapter. This grouping does not necessarily reflect operon structures, although nif genes are often clustered in purple photosynthetic bacteria. For example, in Rba. capsulatus most of the nitrogen fixation genes are localized in four gene clusters (http://www.ergo-light.com). Regulatory genes controlling nitrogen fixation in purple bacteria are listed in Table 2. The respective gene products are either specific for nitrogen fixation such as the central transcriptional activator of all nif
genes, NifA, or they play global roles like the twocomponent regulatory system, RegAB, that integrates the control of photosynthesis, carbon dioxide assimilation, hydrogen oxidation, and nitrogen fixation (Elsen et al., 2000). Beside DNA-binding proteins such as NifA or RegA, which act at the transcriptional level, other proteins like the PII-like signal transduction proteins GlnB and GlnK or ADP-ribosyl transferase DraT act at the protein level (see below). Throughout this chapter we will use the term PII as a generic term for GlnB, GlnK, and GlnZ. B. Environmental Factors Regulating Nitrogen Fixation Because nitrogen fixation is a highly energy-demanding process, expression of nitrogenase is inhibited by NH4+ in all free-living diazotrophs. Beside NH4+, oxygen plays an important role in nitrogen fixation, since all nitrogenases are highly sensitive towards oxygen at the active site. To cope with this problem, bacteria have developed different strategies to protect nitrogenase against O2. Some bacteria, including the best-characterized diazotrophic bacterium, Klebsiella (K.) pneumoniae, and all purple photosynthetic bacteria, synthesize nitrogenase only when oxygen partial pressure is low. Other bacteria, such as Azotobacter (A.) vinelandii are able to fix dinitrogen under aerobic conditions by protecting nitrogenase with increased respiration. A third strategy has been developed by some filamentous cyanobacteria, which form specialized cells called heterocysts, where N2-fixation takes place. At least at moderate temperatures, Monitrogenases exhibit higher specific activities than the non-Mo nitrogenases with respect to N2 reduction rates (for a review, see Masepohl et al., 2002a), and consequently, Mo-nitrogenase is the preferred enzyme as long as the Mo-requirement is fulfilled. Due to these conditions, diazotrophic bacteria typically regulate synthesis and activity of nitrogenase in response to a number of environmental factors including (i) availability of ammonium, (ii) oxygen partial pressure, (iii) availability of molybdenum, (iv)
Bernd Masepohl and Robert G. Kranz
762 Table 2. Genes involved in the regulation of nitrogen fixation Gene
Function of gene product and relevant characteristics
nifA
transcriptional activator of Mo-nitrogenase genes; duplication (nifA1, nifA2) in R. capsulatus transcriptional activator of V-nitrogenase genes in Rps. palustris transcriptional activator of Fe-nitrogenase genes sigma factor 54 specific for nitrogen control; alias NtrA or NifR4 response regulator of 2-component regulatory system; alias NifR1 sensor kinase of 2-component regulatory system; phosphodonor towards NtrC sensor kinase of 2-component regulatory system NtrYX; cross-talk phosphodonor towards NtrC Hfq-like protein controlling nifA and anfA expression PII-like signal transduction protein controlling activity of NtrB, NifA, and DraT PII-like signal transduction protein controlling activity of NifA and DraT PII-like signal transduction protein from Rsp. rubrum controlling activity of DraG transport of NH3 into the cell; control of PII function by sequestration to the cell membrane; duplication (amtB1, amtB2) in Rsp. rubrum uridylyltransferase / UMP-removing enzyme controlling reversible modification of PII
vnfA anfA rpoN ntrC ntrB ntrY nrfA glnB glnK glnY amtB glnD draT draG mopA mopB ranR himA hip hvrA regB regA
dinitrogenase reductase ADP-ribosyl transferase controlling activity of nitrogenase reductase dinitrogenase reductase activating glycohydrolase molybdate-dependent repressor controlling transcription of anfA and the Mo-transport operon mopA-modABC molybdate-dependent repressor controlling transcription of anfA and the Mo-transport operon mopA-modABC putative TetR-like regulator essential for diazotrophic growth via Fe-nitrogenase IHF (integration host factor) α-subunit involved in DNA-bending of RpoNdependent promoters IHF (integration host factor) β-subunit involved in DNA-bending of RpoNdependent promoters histone-like (H-NS) protein modulating expression of photosynthesis and nitrogen fixation genes sensor kinase of 2-component regulatory system; phosphodonor towards RegA; alias PrrB response regulator of 2-component regulatory system; alias PrrA; integrates the control of photosynthesis, carbon dioxide assimilation, uptake hydrogenase, and nitrogen fixation
the C/N ratio at which carbon and nitrogen sources are consumed, (v) availability of iron, (vi) temperature, and (vii) in photosynthetic bacteria, light intensity. In contrast, the substrate of nitrogenase, N2, does not seem to be involved in regulation of nitrogen fixation. Most regulatory studies have focused on the key signals of NH4+-, Mo-, O2-, and light-regulation of N2-fixation.
II. Three Regulatory Levels of Nitrogen Fixation and Molecular Mechanisms Studied in Rhodobacter capsulatus Nitrogen fixation in Rba. capsulatus has been studied for more than two decades, and central regulatory proteins like NtrC, GlnB, GlnK, NifA, RpoN and DraT were found to be highly conserved between
Chapter 38
Regulation of Nitrogen Fixation
763
Fig. 1. Regulatory cascade controlling nitrogen fixation in Rba. capsulatus (adapted from Drepper et al., 2003). Ammonium control of synthesis and activity of Mo-nitrogenase and Fe-nitrogenase is controlled at three levels involving PII-like signal transduction proteins, GlnB and GlnK. In addition to N-control, transcription of anfA is repressed by MopA and MopB in the presence of molybdenum [+ Mo]. Nitrogen-sufficient conditions (presence of ammonium) or N-limiting conditions (absence of ammonium) are symbolized by [+ N] and [– N], respectively. The putative role of RcNtrC as an energy sensor is indicated by [ATP]. Promoters indicated by solid circles are RpoD-dependent and activated by NtrC~P; promoters indicated by solid squares are RpoN-dependent and activated by NifA or AnfA. For further details, see text.
Rba. capsulatus and other purple photosynthetic bacteria (Table 2 for regulatory proteins described here). Thus, Rba. capsulatus has been a model phototroph for many nitrogen regulatory steps (Kranz, 2000). For Rba. capsulatus regulation of N2-fixation can be visualized at three levels (Fig. 1) (Masepohl et al., 2002a). While we emphasize Rba. capsulatus in this section, other purple bacteria differ in several species-specific aspects as outlined below. Like many other diazotrophic bacteria, N2-fixing purple bacteria have evolved regulatory cascades enabling them to control the nitrogen fixation process at various levels (Fig. 1). At the first level, the Ntr (nitrogen regulation) system senses the fixed nitrogen status. Under nitrogen-limiting conditions, the NtrC protein becomes phosphorylated, thus increasing its ability to activate transcription of nifA, which codes for the central activator of all the other nif genes. The second level of regulation affects the activity of the NifA protein itself, which is modulated by the
N-status as well as by the redox-status. A third level of regulation affects the enzymatic activity of nitrogenase by a reversible ‘switch-off/-on’ mechanism responding to sudden changes in the environment such as addition of ammonium to a nitrogen-fixing culture or changes in light intensity. A. Level 1: Sensing of the Fixed Nitrogen Status by the Ntr System Sensing of the N-status by the Ntr system has been best characterized for Escherichia (E.) coli (Jiang et al., 1998a,b,c). Ntr-like systems are widespread in proteobacteria including purple photosynthetic species, suggesting similar mechanisms of N-sensing in this phylogenetic group, while non-proteobacterial species are devoid of such Ntr systems. Ammonium is the preferred source of fixed nitrogen for most bacteria including diazotrophic species, and consequently, synthesis of nitrogenase is inhibited by
764 NH4+. However, it is not NH4+, which is intracellularly sensed by the Ntr system, but the key metabolites of the glutamine synthetase (GS)-glutamate synthetase (GOGAT) pathway, namely glutamine (Gln) and 2oxoglutarate (2-OG). Hence, the ratio of Gln, with 2 nitrogen atoms, to 2-OG, with none, is a sensitive indicator for a cell’s nitrogen status. In E. coli, the ammonium transporter AmtB and the signal transduction protein GlnK couple the intracellular Ntr system to external levels of ammonium (Javelle et al., 2004; Javelle and Merrick, 2005; Durand and Merrick, 2006; see below). The above-mentioned GlnK belongs to the highly conserved class of PII-like signal transduction proteins. Most proteobacteria including Rba. capsulatus, code for two PII-like proteins, GlnB and GlnK. One exception to this is Rhodospirillum (Rsp.) rubrum, which, in addition to GlnB and GlnK, has the ability to synthesize a third PII-like protein, GlnZ (Zhang et al., 2001). The PII proteins are trimers, which synergistically bind 2-OG and ATP. In addition, PII proteins may be reversibly modified through uridylylation as shown for several bacteria including Rsp. rubrum (Johansson and Nordlund, 1997, 1999). Modification of PII proteins is catalyzed by the bifunctional uridylyltransferase/UMP-removing enzyme (UTase/UR, encoded by the glnD gene), which is the Gln sensor in this system. At low Gln concentrations (resembling N-deficiency), the glnD gene product acts as UTase leading to uridylylation of PII. Depending on the state of uridylylation and binding of 2-OG/ATP, the PII proteins interact with various regulatory proteins controlling nitrogen fixation. Target proteins interacting with PII proteins are NtrB, NifA, and DraT (Pawlowski et al., 2003; Zhu et al., 2006), each of which modulates one of the three regulatory levels controlling N2-fixation in Rba. capsulatus (Fig. 1, and see below). 1. The Two-Component System NtrB-NtrC Typically, Ntr systems consist of UTase/UR, PII, and the two-component regulatory system NtrB/NtrC, in which NtrB acts as a sensor histidine kinase and NtrC functions as a response regulator. Under nitrogen-limiting conditions (depletion of fixed N; [–N] conditions), PII becomes uridylylated, with PII-UMP being unable to interact with NtrB. Under these conditions, NtrB autophosphorylates, and NtrB~P serves as a phophodonor towards NtrC. Upon phosphorylation, NtrC becomes activated and in turn, activates tran-
Bernd Masepohl and Robert G. Kranz scription of its target genes. In Rba. capsulatus, NtrC is absolutely essential for transcriptional activation of nifA and anfA, whose products, in turn, are required for expression of all the other nitrogen fixation genes (Kranz and Foster-Hartnett, 1990; Schüddekopf et al., 1993; for a review, see Masepohl et al., 2004). Members of the NtrC family are modular in structure, consisting of three functional domains. The N-terminal domain contains a conserved aspartate residue, which is the site of phosphorylation essential for activation of the regulator. The central domain is responsible for ATP binding/hydrolysis and interaction with RNA polymerase. The C-terminal domain contains a helix-turn-helix motif, which recognizes and binds to specific upstream activator sequences (UAS) located 100–200 bp upstream of the transcription start codon of the respective target genes. As demonstrated by footprint analyses, Rba. capsulatus NtrC (RcNtrC) binds to tandem binding sites upstream of the respective operons (Bowman and Kranz, 1998; Masepohl et al., 2001). A consensus NtrC-binding site, similar to the UAS from enteric bacteria, has been defined for Rba. capsulatus as CGCC–N2–A(T/A)(T/A)–N–(T/A)T–N2–GC (Foster-Hartnett and Kranz, 1994). NtrB is the cognate phosphodonor for NtrC as demonstrated by in vitro reconstitution of the Rba. capsulatus two-component system (Cullen et al., 1996). Mutational analyses suggest, however, that another histidine kinase, NtrY, can substitute for NtrB in vivo as phosphodonor towards RcNtrC (Drepper et al., 2006). In contrast to ntrB and ntrY single mutant strains, which are able to grow with N2 as the sole nitrogen source, a ntrB ntrY double mutant (like a ntrC-deficient strain) is no longer able to grow diazotrophically, suggesting that no further histidine kinase cross-talks towards NtrC. NtrY is thought to be the cognate histidine kinase of the response regulator NtrX, which is involved in control of nifA expression in Azorhizobium (Ar.) caulinodans (Pawlowski et al., 1991). The Rba. capsulatus ntrY-ntrX genes are located immediately downstream of the nifR3-ntrB-ntrC operon. Expression of ntrY depends partially on the nifR3 promoter (Drepper et al., 2006), which constitutively drives transcription of the nifR3-ntrB-ntrC operon (Cullen et al., 1998). Since the concentration of NtrC protein remains constant, N-regulation of NtrC activity may solely be reflected by its status of phosphorylation. Members of the NtrC family typically activate transcription of their target genes in concert with
Chapter 38
Regulation of Nitrogen Fixation
RNA polymerase (RNAP) containing the alternative sigma factor RpoN. One of the few exceptions is RcNtrC, which activates transcription together with RNAP containing the house keeping sigma factor RpoD (Bowman and Kranz, 1998). This fact reflects a natural deletion in the central domain of RcNtrC, encompassing part of the so-called C3 region that is thought to contact with RpoN (Foster-Hartnett et al., 1994; Hoover, 2000). Additionally, it was shown that RcNtrC requires ATP binding but not hydrolysis (Bowman and Kranz, 1998). Therefore, RcNtrC is allosterically controlled by cellular levels of ATP. This surprise suggested that the RcNtrC protein might represent an energy sensor in addition to mediating the nitrogen status. Upon addition of NH4+ to a N2-fixing culture, the intracellular Gln levels increase. As a consequence, GlnD exhibits UR activity and deuridylylates PII. In contrast to PII-UMP, the unmodified form of PII (resembling N-sufficient conditions) binds to NtrB. In the presence of high 2-OG concentrations (resembling C-surplus conditions), however, the unmodified form of PII does not bind to NtrB. In complex with PII, NtrB no longer phosphorylates NtrC, and instead, NtrB promotes dephosphorylation of NtrC~P. As a result, the activator properties of NtrC diminish, and expression of NtrC-dependent promoters does not occur. Rba. capsulatus GlnB—but not GlnK—is able to interact with NtrB as implicated by yeast twohybrid studies (Pawlowski et al., 2003), suggesting that exclusively GlnB regulates NtrB activity. In line with this finding, mutations in glnB lead to high-level expression of NtrC-dependent genes in the presence of ammonium (Drepper et al., 2003) demonstrating that (i) GlnB is not required for NtrC activity, and (ii) GlnK does not substitute for GlnB with respect to control of NtrC activity. 2. The NtrC Regulon Under nitrogen-limiting conditions, Rba. capsulatus NtrC activates transcription of a number of genes involved in nitrogen metabolism. Among these are: nifA1, nifA2, and anfA (coding for the transcriptional activators of all the other nif and anf genes, Masepohl et al., 2002a), glnB-glnA (coding for a PII-like signal transduction protein and glutamine synthetase, Kranz et al., 1990), glnK-amtB (coding for a GlnB paralogue and an ammonium transporter, Drepper et al., 2003; Masepohl et al., 2002a; Yakunin and
765 Hallenbeck, 2002), amtY (coding for a second putative ammonium transporter; Masepohl et al., 2002a; Yakunin and Hallenbeck, 2002), mopA-modABCD (coding for a molybdenum-dependent regulator and a high-affinity ABC-type molybdenum transporter; Wang et al., 1993; Kutsche et al., 1996) and ureDABCEFG (required for synthesis and activity of urease; Masepohl et al., 2001). There appears to be some additional control on nifA1, nifA2 and anfA promoters. A global regulator called RegA acts on the NtrC-activated nifA2 promoter (see below). Additionally, NrfA may interact directly or indirectly on the expression of these genes (Drepper et al., 2002). Expression of nifA1 and nifA2 is about two-fold lower in an nrfA mutant strain as compared to the parental strain, while anfA expression is more than four-fold lower. The Rba. capsulatus nrfA gene product exhibits extensive similarity to the nif regulatory factor NrfA of Ar. caulinodans and the nucleoid-associated protein Hfq of E. coli. While Ar. caulinodans NrfA is absolutely required for nifA expression, and consequently, nrfA mutants exhibit a Nif – phenotype (Kaminski et al., 1994, 1998), Rba. capsulatus NrfA is not essential for diazotrophic growth of Rba. capsulatus, but it is required for maximal growth rates with N2 via either Mo- or Fe-nitrogenase. As the E. coli Hfq originally has been described as a site-specific RNA-binding protein, one may speculate that NrfA binds to and thereby stabilizes nifA1, nifA2, and anfA mRNAs. Rba. capsulatus can efficiently grow with urea as sole N-source. Under N-limiting conditions (absence of NH4+), expression of the urease genes is activated by NtrC (Masepohl et al., 2001). Interestingly, significant transcription occurs even in the presence of NH4+, approximately ten-fold higher in the wild type than in an ntrC mutant background. Thus, ure gene expression in the presence of ammonium also requires NtrC, suggesting that low levels of phosphorylated NtrC may be present even under nitrogen-sufficient conditions. Alternatively, unphosphorylated RcNtrC might activate ure gene expression. Although urea is effectively used as a nitrogen source in an NtrC-dependent manner, nitrogenase activity is not inhibited by urea. In contrast to the situation for E. coli, where NtrC is required for growth on different amino acids (including glutamine, proline, and arginine), mutations in Rba. capsulatus ntrC do not affect the use of either proline or arginine as sole nitrogen source (Kranz and Haselkorn, 1985; Moreno-Vivian et al., 1992; Keuntje et al., 1995). Synthesis of proline
766 dehydrogenase and arginase is induced by proline or arginine, respectively, regardless of the presence of ammonium. B. Level 2: The NifA Regulon and the Alternative Sigma Factor RpoN 1. Transcriptional Control by the NifA Class of Activators The NifA protein is the central regulator of N2-fixation in proteobacteria, and NifA activates transcription of all the other nif genes in concert with RNA polymerase containing the alternative sigma factor RpoN (Dixon, 1998). Transcriptional activation of nif gene expression involves binding of NifA to UAS, which are located at a distance of 100–200 bp upstream of the transcription start of the respective target genes. The consensus sequence for the NifA-specific UAS, which differs from the NtrC-specific UAS (see above), is TGT–N10–ACA (Morett and Buck, 1988). In Rba. capsulatus, perfect UAS sequences are present upstream of nitrogen fixation genes nifH, nifB1, nifB2, and rnfA, while imperfect UAS sequences differing in one base from the consensus are located upstream of nifE, nifU2, and orf6 (Masepohl et al., 1988; 1993; Moreno-Vivian et al., 1989; Preker et al., 1992; Schmehl et al., 1993; Willison et al., 1993). Binding of NifA to the UAS and interaction with RNAP-RpoN leads to open promoter complex formation and initiation of nif gene transcription. Contact between NifA and RNAP-RpoN may be facilitated by the integration host factor, IHF, an αβ-heterodimer encoded by the himA and hip genes (Toussaint et al., 1993). IHF binds to a sequence motif, ATCAA–N4–TTG (Craig and Nash, 1984), located between the binding sites for NifA (UAS) and RpoN (–24/–12). Binding of IHF induces a sharp bend in the DNA, thus bringing NifA and RNAP-RpoN into line with each other. RpoN proteins are unique in several ways compared to other sigma factors. First, initiation of transcription at RpoN-dependent promoters strictly requires an enhancer binding protein like NifA, which must hydrolyze ATP to activate transcription. ATP hydrolysis is thought to induce conformational changes in NifA required for open complex formation (Martinez-Argudo et al., 2004). Second, RpoN factors recognize well-conserved –24/–12 boxes with the sequence motif (C/T)TGG–N8–TTGC (Morett and Buck, 1989) with much more stringency than the family of RpoD factors show for the –35/–10 boxes. In Rba. capsula-
Bernd Masepohl and Robert G. Kranz tus, perfect –24/–12 sequences are present upstream of nitrogen fixation genes nifH, nifB1, nifB2, rnfA, nifE, nifU2, orf6, and fprA (Schmehl et al., 1993). RpoN proteins are modular in structure, consisting of an N-terminal glutamine-rich region required for response to the activator, an acidic region involved in open complex formation, and a C-terminal domain containing a helix-turn-helix motif and the so-called RpoN-box (Hoover, 2000). Rba. capsulatus RpoN exhibits strong similarity to other members of the RpoN family especially in the DNA-binding domain, thus reflecting the high degree of conservation of –24/–12 DNA binding sites. However, Rba. capsulatus RpoN contains a natural deletion of the acidic region that may be responsible for differences in DNA melting properties compared to K. pneumoniae RpoN (Cullen et al., 1994). The Rba. capsulatus rpoN gene forms part of the nifU2-rpoN operon, which belongs to the NifA regulon (Preker et al., 1992; Cullen et al., 1994). Since NifA requires RpoN for transcriptional activation of its target genes, expression of rpoN is autoregulated. In addition to NifA-dependent expression from the promoter upstream of the nifU2-rpoN operon, another promoter directly upstream of rpoN appears to be required for initial levels of RpoN (Cullen et al., 1994). 2. Post-Translational Control of NifA Activity NifA is modular in structure, consisting of an N-terminal domain involved in the control of NifA activity, a central domain for ATP hydrolysis and interaction with RNAP/RpoN, and a C-terminal DNA binding domain. With regard to O2-sensitivity, NifA proteins may be divided into two classes. Proteins such as Bradyrhizobium japonicum NifA have a cysteinerich, oxygen sensitive inter domain linker between the central and C-terminal domains (Fischer, 1994; Dixon, 1998). In contrast, O2-tolerant NifA proteins like K. pneumoniae and A. vinelandii NifA lack this linker, and instead are controlled by another protein, NifL. NifA proteins from Rba. capsulatus and other purple photosynthetic bacteria belong to the O2-sensitive class of rhizobial-type regulators as suggested by the presence of the inter domain linker and the absence of a nifL-like gene (Masepohl et al., 1988; Kern et al., 1998). NifL inhibits NifA activity by protein-protein interaction, modulated by redox changes, ligand binding, and interaction with the signal transduction protein
Chapter 38
Regulation of Nitrogen Fixation
GlnK (Martinez-Argudo et al., 2004). However, GlnK has different roles in K. pneumoniae and A. vinelandii concerning NifL-NifA interaction. While K. pneumoniae GlnK is involved in separation of the NifL-NifA complex under N-deficient conditions, A. vinelandii GlnK supports NifL-NifA interaction under N-sufficient conditions. In contrast to NifLcontrolled NifA proteins, other NifA proteins may directly interact with PII-like proteins as shown by yeast two-hybrid studies for Rba. capsulatus GlnB and GlnK (Pawlowski et al., 2003). In some species including Rsp. rubrum, Herbaspirillum seropedicae and Azospirillum brasilense, NifA activity depends on the presence of GlnB under N-limiting conditions. An Rsp. rubrum glnB mutant shows very low NifA activity (Zhang et al., 2000), suggesting that neither of the other two PII-like proteins, GlnK and GlnZ, substitutes for GlnB in this strain. In other species like Rba. capsulatus and Azorhizobium caulinodans, neither GlnB nor GlnK is required for NifA activity, but both GlnB and GlnK can inhibit NifA activity in the presence of NH4+ (Drepper et al., 2003). In contrast to other diazotrophic bacteria, Rba. capsulatus contains two copies of the nifA gene, nifA1 and nifA2 (Klipp et al., 1988; Masepohl et al., 1988; Paschen et al., 2001). Both NifA proteins differ only in their N-terminal 19 or 22 amino acid residues, respectively, whereas the remainders of both proteins (560 residues) are identical to each other. NifA1 and NifA2 can functionally substitute for each other, and either of the two regulators is sufficient for diazotrophic growth. Activity of both NifA proteins is strongly inhibited by ammonium, but the NifA2 protein seems to be slightly more tolerant than NifA1 towards NH4+ inhibition (Paschen et al., 2001). The difference between the extreme N-terminal sequences of the two NifA proteins and the characterization of ammonium-tolerant NifA1 mutants indicate that the N-terminal domain of NifA is involved in post-translational regulation in response to ammonium (Paschen et al., 2001). In line with this assumption, NifA mutants of H. seropedicae and As. brasilense lacking the N-terminal domain are no longer inhibited by ammonium (Arsene et al., 1996; Monteiro et al., 1999). C. Level 3: Post-Translational Control of Nitrogenase Activity by DraT and DraG While most diazotrophic proteobacteria regulate nitrogen fixation solely by controlling transcription of
767 the nifA gene (Level 1) and/or NifA activity (Level 2), few species have a further regulatory system, which rapidly and reversibly controls nitrogenase activity in response to sudden changes in the environment (for a review, see Halbleit and Ludden, 2000; Nordlund, 2000; Nordlund and Ludden, 2004). Such changes incompatible with N2-fixation like addition of NH4+, increase of O2-partial pressure, or — in case of photosynthetic bacteria — darkness, lead to ‘switch-off ’ of nitrogenase. After the negative stimulus has been removed, e.g., by consumption of NH4+ or re-illumination of the culture, nitrogenase is rapidly re-activated. The ‘switch-off/-on’ effect has extensively been studied in Rsp. rubrum, which became a paradigm for this regulatory mechanism (Halbleit and Ludden, 2000; Nordlund, 2000; Nordlund and Ludden, 2004). The molecular basis underlying ‘switch-off/-on’ involves reversible ADP-ribosylation of dinitrogenase reductase mediated by the DraT-DraG system (Fig. 2). After applying a negative stimulus to a nitrogen-fixing culture, DraT (dinitrogenase reductase ADP-ribosyl transferase) becomes transiently active and catalyzes ADP-ribosylation of the arginine residue 101 of one subunit of the dinitrogenase reductase homodimer. Due to steric hindrance, the modified nitrogenase reductase can no longer provide electrons to dinitrogenase. After removal of the negative stimulus, DraG (dinitrogenase reductase activating glycohydrolase) becomes active and removes the ADP-ribose from the modified enzyme, thereby restoring its activity. Activity of DraT and DraG is modulated by PII enabling these proteins to respond rapid and transiently to sudden changes in the environment ( Zhang et al., 2000; Drepper et al., 2003; Pawlowski et al., 2003). Both Rba. capsulatus GlnB and GlnK can directly interact with DraT as shown by yeast twohybrid studies (Pawlowski et al., 2003). In an Rba. capsulatus glnB glnK double mutant, DraT-mediated ADP-ribosylation of dinitrogenase reductase in response to NH4+ addition is abolished (Drepper et al., 2003). In contrast, response of DraT to darkness is not affected in this strain, suggesting that response of DraT to different environmental stimuli is mediated by different and independent mechanisms. Unlike Rsp. rubrum, where ADP-ribosylation seems to be the sole mechanism responsible for ‘switch-off ’ of nitrogenase activity, Rba. capsulatus also modulates its in vivo nitrogenase activity in response to addition of NH4+ by a second ADP-ribosylation-independent mechanism (Masepohl et al.,
768
Bernd Masepohl and Robert G. Kranz
Fig. 2. Post-translational regulation of nitrogenase (adapted from Zhang et al., 1997; Nordlund and Ludden, 2004). Upon addition of a negative stimulus, e.g., NH4+, DraT becomes transiently active and modifies dinitrogenase reductase. After removal of the negative stimulus, e.g., by consumption of NH4+, DraG removes the ADP-ribose.
1993b; Pierrard et al., 1993; Yakunin and Hallenbeck, 1998; Förster et al., 1999; Yakunin et al., 1999). III. Other Factors that Feed into the Nitrogen Regulatory Circuitry A. Molybdenum Transport and Mo-dependent Regulation by MopA and MopB Under N2-fixing conditions, Mo-nitrogenase is one of the most abundant proteins in the cell with more than 20 % of the soluble protein fraction. Consequently, N2-fixing cells have a high demand for molybdenum, which often is a limiting factor in the environment. To fulfill Mo-requirement at low Mo-concentrations, many bacteria (both diazotrophic and non-diazotrophic species) have evolved high-affinity ABC-type molybdenum transport systems comprising three proteins, ModA, ModB, and ModC (Self et al., 2001; Pau, 2004). ModA is the periplasmic molybdate-binding protein, ModB is the integral membrane channel protein, and ModC is the cytoplasmic ATPase. In Rba. capsulatus, modABC-like genes are located in close proximity to the structural genes of Mo-nitrogenase, nifHDK (Wang et al., 1993). While Moconcentrations in the nanomolar range are sufficient for Mo-nitrogenase activity in the wild type, modABC mutants require 500-fold higher Mo-concentrations
for activity of this enzyme. At Mo-concentrations above 1 µM the ModABC transporter is no longer essential for Mo-nitrogenase activity, suggesting the presence of a yet uncharacterized low-affinity Mo-uptake system. The Rba. capsulatus modABC gene region consists of two divergently transcribed operons, mopAmodABCD and mopB. The mopA and mopB genes encode two similar (but not identical) Mo-dependent regulators that are homologous to E. coli ModE. ModE has an N-terminal DNA-binding domain and a C-terminal molybdate-binding domain (Hall et al., 1999). Rba. capsulatus MopA and MopB can functionally substitute for each other in repression of the mopA-modABCD operon and anfA, thus limiting the amount of the Mo-transporter and Fe-nitrogenase at high Mo-concentrations (Kutsche et al., 1996; see below). Besides repression at high Mo-concentrations by MopA and MopB, transcription of the mopAmodABCD operon is activated under N-limiting conditions by NtrC (Kutsche et al., 1996). As both nifA (see above) and modABC belong to the NtrC regulon, expression of Mo-nitrogenase is directly coupled to expression of the Mo-transporter. In other words, it seems likely that the ModABC transport system primarily fulfills a Mo-requirement of Monitrogenase.
Chapter 38
Regulation of Nitrogen Fixation
B. The Alternative Nitrogenase of Rhodobacter capsulatus Rba. capsulatus can also synthesize an alternative Fe-only nitrogenase (Masepohl et al., 2002b; Schneider et al., 1991; Schüddekopf et al., 1993; Fig. 1). Fe-nitrogenase is expressed exclusively under conditions of nitrogen and molybdenum depletion. Under these conditions, NtrC activates transcription of anfA, coding for the transcriptional activator of the structural genes of Fe-nitrogenase, anfHDGK. Like NifA, AnfA activates transcription in concert with RNAP containing the alternative sigma factor RpoN. In addition to anfHDGK, several of the nif genes are required for synthesis and activity of Fe-nitrogenase (Schüddekopf et al., 1993). Among these are rpoN, nifB, nifV, and the rnf genes (involved in regulation, cofactor biosynthesis, and electron transport to nitrogenase, respectively). Since an Rba. capsulatus nifA mutant is still able to grow diazotrophically, one has to assume that AnfA can activate transcription of all these genes. Mo-nitrogenase exhibits higher specific activity than Fe-nitrogenase with respect to N2 reduction rates (for a review, see Masepohl et al., 2002b), and consequently, Mo-nitrogenase is the preferred enzyme as long as Mo-requirement is fulfilled. As mentioned above, MopA and MopB repress transcription of anfA in the presence of Mo thus preventing synthesis of Fe-nitrogenase (Kutsche et al., 1996). The anfA promoter contains a conserved DNA sequence of dyad symmetry (a so-called Mo-box) overlapping the transcription start site. Deletions within this element abolish Mo-repression of anfA, indicating that the Mo-box is the target of both MopA and MopB. In line with this assumption, the Mo-dependent regulator ModE was shown to bind the Mo-box of the E. coli modABC operon (Anderson et al., 1997). Rba. capsulatus strains defective for the ModABC transporter express Fe-nitrogenase at Mo-concentrations of up to 1 µM, unlike the parental strain (Wang et al., 1993). Additionally, in a mopA mopB double mutant anfA transcription is no longer repressed by molybdenum even at concentrations of 1 mM, suggesting that Mo-repression is completely abolished in this mutant. Like activity of NifA, AnfA activity is inhibited by ammonium (Drepper et al., 2003). However, Ncontrol of AnfA activity is not mediated by GlnB or GlnK as shown by analyses of a glnB glnK double mutant. Therefore, although both transcriptional acti-
769 vators respond to the N-status at the post-translational level, the underlying mechanisms clearly differ from each other. In a glnB glnK double mutant, N-regulation of Mo-nitrogenase is completely circumvented, resulting in the synthesis of active Mo-nitrogenase even at high concentrations of NH4+ (Drepper et al., 2003). Remarkably, the amount of Mo-nitrogenase synthesized in the glnB glnK strain in the presence of NH4+ is even higher than in the wild type under nitrogenase-derepressing conditions. In contrast, the yet unknown mechanism controlling AnfA activity prevents synthesis of Fe-nitrogenase in the glnB glnK strain in the presence of NH4+. Rba. capsulatus ranR codes for a putative TetRlike regulator, which interacts with Anf1 as shown by yeast two-hybrid studies (Sicking et al., 2005). The anf1 gene forms part of the anfHDGK-anf1-anf2-anf3 operon, and its product is essential for diazotrophic growth via Fe-nitrogenase. Like an anf1 mutant, a strain defective for ranR is no longer able to grow with N2 via Fe-nitrogenase. A target gene for the putative TetR-like regulator RanR has not yet been identified, but a role in the acquisition and/or processing of iron has been discussed (Sicking et al., 2005). C. Role of the Ammonium Transporter AmtB in Regulation of Nitrogen Fixation At high ammonium concentration in the environment (above 1 mM), passive membrane permeation of NH3 is sufficient to promote cell growth. At lower concentrations, when diffusion becomes limiting for ammonia uptake, high-affinity transport systems like AmtB are needed. In bacteria and archaea, the amtB genes are linked to glnK (Thomas et al., 2000). GlnK of E. coli and A. vinelandii reversibly bind to the membrane in an AmtB-dependent manner, with GlnK acting as a negative regulator of the transport activity of AmtB (Coutts et al., 2002; Javelle et al., 2004). Membrane binding in both species depends on the uridylylation state of GlnK, such that it is maximal in nitrogen-sufficient conditions. As described above, the Ntr system senses the intracellular nitrogen status in response to availability of 2-OG and Gln. In addition, AmtB and GlnK can also sense ammonia levels (Javelle and Merrick, 2005; Javelle et al., 2004). Because AmtB activity is required for GlnK deuridylylation, AmtB directly dictates the response setting of a key PII protein in the cell (Javelle et al., 2004). The product of the Rba. capsulatus amtB gene,
770 which forms part of the glnK-amtB operon, is required for ADP-ribosylation of nitrogenase reductase in response to NH4+ addition (Yakunin and Hallenbeck, 2002), suggesting that AmtB is involved in modulation of DraT activity. Rba. capsulatus contains a second amtB-like gene, amtY, which is not linked to a PII-encoding gene. No function has yet been assigned to AmtY, but it does not seem to substitute for AmtB concerning post-translational control of nitrogenase activity. As described above for a glnB glnK double mutant, ADP-ribosylation in response to darkness is unaltered in an amtB mutant. Rsp. rubrum contains two amtB homologs, amtB1 and amtB2, both of which are associated with glnBlike genes, glnJ and glnK, respectively (Zhang et al., 2006). An amtB1 mutant is altered in posttranslational control of nitrogenase activity not only in response to NH4+ addition, as is the case in Rba. capsulatus, but also in response to darkness. D. Linkage of N2-Fixation to Photosynthesis, CO2 Assimilation, and H2 Metabolism The homologous two-component regulatory systems Rba. capsulatus RegB-RegA and PrrB-PrrA of Rba. sphaeroides are global systems that integrate the control of photosynthesis, CO2 assimilation, H2 oxidation, and N2 fixation (Joshi and Tabita, 1996; Elsen et al., 2000). RegB is a membrane-spanning histidine kinase that autophosphorylates under anaerobic conditions in the presence of ATP (Bird et al., 1999; Chapter 35, Bauer et al.). RegB~P serves as phosphodonor towards RegA, which, upon phosphorylation, shows increased affinity to its target promoters. RegA activates genes involved in photosynthesis, CO2 assimilation, and N2 fixation, while it represses uptake hydrogenase genes and its own expression (Elsen et al., 2000). Activation of N2 fixation is indirect, as RegA binds to and activates the expression of nifA2, which encodes one of the two functional copies of the nif-specific transcriptional activator, NifA. RegA activates expression of nifA2 up to eight-fold, but it acts only as a coactivator of NtrC, which is absolutely required for nifA2 expression. Another link between photosynthesis and N2 fixation involves the hvrA gene of Rba. capsulatus, which forms part of the senC-regA-hvrA operon (Buggy et al., 1994; Du et al., 1999). HvrA exhibits clear sequence similarity to E. coli H-NS, the histone-like nucleoid structuring protein, and functional similar-
Bernd Masepohl and Robert G. Kranz ity of both proteins was shown by complementation of an E. coli hns mutant with Rba. capsulatus hvrA (Bertin et al., 1999). Growth of an Rba. capsulatus hvrA mutant strain is strongly affected under low-light conditions due to its inability to activate light-harvesting-I and reaction center gene expression in response to reduction in light intensity (Buggy et al., 1994). In addition, HvrA modulates ammonium control of nifH expression (about 20-fold in a nifA1 constitutive background) by binding to the nifH promoter (Kern et al., 1998; Raabe et al., 2002). Integration host factor (IHF) forms a further link between photosynthesis, nitrogen fixation, and H2 oxidation by binding to puf, puc, nif, and hup promoters (Hoover et al., 1990; Toussaint et al., 1991; Kirndörfer et al., 1998). As mentioned above, binding of IHF to its target promoters induces a sharp bend in the DNA, thus bringing the respective transcriptional activator and RNAP into contact. IV. Regulation in Other Purple Photosynthetic Bacteria A. Regulation of Nitrogen Fixation in Rhodobacter sphaeroides Although Rba. capsulatus and Rba. sphaeroides are closely related they differ in several aspects concerning nitrogen fixation. First, in contrast to Rba. capsulatus, Rba. sphaeroides does not have a non-Mo nitrogenase (Table 1). Second, these species differ in the number of rpoN-like genes, and third, Rba. sphaeroides appears to have no draT-draG genes. While Rba. capsulatus contains a single rpoN gene (Cullen et al., 1994; Preker et al., 1992), four rpoN-like genes were identified in Rba. sphaeroides (Poggio et al., 2002). Similar to the situation for Rba. capsulatus, one of the four copies, rpoN1, is associated with a nifU-like gene. While Rba. capsulatus rpoN belongs to the nifU2-rpoN superoperon (Cullen et al., 1994), Rba. sphaeroides rpoN1 is part of the nifUSVW–rpoN1 gene cluster (Meijer and Tabita, 1992; Poggio et al., 2002). Like Rba. capsulatus rpoN, Rba. sphaeroides rpoN1 transcription is induced under N-limiting conditions but some residual activity may always be present. Contrary phenotypes have been described for Rba. sphaeroides rpoN1 mutant strains. In one study, the mutant strain was not impaired in diazotrophic growth (Meijer and
Chapter 38
Regulation of Nitrogen Fixation
Tabita, 1992), while a rpoN1 mutant showed a severe growth defect in nitrogen-free medium in the other study (Poggio et al., 2002). In contrast to Rsp. rubrum and Rba. capsulatus, Rba. sphaeroides appears to have no draTG genes and no evidence for ADP-ribosylation of dinitrogenase reductase was found in this bacterium (Yakunin et al., 2001). However, a Rba. sphaeroides strain carrying the Rba. capsulatus draTG genes is able to reversibly ADP-ribosylate nitrogenase reductase in response to NH4+ addition or darkness suggesting that all the other regulatory elements controlling activity of the DraTDraG system are present in Rba. sphaeroides. B. Regulation of Nitrogen Fixation in Rhodopseudomonas palustris Rps. palustris carries nif, vnf, and anf genes enabling this bacterium to synthesize Mo-, V-, and Fe-nitrogenases (Oda et al., 2005). To date, Rps. palustris is the only known photosynthetic purple bacterium to possess a gene encoding a V-containing nitrogenase. As shown for Rba. capsulatus, Mo-nitrogenase is the preferred enzyme in Rps. palustris as long as the Mo-requirement is fulfilled. However, in contrast to Rba. capsulatus, Rps. palustris synthesizes its alternative nitrogenases in the presence of Mo in situations where it is unable to express a functional Mo-nitrogenase (due to a mutation). While synthesis of Fe-nitrogenase in Rba. capsulatus is prevented by repression of anfA via the Mo-dependent regulators MopA and MopB (Kutsche et al., 1996), the preference of Mo- or V-nitrogenase in the presence of Mo or V, respectively, seems to be regulated differently in Rps. palustris. C. Regulation of Nitrogen Fixation in Rhodospirillum rubrum While NtrC is essential for transcriptional activation of nifA in Rba. capsulatus and many other diazotrophic proteobacteria, NtrC is dispensable for synthesis of nitrogenase in Rsp. rubrum (Zhang et al., 1995). Instead, Rsp. rubrum synthesizes NifA in the presence of NH4+ and oxygen, conditions otherwise unfavorable for nitrogen fixation (Zhang et al., 2000). Apparently, NifA exists in either active or inactive forms, and GlnB is required for activation of NifA under N2-fixing conditions. Unlike glnD mutations in A. vinelandii and other bacteria, glnD mutations are not lethal in Rsp. rubrum, but Rsp. rubrum glnD
771 mutants are unable to fix N2 (Zhang et al., 2005). The inability of Rsp. rubrum glnD mutants to fix N2 may arise from its failure to uridylylate GlnB. Lethality of glnD mutations in A. vinelandii may be due to permanent modification (adenylylation) of glutamine synthetase (GS). In fact, a glnD mutation is no longer lethal in a glnE mutant background, preventing synthesis of the GS-modifying ATase (Colnaghi et al., 2001). Assuming that permanent modification of GS would also be lethal for Rsp. rubrum, it seems likely that GS is not completely adenylylated in a Rsp. rubrum glnD mutant. In addition to Mo-nitrogenase, Rsp. rubrum can synthesize Fe-nitrogenase (Lehman and Roberts, 1991). Similar to Rps. palustris (Oda et al., 2005), Mo does not prevent expression of Fe-nitrogenase in Rsp. rubrum. Instead, Fe-nitrogenase is expressed whenever a Rsp. rubrum strain lacks an active Monitrogenase because of physiological or genetic inactivation. V. Future Perspectives Some of the talents of photosynthetic bacteria include growth under a variety of conditions, to fix carbon dioxide and nitrogen gas, and to of course use solar energy as the source for this reduction. Besides continued studies on the molecular mechanisms of control in diverse phototrophic diazotrophs, a valid question is whether these unique capabilities might be harnessed some day for sources of renewable energy? For instance, is there any value or potential in engineering ammonia production via photosynthetic bacterial nitrogen fixation? Much has been learned already on the mechanisms for controlling nitrogen fixation in purple photosynthetic bacteria. Regulatory factors that are defective, or promoters with simple mutations can result in lack of regulatory control in response to the environment. These mutants, for example, can fix nitrogen at high levels even in the presence of plentiful ammonium (Bowman and Kranz, 1998; Drepper et al., 2003; Kranz and Haselkorn, 1986; 1988; Kranz et al., 1990). Overriding regulatory control is often a step necessary to optimize product synthesis. In the future it may become economically attractive to harness solar energy for ammonia production, and so basic science on these bacteria might lead to valuable lessons towards these goals.
772 Acknowledgments BM acknowledges financial support by Deutsche Forschungsgemeinschaft, and RGK acknowledges support from the United States Department of Agriculture. References Anderson LA, Palmer T, Price NC, Bornemann S, Boxer DH and Pau RN (1997) Characterisation of the molybdenum-responsive ModE regulatory protein and its binding to the promoter region of the modABCD (molybdenum transport) operon of Escherichia coli. Eur J Biochem 246: 119–126 Arsene F, Kaminski PA and Elmerich C (1996) Modulation of NifA activity by PII in Azospirillum brasilense: Evidence for a regulatory role of the NifA N-terminal domain. J Bacteriol 178: 4830–4838 Bertin P, Benhabiles N, Krin E, Laurent-Winter C, Tendeng C, Turlin E, Thomas A, Danchin A and Brasseur R (1999) The structural and functional organization of H-NS-like proteins is evolutionarily conserved in Gram-negative bacteria. Mol Microbiol 31: 319–329 Bird TH, Du S and Bauer CE (1999) Autophosphorylation, phosphotransfer, and DNA-binding properties of the RegB/RegA two-component regulatory system in Rhodobacter capsulatus. J Biol Chem 274: 16343–16348 Bowman WC and Kranz RG (1998) A bacterial ATP-dependent, enhancer binding protein that activates the housekeeping RNA polymerase. Genes Development 12: 1884–1893 Buggy JJ, Sganga MW and Bauer CE (1994) Characterization of a light-responding trans-activator responsible for differentially controlling reaction center and light-harvesting-I gene expression in Rhodobacter capsulatus. J Bacteriol 176: 6936–6943 Colnaghi R, Rudnick P, He L, Green A, Yan D, Larson E and Kennedy C (2001) Lethality of glnD null mutations in Azotobacter vinelandii is suppressible by prevention of glutamine synthetase adenylylation. Microbiology 147: 1267–1276 Coutts G, Thomas G, Blakey D and Merrick M (2002) Membrane sequestration of the signal transduction protein GlnK by the ammonium transporter AmtB. EMBO J 21: 536–545 Craig NL and Nash HA (1984) E. coli integration host factor binds to specific sites in DNA. Cell 39: 707–716 Cullen PJ, Foster-Hartnett D, Gabbert KK and Kranz RG (1994) Structure and expression of the alternative sigma factor, RpoN, in Rhodobacter capsulatus; physiological relevance of an autoactivated nifU2-rpoN superoperon. Mol Microbiol 11: 51–65 Cullen PJ, Bowman WC and Kranz RG (1996) In vitro reconstitution and characterization of the Rhodobacter capsulatus NtrB and NtrC two-component system. J Biol Chem 271: 6530–6536 Cullen PJ, Bowman WC, Foster-Hartnett D, Reilly SC and Kranz RG (1998) Translational activation by an NtrC enhancer-binding protein. J Mol Biol 278: 903–914 Dixon R (1998) The oxygen-responsive NIFL-NIFA complex: A novel two-component regulatory system controlling nitrogenase synthesis in γ-proteobacteria. Arch Microbiol 169: 371–380 Drepper T, Raabe K, Giaourakis D, Gendrullis M, Masepohl B and
Bernd Masepohl and Robert G. Kranz Klipp W (2002) The Hfq-like protein NrfA of the phototrophic purple bacterium Rhodobacter capsulatus controls nitrogen fixation via regulation of nifA and anfA expression. FEMS Microbiol Lett 215: 221–227 Drepper T, Groß S, Yakunin AF, Hallenbeck PC, Masepohl B and Klipp W (2003) Role of GlnB and GlnK in ammonium control of both nitrogenase systems in the phototrophic bacterium Rhodobacter capsulatus. Microbiology 149: 2203–2212 Drepper T, Wiethaus J, Giaourakis D, Groß S, Schubert B, Vogt M, Wiencek Y, McEwan AG and Masepohl B (2006) Crosstalk towards the response regulator NtrC controlling nitrogen metabolism in Rhodobacter capsulatus. FEMS Microbiol Lett 258: 250–256 Du S, Kouadio J-L and Bauer CE (1999) Regulated expression of a highly conserved regulatory gene cluster is necessary for controlling photosynthesis gene expression in response to anaerobiosis in Rhodobacter capsulatus. J Bacteriol 181: 4334–4341 Durand A and Merrick M (2006) In vitro analysis of the Escherichia coli AmtB-GlnK complex reveals a stoichiometric interaction and sensitivity to ATP and 2-oxoglutarate. J Biol Chem 281: 29558–29567 Elsen S, Dischert W, Colbeau A and Bauer CE (2000) Expression of uptake hydrogenase and molybdenum nitrogenase in Rhodobacter capsulatus is coregulated by the RegB-RegA twocomponent regulatory system. J Bacteriol 182: 2831–2837 Fischer H-M (1994) Genetic regulation of nitrogen fixation in rhizobia. Microbiol Rev 58: 352–386 Förster B, Maner K, Fassbinder F and Oelze J (1999) Reversible inactivation of nitrogenase in Rhodobacter capsulatus strain W107I deleted in the draTG gene region. FEMS Microbiol Lett 170: 167–171 Foster-Hartnett D and Kranz RG (1994) The Rhodobacter capsulatus glnB gene is regulated by NtrC at tandem rpoN-independent promoters. J Bacteriol 176: 5171–5176 Foster-Hartnett D, Cullen PJ, Monika EM and Kranz RG (1994) A new type of NtrC transcriptional activator. J Bacteriol 176: 6175–6187 Halbleib CM and Ludden PW (2000) Regulation of biological nitrogen fixation. J Nutr 130: 1081–1084 Hall DR, Gourley DG, Leonard GA, Duke EMH, Anderson LA, Boxer DH and Hunter WN (1999) The high-resolution crystal structure of the molybdate-dependent transcriptional regulator (ModE) from Escherichia coli: A novel combination of domain folds. EMBO J 18: 1435–1446 Hoover TR (2000) Control of nitrogen fixation genes in Klebsiella pneumoniae. In: Triplett EW (ed) Prokaryotic Nitrogen Fixation: A Model System for Analysis of a Biological Process, pp 131–147. Horizon Scientific Press, Wymondham Hoover TR, Santero E, Porter S and Kustu S (1990) The integration host factor stimulates interaction of RNA polymerase with NIFA, the transcriptional activator for nitrogen fixation operons. Cell 63: 11–22 Javelle A and Merrick M (2005) Complex formation between AmtB and GlnK: An ancestral role in prokaryotic nitrogen control. Biochem Soc Trans 33: 170–172 Javelle A, Severi E, Thornton J and Merrick M (2004) Ammonium sensing in E. coli: The role of the ammonium transporter AmtB and AmtB-GlnK complex formation. J Biol Chem 279: 8530–8538 Jiang P, Peliska JA and Ninfa AJ (1998a) Enzymological
Chapter 38
Regulation of Nitrogen Fixation
characterization of the signal-transducing uridylyltransferase/uridylyl-removing enzyme (EC 2.7.7.59) of Escherichia coli and its interaction with the PII protein. Biochemistry 37: 12782–12794 Jiang P, Peliska JA and Ninfa AJ (1998b) Reconstitution of the signal-transduction bicyclic cascade responsible for the regulation of Ntr gene transcription in Escherichia coli. Biochemistry 37: 12795–12801 Jiang P, Peliska JA and Ninfa AJ (1998c) The regulation of Escherichia coli glutamine synthetase revisited: Role of 2-ketoglutarate in the regulation of glutamine synthetase adenylylation state. Biochemistry 37: 12802–12810 Johansson M and Nordlund S (1997) Uridylylation of the PII protein in the photosynthetic bacterium Rhodospirillum rubrum. J Bacteriol 179: 4190–4194 Johansson M and Nordlund S (1999) Purification of PII and PIIUMP and in vitro studies of regulation of glutamine synthetase in Rhodospirillum rubrum. J Bacteriol 181: 6524–6529 Joshi HM and Tabita FR (1996) A global two component signal transduction system that integrates the control of photosynthesis, carbon dioxide assimilation, and nitrogen fixation. Proc Natl Acad Sci USA 93: 14515–14520 Kaminski A and Elmerich C (1998) The control of Azorhizobium caulinodans nifA expression by oxygen, ammonia and by the HF-I-like protein, NrfA. Mol Microbiol 28: 603–613 Kaminski PA, Desnoues N and Elmerich C (1994) The expression of nifA in Azorhizobium caulinodans requires a gene product homologous to Escherichia coli HF-I, an RNA-binding protein involved in the replication of phage Qβ RNA. Proc Natl Acad Sci USA 91: 4663–4667 Kern M, Kamp P-B, Paschen A, Masepohl B and Klipp W (1998) Evidence for a regulatory link of nitrogen fixation and photosynthesis in Rhodobacter capsulatus via HvrA. J Bacteriol 180: 1965–1969 Keuntje B, Masepohl B and Klipp W (1995) Expression of the putA gene encoding proline dehydrogenase from Rhodobacter capsulatus is independent of NtrC regulation but requires an Lrp-like activator protein. J Bacteriol 177: 6432–6439 Kirndörfer M, Jäger A and Klug G (1998) Integration host factor affects the oxygen-regulated expression of photosynthesis genes in Rhodobacter capsulatus. Mol Gen Genet 258: 297–305 Klipp W, Masepohl B and Pühler A (1988) Identification and mapping of nitrogen fixation genes of Rhodobacter capsulatus: Duplication of a nifA-nifB region. J Bacteriol 170: 693–699 Kranz RG (2000) Regulation of nitrogen fixation genes in phototrophs: new mechanisms of bacterial gene activation. In: Triplett EW (ed) Prokaryotic Nitrogen Fixation: A Model System for Analysis of a Biological Process, pp 165–175. Horizon Scientific Press, Wymondham Kranz RG and Foster-Hartnett D (1990) Transcriptional regulatory cascade of nitrogen-fixation genes in anoxygenic photosynthetic bacteria: Oxygen- and nitrogen-responsive factors. Mol Microbiol 4: 1793–1800 Kranz RG and Haselkorn R (1985) Characterization of nif regulatory genes in Rhodopseudomonas capsulata using lac fusions. Gene 40: 203–215 Kranz RG and Haselkorn R (1986) Anaerobic regulation of nitrogen-fixation genes in Rhodopseudomonas capsulata. Proc Natl Acad Sci USA 83: 6805–6809 Kranz RG and Haselkorn R (1988) Ammonia-constitutive nitrogen fixation mutants of Rhodobacter capsulatus. Gene 71: 65–74
773 Kranz RG, Pace VM and Caldicott IA (1990) Inactivation, sequence, and lacZ fusion analysis of a regulatory locus required for repression of nitrogen fixation genes in Rhodobacter capsulatus. J Bacteriol 172: 53–62 Kutsche M, Leimkühler S, Angermüller S and Klipp W (1996) Promoters controlling expression of the alternative nitrogenase and the molybdenum uptake system in Rhodobacter capsulatus are activated by NtrC, independent of σ54, and repressed by molybdenum. J Bacteriol 178: 2010–2017 Lehman LJ and Roberts GP (1991) Identification of an alternative nitrogenase system in Rhodospirillum rubrum. J Bacteriol 173: 5705–5711 Madigan MT (1995) Microbiology of nitrogen fixation by anoxygenic photosynthetic bacteria. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 915–928. Kluwer Academic Publishers, Dordrecht Martinez-Argudo I, Little R, Shearer N, Johnson P and Dixon R (2004) The NifL-NifA system: A multidomain transcriptional regulatory complex that integrates environmental signals. J Bacteriol 186: 601–610 Masepohl B, Klipp W and Pühler A (1988) Genetic characterization and sequence analysis of the duplicated nifA/nifB gene region of Rhodobacter capsulatus. Mol Gen Genet 212: 27–37 Masepohl B, Angermüller S, Hennecke S, Hübner P, Moreno-Vivian C and Klipp W (1993a) Nucleotide sequence and genetic analysis of the Rhodobacter capsulatus ORF6-nifUISVW gene region: Possible role of NifW in homocitrate processing. Mol Gen Genet 238: 369–382 Masepohl B, Krey R and Klipp W (1993b) The draTG gene region of Rhodobacter capsulatus is required for post-translational regulation of both the molybdenum and the alternative nitrogenase. J Gen Microbiol 139: 2667–2675 Masepohl B, Kaiser B, Isakovic N, Richard CL, Kranz RG and Klipp W (2001) Urea utilization in the phototrophic bacterium Rhodobacter capsulatus is regulated by the transcriptional activator NtrC. J Bacteriol 183: 637—643 Masepohl B, Drepper T, Paschen A, Groß S, Pawlowski A, Raabe K, Riedel K-U and Klipp W (2002a) Regulation of nitrogen fixation in the phototrophic purple bacterium Rhodobacter capsulatus. J Mol Microbiol Biotechnol 4: 243–248 Masepohl B, Schneider K, Drepper T, Müller A and Klipp W (2002b) Alternative nitrogenases. In: Leigh GJ (ed) Nitrogen Fixation at the Millennium, pp 191–222. Elsevier, Amsterdam Masepohl B, Drepper T and Klipp W (2004) Nitrogen fixation in the photosynthetic purple bacterium Rhodobacter capsulatus. In: Klipp W, Masepohl B, Gallon JR and Newton WE (eds) Genetics and Regulation of Nitrogen Fixation in FreeLiving Bacteria, pp 141–173. Kluwer Academic Publishers, Dordrecht Meijer WG and Tabita FR (1992) Isolation and characterization of the nifUSVW-rpoN gene cluster from Rhodobacter sphaeroides. J Bacteriol 174: 3855–3866 Moreno-Vivian C, Schmehl M, Masepohl B, Arnold W and Klipp W (1989) DNA sequence and genetic analysis of the Rhodobacter capsulatus nifENX gene region: homology between NifX and NifB suggests involvement of NifX in processing of the iron-molybdenum cofactor. Mol Gen Genet 216: 353–363 Moreno-Vivian C, Soler G and Castillo F (1992) Arginine catabolism in the phototrophic bacterium Rhodobacter capsulatus
774 E1F1. Purification and properties of arginase. Eur J Biochem 204: 531–537 Morett E and Buck M (1988) NifA-dependent in vivo protection demonstrates that the upstream activator sequence of nif promoters is a protein binding site. Proc Natl Acad Sci USA 85: 9401–9405 Morett E and Buck M (1989) In vivo studies on the interaction of RNA polymerase-σ54 with the Klebsiella pneumoniae and Rhizobium meliloti nifH promoters. The role of NifA in the formation of an open promoter complex. J Mol Biol 210: 65–77 Nordlund S (2000) Regulation of nitrogenase activity in phototrophic bacteria by reversible covalent modification. In: Triplett EW (ed) Prokaryotic Nitrogen Fixation: A Model System for Analysis of a Biological Process, pp 149–164. Horizon Scientific Press, Wymondham Nordlund S and Ludden PW (2004) Post-translational regulation of nitrogenase in photosynthetic bacteria. In: Klipp W, Masepohl B, Gallon JR and Newton WE (eds) Genetics and Regulation of Nitrogen Fixation in Free-Living Bacteria, pp 175–196. Kluwer Academic Publishers, Dordrecht Oda Y, Samanta SK, Rey FE, Wu L, Liu X, Yan T, Zhou J and Harwood CS (2005) Functional genomic analysis of three nitrogenase isozymes in the photosynthetic bacterium Rhodopseudomonas palustris. J Bacteriol 187: 7784–7794 Pau RN (2004) Molybdenum uptake and homeostasis. In: Klipp W, Masepohl B, Gallon JR and Newton WE (eds) Genetics and Regulation of Nitrogen Fixation in Free-Living Bacteria, pp 225–256. Kluwer Academic Publishers, Dordrecht Paschen A, Drepper T, Masepohl B and Klipp W (2001) Rhodobacter capsulatus nifA mutants mediating nif gene expression in the presence of ammonium. FEMS Microbiol Lett 200: 207–213 Pawlowski A, Riedel K-U, Klipp W, Dreiskemper P, Groß S, Bierhoff H, Drepper T and Masepohl B (2003) Yeast two-hybrid studies on interaction of proteins involved in regulation of nitrogen fixation in the phototrophic bacterium Rhodobacter capsulatus. J Bacteriol 185: 5240–5247 Pawlowski K, Klosse U and De Bruijn FJ (1991) Characterization of a novel Azorhizobium caulinodans ORS571 two-component regulatory system, NtrY/NtrX, involved in nitrogen fixation and metabolism. Mol Gen Genet 231: 124–38 Pierrard J, Ludden PW and Roberts GP (1993) Posttranslational regulation of nitrogenase in Rhodobacter capsulatus: Existence of two independent regulatory effects of ammonium. J Bacteriol 175: 1358–1366 Poggio S, Osorio A, Dreyfus G and Camarena L (2002) The four different σ54 factors of Rhodobacter sphaeroides are not functionally interchangeable. Mol Microbiol 46: 75–85 Preker P, Hübner P, Schmehl M, Klipp W and Bickle TA (1992) Mapping and characterization of the promoter elements of the regulatory nif genes rpoN, nifA1 and nifA2 in Rhodobacter capsulatus. Mol Microbiol 6: 1035–1047 Raabe K, Drepper T, Riedel K-U, Masepohl B and Klipp W (2002) The H-NS-like protein HvrA modulates expression of nitrogen fixation genes in the phototrophic purple bacterium Rhodobacter capsulatus by binding to selected nif promoters. FEMS Microbiol Lett 216: 151–158 Schmehl M, Jahn A, Meyer zu Vilsendorf A, Hennecke S, Masepohl B, Schuppler M, Marxer M, Oelze J and Klipp W (1993) Identification of a new class of nitrogen fixation genes in Rhodobacter capsulatus: A putative membrane complex
Bernd Masepohl and Robert G. Kranz involved in electron transport to nitrogenase. Mol Gen Genet 241: 602–615 Schneider K, Müller A, Schramm U and Klipp W (1991) Demonstration of a molybdenum- and vanadium-independent nitrogenase in a nifHDK-deletion mutant of Rhodobacter capsulatus. Eur J Biochem 195: 653–661 Schüddekopf K, Hennecke S, Liese U, Kutsche M and Klipp W (1993) Characterization of anf genes specific for the alternative nitrogenase and identification of nif genes required for both nitrogenases in Rhodobacter capsulatus. Mol Microbiol 8: 673–684 Self WT, Grunden AM, Hasona A, Shanmugam KT (2001) Molybdate transport. Res Microbiol 152: 311–321 Sicking C, Brusch M, Lindackers A, Riedel K-U, Schubert B, Isakovic N, Krall C, Klipp W, Drepper T, Schneider K and Masepohl B (2005) Identification of two new genes involved in diazotrophic growth via the alternative Fe-only nitrogenase in the phototrophic purple bacterium Rhodobacter capsulatus. J Bacteriol 187: 92–98 Thomas G, Coutts G and Merrick M (2000) The glnKamtB operon. A conserved gene pair in prokaryotes. Trends Genet 16: 11–14 Toussaint B, Bosc C, Richaud P, Colbeau A and Vignais PM (1991) A mutation in a Rhodobacter capsulatus gene encoding an integration host factor-like protein impairs in vivo hydrogenase expression. Proc Natl Acad Sci USA 88: 10749–10753 Toussaint B, Delic-Attree I, De Sury d´Aspremont R, David L, Vincon M and Vignais PM (1993) Purification of the integration host factor homolog of Rhodobacter capsulatus: Cloning and sequencing of the hip gene, which encodes the β subunit. J Bacteriol 175: 6499–6504 Wang G, Angermüller S and Klipp W (1993) Characterization of Rhodobacter capsulatus genes encoding a molybdenum transport system and putative molybdenum-pterin-binding proteins. J Bacteriol 175: 3031–3042 Willison JC, Pierrard J and Hübner P (1993) Sequence and transcript analysis of the nitrogenase structural gene operon (nifHDK) of Rhodobacter capsulatus: Evidence for intramolecular processing of nifHDK mRNA. Gene 133: 39–46 Yakunin AF and Hallenbeck PC (1998) Short-term nitrogenase regulation in Rhodobacter capsulatus: Multiple in vivo nitrogenase responses to NH4+ addition. J Bacteriol 180: 6392–6395 Yakunin AF and Hallenbeck PC (2002) AmtB is necessary for NH4+-induced nitrogenase switch-off and ADP-ribosylation in Rhodobacter capsulatus. J Bacteriol 184: 4081–4088 Yakunin AF, Laurinavichene TV, Tsygankov AA and Hallenbeck PC (1999) The presence of ADP-ribosylated Fe protein of nitrogenase in Rhodobacter capsulatus is correlated with cellular nitrogen status. J Bacteriol 181: 1994–2000 Yakunin AF, Fedorov AS, Laurinavichene TV, Glaser VM, Egorov NS, Tsygankov AA, Zinchenko VV and Hallenbeck PC (2001) Regulation of nitrogenase in the photosynthetic bacterium Rhodobacter sphaeroides containing draTG and nifHDK genes from Rhodobacter capsulatus. Can J Microbiol 47: 206–212 Zhang Y, Cummings AD, Burris RH, Ludden PW and Roberts GP (1995) Effect of an ntrBC mutation on the posttranslational regulation of nitrogenase activity in Rhodospirillum rubrum. J Bacteriol 177: 5322–5326 Zhang Y, Burris RH, Ludden PW and Roberts GP (1997) Regulation of nitrogen fixation in Azospirillum brasilense. FEMS Microbiol Lett 152: 195–204
Chapter 38
Regulation of Nitrogen Fixation
Zhang Y, Pohlmann EL, Ludden PW and Roberts GP (2000) Mutagenesis and functional characterization of the glnB, glnA, and nifA genes from the photosynthetic bacterium Rhodospirillum rubrum. J Bacteriol 182: 983–92 Zhang Y, Pohlmann EL, Ludden PW and Roberts GP (2001) Functional characterization of three GlnB homologs in the photosynthetic bacterium Rhodospirillum rubrum: Roles in sensing ammonium and energy status. J Bacteriol 183: 6159–6168 Zhang Y, Pohlmann EL and Roberts GP (2005) GlnD is essential for NifA activation, NtrB/NtrC-regulated gene expression, and posttranslational regulation of nitrogenase activity in the
775 photosynthetic, nitrogen-fixing bacterium Rhodospirillum rubrum. J Bacteriol 187: 1254–1265 Zhang Y, Wolfe DM, Pohlmann EL, Conrad MC and Roberts GP (2006) Effect of AmtB homologues on the post-translational regulation of nitrogenase activity in response to ammonium and energy signals in Rhodospirillum rubrum. Microbiology 152: 2075–2089 Zhu Y, Conrad MC, Zhang Y and Roberts GP (2006) Identification of Rhodospirillum rubrum GlnB variants that are altered in their ability to interact with different targets in response to nitrogen status signals. J Bacteriol 188: 1866–1874
Chapter 39 Regulation of the Tetrapyrrole Biosynthetic Pathway Jill Helen Zeilstra-Ryalls* Department of Biological Sciences, 217 Life Sciences Building, Bowling Green State University, Bowling Green, Ohio 43403, U.S.A.
Summary ............................................................................................................................................................... 777 I. Introduction..................................................................................................................................................... 778 II. Tetrapyrrole Biosynthesis Genes ................................................................................................................... 778 III. Comparing and Contrasting Oxygen Control of Tetrapyrrole Biosynthesis Genes in Species of Rhodobacter ................................................................................................................................................. 779 A. Regulation of ALA Synthase Genes ................................................................................................. 780 B. Regulation of the hemB Gene .......................................................................................................... 783 C. Regulation of the hemC and hemE Genes....................................................................................... 784 D. Regulation of Genes Coding for the Synthesis of Protoporphyrinogen IX from Coproporphyrinogen III .................................................................................................................... 785 E. Branchpoint Considerations ............................................................................................................. 787 F. Oxygen Control of Tetrapyrrole Biosynthesis in Rhodobacter As We Know It ................................ 787 IV. Other Aspects of Transcriptional Regulation of Tetrapyrrole Biosynthesis Genes in Rhodobacter Species........................................................................................................................................................... 789 V. A Genomics Perspective on the Regulation of Tetrapyrrole Biosynthesis in Other Purple Anoxygenic Photosynthetic Bacteria.................................................................................................................................. 789 A. Distribution of the Regulators ........................................................................................................... 789 B. Distribution of Tetrapyrrole Biosynthesis Genes .............................................................................. 791 C. Motifs................................................................................................................................................ 795 Note Added in Proof .............................................................................................................................................. 795 Acknowledgments ................................................................................................................................................. 795 References ............................................................................................................................................................ 795
Summary This chapter describes what is known, or can be deduced from genomic analyses, about transcriptional regulation of the tetrapyrrole biosynthesis genes in purple anoxygenic photosynthetic bacteria. The major emphasis is on describing regulatory events that occur when oxygen levels change, as it is a key parameter controlling the expression of these genes. An overview of the common biosynthetic pathway is followed by a comparative description of the known regulatory features about each gene in Rhodobacter (Rba.) sphaeroides and Rba. capsulatus, the two purple anoxygenic photosynthetic bacteria in which this topic has been investigated most extensively. Finally, the information available for the Rhodobacter species is used to explore the DNA sequence databases for other bacteria belonging to this group. First, the results of a survey of the DNA sequences for genes encoding the key regulatory proteins are presented. Second, the tetrapyrrole biosynthesis genes in the sequenced strains of the purple anoxygenic photosynthetic bacteria are identified, and the outcome of an inspection of their upstream sequences for the presence of any known target sequences representing DNA binding *Email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 777–798. © 2009 Springer Science + Business Media B.V.
778
Jill Helen Zeilstra-Ryalls
sites for the regulatory proteins are presented. While the significance of the regulatory sequences will require experimental determination, a number of conclusions can be made as to the distribution of the regulatory proteins and the tetrapyrrole biosynthesis genes among these bacteria. I. Introduction Tetrapyrroles are the cell’s ‘color’ pallet of biologically packaged metal ions that perform indispensable functions in both anabolic and catabolic metabolisms, and the kinds of tetrapyrroles that cells synthesize or acquire delimits their metabolic capabilities (for reviews see Lascelles, 1956; Jordan, 1991; Scheer, 1991). In the case of the organisms of interest here, the presence of two species of tetrapyrrole, heme and bacteriochlorophyll (BChl), make possible their most distinctive feature; anoxygenic photosynthesis. An introduction to the tetrapyrrole biosynthesis genes will be followed by a comparative description of what is known about how oxygen controls their expression in two species of Rhodobacter in which this topic has been extensively explored. In keeping with the genomics approach to this topic, a best effort has also been made to take advantage of the DNA sequence information available for other species of purple anoxygenic photosynthetic bacteria, and to consider what the known features of tetrapyrrole biosynthesis gene regulation in Rhodobacter suggest about regulation of these genes in the other members of this group. II. Tetrapyrrole Biosynthesis Genes The tetrapyrrole biosynthesis pathway is shown in Fig. 1 (for more details see, for example, Beale and Weinstein, 1991; Jordan, 1991; Blankenship et al., 1995; Chapter 4, Willows and Kriegel). All tetrapyrroles are derived from 5-aminolevulinic acid (ALA). However, either ALA is formed through the condensation of glycine and succinyl-CoA in a pyridoxal phosphatedependent reaction catalyzed by ALA synthase (the C4 or Shemin pathway), or tRNAglutamate is the starting molecule from which ALA is formed in two steps sequentially catalyzed by the two pyridoxal phosAbbreviations: Acp. – Acidiphilium; ALA – 5-aminolevulinic acid; BChl – bacteriochlorophyll; E. – Escherichia; Erb. – Erythrobacter; Hlr. – Halorhodospira; Mtb. – Methylobacterium; Rba. – Rhodobacter; Rps. – Rhodopseudomonas; Rsb. – Roseobacter; Rsp. – Rhodospirillum; Rva. – Roseovarius; Rvi. – Rubrivivax; Rvu. – Rhodovulum
phate-dependent enzymes glutamyl-tRNA reductase and glutamate-1-semialdehyde aminotransferase (the C5 pathway). Until recently, only one or the other of these two reactions was thought to exist in any prokaryote. First, an isotope labeling study by Iida and Kajiwara suggests that ALA can be derived via both C4 and C5 pathways in Propionibacterium shermanii (as described in Iida and Kajiwara, 2000). While no ALA synthase gene is present in Propionibacterium acnes KPA171202, sequencing of Propionibacterium freudenreichii subsp. shermanii CIP103027 is in progress, and should directly address those observations. Second, the biosynthetic genes for both pathways have been described in Streptomyces nodosus subsp. asukaensis and related species (Petrícek et al., 2006). Inactivating the glutamyl-tRNA reductase gene, gtr, gives rise to ALA auxotrophy, even though the ALA synthase enzyme is fully functional (when expressed in Escherichia (E.) coli),which begs the intriguing question as to how ALA formed by the ALA synthase is unavailable for tetrapyrrole production. With respect to the bacteria of interest here, inspection of the DNA sequences available, including several complete genomes, indicates that only one pathway is represented in any given species. The ability to carry out photosynthesis requires the production of both heme and BChl, and the last shared precursor of these two tetrapyrroles is protoporphyrin IX. Thus, the six reactions, beginning with the cyclization of two ALA molecules to form pyrrole to the production of protoporphyrin IX, are often described as the common steps in the tetrapyrrole biosynthesis pathway. Although in those species that synthesize corrinoids uroporphyrinogen III is the last common precursor, the regulation of genes required for the production of protoporphyrin IX, beginning with ALA formation, will be the focus of discussion here. Before examining regulation in more detail, it is important to point out that, despite the availability of complete genomic DNA sequences for hundreds of organisms, there are two steps in the pathway for which no gene or gene product has been described in many prokaryotes (O’Brian and Thöny-Meyer, 2002). Our more recent survey, applying the same parameters of using only bona fide gene or protein sequences to
Chapter 39
Regulation of Tetrapyrrole Biosynthesis
779
Fig. 1. The common tetrapyrrole biosynthesis pathway. For each step, gene (in bold italics) and enzyme names are indicated. Note that, for the C5 pathway, gene names are as used in E. coli.
the bacteria of interest here, is consistent with their descriptions, with one exception (discussed in Section V.B). Obviously, without knowledge of the genes, there is nothing that can be said as to their regulation. Therefore, while there is much that is known about the regulation of several steps, it is not possible at this point to be able to present any overall model of regulation of tetrapyrrole biosynthesis genes for any of the bacteria of interest here. As will become evident in later sections, there is considerable variability with respect to the organization of the tetrapyrrole biosynthesis genes and the regulatory events associated with them, even within the narrow scope of different strains of the same species. Thus, while there are many examples in the literature where findings for one strain are assumed to be equally true for other strains, it is important to note that such extrapolations are not necessarily valid.
To provide the most accurate information possible, every effort to avoid such assumptions has been made here. Therefore, strain names are included throughout this chapter; if the strain is not known that will also be indicated. III. Comparing and Contrasting Oxygen Control of Tetrapyrrole Biosynthesis Genes in Species of Rhodobacter Among the purple anoxygenic photosynthetic bacteria there are those that have the ability to switch between respiration and photosynthesis. As might be predicted, in several of those bacteria it has been demonstrated that transcription of photosynthesis genes is responsive to oxygen. Transcription of the tetrapyrrole biosynthesis genes is also responsive,
780 since the need for heme and BChl is dictated to a significant degree by what form of energy metabolism is deployed by the cell. In species of Rhodobacter, the change in tetrapyrrole synthesis as oxygen tensions fall is nothing less than spectacular. Consider, for example, that in the presence of high oxygen tensions, while heme biosynthesis is necessary in order to form respiratory cytochromes, the cells have no need for, nor do they produce, BChl. But when oxygen tensions fall BChl levels are estimated to increase more than 100-fold (Lascelles, 1964), while at the same time heme production also increases (described in Schilke and Donohue, 1992), as both are required for photosynthesis. Thus, the regulatory program present in these cells that switches the right sets of genes on and off in order to support one or the other of these energy production options is both elegant and sophisticated. Two species of Rhodobacter, Rba. sphaeroides and Rba. capsulatus, will be compared and contrasted in terms of what is known at the transcription level about the response of known tetrapyrrole biosynthesis genes to oxygen. The scope and size of differences between the genomic sequences of Rba. sphaeroides 2.4.1 and Rba. capsulatus SB1003 was far greater than would be anticipated from their relatedness assigned by rRNA analysis (Haselkorn et al., 2001; Mackenzie et al., 2001), which will be exemplified in the following analysis. The regulatory genes that are known to be involved in mediating the oxygen-responsiveness of the tetrapyrrole biosynthesis genes will be introduced at such time as they are first associated with a biosynthetic gene of the pathway. Most of the descriptions to follow have emanated from studies of responses to lowering oxygen tensions. However, certainly it is true that the cells are also confronted with the reverse challenge of responding to increasing oxygen tensions. At the post-transcriptional level, this situation has already been the focus of at least one investigation (Willows et al., 2003), and will undoubtedly constitute a future direction for studies at the transcription level. Factors other than oxygen that affect the transcription of the tetrapyrrole biosynthesis genes will be described separately. A. Regulation of ALA Synthase Genes Rhodobacter species, as do all members of the αProteobacteria, form ALA by the Shemin pathway. Rba. sphaeroides 2.4.1, whose genome has been completely sequenced, has two ALA synthase genes,
Jill Helen Zeilstra-Ryalls hemA and hemT, coding for proteins whose amino acid sequences are approximately 53% identical. Using hemA and hemT DNA as probes, Nereng and Kaplan demonstrated that, while many do, some wild type strains of Rba. sphaeroides do not have two ALA synthase genes (Nereng and Kaplan, 1999). The DNA sequence data available for strain ATCC17025, (draft form) supports that finding in that only hemA is present. Since in strain 2.4.1 the two ALA synthase genes are located on two different chromosomes, it is interesting to note that apparently the presence of two chromosomes does not always correlate with the presence of two ALA synthase genes, since strain ATCC17025, notwithstanding the absence of a hemT gene, has two chromosomes (Nereng and Kaplan, 1999). In all three wild type strains of Rba. sphaeroides sequenced (2.4.1, ATCC17025, and ATCC17029) the hemA gene neighborhood is similar, and for those strains having hemT (2.4.1 and ATCC17029) that gene neighborhood is similar as well. Both genes are monocistronic (Neidle and Kaplan, 1993b; Fales et al., 2002); however, the genomic context of the hemA genes is completely different from the hemT genes. While it was first thought that disabling either hemA or hemT in Rba. sphaeroides 2.4.1 had no effect on the cell (Neidle and Kaplan, 1993a), it was subsequently determined that disabling hemA alone already confers ALA auxotropy under aerobic, anaerobic-dark (with dimethyl sulfoxide as alternate electron acceptor), and photosynthetic conditions (Zeilstra-Ryalls and Kaplan, 1995b). This was not evident from the first study, because emergence of ALA prototrophy among HemA– mutant strains arises readily through additional mutations that increase transcription of the hemT gene. These can be both cis and trans relative to the hemT gene (Zeilstra-Ryalls and Kaplan, 1995b). Thus, in wild type 2.4.1 cells, hemT is transcriptionally off under aerobic, photosynthetic, and anaerobicdark conditions, and since HemT is fully functional under all those conditions, it is thought that hemT is normally expressed under conditions that remain to be defined (Zeilstra-Ryalls and Kaplan, 1995b). Horne et al. (1996) showed that in Rba. sphaeroides RS630, ALA synthase activity is lower under photosynthetic conditions in the presence of N2O than in its absence. Those workers also compared the effect of nitrate on hemA and hemT transcription using lacZ fusions involving the hemA and hemT upstream sequences from strain 2.4.1, expressed in strain RS630. The levels of β-galactosidase activity indicate that hemA
Chapter 39
Regulation of Tetrapyrrole Biosynthesis
transcription is lower in the presence of nitrate while hemT transcription is higher. In view of the fact that both cis and trans mutations in wild type strain 2.4.1 can activate transcription of hemT the specificity of this effect in Rba. sphaeroides strain RS630 should be determined, given the heterologous nature of the reporters used in Horne et al. (1996), combined with the unknown status of hemT transcription in wild type strain RS630. While the total variation in hemA transcript levels in response to changing oxygen tensions in Rba. sphaeroides 2.4.1 is approximately 3-fold as measured by microarray analyses (Pappas et al., 2004; Roh et al., 2004), that measurement represents the summation of regulated transcription involving at least two DNA binding proteins (Zeilstra-Ryalls and Kaplan, 1995a; Ranson-Olson et al., 2006) and emanating from two promoters (Neidle and Kaplan, 1993b; Fales et al., 2002). Studies that have successfully unmasked the total possible range in hemA transcription that can be achieved in the cell reveal that it can vary as much as 59-fold (Ranson-Olson et al., 2006), and so it would seem that tight regulation of hemA expression is of critical importance to the cell. The first indication of regulation of the gene was suggested at the time the gene was sequenced, as a perfect FNR consensus sequence, TTGAT-N4-ATCAA, was discerned in the upstream sequences of the hemA gene (Neidle and Kaplan, 1993b). The prediction was that Rba. sphaeroides would have an Fnr-type protein (for a review of E. coli Fnr, see Kiley and Beinert, 1998) that could alter the level of hemA transcription in response to changes in oxygen tensions (Neidle and Kaplan, 1993b). Efforts directed towards finding a gene coding for an Fnr-type protein involved the application of transposon mutagenesis, selecting for mutants with increased transcription from hemA upstream sequences under aerobic conditions. In retrospect, that the gene coding for the regulatory protein was not identified in that study makes sense, since it is now known that the Rba. sphaeroides 2.4.1 homolog of Fnr, FnrL, activates hemA in response to lowering oxygen tensions (Zeilstra-Ryalls and Kaplan, 1995a). However, the transposon mutagenesis approach did identify several mutant strains in which not only hemA transcription is higher than in wild type cells under aerobic conditions, but also photosynthetic membranes are detected in cells grown under highly aerobic conditions (Zeilstra-Ryalls and Kaplan, 1996; Fales et al., 2001). One of the transposon insertions
781
is within the ccoNOQP operon, which codes for a cbb3-type cytochrome c oxidase (Toledo-Cuevas et al., 1998). The mechanism by which it affects hemA expression will be discussed in more detail below. Since the genes coding for the cbb3 cytochrome c oxidase are necessary for symbiotic nitrogen fixation in species of Bradyrhizobium japonicum and Sinorhizobium meliloti, they are designated as ‘fix’ genes in those bacteria, and the genes are within a larger fix region that also contains the fixK gene, coding for an Fnr-type transcription factor that regulates their expression (reviewed in Fischer, 1994). That gene arrangement prompted additional sequence analysis of the cco locus in Rba. sphaeroides 2.4.1, which revealed the presence of the fnrL gene, coding for an Fnr-type protein, in that organism as well (Zeilstra-Ryalls and Kaplan, 1995a). This protein has the hallmark features of Fnr-type regulatory proteins, including cysteines within the N-proximal region of the protein corresponding to an effector domain that could coordinate an oxygen-labile 4Fe-4S cluster and a C-proximal DNA binding domain that is predicted to bind to FNR consensus-like sequences within Rba. sphaeroides (reviewed in Zeilstra-Ryalls and Kaplan, 2004). The importance of fnrL in Rba. sphaeroides 2.4.1 extends far beyond regulation of the hemA gene, as disruption of fnrL results in the loss of all anaerobic growth, either in the light or in the dark with dimethyl sulfoxide as alternate electron acceptor (Zeilstra-Ryalls and Kaplan, 1995a). By both in vitro and in vivo analyses, transcription of hemA has been shown to emanate from two promoters (Neidle and Kaplan, 1993b; Fales et al., 2002). Both promoters are under oxygen control, and anaerobic induction of transcription from the upstream and the downstream promoters requires an intact fnrL gene (Fales et al., 2002). Although suitably located for an activation role in transcription from the downstream promoter, the position of the presumed target for FnrL binding, the FNR consensus-like sequence (which overlaps the start site for transcription from the upstream promoter), is inconsistent with an activator role in transcription from the upstream promoter. To satisfy the fact that fnrL is nevertheless required for anaerobic induction from that upstream promoter, an indirect role has been proposed (Fales et al., 2002). The mechanism is currently unknown. The consequences for hemA expression of a transposon insertion within the cco operon, i.e., higher aerobic expression than in wild type 2.4.1, suggested
782 that, in addition to oxygen availability, the redox state of the cell affects hemA expression (Zeilstra-Ryalls and Kaplan, 1996). That the transposon insertion mutation not only alters hemA expression but also the expression of other photosynthesis genes implicated the two-component regulatory protein PrrA (Photosynthesis response regulator; described in Chapter 35, Bauer et al.) in this process. The proposed stimulus leading to phosphorylation of PrrA by its partner membrane-localized sensor protein, PrrB, is the reduction in cbb3 oxidase activity. Both genetic and biochemical studies have firmly established a sensing-signaling relationship between cbb3 oxidase and the Prr two-component system (see Oh, 2006 and references therein, and Oh et al., 2004). With respect to hemA regulation, a difficulty in the cbb3 oxidasePrr model existed in that previous mRNA studies had indicated hemA transcription is not regulated by PrrA (Eraso and Kaplan, 1994). Re-evaluation of the role of PrrA in hemA transcription showed that the upstream promoter requires PrrA, and that phosphorylation of PrrA increases the level of transcription (Ranson-Olson et al., 2006). Those studies also showed that the hemA upstream sequences contain two PrrA binding sites, and that the relative affinity of PrrA towards the two sites changes according to oxygen tension (Ranson-Olson et al., 2006). Although it was already known that other genes have multiple PrrA binding sites (Dubbs and Tabita, 2003), differential binding affinity among the sites had not been previously described and it suggests the existence of additional complexities with respect to PrrA action that remain to be determined. Also, unlike the situation for FnrL (Fales et al., 2002), PrrA regulation of hemA seems to be specifically directed towards the upstream promoter (Ranson-Olson et al., 2006). Within the PrrA binding sites of the hemA gene there are 2 bp differences between Rba. sphaeroides strains 2.4.1 and ATCC17029, but 19 bp differences between strains 2.4.1 and ATCC17025; six of those are at residues that have been shown to be important for PrrA binding. This suggests that the role of PrrA in regulation of hemA transcription may not be the same among all three strains. Whether or not this difference might in some way compensate for the absence of a hemT gene in strain ATCC17025 remains to be tested, but it draws attention again to the need to use caution in assuming that findings for one strain are equally true for other strains. In Rba. capsulatus SB1003, only one ALA synthase
Jill Helen Zeilstra-Ryalls gene is present, based on both the DNA sequence information available (draft form) and on the fact that two HemA– mutant strains of Rba. capsulatus absolutely require an intact hemA gene for ALA prototrophy (Wright et al., 1987; Hornberger et al., 1990). The amino acid sequence of this gene product is approximately 74% identical to Rba. sphaeroides 2.4.1 HemA and approximately 50% identical to HemT. In contrast to the hemA gene of Rba. sphaeroides 2.4.1, the Rba. capsulatus SB1003 hemA upstream sequences have no FNR consensuslike sequence, and hemA is separated by 96 bp from divergently arranged sequences predicted to code for acetylornithine deacetylase (Smart et al., 2004). On the basis of the sequence differences, it was predicted that hemA would not be regulated by FnrL in Rba. capsulatus SB1003, and it posed the larger question as to whether or not Rba. capsulatus has an Fnr-type protein. Besides the difference with respect to regulation of tetrapyrrole biosynthesis that would be implied by the absence of an fnrL gene in Rba. capsulatus, since FnrL is necessary for photosynthetic growth of Rba. sphaeroides 2.4.1 (Zeilstra-Ryalls and Kaplan, 1995a), it would also speak to whether or not oxygen control of photosynthesis gene expression is the same or different between the two species. The question was answered when, using Rba. sphaeroides 2.4.1 fnrL DNA as a probe, the Rba. capsulatus SB1003 fnrL gene was identified and then cloned from a cosmid library (Zeilstra-Ryalls et al., 1997). The amino acid sequences of Rba. sphaeroides 2.4.1 and Rba. capsulatus SB1003 FnrL proteins are predicted to be 78% identical, and absolutely conserved in the recognition helix of their putative DNA binding domains. Thus, both are thought to target the same FNR consensus-like DNA sequence. However, in contrast to Rba. sphaeroides 2.4.1, disabling the Rba. capsulatus SB1003 fnrL gene does not abolish photosynthetic growth, which argues that oxygen control of photosynthesis gene expression does indeed differ between Rba. sphaeroides and Rba. capsulatus (Zeilstra-Ryalls et al., 1997). Whether or not the difference between the two species with respect to the FnrL requirement for photosynthetic growth is limited to regulatory differences among tetrapyrrole genes has not been fully answered yet, but besides hemA, there are other genes known to have functioning target sites for FnrL binding within the upstream sequences in Rba. sphaeroides 2.4.1 but not in Rba. capsulatus SB1003, including the puc operon
Chapter 39
Regulation of Tetrapyrrole Biosynthesis
(Zeilstra-Ryalls and Kaplan, 1998), coding for the LH2 polypeptides, and bchE (Oh et al., 2000), whose product participates in BChl biosynthesis. In Rba. capsulatus SB1003, as determined using a hemA::lacZ reporter plasmid, hemA transcription increases approximately two- to three-fold in response to lowering oxygen tensions (Wright et al., 1991; Smart et al., 2004). As had been anticipated by virtue of the absence of an FNR consensus-like sequence within the hemA upstream sequences this response does not require an intact fnrL gene (Smart et al., 2004). However, an intact regA gene, coding for the Rba. capsulatus homolog of PrrA, is required (Smart et al., 2004). More than one 5´ end was detected among hemA mRNA isolated from wild type Rba. capsulatus SB1003 grown photosynthetically (Smart et al., 2004), suggesting that hemA is transcribed from multiple promoters, as is also true of Rba. sphaeroides 2.4.1 hemA (Neidle and Kaplan, 1993b; Fales et al., 2002). Whether or not the different Rba. capsulatus hemA transcripts are all present under all conditions has not yet been determined. The amino acid sequences of RegA and PrrA proteins are approximately 83% identical overall and absolutely conserved in their respective DNA binding domains (Elsen et al., 2004). But predictions of binding sites are highly uncertain, since the most conserved feature is an inverted repeat of GCG separated by a number of residues that can vary between 0 and 12 nt (Laguri et al., 2003; Mao et al., 2005), and these organisms have GC contents of approximately 68–69% (Haselkorn et al., 2001; Mackenzie et al., 2001). Comparisons of the known PrrA binding sites within the upstream sequences of the Rba. sphaeroides 2.4.1 hemA gene (RansonOlson et al., 2006) with the four predicted RegA binding sites within the upstream sequences of the Rba. capsulatus SB1003 hemA gene (Smart et al., 2004) were not found to be informative. Already with this description of what is known about the genes whose products catalyze the first step in tetrapyrrole biosynthesis in these bacteria, it is clear that there is considerable complexity and variability not only with respect to their regulation but also with respect to the very number of ALA synthase genes present in a given strain, as well as between species of Rhodobacter. Differences are not limited to regulation of ALA formation, but extend to other tetrapyrrole biosynthesis genes in these bacteria, as well.
783
B. Regulation of the hemB Gene A single hemB gene, coding for porphobilinogen synthase, is apparently present in all three sequenced strains of Rba. sphaeroides, and among the DNA sequences available for Rba. capsulatus SB1003. This enzyme catalyzes the reaction that combines two ALA molecules to form the pyrrole ring. Since ALA can be made by two different pathways among purple anoxygenic photosynthetic bacteria, HemB is in point of fact the first common enzyme in the production of all tetrapyrroles in these bacteria. However, based on the evidence available at present, for both Rba. sphaeroides and Rba. capsulatus, transcription of the ALA synthase genes is affected to a greater degree by changes in oxygen tension than transcription of the porphobilinogen synthase genes. In Rba. sphaeroides 2.4.1, as measured by microarray analyses, the change in hemB transcript levels in response to oxygen (and light) availability is only slightly higher than the cut-off of significance (Braatsch et al., 2004; Pappas et al., 2004; Anthony et al., 2005; Moskvin et al., 2005; Zeller et al., 2005). Those measurements are consistent with results obtained using a lacZ reporter involving the hemB upstream sequences, which showed no difference in the level of transcription in the presence or absence of oxygen (J.H. Zeilstra-Ryalls, unpublished). No motifs associated with binding by any known regulatory protein of Rba. sphaeroides could be detected within the hemB upstream sequences. Collectively, these data argue that hemB transcription is not responsive to changes in oxygen tension. Biel et al. also found no evidence of oxygen control of hemB transcription in Rba. capsulatus strain PAS100 (Biel et al., 2002), as reported from a hemB:: cat transcriptional fusion plasmid. A similar result was obtained by Smart et al. (2004) using a hemB:: lacZ reporter in Rba. capsulatus SB1003, but those investigators also found that hemB transcription is apparently higher under low or no oxygen (photosynthetic) conditions in an FnrL– mutant background versus the wild type SB1003 background. The sequence TTGca-N4-gTCcA within the hemB upstream sequences was suggested as a potential FnrL binding site (matches to the FNR consensus sequence are in capital letters). Although all available data for Rba. sphaeroides 2.4.1 FnrL regulated genes indicate that the sequence TTG-N8-CAA is highly conserved (Zeilstra-Ryalls and Kaplan, 1995a, 1998; Mouncey
784 and Kaplan, 1998a,b; Oh et al., 2000), there is no in vitro evidence for what residues are actually critical for FnrL binding in either Rba. sphaeroides or Rba. capsulatus. However, a direct role for FnrL in hemB regulation would seem to be difficult to reconcile with the available evidence, since the position of the putative binding site is upstream of the promoter identified for hemB and predicts that FnrL would function as an activator, while the transcription data suggest that fnrL has a negative role in hemB transcription (Smart et al., 2004). The upstream sequences reported by Indest and Biel (1995) include two additional nucleotides that change the sequence of the postulated FnrL binding site of hemB as described by Smart et al. (2004) to TTGca-N4-ccaAg. It could not be determined from the literature (Indest and Biel, 1995) whether or not these differences might be due to strain differences, but they may indicate that FnrL has an indirect role in hemB expression. The role of FnrL in hemB transcription has only been examined thus far in strain SB1003, and further studies will likely resolve these issues. The genomic context of hemB, divergently transcribed from a gene of unknown function, is conserved between all three sequenced strains of Rba. sphaeroides and Rba. capsulatus SB1003 (note that this conserved gene is between the orf1148 and hemB coding sequences described in Smart et al., 2004). As this arrangement is also conserved in other purple anoxygenic photosynthetic bacteria, it may have some regulatory significance that is as yet unknown. C. Regulation of the hemC and hemE Genes Four porphobilinogen molecules are linked together to form a linear tetrapyrrole by a reaction that is catalyzed by the hemC gene product. As was already mentioned, the subsequent cyclization reaction required to form uroporphyrinogen III is catalyzed by an unknown gene product in both Rba. sphaeroides and Rba. capsulatus. In those organisms that synthesize corrinoids such as Rhodobacter species, uroporphyrinogen III is the last common precursor in the tetrapyrrole biosynthesis pathway, and the action of the cobA gene product commits uroporphorphyrinogen III to corrinoid synthesis (Chapter 5, Warren and Deery), while that of the hemE gene product commits the tetrapyrrole to protoporphyrin IX formation. Interestingly, while none of the other tetrapyrrole biosynthesis genes are transcriptionally linked in either species, in both Rba. sphaeroides 2.4.1 and
Jill Helen Zeilstra-Ryalls Rba. capsulatus SB1003, the hemC and hemE genes are divergently transcribed. In Rba. sphaeroides 2.4.1, a study directed towards defining the set of genes belonging to the PpsR regulon revealed that, besides being responsible for oxygen- and light-dependent repression of photosynthesis genes, this DNA binding protein also unexpectedly functions as an aerobic repressor of hemC and hemE transcription (Moskvin et al., 2005). A description of the role of PpsR in photosynthesis gene expression in Rba. sphaeroides, as well as other characteristics of the protein, can be found in Chapter 35, Bauer et al. and Chapter 36, Klug and Masuda. As defined by microarray data, and supported by other measurements, the target sequence for PpsR is TGTc-N10gACA (lower case letters indicate less conservation), and all PpsR-dependent genes in Rba. sphaeroides 2.4.1 contain two repressor binding sites (Moskvin et al., 2005). One of the two sites invariably has no mismatches to the TGT-N12-ACA core consensus but may have a single mismatch to the refined consensus (above), while the second site can contain up to three mismatches to the refined consensus (Moskvin et al., 2005). For hemC and hemE, one perfect consensus sequence is located between the coding sequences of the two genes, while a second perfect match is located within the coding sequences of each gene, and all are absolutely conserved among the three sequenced strains of Rba. sphaeroides. Whether or not other oxygen-responsive regulators are involved in the expression of these genes is not known, although direct regulation by FnrL is unlikely since there is no FNR consensus-like sequence between the hemC and hemE coding sequences. Also, the oxygen responsiveness of these genes appears to be completely accounted for by PpsR-mediated regulation, since the increase in the levels of hemC and hemE transcripts present in wild type cells at reduced oxygen tensions is similar to the levels present in the PpsR-defective mutant strain PPS2-4 (Moskvin et al., 2005). That hemC and hemE are members of the PpsR regulon means all three of the DNA binding proteins that have been shown to regulate photosynthesis gene expression in response to changes in oxygen tensions, PpsR, PrrA, and FnrL (reviewed in Zeilstra-Ryalls and Kaplan, 2004), are also involved in regulating that response in tetrapyrrole production. While not all aspects of oxygen control of hemE and hemC transcription are resolved, it appears that regulation of these genes in Rba. capsulatus SB1003 is quite different from how this takes place
Chapter 39
Regulation of Tetrapyrrole Biosynthesis
in Rba. sphaeroides 2.4.1. Smart et al. (2004) found that hemE and hemC transcription is regulated by multiple regulatory proteins that also mediate the oxygen responsiveness of photosynthesis genes in that organism. Among the known genes coding for enzymes of the common tetrapyrrole biosynthetic pathway in Rba. capsulatus, hemE transcription as reported from a hemE::lacZ transcription fusion plasmid is most responsive to changes in oxygen tensions, with approximately 112-fold higher levels under low oxygen versus high oxygen conditions. Although no likely DNA binding sites for FnrL were identified, both functional fnrL and regA genes appear to be critical for that response. Two CrtJ binding motifs were identified among the hemE-hemC intergenic sequences, but an intact crtJ gene was found to be less important in hemE oxygen responsiveness (Smart et al., 2004). The hemC transcription levels reported from a hemC::lacZ fusion plasmid also increase in response to lowering oxygen tensions, but by less than two-fold. However, in contrast to hemE, the levels are higher in the absence of a functional fnrL gene or crtJ gene than in wild type cells under semiaerobic conditions, and expression is little affected in a regA mutant. The amino acid sequence of CrtJ is 52% identical to that of PpsR of Rba. sphaeroides 2.4.1, and the CrtJ and PpsR target sequences are the same (reviewed recently by Elsen et al., 2005). But there is increasing evidence that there are differences as to their mechanisms of action; for example in Rba. sphaeroides 2.4.1, as was first demonstrated by Gomelsky and Kaplan (1997), the DNA binding activity of PpsR is regulated at the protein level by another protein AppA, while Rba. capsulatus SB1003 lacks an appA gene. The oxygen responsiveness of the hemC and hemE genes in Rba. capsulatus SB1003 is also affected by the presence or absence of an intact aerR gene, but in opposite directions; relative to levels in wild type cells, hemC transcription is higher under low oxygen conditions in the AerR– mutant strain while hemE transcription is lower (Smart et al., 2004). Certainly one direction for future studies could be efforts towards the reconciliation of the disparate effects of the individual regulatory proteins that would account for an overall increase in transcription of both hemC and hemE in response to lowering oxygen tensions in Rba. capsulatus SB1003. AerR has been proposed to stabilize CrtJ binding at pairs of target sequences in which one site is within
785
the promoter of the regulated gene and the second site is located some distance away (e.g., 76 bp; Dong et al., 2002), and thus function as a second aerobic repressor of photosynthesis genes. Interestingly, the two CrtJ binding sites associated with hemE and hemC are separated by 35 bp, which would presumably be considered close together, and suggests previously unsuspected complexities with respect to the role of AerR. A gene coding for a product that is similar (30% amino acid sequence identity) to AerR and called ppaA was identified in a genetic screen of cosmids representing the Rba. sphaeroides 2.4.1 library that up-regulatephotosynthesis gene expression in PrrB or PrrA mutants (Gomelsky and Kaplan, 1994). Unlike the activity ascribed to AerR, PpaA acts as an apparent activator of photosynthesis gene expression under aerobic conditions (Gomelsky et al., 2003). Whether or not PpaA participates in regulation of tetrapyrrole biosynthesis genes in Rba. sphaeroides 2.4.1 is not known. Further details about PpsR, CrtJ, PpaA, AerR, and AppA are presented in Chapter 35, Bauer et al. and Chapter 36, Klug and Masuda. D. Regulation of Genes Coding for the Synthesis of Protoporphyrinogen IX from Coproporphyrinogen III E. coli and other facultative bacteria have two different enzymes, HemF and HemN, that catalyze the oxidative decarboxylations required to form protoporphyrinogen IX from coproporphyrinogen III (reviewed in O’Brian and Thöny-Meyer, 2002). Through a combination of structural and biochemical studies, many mechanistic details of the reactions catalyzed by both enzymes are now known, and account for the oxygen requirement of the HemF-mediated reaction as well as the oxygen independence of the reaction catalyzed by HemN (Breckau et al., 2003; Layer et al., 2003, 2006; Phillips et al., 2004; Lee et al., 2005). While Rba. sphaeroides and Rba. capsulatus also have both aerobic and anaerobic lifestyles, each organism has a different complement of these enzymes. In Rba. sphaeroides wild type strain NCIB8253, the first gene whose product was shown to catalyze this reaction was identified by Coomber et al. (1992), who also found that inactivation of the gene leads to an inability to grow photosynthetically. Since the mutant strain retains the ability to grow aerobically, they concluded that the gene must code for an enzyme that is necessary for catalysis of the oxygenindependent reaction (note that, while that gene
786 was first called hemF, the name was subsequently changed to hemN; as it is also called in all other strains of Rba. sphaeroides in which the gene has been identified). The hemN gene is in proximity to the photosynthesis genes in all strains of Rba. sphaeroides in which the genome or the hemN gene has been sequenced. When the sequence of the fnrL gene was determined in Rba. sphaeroides 2.4.1, a gene coding for a polypeptide with 45% identity to the product identified by Coomber et al. and with similarity to HemN, but not to HemF, of E. coli was found to be divergently transcribed from fnrL. This second hemN-type gene is called hemZ (Zeilstra-Ryalls and Kaplan, 1995a). Unlike the distribution of the two ALA synthase genes over the two chromosomes in Rba. sphaeroides 2.4.1, the hemZ gene is located on the same chromosome (chromosome I) as hemN, but the two genes are separated by approximately 395 kbp. These genes are highly responsive to changes in oxygen tension, and, as reported from hemZ::lacZ and hemN::lacZ transcription fusion plasmids, hemZ expression is more than 23-fold and hemN more than 9-fold higher under low oxygen conditions versus high oxygen conditions (Oh et al., 2000). Consistent with the presence of an FNR consensus-like sequence within the upstream sequences of hemN and hemZ, transcription of both genes is up-regulated by FnrL in response to lowering oxygen tensions. However, for hemZ induction an intact fnrL gene is indispensable, while even in the absence of a functional fnrL gene hemN transcription is responsive to changes in oxygen tensions. PrrA is also reported to contribute to anaerobic induction of both genes, but evidently it is less important for the induction of hemZ than for hemN (Oh et al., 2000). Completion of the DNA sequence of Rba. sphaeroides 2.4.1 genome revealed that, besides hemN and hemZ, a hemF gene is also present, coding for an oxygen dependent enzyme (Mackenzie et al., 2001). In contrast to the situation for the genes coding for the oxygen-independent enzymes, the hemF gene is maximally expressed under aerobic conditions, and down-regulation in the absence of oxygen requires an intact prrA gene (Zeilstra-Ryalls and Schornberg, 2006). Since the putative PrrA binding sites have not been defined, the regulatory significance of the variation among the hemF upstream sequences of the Rba. sphaeroides sequenced strains is not known. However, the FNR consensus-like sequences within the hemZ and hemN upstream sequences are com-
Jill Helen Zeilstra-Ryalls pletely conserved. No FNR consensus-like sequence is present in the upstream sequences of the hemF gene, and based on microarray data (Moskvin et al., 2005), neither hemF nor hemN and hemZ belong to the PpsR regulon. A survey of the distribution of the two types of enzymes indicated that all bacteria having hemF genes also have hemN genes, and for several species of bacteria that have hemF the hemN gene alone is capable of meeting the cellular needs for protoporphyrinogen IX, regardless of the presence or absence of oxygen (O’Brian and Thöny-Meyer, 2002). However, disruption of hemF in Rba. sphaeroides 2.4.1 leads to an inability to grow aerobically (Zeilstra-Ryalls and Schornberg, 2006), which suggests that either there is insufficient expression or insufficient activity of the two oxygen-independent enzymes to support aerobic growth. No HemZ– mutants are available as yet, and so its contribution to protoporphyrinogen IX formation in Rba. sphaeroides is not known. Only one gene coding for a HemN-type enzyme is in evidence in Rba. capsulatus SB1003, and no hemF gene is present. Because the gene is divergently transcribed from fnrL, and because the putative amino acid sequence of its product is somewhat more similar to Rba. sphaeroides 2.4.1 HemZ (approximately 55% identity) than HemN (approximately 51% identity), it is called hemZ (Zeilstra-Ryalls et al., 1997). It must be that either hemZ transcription or HemZ activity in Rba. capsulatus is suitably different from Rba. sphaeroides to support both aerobic and anaerobic growth, since Rba. sphaeroides HemF– mutants cannot grow aerobically (Zeilstra-Ryalls and Schornberg, 2006) and HemN– mutants cannot grow photosynthetically (Coomber et al., 1992). Consistent with that argument, and in contrast to the 23-fold induction reported for hemZ in Rba. sphaeroides 2.4.1 (Oh et al., 2000), Rba. capsulatus SB1003 hemZ transcription as reported from a lacZ fusion plasmid apparently varies by at most 3-fold in aerobically versus semiaerobically grown cells (Smart et al., 2004). Intact fnrL, regA, crtJ, and aerR genes are necessary for that response. An FNR consensus-like sequence is present within the hemZ upstream sequences (Zeilstra-Ryalls et al., 1997), but Smart et al. (2004) found no target sequences for RegA or CrtJ binding. We have now arrived at the penultimate step in the common pathway, but it is the last step for which genes are known in these bacteria. While spontaneous oxidations could support aerobic production of
Chapter 39
Regulation of Tetrapyrrole Biosynthesis 787 et al. (2004) examined hemH transcription using a protoporphyrin IX from protoporphyrinogen IX, how lacZ transcription reporter plasmid in wild type and Rhodobacter species carry out this reaction in the several mutant strains of Rba. capsulatus SB1003, absence of oxygen is not known. and found that the levels of transcription decreases by approximately 2–3 fold when oxygen tensions E. Branchpoint Considerations are reduced. That hemH transcription is apparently higher in photosynthetically grown cells than in cells The first branchpoint in tetrapyrrole biosynthesis that grown under low oxygen conditions underscores the exists in these bacteria directs uroporphoryinogen III importance of evaluating the roles of oxygen and light towards corrinoids or towards coproporphyrinogen separately. Normal levels of transcription of hemH III, and the first reaction of the corrinoid branch is under any condition appear to require an intact regA catalyzed by the cobA gene product (see Chapter 5, gene, and disabling crtJ also reduces hemH transcripWarren and Deery). Oxygen responsiveness data tion under all conditions, although no CrtJ target for this gene in Rba. sphaeroides 2.4.1 is available sequence was identified (Smart et al., 2004). from microarray studies (Braatsch et al., 2004), and show that the levels of cobA transcripts do not vary F. Oxygen Control of Tetrapyrrole Biosynthesis in Rhodobacter As We Know It significantly in cells grown under high or low oxygen tension conditions; no data are currently available for Having presented details about transcriptional regulathe Rba. capsulatus SB1003 cobA gene. tion of each gene, this discussion of oxygen control The Mg and Fe chelatases determine the fate of of the known tetrapyrrole biosynthesis genes in the protoporphyrin IX, the last precursor common to two Rhodobacter species concludes by comparing both heme and BChl biosynthesis. A comparison the overall changes in transcription of the genes. It of the transcriptional response of the genes coding is important to emphasize the fact that the changes for these enzymes to oxygen availability in Rba. represent the net outcomes of what are known to be sphaeroides 2.4.1 based on microarray analyses multiple regulatory events associated with several (Braatsch et al., 2004) shows that while transcript of these genes. levels of the ferrochelatase gene, hemH, are little afFor Rba. sphaeroides 2.4.1, the relevant microfected by oxygen availability, bchH, bchI, and bchD array data pertaining to oxygen responsiveness of transcripts, whose products comprise subunits of the tetrapyrrole biosynthesis genes have been assembled Mg chelatase (Gibson et al., 1995), are significantly and presented in Fig. 2 (reliably measured transcript higher when oxygen tensions are reduced, as befits levels are considered to be those whose standard detheir designation as photosynthesis genes. Other viations among three replicates are less than 15%). microarray data show that PpsR regulates bchH, I, Depicted are relative changes in transcript levels and D expression, but not hemH (Moskvin et al., when comparing cells grown under highly aerobic 2005). No other details about the regulation of these conditions in the dark to those present in cells grown genes in Rba. sphaeroides 2.4.1 are available as yet, under semi-aerobic conditions, and when comparbut none have FNR consensus-like sequences within ing cells grown under highly aerobic conditions to their upstream sequences. those present in cells grown under photosynthetic In Rba. capsulatus SB1003 bchH and bchD tran(no oxygen, in the light) conditions. Since there is scription may be responsive to changes in oxygen little difference between the two ratios for most of tensions, but data are only available that compare the genes, oxygen availability appears to be the most transcription levels in aerobically and photosynthetiimportant parameter. The exceptions are the hemC cally grown cells, and so they would represent the and hemF genes. But transcript levels for these sum of both oxygen and light effects. Using reporter genes do not differ significantly in cells grown in plasmids involving translational fusions between a the presence of high versus low light intensities (the promoterless lacZ gene and bchH or bchD (which threshold for significant differences in expression also included the upstream bchI gene) expression in is 1.5-fold or more difference in the average of the photosynthetically grown cells was found to be 3.4transcript levels measured in cells grown under the fold higher and 6.3-fold higher, respectively, than in two conditions). Therefore, rather than light responaerobically grown cells, and that CrtJ represses the siveness, the differences in the ratios of hemC and genes aerobically (Ponnampalam et al., 1995). Smart
788
Jill Helen Zeilstra-Ryalls
Fig. 2. Expression changes (plotted as fold differences) of tetrapyrrole biosynthesis genes as measured by genechip microarrays (data are from the Gene Expression Omnibus at the National Center for Biotechnology Information; http://www.ncbi.nlm.nih.gov/geo/). The expression of every gene in Rba. sphaeroides 2.4.1 grown in high (30%) oxygen is assigned a value of 1. Dark gray fill, the ratio of the average levels of transcripts detected in 2.4.1 grown in low (3% oxygen) versus high oxygen conditions; light gray fill, the ratio of the average levels of transcripts in 2.4.1 grown photosynthetically (0% oxygen, in front of lights, 10W/m2) versus high oxygen conditions. Error bars represent ranges of the values. The threshold for significant changes in expression is 1.5-fold (either positive or negative), and is indicated by dashed lines.
hemF transcripts probably reflect the further lowering of oxygen tensions when the cells are grown photosynthetically compared to semi-aerobically. Although measurements of the transcript levels of those genes in cells grown under anaerobic-dark conditions (with dimethyl sulfoxide as alternate electron acceptor) could be useful in determining the validity of that conclusion, the transcripts were not reliably detected in those cells, and so validation will require other experimentation. The changes in transcript levels measured for the Rba. sphaeroides 2.4.1 tetrapyrrole biosynthesis genes differ from those indicated by the transcription reporter studies in Rba. capsulatus SB1003 of Smart et al. (2004) in two ways. First, while the oxygen responsiveness of the hemF, hemN and hemZ genes in Rba. sphaeroides 2.4.1 far exceeds that of any of the other genes in the common pathway, the hemE gene is most responsive in Rba. capsulatus SB1003. Second, the levels of transcription in photosynthetically grown cells differ from those in semi-aerobically grown cells for all but the hemB gene; for hemA and hemC the levels are higher, but for hemE and hemZ
the levels are lower. With respect to the hemH gene, in both species, transcription is lower in semi-aerobically grown cells than in aerobically grown cells. However, unlike in Rba. sphaeroides 2.4.1, hemH transcription is nearly the same (approximately 1.2-fold different) in both photosynthetically and aerobically grown Rba. capsulatus SB1003. Although the culture conditions that were used in the studies are similar but not identical, the data argue that, with respect to the tetrapyrrole biosynthesis genes, the transcriptional responses to oxygen and light are not as tightly integrated in Rba. capsulatus SB1003 as they are in Rba. sphaeroides 2.4.1. One can point to the reliance on one gene for all ALA synthase activity and one gene for all coprophyrinogen III oxidase activity in Rba. capsulatus versus the two ALA synthase genes and the three coproporphyrinogen III oxidase genes in Rba. sphaeroides for at least one reason as to why regulation might be different. Additionally, in Rba. sphaeroides high and low oxygen tensions define different catabolic metabolisms, because that organism has both the aa3 mitochondrial-type and high oxygen affinity cbb3
Chapter 39
Regulation of Tetrapyrrole Biosynthesis
789
cytochrome c oxidases. Therefore, the differences in regulation of tetrapyrrole biosynthesis genes may be one manifestation of the absence of the aa3 cytochrome c oxidase in Rba. capsulatus.
cell mediates changes in tetrapyrrole gene expression in response to those environmental factors, as well (Joshi and Tabita, 1996). However, this has not yet been demonstrated directly.
IV. Other Aspects of Transcriptional Regulation of Tetrapyrrole Biosynthesis Genes in Rhodobacter Species
V. A Genomics Perspective on the Regulation of Tetrapyrrole Biosynthesis in Other Purple Anoxygenic Photosynthetic Bacteria
The existence of additional regulators of tetrapyrrole biosynthesis genes has already been demonstrated in Rba. capsulatus SB1003. The hbrL gene, coding for a LysR-type transcriptional regulatory protein, was identified by Smart and Bauer (2006) in a visual screen for cosmids that affect β-galactosidase activity emanating from a hemB::lacZ plasmid in Rba. capsulatus SB1003. As is true of other members of the LysR family of proteins (reviewed in Schell, 1993), HbrL activity appears to be modulated by binding of a coeffector that is thought to be heme, since purified HbrL binds heme (Smart and Bauer, 2006). While the screen relied on detecting differences in hemB transcription, HbrL appears to have a more significant role in hemA and hemZ expression, functioning as an activator of those genes when heme levels are low. Intriguingly, while Smart and Bauer found that hemE and hemC are also responsive to exogenous heme, this response does not require an intact hbrL gene, suggesting the existence of at least one additional protein involved in regulation of tetrapyrrole biosynthesis genes (Smart and Bauer, 2006). Identification of the DNA target sequences for HbrL will no doubt be useful in further understanding the role of HbrL in Rba. capsulatus. Examining the effect of heme on tetrapyrrole biosynthesis gene expression in Rba. sphaeroides 2.4.1 will require developing a suitable means to circumvent the apparent inability of that organism to take up heme; growth of mutant strain HemAT1, a 2.4.1 derivative disabled in hemA and hemT genes (Neidle and Kaplan, 1993a), is not supported by the addition of hemin to the medium (J.H. Zeilstra-Ryalls, unpublished). In any event, the absence of an obvious HbrL homolog in Rba. sphaeroides 2.4.1 requires that any heme-associated transcriptional effects on tetrapyrrole biosynthesis gene expression operate through a different mechanism. Since, besides oxygen tension, carbon and nitrogen availability also influence cellular redox poise, PrrA/RegA are logical connects between how the
The availability of DNA sequences of several genomes of purple anoxygenic photosynthetic bacteria other than the Rhodobacter species (deposited at http://www.ncbi.nlm.nih.gov/) made to it possible to survey them for the presence or absence of the known regulators associated with oxygen control of tetrapyrrole biosynthesis genes in Rhodobacter. Tetrapyrrole biosynthesis genes present in each sequenced bacterium were also identified, and their upstream sequences were inspected for the presence of patterns that represent detectable target sequences for regulators. Other than Bradyrhizobium sp. BTAi1, this survey includes only those organisms that belong to the purple anoxygenic photosynthetic bacteria as identified by the International Committee on Systematics of Prokaryotes (Imhoff and Madigan, 2004), and among them only those organisms whose sequences were found to have photosynthesis genes. Since several of the genomic sequences are still in draft stage the absence of a gene, either biosynthetic or regulatory, is not conclusive. A. Distribution of the Regulators Table 1 lists the distribution of genes coding for proteins that are similar to those that mediate oxygen responsiveness of tetrapyrrole biosynthesis genes in Rhodobacter species. Thus, both DNA and predicted amino acid sequences of prrA/regA, prrB/regB, ppsR/crtJ, ppaA/aerR, and fnrL were used to probe the genomic sequences in silico. Although some among the genes identified in this survey have been defined as encoding regulatory proteins, the participation of others in the regulation of those tetrapyrrole biosynthesis genes of interest here has not been described as yet. Apparently, Rhodospirillum (Rsp.) rubrum ATCC11170 has no sequences coding for PrrA/RegA or PrrB/RegB proteins. Likewise, no sequences coding for PrrA/RegA or PrrB/RegB proteins could be
Jill Helen Zeilstra-Ryalls
790
Table 1. Genome sequence status and distribution of regulatory proteins among purple anoxygenic photosynthetic bacteria. PrrB/ RegB no
PpsR/CrtJ1
PpaA/AerR
yes2
PrrA/ RegA no
not detected
not detected
draft
yes
yes
yes
49 (1), 41 (2)
maybe3
Bradyrhizobium sp. ORS278
partial
yes
yes
yes
50 (1), 42 (2)
maybe
Erb. sp NAP1
draft
yes
no
no
25
maybe
Organism Acp. cryptum JF-5
Sequence Status draft
Bradyrhizobium sp. BTAi1
FnrL
4
4
Hlr. halophila SL1
draft
yes
40
22
70/218 identities
not detected
Jannaschia sp. CCS1
complete
yes
yes
yes
48
yes
Loktanella vestfoldensis SKA53
draft
yes
yes
yes
43
yes
Mtb. extorquens AM1
draft
yes
yes
yes
fragmented
maybe
Rba. capsulatus SB1003
draft
yes
yes
yes
50
yes
Rba. sphaeroides 2.4.1
complete
yes
yes
yes
100
yes
Rba. sphaeroides ATCC17025
draft
yes
yes
yes
88
yes
Rba. sphaeroides ATCC17029
draft
yes
yes
yes
100
yes
Rps. palustris BisA53
draft
yes
yes
yes
54 (1), 74 (2)
maybe
Rps. palustris BisB5
complete
yes
yes
yes
67 (1), 82 (2)
maybe
Rps. palustris BisB18
complete
yes
yes
yes
61 (1), 74 (2)
maybe
Rps. palustris CGA009
complete
yes
yes
yes
100 (1), 100 (2)
maybe
Rps. palustris HaA2
complete
yes
yes
yes
71 (1), 83 (2)
maybe
Rsp. rubrum ATCC11170
complete
yes
no
no
41 (2)
not detected
Rhodovulum sulfidophilum W4
partial
unknown
yes
yes
unknown
unknown
Rsb. denitrificans OCh114
complete
yes
yes
yes
43
yes
Rva. sp. 217
draft
yes
yes
yes
42
yes
Rvi. gelatinosus S1 and IL144
partial
unknown
unknown
unknown
30
maybe
5
Thiocapsa roseopersicina BBS partial yes unknown unknown 43 (1), 45 (2) unknown Approximate percentage identities with Rba. sphaeroides 2.4.1 PpsR for those bacteria having one PpsR protein; for those bacteria having two PpsR proteins, percentage identities are with PpsR1 of Rps. palustris CGA009 (1), and with PpsR2 of Rps. palustris CGA009 (2). 2A protein is considered to be present if, for FnrL it has 20% or more amino acid sequence identity with FnrL of Rba. sphaeroides 2.4.1, either of two cysteine motifs (CX2CX7CX88–92C or CX3CX7CX88–92C) are present, and the sequence RX6LGLX2ETVSR within the C-proximal DNA binding domain of the protein is absolutely conserved; for PrrA/RegA, it has 50% or more amino acid sequence identity with PrrA of Rba. sphaeroides 2.4.1; for PrrB/RegB it has 35% or more amino acid sequence identity with PrrB of Rba. sphaeroides 2.4.1 and also has a PrrA protein; and for PpaA/AerR the amino acid sequences are more than 20% identical to Rba. sphaeroides 2.4.1 PpaA. 3Denotes that sequences code for a protein with more than 13 but less than 20% identity to Rba. sphaeroides 2.4.1 PpaA, but also has conserved residues within the vitamin B12-binding motif and the gene is within the cluster of photosynthesis genes. 4Appproximate percentage identities of the best hit with Rba. sphaeroides 2.4.1 PrrA or B proteins, respectively, are indicated. 5The cysteine motif in this protein has the same spacing as E. coli K-12 Fnr; CX2CX5CX92C. 1
identified in either Acidiphilium (Acp.) cryptum JF-5 or Erythrobacter (Erb.) sp. NAP1. Despite the different oxygen-responsiveness of gene expression in Rhodovulum sulfidophilum W4 or Roseobacter (Rsb.) denitrificans OCh114, their prrA/regA and prrB/regB genes are able to appropriately modulate photosynthetic membrane formation in the corresponding mutant strains of Rba. capsulatus SB1003 (Masuda et al., 1999). Clearly, there remain as yet unresolved
complexities associated with this two-component system and how it regulates gene expression. On the other hand, although the bacteria are from very different phylogenetic groups, all appear to have genes coding for proteins with features similar to FnrL (having overall similarity, one or the other of two absolutely conserved cysteine-containing motifs, and conservation within the recognition helix of the DNA binding domains), and with the exception of
Chapter 39
Regulation of Tetrapyrrole Biosynthesis
Acp. cryptum JF-5 all have at least one PpsR/CrtJ protein (note that since the genome sequence of Acp. cryptum JF-5 is incomplete a ppsR gene may nevertheless be present). It should be noted that PpsR/CrtJ-type regulators are known to function in both oxygen and light control of photosynthesis gene expression (Giraud et al., 2002; Jaubert et al., 2004; Elsen et al., 2005; Happ et al., 2005; Kovacs et al., 2005). Because knowledge of PpsR/CrtJ regulation of the tetrapyrrole biosynthesis genes seems to be limited so far to the reports already discussed for Rba. sphaeroides and Rba. capsulatus there was insufficient knowledge to prognosticate about the possible role of PpsR/CrtJ in both oxygen and light regulation of tetrapyrrole biosynthesis genes in other species. Thus, only the presence or absence of putative ppsR genes is reported here, and in Table 2 the presence or absence of PpsR/CrtJ binding motifs among the tetrapyrrole biosynthesis genes. In searching among the genomic sequences for PpaA/AerR-encoding genes using a stringency of 20% amino acid sequence identities, the proteins are not widely represented. However, there are several examples of sequences that have some similarity to PpaA/AerR within a vitamin B12 binding motif (Gomelsky et al., 2003), and that are within the cluster of photosynthesis genes. As important features for the proteins become known, better criteria can be used to reevaluate the distribution of PpaA/AerR among these bacteria. B. Distribution of Tetrapyrrole Biosynthesis Genes Table 2 lists the tetrapyrrole biosynthesis genes that could be identified by searches among the bacteria whose genomic sequences have achieved at least draft assembly status. For nearly two decades, Rba. sphaeroides 2.4.1 was the only bacterium known to have two ALA synthase genes. It is now quite clear from the available genomic sequences that many other α-Proteobacteria also have more than one ALA synthase gene. In fact, this survey revealed that two species have at least three, and one has four genes. On the other hand, only one ALA synthase gene could be detected among the Acp. cryptum JF-5 and Rba. capsulatus SB1003 sequences. Also, as previously discussed, the genome of Rba. sphaeroides strain ATCC17025 appears to contain only one ALA synthase gene but nevertheless has two chromosomes, which is supported by the DNA hybridization studies
791
of Nereng and Kaplan (1999). There is no obvious indication of the existence of protein families from amino acid sequence alignments of the ALA synthases. The survey indicates that the C4 pathway of ALA formation is only represented in species belonging to the α-Proteobacteria, while all others only have genes coding for enzymes of the C5 pathway. No information was found in the literature as to how the genes coding for enzymes of the C5 pathway are regulated in the purple anoxygenic photosynthetic bacteria. However, although beyond the scope of the present discussion, that different precursor molecules are used for ALA formation among these bacteria raises the question as to whether or not formation of those precursor molecules is in any way regulated in order to accommodate the changing needs for tetrapyrrole production. Several bacteria in the survey have two hemB genes; Rhodopseudomonas (Rps.) palustris strains BisB18 and BisA53, and Rsp. rubrum ATCC11170. Amino acid sequence comparisons allowed the prediction that these probably represent different kinds of enzymes, based on the presence of certain residues that are also present in biochemically characterized porphobilinogen synthases (Jaffe, 2003; Bollivar et al., 2004). Thus, it is possible to distinguish putative Zn-dependent enzymes having the motif DXCXCX(Y/F)X3G(H/Q)CG versus Zn-independent enzymes which lack the motif, enzymes whose activity is regulated by an allosteric magnesium having the motif RX~164DX~65EXXXD, and enzymes that have neither of those motifs. For those bacteria having two hemB genes, one codes for a product having both the Zn-binding motif and the allosteric Mg-binding motifs, while the second gene codes for an enzyme having the allosteric motif alone. Among those bacteria having only one hemB gene, none have Zn-binding motifs. Further, the HemB proteins coded for in all three sequenced strains of Rba. sphaeroides, Rba. capsulatus SB1003, Jannaschia sp. CCS1, Loktanella festfoldensis SKA53, Roseovarius (Rva.) sp. 217 and Rsb. denitrificans OCh114 lack all but the last conserved aspartate residue of the allosteric Mg-binding motif; those coded for in Erb. sp. NAP1 and Halorhodospira (Hlr.) halophila SL1 have incomplete motifs (they lack the first aspartate of the motif); and the HemB proteins coded for in all others (including the other three sequenced strains of Rps. palustris) have intact allosteric Mg-binding motifs. It is interesting to note that Acp. cryptum JF-5
hemA none
hemB none (F)6
hemC FNR7
hemE none within 37 bp of upstream sequence present on contig
hemF none
hemN none
hemH none, may be co-transcribed with hemE
bchE present
acsF absent
Bradyrhizobium sp. BTAi1
12-PpsR, A3 2-none 3-B, none
FNR (F)
none, hemC convergent towards hemE
none, hemE convergent towards hemC
PpsR
FNR
none
present
present
Erb. sp NAP1
1-none 2-B, none
none (incomplete F)
none
no motif
2PpsR
FNR
none
absent
present
Hlr. halophila SL1
C54, none
none (incomplete F)
none
none
none
none
none
present
absent
Jannaschia sp.strain CCS1
1-FNR 2-B, 3PpsR, FNR
PpsR
PpsR, hemC divergent from hemE
PpsR, hemE divergent from hemC
none
1-FNR 2-I, FNR (J)
none
absent
present
Loktanella vestfoldensis SKA53
1-C, PpsR 2-none
none
PpsR within divergent hemE gene
none, hemE divergent from hemC
none
FNR
none
absent
present
Mtb. extorquens AM1
1-none 2-none 3-none
none (F)
none
none
none
absent
none
absent
present
Rba. capsulatus SB1003
none [PrrA]5
none [FnrL]
PpsR [PpsR, PrrA, PpaA]
PpsR [Ppsr, FnrL. PrrA, PpaA]
absent
I, FNR [FnrL]
none [PrrA, PpsR, PpaA]
present
present
Rba. sphaeroides 2.4.1
1-FNR [FnrL, PrrA] 2-none
none
2PpsR [PpsR]
2PpsR [PpsR]
none [PrrA]
1-FNR [FnrL, PrrA] 2-I, FNR [FnrL, PrrA]
none
present
present
Rba. sphaeroides ATCC17025
FNR
none
2PpsR
2PpsR
none
1-FNR 2-I, FNR
none
present
present
Rba. sphaeroides ATCC17029
1-FNR 2-none
none
2PpsR
2PpsR
none
1-FNR 2-I, FNR
none
present
present
Rhodoblastus acidophilus1
FNR
?
?
?
?
?
?
?
?
Rps. palustris BisB18
1-2PpsR, FNR 2-B, none
1-none (F, G) 2-FNR (F)
H, none
none
none
K, FNR
none
present
present
Rps. palustris BisB5
1-B, FNR-D 2-FNR
FNR (F)
H, 2FNR
FNR
FNR
K, PpsR, FNR
none
present
present
s
1-B, 2PpsR 2-FNR
1-none (F, G) 2-none (F)
H, none
none
none
K, PpsR, FNR
K,none
present
present
Jill Helen Zeilstra-Ryalls
Organism Acp. cryptum JF-5
792
Table 2. Distribution of tetrapyrrole biosynthesis genes and regulatory motifs within their upstream sequences.
hemA
hemB
hemC
hemE
hemF
hemN
hemH
bchE
acsF
Rps. palustris CGA009
1-B, PpsR, FNR-D 2-FNR
PpsR, FNR (F)
H, FNR
none
none
K, PpsR, FNR
K,none
present
present
Rps. palustris HaA2
1-B, PpsR, FNR-D 2-FNR
FNR (F)
H, 2FNR
FNR
none
K, FNR
K;none
present
present
Rsp. rubrum ATCC11170
1-FNR 2-none
1-none (F, G) 2-none (F)
none
none
absent
none
none
present
absent
Rsb. denitrificans OCh114
1-B, 2PpsR-E 2-none 3-FNR 4-none
PpsR
PpsR, hemC divergent from hemE
PpsR, hemE divergent from hemC
PpsR
1-none 2-I, FNR
none
present
present
Rva. sp 217
1-PspR, FNR 2-B, 2PpsR
none
none, hemC divergent from hemE
none
2FNR
H, FNR
none
present
present
Rvi. gelatinosus S11
C5
?
?
?
?
1-FNR (J) 2-I, PpsR, FNR
?
present
present
Thiocapsa roseopersicina strain BBS1
C5
?
?
?
?
none, may be cotranscribed with bchE
?
present
?
1 Partial sequences available. 2When more than one copy of a gene is present, they are each denoted using numbers in italics followed by a dash. 3Letters used throughout the table are used to denote the following. A: the motif is within the coding sequence of the upstream cysteine synthase gene, B: may be in an operon beginning with bchF; the motifs are upstream of bchF, C: may be co-transcribed with acsF and another open reading frame protein, PpsR site is upstream of acsF, D: the motif is within the PpaA-like protein coding sequences, E: may be in an operon; the motif is upstream of the first gene, F: allosteric Mg-binding motif present, G: Zn-binding motif present, H: may be co-transcribed as hemCbchID, I: divergently transcribed from fnrL, J: incomplete HemN-type motif, K: may be co-transcribed with bchEJ; any motifs are upstream of bchE. 4Indicates the absence of sequences coding for ALA synthase and the presence of sequences coding for glutamyl-tRNA reductase and glutamate-1-semialdehyde aminotransferase. 5Brackets enclose known regulators for the Rba. sphaeroides 2.4.1 and Rba. capsulatus SB1003 genes (see text for details). 6Parentheses enclose information about the gene product (see text for details). 7 This FNR consensus-like sequence is TTGAT-N4-ATGAA.
Regulation of Tetrapyrrole Biosynthesis
Organism
Chapter 39
Table 2. Continued.
793
794 synthesizes Zn-BChl (for a review, see Hiraishi and Shimada, 2001), but its HemB protein appears to be allosterically regulated by Mg. As already described for the Rhodobacter species, the number and distribution of genes coding for oxygen-dependent HemF-type, and oxygen-independent HemN-type coproprophyrinogen III oxidases does not correlate with the facultative nature of the bacteria. The survey shows that there is also no correlation with whether or not the bacteria carry out anoxygenic photosynthesis in the presence or absence of oxygen. Thus, neither Rba. capsulatus SB1003 nor Rsp. rubrum ATCC11170 have a hemF gene, but with the exception of Methylobacterium (Mtb.) extorquens AM1 (in draft stage, so this may change as the genomic sequences become more complete), all of the bacteria have at least one gene coding for a HemN-type enzyme, and Rba. sphaeroides (all three strains), Rsb. denitrificans OCH114, and Jannaschia sp. CCS1 have two HemN-type enzymes. The rule established by O’Brian and Thöny-Meyer (2002) of having a hemN-type gene if the organism has a hemF gene also holds true, other than Mtb. extorquens AM1. The recent identification of the acsF gene product as an oxygen-dependent enzyme that catalyzes the oxidative cyclization of Mg-protoporphyrin IX monomethylester in BChl biosynthesis (Pinta et al., 2002), together with the oxygen-independent activity ascribed to the bchE gene product (Bollivar et al., 1994; Gough et al., 2000), establishes the existence of an enzymic situation that parallels that of the HemFand HemN-type coproporphyrinogen III oxidases. In light of the lack of correlation of the distribution of the hemF, hemN, and hemZ genes with whether or not photosynthesis takes place in a given organism in the presence of oxygen, it was of interest to evaluate the distribution of these genes as well. Whereas Rsp. rubum ATCC11170, which lacks hemF, also lacks an acsF gene Rba. capsulatus has no hemF gene but has both acsF and bchE. Acp. cryptum JF-5 (genome in draft form) has hemF and hemN but no acsF gene, which is also true of Hlr. halophila SL1. By contrast, no bchE homolog was detected in Erb. sp. NAP1 and Jannaschia sp. CCS1, but they do have genes coding for both HemN- and HemF-type enzymes. Finally, Mtb. extorquens AM1 appears to lack both hemN and bchE genes, while acsF and hemF are present. Therefore, there seems to be no correlation between the distribution of these genes and under what conditions photosynthesis takes place among these bacteria.
Jill Helen Zeilstra-Ryalls BchE and the HemN-type enzymes belong to the ‘radical SAM (S-adenosylmethionine)’ superfamily of proteins described by Sofia et al. (2001), which are so named because they are all thought to generate a radical species by reductive cleavage of SAM through a Fe-S center (see Chapter 4, Willows and Kriegel). The distinction between the HemN-type enzymes and other members of the family is that all HemN-type proteins have the absolutely conserved sequence HxxCxxCxxxxCxxC, while other members of the superfamily lack the fourth cysteine of the motif and may or may not have the histidine. Studies of E. coli HemN protein by Layer et al. (2002) indicate that while the histidine, together with the first three cysteines of the motif, provide ligands in the 4Fe-4S cluster necessary for catalysis by the enzyme, the fourth cysteine is also indispensable for activity. Thus, only those genes that code for proteins having all four cysteines were considered hemN-type genes in the survey results listed in Table 2. Both hemN and hemZ genes of Rba. sphaeroides code for proteins that have the HemN-type motif, but their putative amino acid sequences are otherwise only 45% identical. From alignments of the predicted amino acid sequences no obvious distinction among any of the proteins coded for by organisms having two hemN-type genes could be deduced. However, two examples of genes were found that suggests that they would encode HemNtype proteins. These findings were based firstly on genomic context for sequences within the cluster of photosynthesis genes in Rubrivivax (Rvi.) gelatinosus strains S1 and IL144 and for Jannaschia sp. strain CCS1 sequences divergent from a putative fnrL gene, and secondly on having FNR consensus-like sequences within their upstream regions. However, they lack the fourth essential cysteine of the conserved motif and so presumably could not be functional as coproporphyrinogen III oxidases. Their role in tetrapyrrole biosynthesis is therefore uncertain. Unlike the diversity of numbers of other tetrapyrrole biosynthesis genes, only one hemC and hemE gene (and one hemH gene) could be identified in each organism. With one possible exception, genes coding for products that would function as uroporphyrinogen III synthases and protoporphyrinogen IX oxidases could not be identified in any of these bacteria. The exception is that sequences coding for a product having 37% identity at the amino acid sequence level to the protein coded for by hemG of Mesorhizobium loti MAFF303099 were identified in Jannaschia sp. CCS1. However, the putative amino acid sequence of the Jannaschia sp. CCS1 product
Chapter 39
Regulation of Tetrapyrrole Biosynthesis
is only 21% identical to E. coli K-12 HemG and the function of the Mesorhizobium loti gene product (28% identical to E. coli K-12 HemG) has not been confirmed thus far. As will be discussed in Section V.C, since several among the multiple genes present in a given bacterium have upstream motifs associated with one or another DNA binding protein, these would appear to all contribute to tetrapyrrole production, with transcriptional regulation being a prominent means by which the levels and flow of intermediates is directed. C. Motifs The studies of regulation by FnrL and PpsR/CrtJ consistently demonstrate that target sequences for these proteins can be detected with good confidence in Rhodobacter species, but PrrA/RegA binding sites are not reliably detected (Laguri et al., 2003; Mao et al., 2005; Ranson-Olson et al., 2006). The presence of FNR consensus-like sequences and the PpsR/CrtJ core consensus sequence (TGT-N12-ACA) associated with tetrapyrrole biosynthesis genes are indicated in Table 2. As a conservative estimate, since critical residues for FnrL binding (or related proteins) in any of these bacteria have not been defined, only those sequences were included that have matches to at least TTG-N8-CAA. The distribution of these motifs suggests that tetrapyrrole biosynthesis gene expression is distinctive in each organism, which is consistent with the variation already known to exist between Rhodobacter species. It implies that the regulation of these genes is specifically suited to the complete physiology of these bacteria that nevertheless all have the ability to carry out anoxygenic photosynthesis. If the studies of regulation of tetrapyrrole biosynthesis genes in Rhodobacter are any indication, then examining the regulation of these genes in other bacteria will undoubtedly add to our understanding as to how these other bacteria sense and respond to their environment. Note Added in Proof Ouchane et al. (2007) found that, as is true of Rba. sphaeroides 2.4.1, but not Rba. capsulatus SB1003, Rvi. gelatinosus S1 requires an intact fnrL gene for photosynthetic growth. They also showed that, while both hemN2 and bchE transcription is regulated by fnrL, the fnrL requirement for photosynthetic growth
795
of Rvi. gelatinosus and perhaps other but not all purple anoxygenic photosynthetic bacteria is not limited to tetrapyrrole biosynthesis. Acknowledgments The author has attempted to limit this presentation to information that is new to the field since the previous edition of this volume, and apologizes to one and all for any inadvertent omissions or oversights. The author thanks M. O’Brian, J. Smart, D. Bollivar, K. Kovacs, T. Magnuson, T. Gresham, and G. Gussin for helpful discussions, and for clarifications. Their prompt responses were instrumental in completing this manuscript in a timely fashion. Also, the author wishes to gratefully acknowledge all of the work involved in generating the vast amounts of data that are freely available to anyone who has the desire or curiosity to examine them. Support from the National Science Foundation (MCB-0320550) is gratefully acknowledged. References Anthony J, Warczak K and Donohue T (2005) A transcriptional response to singlet oxygen, a toxic byproduct of photosynthesis. Proc Natl Acad Sci USA 102: 6502–6507 Beale SI and Weinstein JD (1991) 2.2 Biosynthesis of 5-aminolevulinic acid in phototrophic organisms. In: Scheer H (ed) Chlorophylls. pp 385–406 CRC Press, Inc., Boca Raton Biel A, Canada K, Huang D, Indest K and Sullivan K (2002) Oxygen-mediated regulation of porphobilinogen formation in Rhodobacter capsulatus. J Bacteriol 184: 1685–1692 Blankenship R, Madigan M and Bauer C (eds) (1995) Anoxygenic Photosynthetic Bacteria. Kluwer Academic Publishers, Dordrecht Bollivar D, Suzuki J, Beatty J, Dobrowolski J and Bauer C (1994) Directed mutational analysis of bacteriochlorophyll a biosynthesis in Rhodobacter capsulatus. J Mol Biol 237: 622–640 Bollivar DW, Clauson C, Lighthall R, Forbes S, Kokona B, Fairman R, Kundrat L and Jaffe EK (2004) Rhodobacter capsulatus porphobilinogen synthase, a high activity metal ion independent hexamer. BMC Biochemistry 5: 17 Braatsch S, Moskvin O, Klug G and Gomelsky M (2004) Responses of the Rhodobacter sphaeroides transcriptome to blue light under semiaerobic conditions. J Bacteriol 186: 7726–7735 Breckau D, Mahlitz E, Sauerwald A, Layer G and Jahn D (2003) Oxygen-dependent coproporphyrinogen III oxidase (HemF) from Escherichia coli is stimulated by manganese. J Biol Chem 278: 46625–46631 Coomber SA, Jones RM, Jordan PM and Hunter CN (1992) A putative anaerobic coproporphyrinogen III oxidase in Rhodobacter sphaeroides. I. Molecular cloning, transposon mutagenesis and sequence analysis of the gene. Mol Microbiol 6: 3159–3169
796 Dong C, Elsen S, Swem L and Bauer C (2002) AerR, a second aerobic repressor of photosynthesis gene expression in Rhodobacter capsulatus. J Bacteriol 184: 2805–2814 Dubbs J and Tabita F (2003) Interactions of the cbbII promoteroperator region with CbbR and RegA (PrrA) regulators indicate distinct mechanisms to control expression of the two cbb operons of Rhodobacter sphaeroides. J Biol Chem 278: 16443–16450 Elsen S, Swem L, Swem D and Bauer C (2004) RegB/RegA, a highly conserved redox-responding global two-component regulatory system. Microbiol. and Mol Biol Reviews 68: 263–279 Elsen S, Jaubert M, Pignol D and Giraud E (2005) PpsR: A multifaceted regulator of photosynthesis gene expression in purple bacteria. Mol Microbiol 57: 17–26 Eraso JM and Kaplan S (1994) PrrA, a putative response regulator involved in oxygen regulation in photosynthesis gene expression in Rhodobacter sphaeroides. J Bacteriol 176: 32–43 Fales L, Kryszak L and Zeilstra-Ryalls J (2001) Control of hemA expression in Rhodobacter sphaeroides 2.4.1: Effect of a transposon insertion in the hbdA gene. J. Bacteriol. 183:1568–1576 Fales L, Nogaj L and Zeilstra-Ryalls J (2002) Analysis of the upstream sequences of the Rhodobacter sphaeroides 2.4.1 hemA gene: In vivo evidence for the presence of two promoters that are both regulated by fnrL. Photosynthesis Research 74: 143–151 Fischer H-M (1994) Genetic regulation of nitrogen fixation in rhizobia. Microbiol. Reviews 58:352–386 Gibson, LCD, Willows RD, Kannangara CG, von Wettstein D and Hunter CN (1995) Magnesium-protoporphyrin chelatase of Rhodobacter sphaeroides: Reconstitution of activity by combining the products of bchH, I, and D genes expressed in Escherichia coli. Proc Natl Acad Sci USA 92: 1941–1944 Giraud E, Fardoux J, Fourrier N, Hannibal L, Genty B, Bouyer P, Dreyfus B and Verméglio A (2002) Bacteriophytochrome controls photosystem synthesis in anoxygenic bacteria. Nature 417: 202–205 Gomelsky M and Kaplan S (1994) Identification of transcriptional factors involved in the regulation of photosynthesis gene expression in Rhodobacter sphaeroides 2.4.1. Abstracts of the VIII International Symposium on Phototrophic Prokaryotes. 41B: Sant’Angelo in Vado, Italy: Tipgrafia Grafica Valdese Gomelsky M and Kaplan S (1997) Molecular genetic analysis suggesting interactions between AppA and PpsR in regulation of photosynthesis gene expression in Rhodobacter sphaeroides 2.4.1. J Bacteriol 179: 128–134 Gomelsky L, Sram J, Moskvin OV, Horne IM, Dodd HN, Pemberton JM, McEwan AG, Kaplan S and Gomelsky M (2003) Identification and in vivo characterization of PpaA, a regulator of photosystem formation in Rhodobacter sphaeroides. Microbiology 149: 377–388 Gough S, Petersen B and Duus J (2000) Anaerobic chlorophyll isocyclic ring formation in Rhodobacter capsulatus requires a cobalamin cofactor. Proc Natl Acad Sci USA 97: 6908–6913 Happ H, Braatsch S, Broschek V, Osterloh L and Klug G (2005) Light-dependent regulation of photosynthesis genes in Rhodobacter sphaeroides 2.4.1 is coordinately controlled by photosynthetic electron transport via the PrrBA two-component system and the photoreceptor AppA. Mol Microbiol 58: 903–914 Haselkorn R, Lapidus A, Kogan Y, Vlcek C, Paces J, Paces V,
Jill Helen Zeilstra-Ryalls Ulbrich P, Pecenkova T, Rebrekov D, Milgram A, Mazur M, Cox R, Kyrpides N, Ivanova N, Kapatral V, Los T, Lykidis A, Mikhailova N, Reznik G, Vasieva O and Fonstien M (2001) The Rhodobacter capsulatus genome. Photosynth Res 70: 43–52 Hiraishi A and Shimada K (2001) Aerobic anoxygenic photosynthetic bacteria with zinc-bacteriochlorophyll. J Gen Appl Microbiol 47: 161–180 Hornberger U, Liebetanz R, Tichy HV and Drews G (1990) Cloning and sequencing of the hemA gene of Rhodobacter capsulatus and isolation of a δ-aminolevulinic acid-dependent mutant strain. Mol. Gen. Genet. 221:371–378 Horne I, Pemberton J and McEwan A (1996) Photosynthesis gene expression in Rhodobacter sphaeroides is regulated by redox changes which are linked to electron transport. Microbiology 142: 2831–2838 Iida K and Kajiwara M (2000) Evaluation of biosynthetic pathways to δ-aminolevulinic acid in Propionibacterium shermanii based on biosynthesis of vitamin B12 from D-[1-13C] glucose. Biochemistry 39: 3666–3670 Imhoff J and Madigan M (2004) Minutes: International committee on systematics of prokaryotes, subcommittee on the taxonomy of phototrophic bacteria. International Journal of Systematic and Evolutionary Microbiology 54: 1001–1003 Indest K and Biel AJ (1995) Nucleotide sequence of the Rhodobacter capsulatus hemB gene. Plant Physiol. 108: 1 Jaffe E (2003) An unusual phylogenetic variation in the metal ion binding sites of porphobilinogen synthase. Chemistry & Biology 10: 25–34 Jaubert M, Zappa S, Fardoux J, Adriano J-M, Hannibal L, Elsen S, Lavergne J, Verméglio A, Giraud E and Pignol D (2004) Light and redox control of photosynthesis gene expression in Bradyrhizobium. Dual roles of two PpsR. J Biol Chem. 279: 44407–44416 Jordan P (ed) (1991) Biosynthesis of Tetrapyrroles. Elsevier, Amsterdam Joshi H and Tabita F (1996) A global two component signal transduction system that integrates the control of photosynthesis, carbon dioxide assimilation, and nitrogen fixation. Proc Natl Acad Sci USA 93: 14515–14520 Kiley P and Beinert H (1998) Oxygen sensing by the global regulator, Fnr; the role of the iron-sulfur cluster. FEMS Microbiol Rev 22: 341–352 Kovacs A, Rakhely G and Kovacs K (2005) The PpsR regulator family. Res Microbiol. 156: 619–625 Laguri C, Phillips-Jones M and Williamson M (2003) Solution structure and DNA binding of the effector domain from the global regulator PrrA (RegA) from Rhodobacter sphaeroides: Insights into DNA binding specificity. Nucleic Acids Research 31: 6778–6787 Lascelles J (1956) The synthesis of porphyrins and bacteriochlorophyll by cell suspensions of Rhodopseudomonas spheroides. Biochem J 62: 78–93 Lascelles J (1964) Tetrapyrrole Biosynthesis and its Regulation. W. A. Benjamin, Inc., New York Layer G, Moser J, Heinz D, Jahn D and Schubert W-D (2003) Crystal structure of the coproporphyrinogen III oxidase reveals cofactor geometry of radical SAM enzymes. EMBO Journal 22: 6214–6224 Layer G, Pierik AJ, Trost M, Rigby SE, Leech HK, Grage K, Breckau D, Astner I, Jansch L, Heathcote P, Warren MJ, Heinz DW and Jahn D (2006) The substrate radical of Escherichia coli
Chapter 39
Regulation of Tetrapyrrole Biosynthesis
oxygen-independent coproporphyrinogen III oxidase HemN. J Biol Chem 281: 15727–15734 Layer G, Verfurth K, Mahlitz E and Jahn D (2002) Oxygenindependent coproporphyrinogen-III oxidase HemN from Escherichia coli. J Biol Chem 277: 34136–34142 Lee D, Flachsova E, Bodnarova M, Demeler B, Martasek P and Raman C (2005) Structural basis of hereditary coproporphyria. Proc Natl Acad Sci USA 102: 14232–14237 Mackenzie C, Choudhary M, Larimer FW, Predki PF, Stilwagen S, Armitage JP, Barber RD, Donohue TJ, Hosler JP, Newman JE, Shapleigh JP, Sockett RE, Zeilstra-Ryalls J and Kaplan S (2001) The home stretch, a first analysis of the nearly completed genome of Rhodobacter sphaeroides 2.4.1. Photosynth Res 70: 19–41 Mao L, Mackenzie C, Roh J, Eraso J, Kaplan S and Resat H (2005) Combining microarray and genomic data to predict DNA binding motifs. Microbiology 151: 3197–3213 Masuda, S, Matsumoto, Y, Nagashima, KVP, Shimada, K, Inoue, K, Bauer, CE and Matsuura, K (1999) Structural and functional analyses of photosynthetic regulatory genes regA and regB from Rhodovulum sulfidophilum, Roseobacter denitrificans, and Rhodobacter capsulatus. J Bacteriol 181: 4205–4215 Moskvin O, Gomelsky L and Gomelsky M (2005) Transcriptome analysis of the Rhodobacter sphaeroides PpsR regulon: PpsR as a master regulator of photosystem development. J Bacteriol 187: 2148–2156 Mouncey NJ and Kaplan S (1998a) Cascade regulation of dimethyl sulfoxide reductase (dor) gene expression in the facultative phototroph Rhodobacter sphaeroides 2.4.1T. J Bacteriol 180: 2924–2930 Mouncey NJ and Kaplan S (1998b) Oxygen regulation of the ccoN gene encoding a component of the cbb3 oxidase in Rhodobacter sphaeroides 2.4.1T: Involvement of the FnrL protein. J Bacteriol. 180: 2228–2231 Neidle EL and Kaplan S (1993a) 5-aminolevulinic acid availability and control of spectral complex formation in HemA and HemT mutants of Rhodobacter sphaeroides. J Bacteriol 175: 2304–2313 Neidle EL and Kaplan S (1993b) Expression of the Rhodobacter sphaeroides hemA and hemT genes, encoding two 5-aminolevulinic acid synthase isozymes. J Bacteriol 175: 2292–2303 Nereng K and Kaplan S (1999) Genomic complexity among strains of the facultative photoheterotrophic bacterium Rhodobacter sphaeroides. J Bacteriol 181: 1684–1688 O’Brian M and Thöny-Meyer L (2002) Biochemistry, regulation and genomics of haem biosynthesis in prokaryotes. Adv Microb Physiol 46: 257–318 Oh J-I (2006) Effect of mutations of five conserved histidine residues in the catalytic subunit of the cbb3 cytochrome c oxidase on its function. J Microbiol 44: 284–292 Oh J-I, Eraso J and Kaplan S (2000) Interacting regulatory circuits involved in orderly control of photosynthesis gene expression in Rhodobacter sphaeroides 2.4.1. J Bacteriol 182: 3081–3087 Oh J-I, Ko I-J and Kaplan S (2004) Reconstitution of the Rhodobacter sphaeroides cbb3-PrrBA signal transduction pathway in vitro. Biochemistry 43: 7915–7923 Ouchane S, Picaud M, Therizols P, Reiss-Husson F, and Astier C (2007) Global regulation of photosynthesis and respiration by FnrL: The first two targets in the tetrapyrrole pathway. J Biol Chem 282: 7690–7699 Pappas CT, Sram J, Moskvin OV, Ivanov PS, MacKenzie RC,
797
Choudhary M, Land ML, Larimer FW, Kaplan S and Gomelsky M (2004) Construction and validation of the Rhodobacter sphaeroides 2.4.1 DNA microarray: Transcriptome flexibility at diverse growth modes. J Bacteriol 186: 4748–4758 Petrícek M, Petríceková K, Havlícek L and Felsberg J (2006) Occurrence of two 5-amionolevulinate biosynthetic pathways in Streptomyces nodosus subsp. asukaensis is linked to the production of asukamycin. J Bacteriol 188: 5113–5123 Phillips JD, Whitby FG, Warby CA, Labbe P, Yang C, Pflugrath JW, Ferrara JD, Robinson H, Kushner JP and Hill CP (2004) Crystal structure of the oxygen-dependant coproporphyrinogen oxidase (Hem13p) of Saccharomyces cerevisiae. J Biol Chem 279: 38960–38968 Pinta V, Picaud M, Reiss-Husson F and Astier C (2002) Rubrivivax gelatinosus acsF (previously orf358) codes for a conserved, putative binuclear-iron-cluster-containing protein involved in aerobic oxidative cyclization of Mg-protoporphyrin IX monomethylester. J Bacteriol 184: 746–753 Ponnampalam S, Buggy J and Bauer C (1995) Characterization of an aerobic repressor that coordinately regulates bacteriochlorophyll, carotenoid, and light harvesting-II expression in Rhodobacter capsulatus. J Bacteriol 177: 2990–2997 Ranson-Olson B, Jones D, Donohue T and Zeilstra-Ryalls J (2006) In vitro and in vivo analysis of the role of PrrA in Rhodobacter sphaeroides 2.4.1 hemA gene expression. J Bacteriol 188: 3208–3218 Roh J, Smith W and Kaplan S (2004) Effects of oxygen and light intensity on transcriptome expression in Rhodobacter sphaeroides 2.4.1. J Biol Chem 279: 9146–9155 Scheer H (ed) (1991) Chlorophylls. CRC Press, Inc., Boca Raton Schell M (1993) Molecular biology of the LysR family of transcriptional regulators. Ann Rev Microbiol 47: 597–626 Schilke BA and Donohue TJ (1992) δ-aminolevulinate couples cycA transcription to changes in heme availability in Rhodobacter sphaeroides. J Mol Biol 226: 101–115 Smart J and Bauer C (2006) Tetrapyrrole biosynthesis in Rhodobacter capsulatus is transcriptionally regulated by the heme-binding regulatory protein, HbrL. J Bacteriol 188: 1567–1576 Smart J, Willett J and Bauer C (2004) Regulation of hem gene expression in Rhodobacter capsulatus by redox and photosystem regulators RegA, CrtJ, FnrL, and AerR. J Mol Biol 342: 1171–1186 Sofia H, Chen G, Hetzler B, Reyes-Spindola J and Miller N (2001) Radical SAM, a novel protein superfamily linking unresolved steps in familiar biosynthetic pathways with radical mechanisms: Functional characterization using new analysis and information visualization methods. Nucleic Acids Res 29: 1097–1106 Toledo-Cuevas M, Barquera B, Gennis R, Wikstrom M and Garcia-Horsman J (1998) The cbb3-type cytochrome c oxidase from Rhodobacter sphaeroides, a proton-pumping heme-copper oxidase. Biochim Biophys Acta 1365: 421–434 Willows R, Lake V, Roberts T and Beale S (2003) Inactivation of Mg chelatase during transition from anaerobic to aerobic growth in Rhodobacter capsulatus. J Bacteriol 185: 3249–3258 Wright M, Cardin R and Biel A (1987) Isolation and characterization of an aminolevulinate-requiring Rhodobacter capsulatus mutant. J Bacteriol 169:961–966 Wright M, Eckert J, Biel S and Biel A (1991) Use of a lacZ fusion to
798 study transcriptional regulation of the Rhodobacter capsulatus hemA gene. FEMS Microbiol Lett 78: 339–342 Zeilstra-Ryalls JH and Kaplan S (1995a) Aerobic and anaerobic regulation in Rhodobacter sphaeroides 2.4.1: The role of the fnrL gene. J Bacteriol 177: 6422–6431 Zeilstra-Ryalls JH and Kaplan S (1995b) Regulation of 5-aminolevulinic acid synthesis in Rhodobacter sphaeroides 2.4.1: The genetic basis of mutant H-5 auxotrophy. J Bacteriol 177: 2760–2768 Zeilstra-Ryalls JH and Kaplan S (1996) Control of hemA expression in Rhodobacter sphaeroides 2.4.1: Regulation through alterations in cellular redox state. J Bacteriol 178: 985–993 Zeilstra-Ryalls JH and Kaplan S (1998) Role of the fnrL gene in photosystem gene expression and photosynthetic growth of
Jill Helen Zeilstra-Ryalls Rhodobacter sphaeroides 2.4.1. J Bacteriol 180: 1496–1503 Zeilstra-Ryalls J and Kaplan S (2004) Oxygen intervention in the regulation of gene expression: The photosynthetic bacterial paradigm. Cell Molec Life Sci 61: 417–436 Zeilstra-Ryalls J and Schornberg K (2006) Analysis of hemF gene function and expression in Rhodobacter sphaeroides 2.4.1. J Bacteriol 188: 801–804 Zeilstra-Ryalls JH, Gabbert K, Mouncey NJ, Kranz RG and Kaplan S (1997) Analysis of the fnrL gene and its function in Rhodobacter capsulatus. J Bacteriol 179: 7264–7273 Zeller T, Moskvin O, Li K, Klug G and Gomelsky M (2005) Transcriptome and physiological responses to hydrogen peroxide of the facultatively phototrophic bacterium Rhodobacter sphaeroides. J Bacteriol 187: 7232–7242
Chapter 40 Bacteriophytochromes Control Photosynthesis in Rhodopseudomonas palustris Katie Evans, Toni Georgiou, Theresa Hillon, Anthony Fordham-Skelton and Miroslav Papiz* Science and Technology Facilities Council, Daresbury Laboratory, Daresbury Science and Innovation Campus, Warrington, Cheshire, WA4 4AD, U.K.
Summary ............................................................................................................................................................... 799 I. Introduction..................................................................................................................................................... 800 II. Bacteriophytochrome Gene Organization and Regulation of Photosynthesis in Rhodopseudomonas palustris ......................................................................................................................................................... 800 III. Phytochrome Domain Organization ............................................................................................................... 803 IV. Bilin Chromophore Photo-conversion............................................................................................................. 804 V. Chromophore Binding Domain of Deinococcus radiodurans Bacteriophytochrome ..................................... 805 VI. Small Angle X-ray Scattering Solution Structure of Bph4 from Rhodopseudomonas palustris .................... 805 VII. Conclusions .................................................................................................................................................... 807 Acknowledgments ................................................................................................................................................. 807 References ............................................................................................................................................................ 807
Summary The ability to endure and adapt to a broad array of environmental conditions is a feature of photosynthesizing organisms. This is particularly true of changing light conditions as acclimatization to light is of importance for energy utilization and for the prevention of radiation damage. The available light energy is influenced by several factors such as day/night cycles, shade, water depths, and competition from other photosynthesizing organisms. A more subtle effect is the availability of light in different parts of the electromagnetic spectrum which can influence the chromophores chosen by the photosynthetic organism, and the ways in which these are organized within chromophore-protein complexes. Strategies are available for responding to the presence of solar energy and for optimizing photon capture. In most cases the strategies begin with the initiation of an altered pattern of gene expression arising from light sensing mechanisms that repress or activate the expression of genes required for photosynthesis. A further fine tuning of gene expression, to suit particular light conditions, can achieve additional improvements in efficient utilization of solar energy. This can be as simple as an increase in the quantity of pigmented complexes, in response to altered light intensity, or structural changes to photosynthetic complexes that ensure efficient adaptation to an altered distribution of photon energies. Phytochromes are an important class of macromolecules that sense fluctuating light intensity and spectral quality to alter gene expression. They have been shown to be of central importance in plant acclimatization and in recent years bacterial homologs have been found. A number of bacteriophytochromes have been discovered in the phototrophic non-sulfur purple bacterium Rhodopseudomonas (Rps.) palustris that are involved in the control of photosynthesis and which provide insights into how this organism adapts to light in its environment.
*Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 799–809. © 2009 Springer Science + Business Media B.V.
800
Katie Evans, Toni Georgiou, Theresa Hillon, Anthony Fordham-Skelton and Mirolav Papiz
I. Introduction Phytochromes are photoreceptors that use bilin (linear tetrapyrrole) chromophores to sense the red (R) and far-red (FR) region of the electromagnetic spectrum. A photo-conversion proceeds between the R absorbing Pr form and the FR absorbing Pfr form. In the majority of phytochromes the absorbed light facilitates or inhibits an auto-phosphorylation reaction. In their simplest form phytochromes are part of a two-component light sensory system influencing gene expression by phosphorelay to a response regulator. However, in plants they can be part of a more complex multi-step system by including additional protein modules containing phosphorylatable Asp or His residues (Lohrmann and Harter, 2002). Initially found in plants, phytochromes are now known to be distributed in diverse organisms such as cyanobacteria, fungi and bacteria (Karniol et al., 2005). The function of phytochromes is to initiate adaptation to changing light conditions in the environment. In plants they are known as transcriptional regulators controlling many developmental processes such as seed germination, shade avoidance, flowering and leaf development (Maloof et al., 2000). Several bacteriophytochromes (Bphs) have been characterized and represent the most ancient branch of the phytochrome family (Hughes et al., 1997; Yeh et al., 1997; Davis et al., 1999; Bhoo et al., 2001; Giraud et al., 2002; Lamparter et al., 2002). The discovery of Bphs poses an interesting question concerning probable function as they are found in both photosynthetic and non-photosynthetic bacteria, indicating that they may control a diverse set of biological systems. Bacterial and cyanobacterial phytochromes (Cphs) are known to control chromatic adaptation, Abbreviations: Bph – bacteriophytochrome; BrbphP – Bradyrhizobium ORS278 bacteriophytochrome; BV – biliverdin IXα; CA – catalytic ATPase kinase domain; CBD – chromophore binding domain; Cph – cyanobacterial phytochrome; D. – Deinococcus; Dhp – dimerization and histidine phosphotransfer domain; FR – far-red; GAF – cGMP phosphodiesterases/adenylyl cyclases/ bacterial transcription factor FhlA domain; LH – light-harvesting complex; LH2 – light-harvesting complex 2; LH4 – lightharvesting complex 4; Lx – lux, luminous flux on a surface area (lumen/m2); PAS – period protein (PER)/aryl hydrocarbon receptor nuclear translocator protein (ARNT)/ single-minded protein (SIM) domains; PCB – phycocyanobilin; Pfr – far-red absorbing phytochrome; PHY domain – phytochrome domain; Pr – red absorbing phytochrome; PYP – photoactive yellow protein; PΦB – phytochromobilin; R – red; RC-LH1 – reaction center - LH1 core complex; Rps. – Rhodopseudomonas; RR – response regulator; SAXS – small angle X-ray scattering
carotenoid and chlorophyll synthesis, and phototaxis (Giraud et al., 2004; Stowe-Evans and Kehoe, 2004; Fiedler et al., 2005). The racE gene product from the cyanobacterium Fremyella displosiphon facilitates the complementary chromatic adaptation signal transduction pathway (Yeh et al., 1997). The Cph gene product PixJ1 from Synechocystis sp. PCC 6803 is essential for positive phototaxis (Yoshihara et al., 2004). Some phytochrome proteins are also known to absorb blue light, such as the phytochrome-like Ppr from Rhodospirillum centenum which contains an N-terminal photoactive yellow protein (PYP) domain (Jiang et al., 1999). The Bph called BrbphP is encoded by a gene located downstream from the photosynthesis gene cluster in the Bradyrhizobium ORS278 symbiont of the Aeschynomene plant, and controls the complete synthesis of the Bradyrhizobium photosynthetic apparatus (Giraud et al., 2002). The bacterium also contains a Bph that senses light to regulate the biosynthesis of the carotenoid spirilloxanthin (Giraud et al., 2004). The genes regulated by many Bphs in non-photosynthetic bacteria are unknown but Deinococcus (D.) radiodurans Bph is thought to regulate pigmentation, by regulating deinoxanthin carotenoid biosynthesis, and so affording the bacterium protection from light radiation (Davis et al., 1999). II. Bacteriophytochrome Gene Organization and Regulation of Photosynthesis in Rhodopseudomonas palustris Complete genome sequencing of Rps. palustris strain CGA009 (Larimer et al., 2004) has revealed the organization of genes responsible for photosynthesis, Fig. 1. There are also 6 Bph-like genes situated throughout the genome. Four of these genes are near to genes coding for photosynthetic peptides or pigment biosynthetic enzymes, suggesting that these Bphs form part of a photosynthesis controlling mechanism. Rps. palustris is unusual in containing six Bphs (most bacteria have none or at most two), which indicates that it invests heavily in sensing light quality and intensity in the environment. Between different wild type strains of Rps. palustris there is also considerable variation in the distribution of both Bphs and light-harvesting (LH) complexes, as determined by an analysis of the genomes of strains BisB5, BisA53, BisB18 and HaA2, which are known to vary in their
Chapter 40
Bacteriophytochromes and Photosynthesis
801
Fig. 1. Organization of photosynthesis and Bph genes within the chromosome of Rps. palustris strain CGA009. (A) Representation of the circular chromosome showing the positions of the 6 Bphs (Bph0 – Bph5), LH2 genes (pucBaAa – pucBeAe), LH1 genes pufBA. In strain CGA009 pucAc is a pseudo-gene and is presumed to be inactive. (B) A cluster of genes responsible for photosynthesis are found between 1650 and 1720 kilobases: the cluster includes genes for RC and LH1 peptides, enzymes to synthesise bacteriochlorophyll a (bch*) and carotenoid (crt*) pigments. The genes encoding repressors of photosynthesis are ppsR1 and ppsR2. HemO is the gene synthesizing BV by linearizing heme, and Bph3 (Rpa1537) is the bacteriophytochrome that modulates PpsR2 under the control of light. In strain CGA009 Bph3 is frameshifted and therefore inactive but in all other strains it is expressed. PucBeAe codes for LH2-e and is near Bph2 which does not have the BV-binding Cys but may still function as a histidine kinase. The gene cluster between 3400 – 3420 kilobases contains the genes for LH4 (pucBdAd), the genes responsible for controlling expression, Bph4 and Bph5, and their response regulator RR.
pigmentation (unpublished; see http://genome.jgi-psf. org/finished_microbes). For Rps. palustris strain CGA009 the complexity of Bph organization is mirrored by the presence of five peripheral LH complex (pucBA) operons: three encode LH2 complexes, absorbing at 800 and 850 nm; one encodes an unusual LH4 complex that absorbs only at 800 nm; and one contains a pseudo-gene (pucAc) that is presumed to be non-functional (Tharia et al., 1999). It has been shown that two adjacent Bph genes, Rpa3015 (Bph4) and Rpa3016 (Bph5), regulate the expression of LH4 PucBdAd peptides, expressed under low-light conditions or upon R light illumination (Evans et al., 2005; Giraud et al., 2005), while Rpa1537 (Bph3) controls the expression of a major cluster of photosynthesis genes under the influ-
ence of FR light and low oxygen tension (Giraud et al., 2002; Braatsch et al., 2006). Rpa1490 (Bph2) is near the pucC gene, which is thought to be involved in LH2 biogenesis, and pucBeAe genes coding for LH2e peptides. The genes pucBaAa and pucBbAb, coding for LH2 complexes, appear to be isolated from other photosynthesis genes. However it is known that the pucBaAa genes are down-regulated and pucBbAb completely suppressed in low-light conditions, and these responses are linked to pucBdAd (LH4) up-regulation (Tharia et al., 1999). The near-infrared spectra of LH2/LH4 complexes, purified from bacteria grown under different light intensities, are shown in Fig. 2A. Under high light growth conditions (~3000 lx) the fractional ratio of LH4:LH2 is ~0.33:0.67 and rises to ~0.95:05 under
802
Katie Evans, Toni Georgiou, Theresa Hillon, Anthony Fordham-Skelton and Mirolav Papiz
Fig. 2. (A) Near-infrared spectra of LH2 and LH4 light-harvesting complexes from Rps. palustris, purified by sucrose gradient centrifugation to remove RC-LH1 complexes. LH2 absorbs at ~857 and 802 nm, whereas LH4 absorbs at 804 nm. The arrow denotes decreasing light intensity during cell growth: from top to bottom are spectra of complexes from cells grown with 3000, 1000, 300, 90 lx (luminous flux on a surface area, lumen/m2) and a purified LH4 complex spectrum. (B) Photo-conversion absorption spectra of Bph3 (Rpa1573), Bph4 (Rpa3015), Bph5 (Rpa3016). Dark stable states (black) are photo-converted (grey) by FR light for Bph3, and R light for Bph4 and Bph5. Due to spectral overlaps of Pr and Pfr states of Bph4 the photo-conversion from Pr to Pfr is only 60% complete.
low light (90 lx) conditions (Tharia et al., 1999). Bph4 and Bph5 autophosphorylate and phosphotransfer efficiently to a cognate response regulator (RR) Rpa3017 in their Pr forms (Giraud et al., 2005) and both are required for the production of LH4: the Pr form is created by FR light >740 nm or by incubation in the dark. As R light stimulates LH4 peptide expression this implies that the phosphorylated form of the RR acts as a repressor. The spectrum of Bph4 is typical of many Bphs, photo-conversion proceeds in R/FR light between absorption maxima at 707 nm and 750 nm. Bph5 is unique in its spectroscopic absorption properties, and quenches on R light illumination into a Pnr state, with a slight increase in the absorption shoulder at 650 nm (Fig. 2B) (Giraud et al., 2005). Important differences are the dark reversion to stable states; Bph3 goes into the Pfr form while Bph4 and Bph5 revert to Pr (Fig. 2B). In these Bphs
the light wavelengths that turn on biological activity, photosynthesis and LH4 production are those which force them out of their dark stable forms. The evidence therefore suggests that Bph3, Bph4 and Bph5 remove gene repression when in their non-dark stable states which are Pr for Bph3, Pfr for Bph4 and Pnr for Bph5 (Evans et al., 2005; Giraud et al., 2005). Why there are two Bphs responsible for LH4 production is not clearly understood; however LH4 can be produced in low-light conditions, without changing the R/FR ratio, or by altering the R/FR light ratio. The R/FR effect appears to be insensitive to light intensity (Evans et al., 2005). Although Bph4 possesses a true R/FR spectrum, Bph5 is merely quenched and is otherwise spectrally unchanged, suggesting that it is only sensitive to photon flux rather than light wavelength. These two Bphs may therefore form an intensity/color sensing pair which can cater for
Chapter 40
Bacteriophytochromes and Photosynthesis
a number of environmental eventualities. It has been proposed that as Rps. palustris adapted to an aquatic environment it evolved the ability to sense light wavelength/intensity differences that exist at different depths due to the attenuation properties of water. The dramatic reduction in water transmission of light, in going from R to FR, is the same spectral range covered by the absorption maxima of the Pr/Pfr forms of Bph4 (Evans et al., 2005). LH4 is an octamer containing 32 bacteriochlorophyll a pigments, rather than 27 in the nonameric LH2 (see Chapter 9, Gabrielsen et al.), and has a very different exciton band structure optimized for absorption at only 800 nm (de Ruijter et al., 2004; Hartigan et al., 2002). LH2 contains 9 bacteriochlorophylls a molecules absorbing at 800 nm, and 18 absorbing at 850 nm. In water the 800 nm wavelength transmission is ~10-fold larger than at the 850 nm wavelength at depths of ~ 1 m (Evans et al., 2005), and LH4 may give the bacterium an advantage in such low-light conditions by an increased pigment density at 800 nm. Detailed structural information and energy transfer experiments may reveal additional explanations arising from the novel exciton energy band characteristics of LH4. The adaptation appears to be aimed at achieving specific profiles of LH complex expression, under different light conditions, to achieve optimal photosynthesis. III. Phytochrome Domain Organization Phytochromes share similar domain architecture, although some differ significantly in the C-terminal region. The canonical forms of domain organization in plants, cyanobacteria, bacteria and fungi therefore exhibit close similarities (Fig. 3). One of the modules utilized extensively is the PER/ARNT/ SIM (PAS) domain, which is important in signaling modules such as those that monitor changes in light, redox potential, oxygen, small ligands, and overall energy levels of a cell (Taylor and Zhulin, 1999). The PAS domain is often involved in protein-protein interactions and forms heterodimers (Gomelsky and Kaplan, 1995; Gomelsky et al., 2000). In phytochromes it forms part of the input module of a two-component system but can also be an output module, such as in the repressor of photosynthesis PpsR2 which is acted upon by Bph3 (Braatsch et al., 2007). The N-terminal photosensory region invariably comprises a PAS-cGMP phosphodiesterases/adenylyl cyclases/ bacterial
803
Fig. 3. The canonical domain structures of phytochromes in plants (PHY), bacteria (BPHY), cyanobacteria (CPHY) and fungi (FPHY). PAS1 and PAS2 domains are named after the period protein (PER), the aryl hydrocarbon receptor nuclear translocator protein (ARNT) and the Drosophila single-minded protein (SIM) domains (Zhulin et al., 1997). GAF is named after cGMP-regulated cyclic nucleotide phosphodiesterases, adenylyl cyclases and bacterial transcription factor FhlA domains (Aravind and Ponting, 1997). The phytochrome-specific domain PHY is followed by a histidine kinase domain containing a phosphorylating His in the dimerization domain (Dhp), and an ATP binding kinase domain (Marina et al., 2005). Plant phytochromes are believed to be Ser/Thr kinases (Yeh and Lagarias, 1998). The bilin binding cysteine (C) is found near the N-terminus prior to PAS2 in BPHY and FPHY, and in the GAF domain in PHY and CPHY. In fungi and some bacterial phytochromes a response regulator (RR) is C-terminal to the kinase domain (Karniol et al., 2005).
transcription factor FhlA (PAS-GAF) domain forming the chromophore binding domain (CBD), followed by the phytochrome domain (PHY domain) which is unique to these macromolecules and is believed to regulate the light-initiated photo-conversion between Pr and Pfr. In plants the PHY domain is followed by two PAS domains followed by a histidine kinase-related domain. In cyanobacteria, bacteria and fungi, PHY is followed by a type 1 histidine kinase domain comprising a dimerization domain (Dhp), containing the histidine to be autophosphorylated, and a catalytic kinase ATPase (CA) domain (Hughes and Lamparter, 1999). Fungi also have a receiver domain (RR) C-terminal to the histidine kinase domain, however this is not exclusive to fungi as Bph1 (Rpa0990) in Rps. palustris has this architecture too. The canonical N-terminal organization is PAS-GAF-PHY and distinguishes all phytochrome-like molecules, while the canonical C-terminus has Dhp and CA domains. However an important exception to this architecture, in Rps. palustris, is Bph3 (Rpa1537) which does not autophosphorylate and contains a dimerization and regulatory domain, which is partly formed from a PAS domain followed by a domain of unknown
804
Katie Evans, Toni Georgiou, Theresa Hillon, Anthony Fordham-Skelton and Mirolav Papiz
architecture. This C-terminal region is believed to antagonize DNA-binding by the photosynthesis gene repressor PpsR2 (Braatsch et al., 2007; Jaubert et al., 2004). Other notable exceptions to the canonical C-terminal domain are the stage II sporulation E domain in Kineococcus radiotolerans, GGDEF/EAL in Rhodobacter sphaeroides, and PAS/GGDEF in strain HaA2 of Rps. palustris (unpublished; see http://genome.jgi-psf.org/finished_microbes). IV. Bilin Chromophore Photo-conversion Phytochromes incorporate a linear tetrapyrrole chromophore by ligation to a cysteine residue as a thioether linkage to a group on the A pyrrole ring: the group is an ethylidene in plants and cyanobacteria, and vinyl in bacteria (Fig. 4). In plant and cyanobacterial phytochromes the cysteine is located within the GAF domain, for example at Cys259 in Cph1, but in Bphs it is a highly conserved Cys near to the Nterminus (Lamparter et al., 2003). BPhs and fungal phytochromes incorporate biliverdin IXα (BV) but additional oxidoreductase enzymes are required to make, respectively, phytochromobilin (PΦB) and phycocyanobilin (PCB) in plant and cyanobacterial phytochromes (Lagarias and Lagarias, 1989). In Cph1 the initial assembly conformation of apo-protein and PCB is believed to be C15-Z,syn followed by proton rearrangement which forces the D ring into a C15Z,anti conformation ( Foerstendorf et al., 2001; Otto et al., 2003; Borucki et al., 2005). It has been shown, by resonance Raman spectroscopy, that Pr to Pfr photo-conversion involves the isomerization of ring D (Fig. 4) to the C15-E,anti conformation (Fodor et al., 1988; Fodor et al., 1990). The subsequent dark reversion of Pfr to Pr is proposed to proceed via an additional C15-E,syn isomerization (Andel et al., 2000), however the X-ray structure of the CBD from D. radiodurans, in the Pr state, indicates that C15-E,syn would be sterically unfavorable (Wagner et al., 2005) and recent studies, on locked bilins, that cannot rotate about the C15-C16 bond, show that only C15-E,anti produces an adduct which gives Pfr-like absorption spectra (Lamparter et al., 2003). The reverse photo-conversion of Pfr to Pr proceeds through a different Lumi-F photoproduct adopting a Pr-like conformation which implies a C15-Z,anti type conformation (Foerstendorf et al., 2000). The quantum yields (φpr, φpfr) of wild type Bph and Cph are relatively low (Fischer et al., 2005; Lamparter
Fig. 4. The bilin chromophore is a tetrapyrrole (A, B, C, D). The structure is shown in its ZZZssa conformation. Isomerization proceeds about the C15-C16 bond to the ZZEssa conformation. The chromophores are BV in bacteria and fungi, phycocyanobilin (PCB) in cyanobacteria and phytochromobilin (PΦB) in plants. Coordination to cysteine is via a vinyl-A in BV and an ethylidene-A in PCB and PΦB (inset). R is –CH-CH2 in BV and PΦB, but –CH2-CH3 in PCB (Fodor et al., 1988; Li and Lagarias, 1992). The description of bilin conformations is relative to the circular ZZZsss conformation with N atoms innermost. Z-E and s-a conformation changes are the double bond isomerizations and single bond rotations, respectively, of the bridging methine pyrrole bonds.
et al., 2002): φpr and φpfr are less than 0.1 and φpfr is usually lower than φpr. Elevation of quantum yield for amino acid mutants of Cph1 at Tyr176 show that this residue is essential for determining the extended PCB conformation during photo-conversion, and it is suggested that this is achieved by the stabilization of bilin chromophore protonation (Fischer et al., 2005). The photo-conversion cycle comprises a number of secondary intermediate states that follow the creation of the primary photo-conversion intermediates, LumiR (Pr to Pfr) and Lumi-F (Pfr to Pr), formed on Z-E isomerization (Fig. 5). Femtosecond time-resolved transient absorption spectroscopy has characterized the primary photo-isomerization dynamics of Cph1 from Synechocystis PCC 6803. The Lumi-R state is achieved after an initial relaxation of ~150 fs followed by an E-Z isomerization which is characterized by a redistribution of rate constants with a first moment of ~16 ps. The reverse isomerization from Pfr to Lumi-F is characterized by two shorter time constants, 0.5 and 3.2 ps (Heyne et al., 2002). Secondary intermediates have been observed by cryogenically trapping the meta states. For PhyA from the plant Avena satina, the isomerization reaction into Lumi-R is followed by
Chapter 40
Bacteriophytochromes and Photosynthesis
805
overall picture of photo-conversion is a rapid Z-E isomerization followed by a relaxation of chromophore and protein on longer timescales that involves proton rearrangement. V. Chromophore Binding Domain of Deinococcus radiodurans Bacteriophytochrome
Fig. 5. The cycle of photo-conversion of Pr to Pfr, and Pfr to Pr, comprises a number of intermediate states. The rapid (150 fs–16 ps) isomerization events are Lumi-R and Lumi-F. Thermal relaxation to the meta intermediate states are on longer timescales (0.23 to 260 ms), and are characterized by chromophore torsion angle relaxation and proton rearrangement within the chromophore. The temperatures required to trap the intermediate states are shown.
a thermal decay reaction to meta-Ra which includes a relaxation of the methine bridge double bond joining pyrrole rings C and D. The subsequent formation of meta-Rc is a structural conformation change of the pyrrole rings B and C in the bilin binding pocket (Kneip et al., 1999). Flash photolysis measurements at 133 K on BphAgp1 from Agrobacterium tumefaciens, have determined thermal relaxation times of Lumi-R, meta-Ra and meta-Rc states which are 0.23 ms, 3.1 ms and 260 ms, respectively (van Thor et al., 2001; Borucki et al., 2005). The formation of meta-Rc is accompanied by the release of a proton and then decay to Pfr with an associated proton uptake. The work suggests that Pr is in the ZZZasa conformation and Pfr in the ZZEssa conformation. This is at odds with the X-ray structure of the CBD of a Bph from D. radiodurans which is in the Pr state and clearly shows BV in the ZZZssa conformation (Wagner et al., 2005). Nuclear magnetic resonance measurements of a 13C-substituted chromophore in the truncated Cph1-515 photoreceptor domain shows a chemical shift of C4 in ring A, which is consistent with a hula-twist of C5 (van Thor et al., 2006) and indicates relaxation occurring at the methane bridge between pyrrole rings A and B. The nuclear magnetic resonance data are also consistent with a Pr-BV in the ZZZssa conformation. The plant PhyA meta-F state reveals a further relaxation of the C-D methine following isomerization (Matysik et al., 1995). The
A major breakthrough in phytochrome research has been the crystal structure determination of an Nterminal fragment which contains the chromophore binding domain (CBD) of a Bph from D. radiodurans (Wagner et al., 2005), and which constitutes 40% of a Bph structure. The BV chromophore is connected to the vinyl group of pyrrole A via Cys24, and is buried in a pocket within the GAF domain formed by a 6-stranded β-sheet and 2 α-helices (Fig. 6). An unusual feature is a trefoil knot formed by a peptide strand and the N-terminus passing through a loop of a β-sheet. This loop insertion is specific to phytochromes and helps to lock the relative orientations of the PAS and GAF domains. Mutations of conserved residues surrounding the chromophore indicate that these are essential, as would be expected. However other mutations disrupting function are at the interface between the PAS domain and the trefoil knot, and indicate the importance of this domain for stabilizing the BV pocket (Rockwell et al., 2006). In solution, the CBD fragment used in the structure determination is only stable in the Pr spectral form (Karniol et al., 2005), and therefore shows unambiguously that Pr, as observed by X-ray structure determination, is in a ZZZssa conformation. VI. Small Angle X-ray Scattering Solution Structure of Bph4 from Rhodopseudomonas palustris The low resolution structure of Bph4 from Rps. palustris has been determined in the catalytically active Pr state (Evans et al., 2006) by small angle Xray scattering (SAXS). The macromolecular envelope reveals a dimer at a resolution of 20 Å, which can be modeled using the atomic structures of the CBD from D. radiodurans (Wagner et al., 2005), and the cytoplasmic part of a type 1 histidine kinase (CHK) from Thermotoga maritima (Marina et al., 2005) comprising the Dhp and kinase CA domains. The
806
Katie Evans, Toni Georgiou, Theresa Hillon, Anthony Fordham-Skelton and Mirolav Papiz
Fig. 6. The X-ray structure of the chromophore binding domain (CBD) from D. radiodurans (Wagner et al., 2005). The PAS-2 and GAF domains are aligned, relative to one another, by a trefoil knot formed from a loop of the GAF domain and residues 1–35 N-terminal to PAS-2. BV is covalently bound to Cys24 and is located in a pocket within the GAF domain formed from a β-sheet and two α-helices. See also Color Plate 15, Fig. 24.
structures possess BV in the CBD, an ATP analog bound to the kinase domain (CA), and the autophosphorylating histidine in the Dhp domain. The missing 25% of the structure is the PHY domain for which a 3-D structure is not available. Bph4 is an oblate Y-shaped homo-dimer with the CBDs positioned at the extremes of the arms forming the Y (Fig. 7). The body of Bph4 comprises Dhp, formed from a 4 αhelix bundle, while CA domains form the base. The inter BV distance is ~112–117 Å, the BV to His532 distance is ~70–75 Å, and the distance between the ATP analog to His532 on the opposite protomer is 30–35 Å. An important feature is additional electron density at the Dhp domain which covers the phospho-acceptor residue His532. In the cytoplasmic histidine kinase domain, this residue is exposed, and a mechanism was proposed for auto-phosphorylation that required the CA domain to move through 25 Å to the phosphorylating site on the opposite protomer (Marina et al., 2005). The flexibility of CA is observed in a homolog, CheA (Bilwes et al., 1999), where the two CA conformations deviate from 2-fold symmetry by 25º. Unlike the crystal structure, the CA domains of the Bph4 solution structure are free to move and a
model can be devised, which accounts for the electron density over the phosphorylation site (His532), by assuming two conformations of the CA domain. The two conformers of the CA domain are called the ‘open’ and ‘closed’ states in the non-phosphorylating and phosphorylating state, respectively. The model clearly shows the long distances (~100 Å) over which the motion is transmitted to affect kinase activity. The structure indicates that the CA domain is mobile in the Pr form but it is unknown if it is restrained in the Pfr form, preventing access to the His532. The PHY domain orchestrates signal transduction and it appears that there are different interaction surfaces with CA in the closed and open states, which may explain how the Pfr state can be locked in the open state. An alternative model is that CA is mobile in both the Pr and Pfr states with His532 accessible only in the Pr state: schemes for histidine kinase activity have been proposed which require alterations to αhelix orientations within the Dhp 4-helix bundle that change histidine accessibility (Parkinson and Kofoid, 1992; Williams and Stewart, 1999).
Chapter 40
Bacteriophytochromes and Photosynthesis
807
Fig. 7. A small angle X-ray scattering (SAXS) ab initio envelope (transparent) of Bph4 from Rps. palustris (Evans et al., 2006). The structure is a homodimer and is modeled with the crystal structures of the CBD (upper lobes; Wagner et al., 2005) and histidine kinase (HK) (middle, and lower lobes; Marina et al., 2005). The positions of the BV, phosphorylating histidine and ATP are shown. There is no high resolution structure of a PHY domain that can be fitted, but it is assumed to occupy the envelope between CBD and the HK domains. See also Color Plate 15, Fig. 25.
VII. Conclusions The gene expression-controlling properties of Bphs arise from an ability to use light as an environmental signal rather than a source of physiological energy. These proteins therefore function as a ‘bacterial eye’ giving the ability to sense light, in particular the quality, intensity and, in the case of phototaxis, direction of light. The events that Bphs initiate are complex and may interact with other systems, as is the case for Rps. palustris Bph3 which binds the O2sensing repressor of photosynthesis PpsR2 (Rpa1536) (Braatsch et al., 2006). The evolution of Bphs may have arisen in systems directly connected to photosynthesis, but the presence of Bphs in non-photosynthetic bacteria suggests that a wider range of biochemical systems are influenced by light. In Rps. palustris the influences of Bphs on the expression profile of RC-LH1, LH2 and LH4 are complicated. Nevertheless, our current understanding shows that fine adjustments are employed, by this bacterium, to match the differing energy-trapping properties of its photosynthetic machinery to environmental conditions. Why so many Bphs exist in this bacterium is unknown, but Rps. palustris is metabolically versatile with many biosynthetic and
biodegradation pathways, which require optimal utilization of environmental resources that are directly and indirectly dependent on both the intensity and wavelength of light. Acknowledgments Katie Evans was supported by a Daresbury Laboratory/John Moores University (Liverpool, UK) Ph.D. scholarship. The authors acknowledge STFC for the use of Daresbury laboratory SRS facilities. References Andel F III, Murphy JT, Haas JA, McDowell MT, van der Hoef I, Lugtenburg J, Lagarias JC and Mathies RA (2000) Probing the photoreaction mechanism of phytochrome through analysis of resonance Raman vibrational spectra of recombinant analogues. Biochemistry 39: 2667–2676 Aravind L and Ponting CP (1997) The GAF domain: an evolutionary link between diverse phototransducing proteins. Trends Biochem Sci 22, 458–459 Bhoo SH, Davis SJ, Walker J, Karniol B and Vierstra RD (2001) Bacteriophytochromes are photochromic histidine kinases using a biliverdin chromophore. Nature 414: 776–779 Bilwes AM, Alex LA, Crane BR and Simon MI (1999) Struc-
808
Katie Evans, Toni Georgiou, Theresa Hillon, Anthony Fordham-Skelton and Mirolav Papiz
ture of CheA, a signal-transducing histidine kinase. Cell 96: 131–141 Borucki B, von Stetten D, Seibeck S, Lamparter T, Michael N, Mroginski, MA, Otto H, Murgida DH, Heyn MP and Hildebrandt P (2005) Light-induced proton release of phytochrome is coupled to the transient deprotonation of the tetrapyrrole chromophore. J Biol Chem 280: 34358–34364 Braatsch S, Bernstein JR, Lessner F, Morgan J, Liao JC, Harwood CS and Beatty JT (2006) Rhodopseudomonas palustris CGA009 has two functional ppsR genes, each of which encodes a repressor of photosynthesis gene expression. Biochemistry 45: 14441–14451 Braatsch S, Johnson JA, Noll K and Beatty JT (2007) The O2responsive repressor PpsR2 but not PpsR1 transduces a light signal sensed by the BphP1 phytochrome in Rhodopseudomonas palustris CGA009. FEMS Microbiol Lett 272: 60–64 Davis SJ, Vener AV and Vierstra RD (1999) Bacteriophytochromes: Phytochrome-like photoreceptors from nonphotosynthetic eubacteria. Science 286: 2517–2520 de Ruijter WP, Oellerich S, Segura JM, Lawless AM, Papiz M and Aartsma, TJ (2004) Observation of the energy-level structure of the low-light adapted B800 LH4 complex by single-molecule spectroscopy. Biophys J 87: 3413–3420 Evans K, Fordham-Skelton AP, Mistry H, Reynolds CD, Lawless AM and Papiz MZ (2005) A bacteriophytochrome regulates the synthesis of LH4 complexes in Rhodopseudomonas palustris. Photosynth Res 85: 169–180 Evans K, Grossmann JG, Fordham-Skelton AP and Papiz MZ (2006) Small-angle X-ray scattering reveals the solution structure of a bacteriophytochrome in the catalytically active Pr state. J Mol Biol 364: 655–666 Fiedler B, Borner T and Wilde A (2005) Phototaxis in the cyanobacterium Synechocystis sp PCC 6803: role of different photoreceptors. Photochem Photobiol 81: 1481–1488 Fischer AJ, Rockwell NC, Jang AY, Ernst LA, Waggoner AS, Duan Y, Lei H and Lagarias JC (2005) Multiple roles of a conserved GAF domain tyrosine residue in cyanobacterial and plant phytochromes. Biochemistry 44: 15203–15215 Fodor, S P, Lagarias, J C, and Mathies, R A (1988) Resonance Raman spectra of the Pr-form of phytochrome Photochem Photobiol 48: 129–136 Fodor SP, Lagarias JC and Mathies RA (1990) Resonance Raman analysis of the Pr and Pfr forms of phytochrome. Biochemistry 29: 11141–11146 Foerstendorf H, Lamparter T, Hughes J, Gartner W and Siebert F (2000) The photoreactions of recombinant phytochrome from the cyanobacterium Synechocystis: A low-temperature UV-Vis and FT-IR spectroscopic study. Photochem Photobiol 71: 655–661 Foerstendorf H, Benda C, Gartner W, Storf M, Scheer H and Siebert F (2001) FTIR studies of phytochrome photoreactions reveal the C=O bands of the chromophore: Consequences for its protonation states, conformation, and protein interaction. Biochemistry 40: 14952–14959 Giraud E, Fardoux J, Fourrier N, Hannibal L, Genty B, Bouyer P, Dreyfus B and Verméglio A (2002) Bacteriophytochrome controls photosystem synthesis in anoxygenic bacteria. Nature 417: 202–205 Giraud E, Hannibal L, Fardoux J, Jaubert M, Jourand P, Dreyfus B, Sturgis, JN and Verméglio, A (2004) Two distinct crt gene clusters for two different functional classes of carotenoid in
Bradyrhizobium. J Biol Chem 279: 15076–15083 Giraud E, Zappa S, Vuillet L, Adriano JM, Hannibal L, Fardoux J, Berthomieu C, Bouyer P, Pignol D and Verméglio A (2005) A new type of bacteriophytochrome acts in tandem with a classical bacteriophytochrome to control the antennae synthesis in Rhodopseudomonas palustris. J Biol Chem 280: 32389–32397 Gomelsky M and Kaplan S (1995) Genetic evidence that PpsR from Rhodobacter sphaeroides 2.4.1 functions as a repressor of puc and bchF expression. J Bacteriol 177: 1634–1637 Gomelsky M, Horne IM, Lee HJ, Pemberton JM, McEwan AG and Kaplan S (2000) Domain structure, oligomeric state, and mutational analysis of PpsR, the Rhodobacter sphaeroides repressor of photosystem gene expression. J Bacteriol 182: 2253–2261 Hartigan N, Tharia HA, Sweeney F, Lawless AM and Papiz MZ (2002) The 7.5-Å electron density and spectroscopic properties of a novel low-light B800 LH2 from Rhodopseudomonas palustris. Biophys J 82: 963–977 Heyne K, Herbst J, Stehlik D, Esteban B, Lamparter T, Hughes J and Diller R (2002) Ultrafast dynamics of phytochrome from the cyanobacterium Synechocystis, reconstituted with phycocyanobilin and phycoerythrobilin. Biophys J 82: 1004–1016 Hughes J, Lamparter T, Mittmann F, Hartmann E, Gartner W, Wilde A, and Borner T (1997) A prokaryotic phytochrome. Nature 386: 663 Hughes J and Lamparter T (1999) Prokaryotes and phytochrome. The connection to chromophores and signaling. Plant Physiol 121: 1059–1068 Jaubert M, Zappa S, Fardoux J, Adriano JM, Hannibal L, Elsen S, Lavergne J Verméglio A, Giraud E and Pignol D (2004) Light and redox control of photosynthesis gene expression in Bradyrhizobium: dual roles of two PpsR. J Biol Chem 279: 44407–44416 Jiang Z, Swem LR, Rushing BG, Devanathan S, Tollin G and Bauer CE (1999) Bacterial photoreceptor with similarity to photoactive yellow protein and plant phytochromes. Science 285: 406–409 Karniol B, Wagner JR, Walker JM and Vierstra RD (2005) Phylogenetic analysis of the phytochrome superfamily reveals distinct microbial subfamilies of photoreceptors. Biochem J 392: 103–116 Kneip C, Hildebrandt P, Schlamann W, Braslavsky SE, Mark F and Schaffner K (1999) Protonation state and structural changes of the tetrapyrrole chromophore during the Pr → Pfr phototransformation of phytochrome: A resonance Raman spectroscopic study. Biochemistry 38: 15185–15192 Lagarias JC and Lagarias DM (1989) Self-assembly of synthetic phytochrome holoprotein in vitro. Proc Natl Acad Sci USA 86: 5778–5780 Lamparter T, Michael N, Mittmann F and Esteban B (2002) Phytochrome from Agrobacterium tumefaciens has unusual spectral properties and reveals an N-terminal chromophore attachment site. Proc Natl Acad Sci U S A 99: 11628–11633 Lamparter T, Michael N, Caspani O, Miyata T, Shirai K and Inomata K (2003) Biliverdin binds covalently to Agrobacterium phytochrome Agp1 via its ring A vinyl side chain. J Biol Chem 278: 33786–33792 Larimer FW, Chain P, Hauser L, Lamerdin J, Malfatti S, Do L, Land ML, Pelletier DA, Beatty JT, Lang, AS et al. (2004) Complete genome sequence of the metabolically versatile
Chapter 40
Bacteriophytochromes and Photosynthesis
photosynthetic bacterium Rhodopseudomonas palustris. Nat Biotechnol 22: 55–61 Li L and Lagarias JC (1992) Phytochrome assembly. Defining chromophore structural requirements for covalent attachment and photoreversibility. J Biol Chem 267: 19204–19210 Lohrmann J and Harter K (2002) Plant two-component signaling systems and the role of response regulators. Plant Physiol 128: 363–369 Maloof JN, Borevitz JO, Weigel D and Chory J (2000) Natural variation in phytochrome signaling. Semin Cell Dev Biol 11: 523–530 Marina A, Waldburger CD and Hendrickson WA (2005) Structure of the entire cytoplasmic portion of a sensor histidine-kinase protein. EMBO J 24: 4247–4259 Matysik J, Hildebrandt P, Schlamann W, Braslavsky SE and Schaffner K (1995) Fourier-transform resonance Raman spectroscopy of intermediates of the phytochrome photocycle. Biochemistry 34: 10497–10507 Otto H, Lamparter T, Borucki B, Hughes J, and Heyn MP (2003) Dimerization and inter-chromophore distance of Cph1 phytochrome from Synechocystis, as monitored by fluorescence homo and hetero energy transfer. Biochemistry 42: 5885–5895 Parkinson JS and Kofoid EC (1992) Communication modules in bacterial signaling proteins. Annu Rev Genet 26: 71–112 Rockwell NC, Su YS and Lagarias JC (2006) Phytochrome structure and signaling mechanisms. Annu Rev Plant Biol 57: 837–858 Stowe-Evans EL and Kehoe DM (2004) Signal transduction during light-quality acclimation in cyanobacteria: A model system for understanding phytochrome-response pathways in prokaryotes. Photochem Photobiol Sci 3: 495–502 Taylor BL and Zhulin IB (1999) PAS domains: Internal sensors of oxygen, redox potential, and light. Microbiol Mol Biol Rev
809
63: 479–506 Tharia HA, Nightingale TD, Papiz MZ and Lawless AM (1999) Characterisation of hydrophobic peptides by RP-HPLC from different spectral forms of LH2 isolated from Rps palustris. Photosynth Res 61: 157–167 van Thor JJ, Borucki B, Crielaard W, Otto H, Lamparter T, Hughes J, Hellingwerf KJ and Heyn MP (2001) Light-induced proton release and proton uptake reactions in the cyanobacterial phytochrome Cph1. Biochemistry 40: 11460–11471 van Thor JJ, Mackeen M, Kuprov I, Dwek RA and Wormald MR (2006) Chromophore structure in the photocycle of the cyanobacterial phytochrome Cph1. Biophys J 91: 1811–1822 Wagner JR, Brunzelle JS, Forest KT and Vierstra RD (2005) A light-sensing knot revealed by the structure of the chromophorebinding domain of phytochrome. Nature 438: 325–331 Williams SB and Stewart V (1999) Functional similarities among two-component sensors and methyl-accepting chemotaxis proteins suggest a role for linker region amphipathic helices in transmembrane signal transduction. Mol Microbiol 33: 1093–1102 Yeh KC and Lagarias JC (1998) Eukaryotic phytochromes: Lightregulated serine/threonine protein kinases with histidine kinase ancestry. Proc Natl Acad Sci USA 95: 13976–13981 Yeh KC, Wu SH, Murphy JT and Lagarias JC (1997) A cyanobacterial phytochrome two-component light sensory system. Science 277: 1505–1508 Yoshihara S, Katayama M, Geng X and Ikeuchi M (2004) Cyanobacterial phytochrome-like PixJ1 holoprotein shows novel reversible photoconversion between blue- and green-absorbing forms. Plant Cell Physiol 45: 1729–1737 Zhulin IB, Taylor BL and Dixon R (1997) PAS domain S-boxes in Archaea, Bacteria and sensors for oxygen and redox. Trends Biochem Sci 22: 331–333
Chapter 41 Photoreceptor Proteins from Purple Bacteria Johnny Hendriks1, Michael A. van der Horst1, Toh Kee Chua2, Marcela Ávila Pérez1, Luuk J. van Wilderen2, Maxime T.A. Alexandre 2, Marie-Louise Groot 2, John T. M. Kennis2 and Klaas J. Hellingwerf 1,2* 1
Microbial Physiology Group, Swammerdam Institute for Life Sciences, BioCentrum, University of Amsterdam, The Netherlands; 2Biophysics Group, Department of Physics and Astronomy, Vrije Universiteit Amsterdam, The Netherlands
Summary ............................................................................................................................................................... 811 I. Introduction..................................................................................................................................................... 812 II. Light, Oxygen, or Voltage Domains................................................................................................................ 813 A. Occurrence, Structure and Function of Light, Oxygen, or Voltage Domains.................................... 813 B. (Ultra)fast Spectroscopy and the LOV Reaction Mechanism ........................................................... 814 C. Signaling State Formation in LOV Domains..................................................................................... 815 III. The BLUF Domain Containing Family of Photoreceptors .............................................................................. 816 A. Occurrence, Structure and Biological Function ................................................................................ 816 B. (Ultra)fast Spectroscopy and the BLUF Reaction Mechanism ......................................................... 820 C. Signaling State Formation in BLUF Domains................................................................................... 821 IV. Comparison Between LOV and BLUF Domains ............................................................................................ 822 V. The Xanthopsins ............................................................................................................................................ 823 A. Introduction and Function ................................................................................................................. 823 B. (Ultra)fast Spectroscopy ................................................................................................................... 824 C. Xanthopsin Signaling ....................................................................................................................... 825 VI. Bacteriophytochromes.................................................................................................................................... 827 A. Occurrence, Structure and Function of (Bacterio)phytochromes ..................................................... 827 B. (Ultra)Fast Spectroscopy of Bacteriophytochromes ......................................................................... 829 C. Signaling-State Formation in Bacteriophytochromes ....................................................................... 830 VII. Concluding Remarks ...................................................................................................................................... 831 Acknowledgments ................................................................................................................................................. 832 References ............................................................................................................................................................ 832
Summary Purple bacteria contain representatives of four of the six main families of photoreceptor proteins: phytochromes, BLUF domain containing proteins, xanthopsins (i.e., photoactive yellow proteins), and phototropins (containing one or more light, oxygen, or voltage (LOV) domains). Most of them have a function in adjusting the cellular transcript profile to the ambient light climate. Here we will discuss, with examples of the most important representative(s) of each of these four families, the interdependent topics: (i) the proteins’ biological functions and their molecular context, (ii) the results of ultra-fast and static spectroscopic studies, and (iii) structural alterations required for initiation of signal transfer, as resolved with transient spectroscopy and the methodology of structural biology. *Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 811–837. © 2009 Springer Science + Business Media B.V.
812
J. Hendriks, M. A. van der Horst, T. K. Chua, M. Á. Pérez, L. J. van Wilderen, M. T. A. Alexandre, M.-L. Groot, J. T. M. Kennis and K. J. Hellingwerf
For all four of these photoreceptor protein families, detailed insight is available about the structural basis of signal generation in the light-input domain. The next main challenge is to understand how this information is transmitted to the downstream partner in the signal transduction chain. This determination will generally require more detailed insight into the spatial arrangement of multi-domain proteins. I. Introduction All phototrophic organisms, including purple bacteria, are confronted with daily variations in the intensity and color of the available amount of ambient electromagnetic radiation in the visible part of the energy spectrum. They will use this radiation primarily for energy conversion in photosynthesis. However, because of these daily variations, the photosynthesis process requires continuous adaptation, e.g., with respect to the reaction center/antenna pigment ratio. And in many organisms even much more profound adaptations are required, like the circadian on/off switching of primary metabolic pathways. The nitrogenase-catalyzed fixation of ammonia is one of the main examples of the latter. Besides its function as an energy source, visible radiation is also a rich source of information for the above-mentioned physiological adaptations. Accordingly, phototrophic bacteria — and other bacteria (see below) — have developed a wide variety of photoreceptor proteins that initiate regulatory cascades in response to (changes in) the ambient light climate. These regulatory cascades may lead to the adjustment of enzyme activity and/or gene-expression level. Mostly, these responses are a direct result of alteration in the intensity and/or color of the light available, but also more advanced responses in which the organism can sense the degree of polarization of the light and/or the direction from which the light is emitted have developed. Abbreviations: BLUF – sensors of blue light by using flavin adenine dinucleotide; CBD – chromophore-binding domain; c-di-GMP – cyclic diguanosine monophosphate; E – Escherichia; EAL – glutamate-alanine-leucine; FAD – flavin adenine dinucleotide; FMN – flavin mononucleotide; FTIR – Fourier transform infrared; GAF domain – domain present in phytochromes and cyclic guanosine monophosphate specific phosphodiesterases; H-bonds – hydrogen bonds; Hlr – Halorhodospira; ISC – intersystem crossing; LOV – light, oxygen, or voltage; MCP – methyl-accepting chemotaxis protein; NMR – nuclear magnetic resonance; PAS – acronym formed from the names of three proteins: Per, Arnt and Sim; phot – phototropin; PHY – phytochrome; phy3 – phytochrome 3; PLD – PAS like domain; PYP – photoactive yellow protein; Rba. – Rhodobacter; Rps. – Rhodopseudomonas; Rvi. – Rubivivax; STAS – sulfate transporter and anti-sigma factor antagonist; UV/Vis – ultraviolet and visible
Photoreceptor proteins therefore are an important target in our quest to understand the physiological regulation in phototrophic organisms. In addition, they present a challenge for structural characterization because besides the apo-protein they must contain an additional, highly conjugated, elemental chemical structure (usually referred to as the chromophore) that will facilitate light absorption. On the other hand, the light-absorbing chromophore allows the application of a wide range of (ultra-fast) spectroscopic and structural techniques to study the dynamical changes in both chromophore and protein structure, which are of critical importance for understanding photoreceptor function. This latter aspect has brought such progress that the photoreceptor proteins have even become ‘star actors’ in studies on structure/function relationships of proteins in general. The various visual rhodopsins are good examples; we not only know their structure, but also how this structure relates to function. Genome sequencing during the last few years has brought yet one additional dimension to photoreceptor research. Many genomes from chemotrophic organisms also encode one or more photoreceptor proteins (see e.g., the phylogenetic distribution of the LOV domains). Subsequent physiological studies are now beginning to reveal the physiological role of some of these photosensing systems, thereby further enhancing the scientific impact of studies of photoreceptor function. Six main photoreceptor families, each with a unique type of structure and primary photochemistry, have so far been characterized. These are the: (1) rhodopsins, (2) phytochromes, (3) xanthopsins (i.e., photoactive yellow proteins), (4) cryptochromes, (5) phototropins (containing one or more LOV domains), and (6) BLUF domain containing photoreceptors. Representatives of these photoreceptor families in purple bacteria have convincingly been demonstrated only for families 2, 6, 3 and 5 (in decreasing order of quantitative importance). The only prokaryotic protein claimed so far to be a cryptochrome was identified in the cyanobacterium Synechocystis (note, however, that this protein may well be a photolyase instead). Also rhodopsins might be present in purple bacteria: a gene encoding a proteorhodopsin has been
Chapter 41
Photoreceptor Proteins from Purple Bacteria
identified in a large genomic fragment of an organism that shows striking homology to Allochromatium vinosum, but the exact host organism remains to be characterized. In this chapter we will only discuss the (bacterio)phytochromes, the BLUF domain containing photoreceptors, the xanthopsins, and the LOV domain containing bacterial phototropins. For (a) representative member(s) of each of these four families we will discuss the interdependent issues of: (1) (biological) function and molecular context, (2) ultra-fast and static spectroscopic studies, and (3) structural alterations required for (initiation of) signal transfer, as resolved with transient spectroscopy and the methodology of structural biology. II. Light, Oxygen, or Voltage Domains A. Occurrence, Structure and Function of Light, Oxygen, or Voltage Domains Light, Oxygen or Voltage (LOV) domains were first identified in the phototropin (phot) family of plant blue-light photoreceptors. The phototropins are serine/threonine kinases that undergo autophosphorylation in response to absorption of blue light. Two LOV domains, LOV1 and LOV2, constitute the light-sensitive input domains of the phot photoreceptors. Both LOV1 and LOV2 bind a flavin mononucleotide (FMN) chromophore. Higher plants utilize two phototropins, phot1 and phot2, to elicit a variety of physiological responses including phototropism, light-mediated chloroplast movement and stomatal opening (Christie and Briggs, 2005). The lower plant chimeric photoreceptor phytochrome 3 (phy3) contains both a phytochrome-derived GAF domain (domain present in phytochromes and cyclic guanosine monophosphate specific phosphodiesterases) and a phot LOV domain, and is sensitive to blue and red light simultaneously (Suetsugu and Wada, 2005). LOV-based photoreceptors were also discovered in eukaryotic microorganisms. Chlamydomonas reinhardtii possesses a phototropin homolog involved in its sexual life cycle (Huang and Beck, 2003). In Neurospora crassa, a filamentous fungus, White Collar 1 (WC1) is involved in circadian cycling. Its LOV domain binds flavin adenine dinucleotide (FAD) instead of FMN. VIVID (VVD), a single short LOV domain without a readily detectable output module, modulates light responses in Neurospora, result-
813
ing in photoadaptation (Dunlap and Loros, 2005). Analysis of prokaryotic genomes has revealed several additional LOV domain containing proteins. These ‘LOV proteins’ were primarily found in heterotrophic bacteria (e.g., Bacillus, Pseudomonas, etc.) and in a limited number of phototrophic bacteria. They are linked to a large variety of output domains such as histidine kinases, glutamate-alanine-leucine (EAL) domains, DNA binding domains, etc. (Crosson et al., 2003; Crosson, 2005; Losi, 2006). The Rhodobacter (Rba.) sphaeroides and Chloroflexus aurantiacus genomes encode a short LOV protein without an output domain, whereas in the genome of Rubrivivax (Rvi.) gelatinosus and in Roseobacter denitrificans a multi-domain LOV-kinase-response regulator was identified. Strikingly, none of the prokaryotic LOV proteins contains more than one LOV domain. A number of prokaryotic LOV proteins have been cloned and characterized. All displayed typical LOV photochemistry (Losi, 2006). The YtvA protein of Bacillus subtilis, consisting of a LOV and a STAS (sulfate transporter and antisigma factor antagonist) domain, was shown to be involved in a light-dependent stress response (Avila-Pérez et al., 2006). A physiological function has not been established for any of the other prokaryotic LOV domains. At present, three LOV domain structural models are available: X-ray structures of Adiantum (Maidenhair fern) phy3 LOV2 (Crosson, 2005), Chlamydomonas phot LOV1 (Fedorov et al., 2003), and a nuclear magnetic resonance (NMR) structure of Avena sativa (oat) phot1 LOV2 (Harper et al., 2003). The structures revealed a typical PAS (Per, Arnt, Sim)-fold, consisting of a five-stranded antiparallel β-sheet, flanked by a helix-turn-helix motif, a single helical turn, and a connector helix. A close-up of the FMN binding pocket is shown in Fig. 1a. The FMN is noncovalently held in place by polar interactions with its pyrimidine moiety, nonpolar interactions with the dimethylbenzene ring, and via a number of H-bonds with the apoprotein. A conserved cysteine is located within 4 Å of the FMN isoalloxazine ring. The receptor state of LOV domains, generally referred to as D447, exhibits an absorption with electronic transitions near 450 nm and 360 nm. Upon blue-light illumination, LOV domains undergo a photocycle that leads to the formation of a long-lived species with a prominent absorption band at 390 nm, referred to as S390. Various lines of evidence have convincingly shown that S390 corresponds to a state wherein the FMN has formed a covalent C(4a)-thiol adduct with
814
J. Hendriks, M. A. van der Horst, T. K. Chua, M. Á. Pérez, L. J. van Wilderen, M. T. A. Alexandre, M.-L. Groot, J. T. M. Kennis and K. J. Hellingwerf
Fig. 1. LOV photocycle and active site structure. a: Active site receptor state. b: Active site signaling state. c: Typical photocycle of a LOV domain. Panels a and b were prepared using the program PyMOL (http://www.pymol.org). The structure coordinate file of LOV2 from Adiantum capillus-veneris with PDB ID: 1G28 (Crosson and Moffat, 2001) was used to create panel a, and with PDB ID: 1JNU (Crosson and Moffat, 2002) to create panel b.
the conserved cysteine, as shown in Fig. 1b (Fedorov et al., 2003; Crosson, 2005; Swartz and Bogomolni, 2005). B. (Ultra)fast Spectroscopy and the LOV Reaction Mechanism Femtosecond to nanosecond transient absorption and time-resolved fluorescence experiments performed on various LOV domains indicated a single-exponential decay (2–3 ns) of the FMN singlet excited state. The singlet-excited state was found to evolve into the FMN triplet excited state, with an estimated quantum
yield of 0.6 to 0.8. No additional time constants or spectral evolution were observed, which indicates that the FMN triplet states are formed directly from the FMN singlet excited state via intersystem crossing (ISC) (Holzer et al., 2002; Schuttrigkeit et al., 2003; Kennis and Alexandre, 2006). Remarkably, despite a two-fold shortening of the singlet-excited state lifetime of LOV-bound FMN as compared to solution (4.7 ns), the triplet yield was essentially the same. This implied that the ISC rate in LOV domains is enhanced, which presumably results from the cysteine’s sulfur in close vicinity to FMN (Holzer et al., 2002; Kennis and Alexandre, 2006). Further evidence that the primary photoproduct corresponds to an FMN triplet state comes from femtosecond to nanosecond infrared spectroscopy, where frequency shifts of the FMN C=O and C=N stretches were observed that corresponded well with those of a flavin model compound triplet state (Kennis and Alexandre, 2006). Nanosecond to microsecond spectroscopy on various LOV domains indicated formation of the FMN triplet at high yields, consistent with the ultrafast results. The FMN triplet was found to evolve into the covalent FMN-cysteinyl adduct state S390 on a µs timescale without an apparent intermediate (Kottke et al., 2003a; Swartz and Bogomolni, 2005; Losi, 2006). The quantum yield of covalent FMN-cysteinyl adduct formation varies widely among LOV domains, with reported values of 0.05 for phot1 LOV1, 0.30–0.45 for phot1 LOV2, and 0.6 – 0.8 for Chlamydomonas LOV1 (Kasahara et al., 2002; Kottke et al., 2003a; Kennis and Alexandre, 2006; Losi, 2006). Near-ultraviolet light is capable of breaking the covalent adduct and regenerating the dark state D447, as shown by femtosecond transient absorption spectroscopy on the photoaccumulated adduct state of phot1 LOV2. Dark-state regeneration was complete within 100 ps, at an estimated quantum yield of 0.2–0.25 (Kennis and Alexandre, 2006). A similar light-driven back reaction was proposed for the Chlamydomonas LOV1 and LOV2 domains (Kottke et al., 2003a; Holzer et al., 2005). Thus, LOV domains act as reversible photochromic switches, a property shared with other classes of photoreceptor proteins such as rhodopsins, phytochromes and xanthopsins (van der Horst and Hellingwerf, 2004). However, the physiological significance of the light-driven adduct cleavage remains uncertain. In contrast to a consensus on spectroscopically distinguishable intermediates in the LOV photocycle, the mechanism by which covalent FMN-cysteinyl ad-
Chapter 41
Photoreceptor Proteins from Purple Bacteria
duct formation proceeds is a matter of considerable debate. Broadly speaking, two reaction mechanisms have been put forward: an ionic and a radical-pair mechanism. According to the ionic model, the sharply increased basicity of the flavin N(5) upon promoting FMN to the triplet state triggers its protonation. The proton would either be donated by the conserved cysteine itself (Fedorov et al., 2003; Crosson, 2005; Kennis and Alexandre, 2006) or by another nearby group (Swartz and Bogomolni, 2005). This event would change the double-bond of N(5)=C(4a) to a single bond, leaving a very reactive carbo-cation at the C(4a) position. This site would have an sp3 hybridization, which would decrease the distance to the cysteine and could be important in the progress of the reaction. A nucleophilic attack by the cysteine thiolate on the C(4a) carbo-cation would follow, leading to formation of the covalent FMN-C(4a)-thiol adduct. These events could occur sequentially (Kennis and Alexandre, 2006) or in a concerted fashion (Fedorov et al., 2003; Crosson, 2005). Evidence for the ionic mechanism was presented by Kennis and Alexandre (2006), who interpreted the FMN triplet spectra in LOV2 as showing a partially protonated N(5). This protonation would occur on a nanosecond timescale. Further evidence for the ionic model was provided by quantum-chemical calculations which showed that upon excitation of the FMN triplet, the N(5) atom of FMN could abstract a proton from proximal cysteine conformers (Fedorov et al., 2003). Kay et al. (2003) noted that with an ionic mechanism, the covalent adduct has to be formed in the triplet state which would be energetically unfavorable. Instead, they proposed a radical-pair mechanism. Upon promotion of FMN to the triplet state, a hydrogen would be transferred from the cysteine to the N(5) of the flavin, resulting in a FMNH•–S•–CH2– radical pair, with the unpaired electron density on FMN residing at C(4a). Such a radical pair would be formed in a triplet state; however, the close proximity of the (heavy-atom) sulfur radical to FMN causes a strong spin-orbit coupling, inducing a rapid triplet-singlet interconversion. Once the radical pair obtains an appreciable singlet character, radical-pair recombination may take place to result in the FMN-C(4a)-thiol adduct. The trigger for adduct formation could either be electron transfer from cysteine to FMN, followed by proton transfer, or a concerted mechanism with a net hydrogen transfer from the thiol to FMN (Kay et al., 2003; Kottke et al., 2003b; Schleicher et al., 2004).
815
The thermal recovery of the dark state proceeds rather slowly in LOV domains, and varies between a few seconds for phot2 LOV2, minutes in phot1 LOV domains, to several hours for Bacillus subtilis YtvA (Kasahara et al., 2002; Kottke et al., 2003a; Swartz and Bogomolni, 2005; Losi, 2006). A base-catalyzed mechanism for adduct cleavage was proposed, in which the proton at the N(5) atom of FMN would be abstracted by a basic group (Fedorov et al., 2003; Kottke et al., 2003a; Swartz and Bogomolni, 2005). The adduct decay is slowed down by a factor of 3 upon H/D exchange, suggesting that the active base could be in contact with FMN via a network of hydrogen bonds. The pronounced pH dependence of adduct decay in Chlamydomonas LOV1 and phot1 LOV2 supports the base-catalyzed hypothesis, and histidines at the surface of the domain and the phosphate group of FMN were suggested to act as the base. Further support was provided by the observation that chemical modification of surface histidines in LOV2 leads to a significant increase of the recovery time. Moreover, the addition of millimolar amounts of imidazole greatly enhanced the dark recovery rate, presumably by base abstraction at the FMN N(5) position by imidazole (Alexandre et al., 2007). C. Signaling State Formation in LOV Domains The crystal structures of phy3 LOV2 and Chlamydomonas LOV1 indicated little overall structural changes upon cysteinyl-FMN adduct formation (Fedorov et al., 2003; Crosson, 2005). The changes are mainly confined to the immediate environment of the flavin chromophore, and involve an altered orientation of a conserved glutamine; this glutamine (i.e., Q1029 in phy3 LOV2) becoming a H-bond acceptor of the FMN N5 proton, rather than a Hbond donor to the C(4)=O carbonyl, as shown in Fig. 1a,b. In phy3 LOV2 a conserved volume was identified that stretched from FMN to a conserved salt bridge at the surface of the domain. The residues constituting the conserved volume showed a small but significant concerted movement towards the surface upon photoactivation. A light-induced modulation of the salt bridge could influence the binding affinity for an interaction partner, either by conformational changes at the surface or entropically by changes of molecular flexibility (Crosson et al., 2003). Using NMR spectroscopy, a C-terminal helical secondary structure termed Jα was identified outside the PAS core in Avena sativa phot1 LOV2. This helix,
816
J. Hendriks, M. A. van der Horst, T. K. Chua, M. Á. Pérez, L. J. van Wilderen, M. T. A. Alexandre, M.-L. Groot, J. T. M. Kennis and K. J. Hellingwerf
40 amino acids long, is amphipatic and binds to the solvent exposed β-sheet surface of LOV2, burying hydrophobic surfaces on both interaction partners. No such C-terminal helical structure is found for LOV1. Upon illumination of the LOV2 domain, the LOV2/Jα interaction was perturbed whereby Jα lost its secondary structure and partially unfolded (Harper et al., 2003). These observations were in agreement with circular dichroism studies which indicated a reversible loss of α-helicity (Swartz and Bogomolni, 2005) and with Fourier Transform Infrared (FTIR) studies which showed a loosening of helical H-bonds and a tightening of the β-sheet (Iwata et al., 2003) during the photocycle. Such changes in the tertiary structure following signaling-state (S390) formation were not detected with transient Ultraviolet/Visible (UV/Vis) spectroscopy (Kottke et al., 2003a; Guo et al., 2005; Swartz and Bogomolni, 2005), because only the physico-chemical state of the FMN is probed with this technique. Utilizing transient grating spectroscopy a change of diffusion coefficient occurring on a ms timescale after photoactivation was detected in a phot1 LOV2/Jα construct, which was interpreted as a volume increase resulting from Jα unfolding (Eitoku et al., 2005). In LOV1, only signals that could be assigned to FMN or its immediate vicinity were detected with FTIR spectroscopy (Ataka et al., 2003; Iwata et al., 2005). Moreover, a volume contraction was detected upon cysteinyl-FMN adduct formation (Losi, 2006). The absence of extensive structural dynamics in LOV1 as compared to LOV2 very likely is related to its yet unknown function. Recently, molecular dynamics simulations revealed significant differences in the response of LOV1 and LOV2 domains upon activation (Freddolino et al., 2006): in LOV1, activation presumably results in destabilization of the highly conserved salt bridge, whereas in LOV2 activation results in a changed flexibility in various protein loops, leading to disruption of the interactions with the Jα helix and ultimately melting of the Jα helix. Nozaki et al. (2004) showed that replacement of the conserved glutamine (Q1029) in LOV2 (shown in Fig. 1a,b) leads to cancellation of the light-induced conformational change of the α-helices and β-sheets, as compared to wild type. FMN is sandwiched between the helix containing the conserved cysteine and the β-strand that contains the conserved glutamine. Thus, the FMN-cysteinyl adduct formation causes the H-bond switch of the conserved glutamine, and it is conceivable that this event in turn leads to β-sheet
motion and subsequent Jα unbinding and unfolding by loss of van der Waals complementarities (Harper et al., 2004; Nozaki et al., 2004). In vivo, support for the functional role of Jα unbinding has been obtained (Harper et al., 2004). In vitro, relay of conformational change from an N-terminal LOV domain to a C-terminal STAS domain has been demonstrated through altered fluorescence of a modified guanine nucleotide bound to the STAS domain (Buttani et al., 2006). It has been argued, based on the analysis of tryptophan fluorescence, that the region which is C-terminal to the LOV domain is involved in this signal relay in two bacterial LOV domain containing proteins, i.e., YtvA and in PpSB2-LOV from Pseudomonas putida (Losi, 2006). Figure 1c depicts a summarizing view of the photocycle scheme of LOV domains. Upon photon absorption by FMN, ISC to the triplet state (FMNT) takes place in a few nanoseconds at a substantial quantum yield. Next, the covalent FMN-cysteinyl adduct S390 is formed on a microsecond timescale by either an ionic or radical-pair mechanism. Although no absorbance changes can be observed in the UV/Vis between S390 formation and thermal decay back to the dark state D447, a putative intermediate S390’ is included and may represent unfolding of the Jα helix on a millisecond timescale. Such intermediate would only be relevant for LOV2 and LOV2-like domains. Sophisticated time-resolved methods will be required to unambiguously resolve the dynamic functional-structural changes that follow covalent FMN-cysteinyl adduct formation in LOV domains. III. The BLUF Domain Containing Family of Photoreceptors A. Occurrence, Structure and Biological Function The BLUF (for: ‘sensors of blue-light using FAD’) domain is about 90-amino acids in size and binds FAD non-covalently in vivo (Gomelsky and Klug, 2002; Hasegawa et al., 2006); upon heterologous expression in Escherichia (E.) coli it may bind FMN or riboflavin instead (Laan et al., 2004; Ito et al., 2005). Using its conserved sequence and the BLASTp search software, one finds this domain mainly in two branches of the Bacteria: the Proteobacteria (more specifically in the purple-non-sulfur bacteria, the enterobacteria and in unclassified proteobacteria) and
Chapter 41
Photoreceptor Proteins from Purple Bacteria
in the Cyanobacteria. In the domain of the Eucarya, BLUF domains have been found in unicellular flagellates such as Euglena and Eutreptia and recently also in the basidiomycete Ustilago maydis 521 (Table 1). BLUF domain containing proteins can be divided into ‘complex’, multi-domain proteins, and ‘short’ proteins, composed of a BLUF domain plus 30 to 100 additional amino acids, without significant sequence similarity to known protein motifs. In complex BLUF domain containing proteins the BLUF domain can transfer the information that it has absorbed a blue photon, to a limited number of output domains, i.e., the adenylate cyclase domain in Eucarya, and in
817
Bacteria to (a) PAS domain(s), a protein-interaction domain showing similarities with vitamin B12-binding proteins, or to a GGDEF or EAL domain, involved in cyclic diguanosine monophosphate (c-di-GMP) synthesis and hydrolysis, respectively (Gomelsky and Klug, 2002; Romling et al., 2005). Among the BLUF domain containing proteins only AppA and BlrB from Rba. sphaeroides, Slr1694 from Synechocystis sp. PCC 6803, Tll0078 from Thermosynechococcus elongatus, YcgF from E. coli, and the α and β PAC subunits from adenylate cyclase from Euglena gracilis have been partially characterized with respect to biological function.
Table 1. Domain architecture, abundance, function and primary photochemistry of BLUF containing proteins. PDE-A: c-di-GMP phosphodiesterase. N.A.: Not available. LivK: domain that contains the periplasmic binding protein involved in the ABC-type branched-chain amino acid transport system. The basidiomycete Ustilago Maydis 521 BLUF domain shows 36% amino acid identity (E value Blastp = 2·10–7) with the BLUF domain of AppA and 40% (E value Blastp = 6·10–9) with PACβ-F1 from Euglena gracilis. Domain Architecture
Occurence Unicellular flagellates: Colacium sideropus ; Euglena gracilis, longa, quartana, atellata ; Eutreptia viridis ; Eutreptiella gymnastica Purple bacteria: Rba. sphaeroides Klebsiella pneumoniae, Alteromonas macleodii ‘Deep ecotype’ and the enterobacteria: E. coli and Shigella boydii
+ 30 – 200 aa
Purple bacteria: Rba. sphaeroides, Rps. palustris, Rvi. gelatinosus and additional α and β proteobacteria. Basidiomycete: Ustilago maydis 521 (fungal pathogen of maize), Cyanobacteria: Theromosynechococcus elongatus BP1, Synechocystis spp.
Magnetococcus sp. mc-1
Example(s) (Protein & Organism) Photoactivated adenylyl cyclase (PAC) α and β subunit. Euglena gracilis
Function or Speculated Function Phototaxis (Ntefidou et al., 2003)
AppA (450 aa). Rba. sphaeroides
Transcriptional anti- ~1000 represor (Masuda and (Laan et Bauer, 2002) al., 2003; Hasegawa et al., 2006) Predicted to be a light ~600 regulated PDE-A (Rajagopal et (Rajagopal et al., al., 2004) 2004)
YcgF. E. coli
BLUF recovery time (s) 34–44 (Ito et al., 2005; Hasegawa et al., 2006)
Slr1694 (150 aa) Synechocystis sp. PCC6803 BlrB (140 aa) Rba. sphaeroides BlrA (136 aa) Rba. sphaeroides
Phototaxis (Fiedler et 7.2 (Okajima al., 2005) et al., 2005; Hasegawa et N.A. al., 2006)
Single example so far
N.A.; the EAL and GGDEF domain are involved in c-di-GMP turnover (Romling et al., 2005)
N.A.
2 (Zirak et al., 2006) No FAD binding (Jung et al., 2005) N.A.
818
J. Hendriks, M. A. van der Horst, T. K. Chua, M. Á. Pérez, L. J. van Wilderen, M. T. A. Alexandre, M.-L. Groot, J. T. M. Kennis and K. J. Hellingwerf
The photochemistry and mechanism of signaling state formation for these proteins is under intense investigation in several different labs. AppA is composed of an N-terminal light sensing BLUF domain and a C-terminal domain that is required for redox regulation and for the transmission of redox and light signals (Han et al., 2004; Chapter 35, Bauer et al.). Furthermore, part of this C-terminal domain is similar to vitamin B12 binding domains. Indeed AppA has a low affinity for vitamin B12, although it has a higher affinity for the structurally similar tetrapyrrole haem (Han et al., 2007). The possible role of heme in the function of AppA is still unclear, but it likely plays a role in the redox-sensing functionality of AppA and may (in)directly be involved in the interaction of the C-terminus with the BLUF domain. During the past two years the molecular structures of several BLUF domains have been reported. The structure of the BLUF domain of AppA (Anderson et al., 2005), BlrB (Jung et al., 2005) and Tll0078 (Kita et al., 2005) have been determined with X-ray diffraction experiments and for the corresponding domain from AppA an NMR structure is also available (Grinstead et al., 2006a). All BLUF domains show the same overall fold: an α/β-sandwich fold with βαββαββ topology. The structure shows a five-stranded β-sheet structure (four antiparallel strands and one parallel strand), with two antiparallel α-helices, that have the same orientation as the strands, on top of the sheet. Fig. 2a,b shows a close-up of the FAD binding pocket of the AppA BLUF domain. The FAD chromophore binds on top of the β-sheet, between the two α-helices. Its ribose, phosphate, and adenine moieties extend out into solution. The flavin-binding pocket is very hydrophobic, with a few polar or charged residues in positions where they make specific contacts with the flavin N(5), N(3), O(4), and O(2) atoms. Of the residues making contact with the flavin, almost all are completely conserved within the BLUF domain sequence family. The methyl groups of the flavin ring are packed against the aromatic rings of nearby aromatic residues and hydrophobic side chains from the sheet and from both helices. The flavin isoalloxazine ring extends along the two helices, with the O(2) and H(3) groups making hydrogen-bonding contacts with H44 and N45 in the solution structure of AppA (Grinstead et al., 2006a). A strongly conserved glutamine/tyrosine couple (E63/Y21 in AppA) is in hydrogen-bonding contact with the N(5) of the flavin ring and is crucial for the mechanism of photoacti-
Fig. 2. BLUF photocycle and active site structure. Panel a: active site receptor state. Panel b: active site signaling state. Panel c: typical photocycle of BLUF-domains. Panels a and b were prepared using the program PyMOL (http://www.pymol.org). The structure coordinate file of the BLUF domain of AppA from Rba. sphaeroides (PDB ID: 1YRX) (Anderson et al., 2005) was used to create both panels. Note that for panel b the position of the side-chain nitrogen and oxygen of Q63 were manually switched to indicate the Q-flip. See also Color Plate 14, Fig. 23.
vation of BLUF domains (see further below). Upon illumination with blue light, BLUF domains show a characteristic spectral red-shift of their absorption spectrum by approximately 10 nm, which is a unique feature of this photoreceptor family and is thought to correspond to the signaling state (Masuda and Bauer, 2002). The photochemical mechanism by which BLUF domains are activated by blue light is under considerable debate. Early studies suggested that an increased π-π stacking between tyrosine and FAD, or deprotonation of FAD would induce signaling state formation (Kraft et al., 2003; Laan et al., 2003). Jung et al. (2005) proposed that in the BlrB
Chapter 41
Photoreceptor Proteins from Purple Bacteria
BLUF domain, proton transfer takes place from an arginine residue to the O(2) atom of the isoalloxazine ring upon blue-light absorption, triggered by the increased basicity of O(2) upon promotion of FAD to the singlet excited state. Masuda and co-workers interpreted their results from FTIR spectroscopy by a hydrogen-bond rearrangement around the FAD chromophore upon signaling state formation (Masuda et al., 2004). Anderson et al. (2005) have proposed a photoactivation model that involves such a hydrogenbond rearrangement accompanying a 180o rotation of the conserved glutamine, as shown in Fig. 2a,b. In this view, photon absorption leads to the breaking of hydrogen bonds from the glutamine amino group to the N5 of flavin and to the tyrosine, and formation of a hydrogen bond to the O4 of the flavin. In the various BLUF crystal structures, the dark state orientation of the conserved glutamine remained ambiguous: according to Anderson et al. (2005) the amino group of the conserved glutamine would donate hydrogen bonds to N5 and the conserved tyrosine, while Jung et al. (2005, 2006) and Kita et al. (2005) favored an orientation where the glutamine’s amino group donates a hydrogen bond to O(4) and receives a hydrogen bond from the conserved tyrosine. NMR data indicated that the hydroxyl group of the conserved tyrosine became strongly hydrogen bonded only in the light-activated state (Grinstead et al., 2006b), whereas Raman experiments showed that mutation of the conserved glutamine in the AppA BLUF domain did not alter hydrogen bonding of the FAD O(4) (Unno et al., 2006). The latter observations are consistent with the dark orientation proposed by Anderson et al. (2005), as shown in Fig. 2a. Another controversial issue concerns the conformation of the conserved tryptophan. The wild type AppA crystal structure and solution NMR structure locate the tryptophan in the vicinity of FAD in a buried conformation (Anderson et al., 2005; Grinstead et al., 2006a,b). In contrast, the AppA1–124 C20S mutant, the BlrB and the cyanobacterial BLUF crystal structures show the tryptophan in a ‘flipped out’, solvent-exposed conformation (Jung et al., 2005; Kita et al., 2005; Jung et al., 2006). The genome of Rba. sphaeroides 2.4.1 encodes three BLUF domain containing proteins: AppA (450 aa), and twoshort BLUF-domain containing proteins, BlrA (134 aa) and BlrB (140 aa) (Table 1).AppA is a transcriptional anti-repressor, controlling photosynthesis gene expression in the purple bacterium Rba. sphaeroides in response to blue light and oxygen,
819
through interaction with the transcriptional repressor PpsR (Masuda and Bauer, 2002; Chapter 35, Bauer et al.). At low oxygen tension, two cysteines from the Cterminal domain are reduced and subsequently AppA binds to and reduces PpsR. Accordingly, reduced AppA prevents PpsR from repressing the promoters of structural genes in the photosynthesis cluster, i.e., puc and puf. Fully oxidized AppA (present at high oxygen tension) and fully reduced AppA mediate the redox signal, independently of light. However, at intermediate oxygen concentrations, high intensities of blue light produce a conformational change in this photoreceptor protein (see below) that decreases its affinity for PpsR. Whereas the BLUF domain is required for sensing the light signal, the C-terminal part of AppA is sufficient for normal redox regulation. However, the C-terminus is required for the transmission of both redox and light signals, implying that this domain interacts with PpsR (Han et al., 2004). Furthermore, the AppA-BLUF domain and the remainder of the protein can transmit the light signal even when expressed as separated domains (Han et al., 2004). Currently, AppA is the only known proteinthat transduces and integrates light and redox signals. The short BLUF domain containing proteins from Rba. sphaeroides have also been purified, but only BlrB binds flavin upon expression in E. coli (Jung et al., 2005). BlrB has been photochemically characterized, but no evidence is available about its biological function. Upstream of blrB a gene is located with similarity to the response regulators (i.e., a signal receiver domain, RSP_1262). However, it has not been reported whether or not BlrB interacts with this protein. BLUF protein homologs have also been identified in the purple bacteria Jannaschia sp. CCS1, Rhodopseudomonas (Rps.) palustris, and Rvi. gelatinosus; however, at this time no further information regarding their structures or functions is available. Slr1694 (PixD) from the cyanobacterium Synechocystis sp. PCC 6803 is one of the photoreceptors involved in positive phototaxis (Okajima et al., 2005). This protein binds the Slr1693 polypeptide, which is encoded directly upstream of PixD (Okajima et al., 2005). Slr1693 also shows similarities with response regulators, more specifically with members of the CheY family. Moreover, the gene slr1692 encodes an EAL-domain containing protein. EAL domains, together with GGDEF domains, are involved in cdi-GMP turnover (Romling et al., 2005; Romling and Amikam, 2006). This cyclic nucleotide is an
820
J. Hendriks, M. A. van der Horst, T. K. Chua, M. Á. Pérez, L. J. van Wilderen, M. T. A. Alexandre, M.-L. Groot, J. T. M. Kennis and K. J. Hellingwerf
important bacterial second messenger that is related to complex biological processes such as biofilm formation, virulence gene expression and photosynthesis (Romling and Amikam, 2006). This characteristic suggests several hypothetical functions for these BLUF-domain containing proteins, of which the genes are located close to EAL- or GGDEF-domain encoding genes in bacteria. Tll0078 from the cyanobacterium Thermosynechococus elongatus BP1 has only been characterized photochemically (Fukushima et al., 2005), thus no information is available about its biological function, although its similarity with Slr1694 is very high. The BLUF-domain containing protein from E. coli, YcgF, is predicted to be a light regulated c-di-GMP phosphodiesterase, as an EAL domain is present at its C-terminus (Table 1). The photoactivated adenylylcyclase (PAC) from Euglena gracilis (Table 1) is the photoreceptor for negative and positive phototaxis and step-up photophobic responses (photoresponse that is activated by a light intensity increment) (Iseki et al., 2002; Ntefidou et al., 2003). It consists of two α subunits (PACα) and two β subunits (PACβ). Each subunit contains two BLUF domains (F1, F2) and two adenylyl cyclase catalytic domains, which form cAMP only in the presence of blue light (Iseki et al., 2002). Only F2, and not F1 (single domain constructs from PACα expressed in E. coli), seems to bind FAD, FMN, or riboflavin and to show BLUF photocycle features (Ito et al., 2005). Nevertheless, the exchange of the BLUF domain in the full length AppA protein by F1 does not impair blue-light sensing (Han et al., 2004). This is even more surprising if one realizes that only approximately 30% sequence identity exists between PACα-F1 and this bacterial BLUF domain. B. (Ultra)fast Spectroscopy and the BLUF Reaction Mechanism Ultrafast studies on the Slr1694 BLUF domain from Synechocystis revealed several FAD radical species in the photoreaction (Gauden et al., 2006). Upon blue-light excitation, a multi-exponential decay of the FAD excited state was observed with a dominant contribution by a 7 ps component. The excited state decay was assigned to electron transfer from the protein to FAD, and resulted in the anionic semiquinone FAD•−. The anionic species was subsequently protonated in a few picoseconds to form the neutral semiquinone FADH•. From FADH•, the red-shifted BLUF signaling state was formed in 65 ps at an
estimated quantum yield of 40%. AppA showed a significantly longer FAD excited-state lifetime as compared with Slr1694. Remarkably, the red-shifted signaling state was formed directly from the FAD excited state without any apparent intermediate, with a dominant time constant of 590 ps, at a quantum yield of 24% (Gauden et al., 2005). Given the similarities in structure and sequence between all BLUF domains studied so far, a similar reaction mechanism may very likely hold for all BLUF domains. The reason that the transient FAD radical intermediates seen in the Slr1694 photoreaction were not observed in AppA probably stems from a rate-limiting initial electron transfer rate from the FAD excited state. Alongside the red-shifted signaling state, FAD triplet states were formed as a side product at a quantum yield of 9 %, which decayed on the microsecond timescale. No spectroscopically distinguishable changes in UV/Vis spectra were detected over a micro- to millisecond time scale. The role of the conserved aromatic residues (Y21 and W104 in the AppA BLUF domain) was investigated by applying ultrafast spectroscopy to site-directed mutants of AppA. Remarkably, removal of W104 leads to an increase of the FAD excited-state lifetime and an increase of the quantum yield of signaling state formation (Laan et al., 2006; Gauden et al., 2007). In the Y21 mutants, transient FADH• species were formed upon excitation, which recombined to the ground state on a (sub)nanosecond timescale. Furthermore, all light-driven electron transfer processes were abolished upon removal of both Y21 and W104. These results were consistent with a photoactivation model where light-driven electron and proton transfer from the conserved tyrosine leads to signaling-state formation, whereas light-driven electron transfer from W104 provides a competing excited-state deactivation pathway that lowers the quantum yield of signaling state formation. A photoactivation mechanism for BLUF domains was proposed that involves a successive rupture of hydrogen bonds between the conserved tyrosine and glutamine by light-induced electron transfer from tyrosine to flavin (resulting in formation of FAD•−), and between the glutamine and flavin by subsequent protonation at flavin N(5) (resulting in formation of FADH•). The conserved tyrosine would very likely act as proton donor, resulting in the radical pair FADH•–Tyr•. These events would leave the conserved glutamine unhinged and free to rotate by ~180o, resulting in formation of a new hydrogen bond between
Chapter 41
Photoreceptor Proteins from Purple Bacteria
the amino group of the glutamine and the C(4)=O of FAD. Radical-pair recombination of FADH•-Tyr• would take place, whereby a hydrogen is shuttled back from FAD to tyrosine. The reprotonated tyrosine then donates a hydrogen bond to the carbonyl group of the conserved glutamine, locking it in place and stabilizing the signaling state (Gauden et al., 2006). In support of this model, removal of the conserved tyrosine or glutamine by site-directed mutagenesis abolishes the photocycle (Kraft et al., 2003; Laan et al., 2003; Kita et al., 2005). Ultrafast experiments on the photoactivated (signaling) state of the AppA BLUF domain showed that upon excitation, FAD undergoes electron and proton transfer to result in the neutral semiquinone FADH• on a 10 ps timescale. FADH• recombines to the original ground state of the signaling state in about 50 ps (K.C. Toh, I.H.M. van Stokkum, J. Hendriks, M.T.A. Alexandre, J.C. Arents, M. Ávila-Pérez, R. van Grondelle, K.J. Hellingwerf, J.T.M. Kennis, unpublished), as indicated in Fig. 2c. The results on the W104F mutant are essentially identical, indicating that initial electron/hydrogen transfer would arise from the conserved tyrosine. Thus, the net effect of photon absorption by the signaling state in BLUF domains is zero, and it may be concluded that BLUF domains are not photoreversible, in contrast to many other photoreceptor proteins such as phytochromes, rhodopsins, PYP, and LOV domains (Kennis et al., 2004; van der Horst and Hellingwerf, 2004). BLUF domains show a unique photocycle compared to other photoreceptors like PYP, LOV domains or the rhodopsins. In these photoreceptors, multiple spectroscopically distinguishable photocycle intermediates follow one another on increasing timescales from nano- micro- to millisecond timescales, thereby altering the physico-chemical state of the chromophore and gradually decreasing the free energy of the system until a stable, long-lived signaling state is reached (Losi and Braslavsky, 2003). In BLUF domains, there is just one clearly distinct photocycle intermediate, generated from the FAD singlet excited state via anionic and neutral semiquinone radicals on the picosecond timescale. After this intermediate state is formed on the ultrafast timescale, no spectroscopically distinguishable changes over 10–12 decades of time can be detected with UV/Vis spectroscopy. This presents a paradox: on one hand, formation of the signaling state in BLUF domains is faster than in any other photoreceptor, while on the other hand, with a lifetime of 1800 s the AppA signaling state
821
ranks among the most thermally stable. This raises fundamental questions as in what way the BLUF signaling state is energetically stabilized, and how fast back-reactions to the dark state are prevented. C. Signaling State Formation in BLUF Domains The mechanism by which the initial local structural changes in the vicinity of FAD are transferred to the molecular surface is poorly understood. Anderson et al. (2005) proposed that the conserved tryptophan is involved in a light-induced change of mobility in the β-sheet, mediated through light-dependent hydrogen bonding with the conserved glutamine. Mutagenesis of the conserved tryptophan (W104 in AppA) reveals interesting aspects of BLUF photoactivation. W104 mutants exhibit reactions upon light activation that are similar to wild type but exhibited a dark-recovery rate 2–150 times faster than that of wild type (Masuda et al., 2005b; Laan et al., 2006). Moreover, the addition of a large excess of imidazole accelerates the thermal decay of the W104 mutants more efficiently than in wild type (Laan et al., 2006). Interestingly, significantly decreased light-induced structural changes are observed in the apoprotein of the W104A mutant (Masuda et al., 2005b). Thus, although W104 mutants exhibit a photocycle and formation of the red-shifted intermediate, it is not clear whether they function as a transducer. Jung et al. (2005) suggested that in BlrB, light-induced structural changes would be mediated through proton transfer to C(2)=O of FAD from a nearby arginine. These events would influence the adenine moiety of FAD, which protrudes from the molecular surface and is fully solvent-exposed. The solvent-exposed adenine moiety would function as a recognition site for different effector domains in the C-terminus of the BLUF proteins (Jung et al., 2005). Resonant Raman spectroscopy on AppA BLUF indicated changes on the N10 ribityl chain upon illumination (Unno et al., 2005). However, FTIR studies on AppA BLUF domains reconstituted with FAD, FMN or riboflavin showed identical structural changes in the apoprotein upon illumination (Masuda et al., 2005a). In the X-ray structure of the C20S AppA mutant, a methionine is proposed to interact with the conserved glutamine instead of the conserved tryptophan. The latter residue was found in a solvent-exposed configuration in the dark-adapted state (Jung et al., 2006). It was suggested that a light-induced glutamine
822
J. Hendriks, M. A. van der Horst, T. K. Chua, M. Á. Pérez, L. J. van Wilderen, M. T. A. Alexandre, M.-L. Groot, J. T. M. Kennis and K. J. Hellingwerf
flip, followed by a concerted swing of the conserved tryptophan into the protein core and methionine to the solvent would constitute formation of the signaling state. Figure 2c shows a summary of the photocycle scheme of BLUF domains. Upon photon absorption by FAD, electron transfer takes place from the conserved tyrosine to FAD on a picosecond timescale. Next, the anionic FAD radical is rapidly protonated to result in the neutral semiquinone radical, which evolves on a sub-100 picosecond timescale to the red-shifted signaling state. No further spectral changes are detected with UV/Vis spectroscopy up to the timescale of seconds or minutes. Undoubtedly, structural changes will take place on timescales longer than picoseconds or nanoseconds. Time-resolved IR or X-ray methods may be used to detect such structural changes. IV. Comparison Between LOV and BLUF Domains In view of some recently published results, it is of interest to discuss the parallels between the photochemistry and signal generation in LOV and BLUF domains. In short, this photochemistry for the LOV domains involves intersystem crossing to the triplet state, followed by formation of a covalent adduct between the C(4) of the FMN chromophore and the sulfur atom of a nearby cysteine. In the BLUF domains, the light-induced electronically singlet excited state initiates reversible proton and electron transfer to the flavin from a nearby tyrosine, which in turn leads to a ‘Q-flip’ (i.e., a ~180° rotation of glutamine, see above). The first surprise with respect to similarities was the observation that also in LOV domains a ‘Q-flip’ may occur, i.e., Q1029 in AvPhy3-LOV2 (Crosson and Moffat, 2002; Nozaki et al., 2004). This ‘Q-flip’ may be crucial in signal generation and transmission, because of the β-5 strand-sliding that it may lead to, although de-stabilization of a highly conserved salt bridge between E51 and K92 in LOV1, or more general loop destabilization in LOV2 may also be involved (Crosson and Moffat, 2002; Nozaki et al., 2004; Freddolino et al., 2006). Next, it was reported that when a cysteine is engineered into the BLUF domain, at the proper position with respect to the flavin, the latter domain also displays lightinduced adduct formation (unpublished observation in Okajima et al., 2006, unpublished).
This of course brings up the question whether just a ‘Q-flip’ would also suffice in a LOV domain to initiate signaling. Although cysteine knock-out mutants of LOV domains have been considered for several years as proper vehicles for ‘blind’ LOV domain photosensor controls (e.g., Christie et al., 2002), a recent study revealed that there is remaining activity of phototropin variants of which the catalytic cysteine in both LOV domains was replaced by an alanine, through site-directed mutagenesis. This effect was particularly noticeable at higher light intensities, i.e., >10 µE m–2 s–1 (Cho et al., 2007). Although the authors explain their observation by leakiness of their phot2-1 mutant, in combination with cross-phosphorylation between functional phot2 and inactivated phot1 molecules, it is of interest to test rigorously whether LOV domains of which the catalytic cysteine has been eliminated are indeed fully inactive with respect to light-induced signal generation. A technically straightforward way to test this would be through the use of optical spectroscopy. In BLUF domains the ‘Q-flip’ is accompanied by a 10 nm red-shift of the visible absorption band of the flavin (Masuda and Bauer, 2002). This red-shift is presumably caused by the additional hydrogen bond between the flavin and the surrounding amino acids in the light-induced signaling state (Grinstead et al., 2006b). Significantly, no similarly red-shifted signaling state has been reported for any LOV domain, not even in the variants in which the catalytic cysteine was eliminated. However, this may be explained by the fact that the ‘Q-flip’ in LOV domains is not accompanied by formation of an additional hydrogen bond (Crosson and Moffat, 2002; Freddolino et al., 2006). The glutamine amide group and flavin are parallel in the BLUF domain, both in the receptor and in the signaling state. The ‘Q-flip’ then generates an extra hydrogen bond between the Q-γ-NH2 as the donor and the N(5) of the flavin as the H-bond acceptor. In the LOV domain the amide group of Q1029 is perpendicular to the isoalloxazine ring of the FMN. Light absorption converts the catalytic glutamine from a H-bond donor to O(4) of the flavin to a H-bond acceptor of N(5). Crucial for photo-activation of the LOV domains without a catalytic cysteine, will be the question whether or not a suitable e–/H donor (e.g., the side chain of a tryptophan, tyrosine or histidine) is sufficiently nearby to transiently reduce the flavin to the semi-quinone form. This would sufficiently increase the acidity of the N(5) atom and allow it to act as an
Chapter 41
Photoreceptor Proteins from Purple Bacteria
H-bond donor for the C=O group of the glutamine. For some LOV domains this is the case (Kay et al., 2003) and significantly, the neutral flavin radical has a lifetime on the order of tens of minutes. The next important issue is whether or not the ensuing ‘Q-flip’ will sufficiently displace the β-5 strand. Another yet unsolved issue is the question whether these observed parallels between BLUF and LOV domains are the result of divergent or of convergent evolution. At this moment we do not find convincing arguments for one or the other point of view, although it is attractive to think that a primordial BLUF domain ‘invented’ a nearby cysteine to increase its quantum yield of signaling state formation. Following this line of thinking we consider it less likely that the ‘goal’ in their joint evolutionary history was to increase the lifetime of the signaling state, because both among the BLUF and among the LOV domains there are some that recover only very slowly (Masuda and Bauer, 2002; Losi et al., 2003); furthermore there are concrete competitive disadvantages in extending this lifetime too long (Hellingwerf, 2005). A meaningful comparison between LOV and BLUF domains and Cryptochromes requires more detailed insight into the photochemistry and signal generation of the latter. As explained in the introduction to this chapter, photosensory proteins have acquired an important role in the understanding of structure/function relations in proteins in general, i.e., an impact that extends far beyond the realm of only the purple bacteria. The scientific impact of this class of proteins, however, has also increased for another reason: As a result of many of the genome sequencing projects, many genes encoding a putative photoreceptor protein have also been detected in many chemotrophic organisms. Recently, we were able to demonstrate a physiological function for two of these, the LOV domain of YtvA from Bacillus subtilis (Avila-Pérez et al., 2006) and the BLUF domain of YcgF of E. coli (J. Key, S. Crosson, R. van Grondelle, J.T.M. Kennis, K.J. Hellingwerf, unpublished). V. The Xanthopsins A. Introduction and Function The first discovered member of the Xanthopsin family is the Photoactive Yellow Protein (PYP) from the purple sulfur bacterium Halorhodospira (Hlr.)
823
halophila (then Ectothiorhodospira halophila). In a fractionation study, among various soluble, colored proteins, such as cytochromes and ferredoxins, a yellow protein was also found (Meyer, 1985). Subsequent experiments showed that this protein absorbs maximally at 446 nm, and that it is photoactive, hence Photoactive Yellow Protein (Meyer et al., 1987). The chromophore that gives the protein its yellow color is 4-hydroxy cinnamic acid (or p-coumaric acid) (van Beeumen et al., 1993; Baca et al., 1994; Hoff et al., 1994). In the dark state of the protein, this chromophore is deprotonated and in the trans configuration (see below). The protein structure of PYP has been determined in extremely high detail. It contains a central antiparallel β-sheet (β-scaffold), surrounded by α-helices (Borgstahl et al., 1995; Getzoff et al., 2003). The chromophore is bound via a thio-ester linkage to the protein’s single cysteine in a hydrophobic pocket. The negative charge on the phenolic oxygen of the chromophore is stabilized through a hydrogen-bond with E46. Other residues that have been shown to be important in the tuning of the color and/or in the functioning of the protein are Y42 and T50 (both also form part of the hydrogen-bonding network connecting the chromophore and E46), R52 (shielding the chromophore from the solvent), and M100 (possibly involved in the thermal re-isomerization of the chromophore). PYP has become the structural prototype for the so-called PAS-fold. This module is found in all kingdoms of life and is often involved in signaling and/or protein-protein interactions. The PAS-domain containing proteins are structurally very similar, although PAS members are not very similar at the level of primary sequence. In addition, a variety of cofactors have been identified from those PAS members that require one. Upon blue-light illumination, PYP enters a photocycle, in which three main steps can be distinguished: (i) ultra-fast isomerization of the chromophore vinyl bond, yielding the red-shifted intermediate pR, (ii) protonation of the phenolic oxygen of the chromophore and structural changes in the protein, resulting in the blue-shifted intermediate pB, and finally (iii), deprotonation and re-isomerization of the chromophore to get back to the groundstate pG. Details about the early events in the photocycle and about structural changes in the protein will be discussed below. In addition to being found in Hlr. halophila, PYP was also discovered in the closely related phototrophic purple bacteria Chromatium salexigens and
824
J. Hendriks, M. A. van der Horst, T. K. Chua, M. Á. Pérez, L. J. van Wilderen, M. T. A. Alexandre, M.-L. Groot, J. T. M. Kennis and K. J. Hellingwerf
Rhodothalassium salexigens (Meyer et al., 1990; Koh et al., 1996). The protein from the latter organism was characterized in more detail. Its spectroscopic properties, such as UV/Vis absorption maximum and kinetics of bleaching and recovery were remarkably similar to those of PYP (In this review, ‘PYP’ always refers to the PYP from Hlr. halophila, unless stated otherwise). Over the past few years, PYP has been detected in more and more organisms, in large part due to various genome sequencing projects. Most of these Xanthopsin proteins can be classified, based on a phylogenetic analysis, into 3 main groups. Besides the first group, to which the PYPs mentioned above belong, a well-defined second group is formed by the PYPs from Rba. sphaeroides and Rba. capsulatus. Both have been characterized spectroscopically; under physiological conditions two photoactive states are present in both. Besides the absorbance at 446 nm, there is a second absorbance maximum at 360 nm (435 and 375 nm for PYP from Rba. capsulatus) (Haker et al., 2000; Kyndt et al., 2004b). It was shown that in the dark the 360 nm form of the protein has a protonated cis-chromophore. Excitation of either peak results in the initiation of a photocycle. Recovery of the ground state, after excitation at 446 nm, was shown to be about 100-fold faster in this Rhodobacter protein compared to PYP (Haker et al., 2000; Haker et al., 2003). The physiological relevance of these differences, however, remains to be established. The Rhodobacter PYPs have been implicated in lightregulation of cell buoyancy, because in Rba. capsulatus, the pyp gene is flanked by genes encoding gas vesicle proteins. A genetic proof for this proposed function, however, remains to be given. A very interesting finding was the presence of multi-domain signal transduction proteins in which PYP is one of the signal-input domains. Two examples have been described in which PYP is fused to a bacteriophytochrome-like domain, which in turn is connected to a C-terminal output domain: a histidine protein kinase in Ppr (first named Pph) from Rhodocista centenum (Jiang et al., 1999), and EAL/GGDEF domains in Ppd from Thermochromatium tepidum (Kyndt et al., 2004a). These latter Xanthopsins make up the third group. The PYP domain from Ppr was crystallized and shown to have a structure very similar to PYP (Rajagopal and Moffat, 2003). Its photocycle properties, however, differ significantly from PYP: ground state recovery is slowed down to ~100 s. The PYP domain from Ppd lacks the conserved
E46 residue and as a result of this, has a wavelength maximum at 358 nm, at physiological pH (Kyndt et al., 2005). Kinetically, this Ppd is similar to Ppr, even though Ppd may act as a UV-sensor. The GGDEF and EAL domains, respectively, catalyze the synthesis and hydrolysis of the recently discovered signaling molecule c-di-GMP. This molecule has been shown to be involved in bacterial biofilm formation (Romling et al., 2005), implying the involvement of PYP in lightregulation of biofilm formation. These multi-domain Xanthopsins open up the possibility of investigating light-induced domain-domain interactions, and resulting light-regulated enzyme activity. Another striking finding is the presence of a second PYP in Hlr. halophila, PYP(b), (van der Horst et al., 2005b). This protein was shown to be an authentic yellow protein, and to have a photocycle comparable to that of PYP(a), albeit with a recovery rate that is ~100 times slower. This might be taken as an indication for a different function. Furthermore, the genomic organization around PYP(b) is quite different from that of PYP(a). Directly downstream of PYP(b) a cluster of genes is found all of which encode proteins with high similarity to two-component signal transduction proteins. Other new Xanthopsin members are found in the organisms Rps. palustris and Salinibacter ruber. Both organisms contain, besides PYP, several other photoreceptor proteins. PYP has been found in one additional non-phototrophic organism besides Salinibacter ruber, the deep-sea bacterium Idiomarina loihiensis. These latter three PYP proteins can not easily be categorized in one of the three groups, based on phylogenetic analyses, and have not yet been analyzed spectroscopically. B. (Ultra)fast Spectroscopy Since PYP is a small, stable and well characterized protein, it is an excellent model for studying the initial steps of signal generation and transduction. A fairly large number of spectroscopic studies on native PYP, genetically modified PYP, and on modified chromophores both free in solution and bound to the protein, have been performed, and have led to a detailed picture of the photocycle. The protein environment has been found to have a profound effect on the chromophore dynamics, as the ultrafast (femto- to picosecond) dynamics of photo-excited model chromophores in solution have revealed. It appears that the chemical structure at the distal side
Chapter 41
Photoreceptor Proteins from Purple Bacteria
of the chromophore determines whether or not it will isomerize. For example, the chromophore with a phenyl group attached to the thio-ester moiety (i.e., thiophenyl-p-hydroxycinnamate) isomerizes (in the denatured protein the chromophore isomerizes too) (Changenet-Barret et al., 2004). In contrast, when a methyl group replaces the phenyl group (i.e., in thiomethyl p-hydroxycinnamate) it does not (Larsen et al., 2004b). These observations are still under debate and the discussion extends also to the gas phase in which both the neutral and the anionic form of trans-coumaric acid are found to isomerize (Ryan et al., 2002; Lee et al., 2006). The hydrophobic protein pocket not only tunes the absorption properties of the protein (Chosrowjan et al., 1998), but also channels the vibrational motions of the chromophore into specific modes. The photocycle of wild type PYP is known to span 14 decades of time at room temperature. Up to the millisecond timescale several transient (electronic) ground-states play a role. Depending on the environmental conditions, such as temperature, pH and humidity, different photocycle dynamics have been observed and thus a range of different photocycle schemes and nomenclatures have been proposed to fit these data. Here, we use the notation as described by Hellingwerf et al. (2003). In the dark-adapted ground state at neutral pH the p-coumaric acid chromophore is negatively charged and embedded in a hydrogen bonding network. Upon electronic excitation, an immediate charge translocation from the phenolic moiety towards the carbonyl of the chromophore is observed (Groot et al., 2003). This rearranges the conjugation across the chromophore and facilitates the subsequent trans-cis isomerization of the chromophore. The excited state lifetime of PYP has been determined by fluorescence spectroscopy. This state decays multi-exponentially in a period of the order of picoseconds (Chosrowjan et al., 1997; Changenet et al., 1998). This is also the timescale (~2 ps) on which, with time-resolved vibrational spectroscopy, a stable cis-ground state has been observed to form (Groot et al., 2003; Heyne et al., 2005). In conjunction with presumed isomerization, a red-shifted intermediate I0 appears in the visible part of the spectrum, exhibiting a main absorption band at 500 nm, instead of at 446 nm. The next stable intermediate pR is formed in 1–3 ns and has a peak absorption at 460 nm (Baltuška et al., 1997; Ujj et al., 1998; Devanathan et al., 1999; Imamoto et al., 2001; Gensch et al., 2002; Larsen et al., 2004a). This state is subsequently followed by large structural changes on the millisecond time scale,
825
due to formation of the signaling state (Cusanovich and Meyer, 2003; Hellingwerf et al., 2003; Larsen and van Grondelle, 2005). Extensive dynamical rearrangements of the hydrogen bonding network in the chromophore-binding pocket have been observed on to the pico- to nanosecond timescale: upon electronic excitation the chromophore isomerizes from trans to cis, and the hydrogen-bond between the carbonyl of the chromophore and the backbone nitrogen of C69 is broken. Recently, with time-resolved vibrational spectroscopy, it was shown that breaking of this hydrogen bond within a few picoseconds is a decisive step for successful entry into the photocycle. Upon breaking of this hydrogen bond, the chromophore fully isomerizes and forms the I0 intermediate. When, in contrast, the bond is not broken the chromophore falls back to the ground-state, through a very shortlived ground-state intermediate (van Wilderen et al., 2006). The relatively low success rate of the hydrogen-bond breaking causes the quantum yield of the photocycle to be ~0.3 (van Brederode et al., 1995; Groot et al., 2003; Larsen et al., 2004a). The hydrogen bond network around the phenol ring of the chromophore, formed with residues E46, Y42, T50 follows a different dynamic behavior. Initially the hydrogen bond between the phenolic oxygen and the Glu46 residue weakens, but it strengthens as soon as the stable intermediate pR is formed (Groot et al., 2003). The strength of this hydrogen bond is, however, not essential for a successful entry into the photocycle, as the introduction of a weaker hydrogen bond donor glutamine causes the hydrogen bond to break on the nanosecond timescale, but hardly affects the initial part of the photocycle in other ways (van Wilderen et al., 2006). Different photocycle pathways become accessible when the level of hydration of a film of PYP protein is varied (van der Horst et al., 2005a). A low level of hydration (< 50% relative humidity) leads to additional pathways to the ground state from I0 without forming pR, showing the plasticity of the photocycle transitions. C. Xanthopsin Signaling To date, a molecular model of signal transduction has not been described for any of the Xanthopsins (i.e., the interaction of a signal-transducer with the Xanthopsin signaling state, or ground state). However, a great deal of information is available on the photocycle
826
J. Hendriks, M. A. van der Horst, T. K. Chua, M. Á. Pérez, L. J. van Wilderen, M. T. A. Alexandre, M.-L. Groot, J. T. M. Kennis and K. J. Hellingwerf
of the best studied of all Xanthopsins, the PYP from Hlr. halophila.The primary photoreactions leading to formation of the pR intermediate were discussed above. Basically those events determine/describe the photosensitivity of the protein. Here we will continue the description of the photocycle with signaling state formation and the subsequent ground state recovery. These are the events that are important for the actual transduction of a signal to the cell, enabling the cell to respond appropriately to the detected light. In PYP many photocycle-related events/characteristics are pH dependent (Genick et al., 1997b; Hoff et al., 1997; Hendriks et al., 1999a, 2003; Shimizu et al., 2006), which suggests that protonation changes and/ or states of specific residues play an important role in this cycle. Traditionally, photocycle intermediates have been defined on the basis of their spectroscopic characteristics. However, due to the pH-dependent nature of most of the photocycle characteristics, such a description has become more and more complicated as different pH values seem also to result in different photocycle reaction mechanisms (Hendriks et al., 2003). Not only pH but also other mesoscopic conditions such as degree of hydration (van der Horst et al., 2005a), ionic strength (Borucki et al., 2005b), buffer composition (Shimizu et al., 2002) and phase (crystalline or solution) (Xie et al., 2001) seem to have marked effects on the photocycle. This has also led to a very complicated and sometimes confusing nomenclature in the literature. In an attempt to describe these intricacies of the photocycle more clearly we have described the PYP photocycle in a tree form, defining each intermediate by the state of some key characteristics, i.e., the chromophore isomerization state, the protonation state of the chromophore and residue E46, and the protein folding state (see Fig. 3). Note, that in the tree an intermediate may have several possible states, and that conditions such as pH determine which of these states dominates. We feel that taking this into account will help to explain the sometimes seemingly contradictory results found in the literature. In addition such a description may be used as a starting point to prepare a systems biology model of the photocycle taking into account the many mesoscopic conditions with which PYP has been studied. After the primary photocycle events pR is formed. Structurally pR is very similar to the ground state with one key difference: The chromophore has isomerized from trans to cis (Perman et al., 1998). This upsets the hydrogen-bonding network in the chromophore-
Fig. 3. PYP Photocycle tree. Representation of the PYP photocycle in tree form. Note that a photocycle intermediate may have several states. Below the name of an intermediate, the states of several key characteristics are shown; top left: isomerization state of the chromophore; top right: protonation state of the chromophore; bottom right: protonation state of the Glu46 residue; bottom left folding state of the protein (Fi: initial fold (ground state fold); Fx: undefined relatively folded structure; Fr: folded structure enabling cis to trans isomerization of the chromophore; U: undefined unfolded structure). Picosecond intermediates (like I0) have been omitted from this scheme.
binding site enough to favor protonation of the chromophore. FTIR measurements have shown that before anything else can happen the chromophore is protonated and E46 is simultaneously deprotonated (Xie et al., 2001). E46 apparently donates its proton to the chromophore, which is a reversible event (Hendriks et al., 2003). Notably, the pH dependence of the forward and backward reaction is different (Hendriks et al., 2003), suggesting more than one possible reaction mechanism. Protonation of the chromophore
Chapter 41
Photoreceptor Proteins from Purple Bacteria
results in a large blue-shift in its UV/Vis absorption spectrum. In pB’, which is structurally highly similar to pR, the negative charge on E46 that results from the intra-molecular proton transfer is buried and not stabilized via a hydrogen-bonding network. Besides the reverse reaction to pR, there are two other possible methods to relieve the stress caused by the buried negative charge on E46. One is large structural change, as indicated by FTIR measurements (Xie et al., 2001). A second possibility would be to protonate E46 from solution. The latter mechanism will likely dominate at low pH and would not result in large structural changes within the protein (Hendriks et al., 2003). Alternatively, at low pH the chromophore might get protonated directly via the solvent resulting in the same situation, in this case pB’ and pB might be considered the same species. The absorption spectrum of pB’ is slightly red-shifted compared to the absorption spectrum of pB, a shift which is itself pH dependent (Hendriks et al., 2003; Shimizu et al., 2006). Recent work has shown that the absorption spectrum of pB can be resolved into 2 species, one of which has an absorption spectrum very similar to pB’ (Shimizu et al., 2006) but an FTIR spectrum different from pB’ (Xie et al., 2001; Shimizu et al., 2006). We think it likely that pB actually consists of a myriad of different sub-states, as shown in Fig. 3, several of which will have a relatively folded and several a relatively unfolded structure. Very little is known about the details of ground state recovery. However, as light-induced re-isomerization speeds up ground state recovery tremendously (Hendriks et al., 1999b; Joshi et al., 2005), it is fair to assume that the rate of cis to trans isomerization of the chromophore is rate limiting for the recovery. Furthermore, to facilitate thermal re-isomerization it is highly likely that a deprotonated chromophore is required (Demchuk et al., 2000; Hendriks et al., 2003). This is supported by the observed dependence of receptor-state recovery rate on the hydroxide concentration (Hendriks et al., 2003). Furthermore, in order for the protein to catalyze the thermal isomerization reaction, presumably a specifically folded state of the protein has to be present. This may well be a relatively folded structure as observed in crystalline PYP, where fairly small structural changes occur upon photoactivation (Genick et al., 1997a; Xie et al., 2001; Ihee et al., 2005). In agreement with this, receptor state recovery is comparatively fast in crystalline PYP (Yeremenko et al., 2006). Furthermore, it is likely that E46 will be protonated in such a folded structure. A hypothetical
827
intermediate that has all these characteristics has been included in Fig. 3 as pG’. One would expect that this intermediate is red-shifted compared to pB (due to the deprotonated chromophore) and indications are that this intermediate may have an absorption spectrum very similar to pG (Hendriks et al., 2003). Since Hlr. halophila is an alkaliphilic organism, and such organisms tend to maintain a cytoplasmic pH of around 8 when grown under alkaline conditions, it is fair to assume that physiologically PYP will function at around pH 8. This assumption is supported by the fact that PYP has optimal receptor state recovery kinetics around pH 8 (Genick et al., 1997b). Therefore the photocycle route with the highest physiological relevance is the one straight down the middle of the tree in Fig. 3. This suggests that the major structural change, and possibly protonation changes of specific residues, is important for signaling the cell that a blue photon was absorbed. VI. Bacteriophytochromes A. Occurrence, Structure and Function of (Bacterio)phytochromes Phytochromes were discovered in plants as the photoreceptor family responsible for red/far-red reversible light responses of leaf movement, stem growth, stomatal opening, etc.. They exist as dimers in the eukaryotic cytoplasm, and translocate to the nucleus upon far-red light-activation. A linear tetrapyrrole (phytochromobilin) is their light-sensitive chromophore, bound via a thio-ether linkage. Red light triggers a cis to trans change in the configuration of the extended ‘all-cis’ chromophore, converting it into the far-red light absorbing Pfr form, which is the biologically active form. Subsequently, it very slowly reverts back in the dark (on a timescale of hours) or almost instantaneously upon absorption of far-red light. During these transitions proton uptake and release reactions take place, as well as structural changes in the protein, on the micro- and millisecond timescales; for reviews see Quail et al. (1995) and Rockwell et al. (2006). In the sequencing project of Synechocystis PCC6803 (Kaneko and Tabata, 1997) the first phytochrome-like gene was discovered in a prokaryote, which has led to the identification of many (>100) cyanobacterial and bacterio-phytochromes, in both phototrophic and in chemotrophic bacteria (e.g., in
828
J. Hendriks, M. A. van der Horst, T. K. Chua, M. Á. Pérez, L. J. van Wilderen, M. T. A. Alexandre, M.-L. Groot, J. T. M. Kennis and K. J. Hellingwerf
Pseudomonas aeruginosa (Davis et al., 1999), Rps. palustris (Giraud et al., 2002), and Rhodospirillum centenum (Jiang et al., 1999)) and in lower eukaryotes like the fungus Neurospora crassa. Typically, bacteriophytochromes will show red (~650 nm)/far-red (~750 nm) photoreversibility, but a lot of variation on this theme can be encountered, up to shifts to the near-red and green/red-light photoreversibility (see further below). Phytochrome structure is strikingly modular (Rockwell et al., 2006): For detection of the light input signal a central GAF domain binds the bilin chromophore, surrounded by an upstream PAS-like domain (PLD) and a downstream phytochrome (PHY) domain, necessary to stabilize the Pfr state. In plant phytochromes an additional P1 domain is present N-terminally. The signal output domains show considerable diversity: in phytochromes from plants and algae this is a combination of two PAS domains and a histidine-kinase-like domain that presumably displays serine/threonine kinase activity. In the bacterio-phytochromes (and those from fungi) the most frequently found signal output domain is a histidine kinase domain, of the classical- or of the phospho-relay type (Karniol et al., 2005; Chapter 40, Evans et al.). As an alternative, one of various domains involved in bacterial signal transduction may be present, like an MCP (methyl-accepting chemotaxis protein) domain (Zhulin, 2000; Bhaya, 2004), a PP2C phosphatase domain, an adenylate cyclase domain, or (a) domain(s) involved in the metabolism of c-di-GMP (i.e., the GGDEF/EAL or DUF1/2 domains). Often, multiple (bacterio)phytochromes are present in a single organism. The Rba. sphaeroides genome encodes two bphy sequences; the Rps. palustris genome six (Giraud et al., 2005; Chapter 40, Evans et al.). The light-input domain of the phytochrome family can also join forces with other light-sensitive signal receiving domains, such as the LOV-domain(s) of phototropin in ferns (Nozue et al., 1998) and the PYP domain in purple bacteria (Jiang et al., 1999). The biological function of the bacterial phytochromes — just like the corresponding function in eukaryotic microorganisms and plants — is to modulate or adjust the physiology and gene expression of the organism to the ambient light climate (van der Horst and Hellingwerf, 2004). The precise nature of the output domain often provides more detail: presence of an MCP domain suggests involvement in phototaxis; the presence of a DUF domain (like in Thermochroma-
tium tepidum) is indicative of involvement in biofilm formation (Jenal and Malone, 2006) and the linkage to a helix-turn-helix domain containing response regulator suggests a direct involvement in the regulation of gene expression. Bacteriophytochrome functions that have clearly been demonstrated in purple bacteria are: regulation of pigment (i.e., chalcone) synthesis in Rhodospirillum centenum (Jiang et al., 1999) and regulation of antenna synthesis by at least three of the bacteriophytochromes of Rps. palustris (Giraud et al., 2002; Giraud et al., 2005; Chapter 40, Evans et al.). BPhyP1 controls expression of RC-LH1 complexes via protein/protein interactions that modulate two transcriptional anti-repressors, whereas BPhyP2 and BPhyP3 jointly affect phosphorylation of the common response regulator Rpa3017 to modulate expression of LH4 (which is highly significant in the ecosystem in which this organism thrives). The latter couple of bacteriophytochromes shows very interesting photochemistry (see below). Two cyanobacterial phytochromes (TaxD1 and Cph2) are involved in phototaxis (Bhaya, 2004; Fiedler et al., 2005). The bilin chromophore of phytochrome can be bound to either a cysteine (via a thio-ether bond), or to a histidine (via a Schiff base linkage) and can even be functionally associated in a non-covalent way (see Lamparter et al. (2002) and references therein). Mostly, the residue for covalent attachment would be found in the GAF domain, but recently, an interesting new member of the bacteriophytochrome family was found in which this was a cysteine from the PLD domain (Lamparter et al., 2002). The bilin chromophore is bound through the C-3 atom of pyrrole ring A. The strong sequence conservation of many of the structural domains that play a role in the phytochrome family allows prediction of apoprotein structure through homology modeling. This is particularly true for bacteriophytochromes of which the PLD, PAS, GAF, and histidine kinase domains each have at least one homologous domain of which the spatial atomic structure has been resolved. The only domain for which this information is lacking is the PHY domain, although this domain has been proposed to be a member of the PAS family too. However, such modeling leaves the question unresolved as to how the various domains are arranged in space with respect to each other. Nevertheless, additional information is available to tackle this latter problem, like the interchromophore distance in Cph1 dimers, as determined with fluorescence resonance energy transfer (Otto et
Chapter 41
Photoreceptor Proteins from Purple Bacteria
al., 2003). Using this approach, homology modeling of Cph1 has led to a proposal for the structure of the complete chromophore-binding domain (CBD) (Laan, 2005). A breakthrough in our understanding of (bacterio) phytochrome structure has been achieved with the resolution of the spatial structure of the N-terminal PLD plus GAF domain (i.e., the first 321 residues) of the bacteriophytochrome (DrBphP) from Deinococcus radiodurans (Wagner et al., 2005). This part of this light-activated histidine-protein kinase is composed of a disordered N-terminal tail (amino acid 1–37), a PLD domain (38–128) and a GAF domain (139–321). The two domains contain a five- and six-stranded anti-parallel central β-sheet, respectively, and, in both, the sheet is surrounded by several α-helices. They are connected by a 10-residue polypeptide linker and by a complex structure called the ‘trefoil knot’ plus the chromophore that links the β-sheets of the PDL and GAF domains. The 37 N-terminal residues (of which C24 binds the biliverdin chromophore through the C-32 atom of the A ring of the bilin) form a random coil that is folded as a knot with the trefoil structure (Fig. 4). The bilin chromophore is bound in an extended 5Zsyn, 10Zsyn, 15Zanti configuration within the GAF domain, between the central β-sheet and α-helices 6 and 7, and is protonated (van Thor et al., 2001). It is held in position through ionic (particularly with the propionic acid side chains of rings B and C) and by van der Waals interactions between the pyrrole rings and a series of key conserved aromatic residues from this domain. The PLD domain contributes little in this except for those bacteriophytochromes in which the chromophore is covalently linked to the N-terminus, that forms part of the trefoil knot (as in DrBphP) (Wagner et al., 2005). It would be of great interest (and presumably helpful for understanding the molecular basis of photoactivation of phytochromes) to also resolve the spatial structure of a bacteriophytochrome family member of which not the Pr, but rather the Pfr absorbing form is the most stable state in the dark (see below). B. (Ultra)Fast Spectroscopy of Bacteriophytochromes With the discovery of the bacteriophytochromes the diversity in the phytochrome family has strongly increased. This applies first of all to the involvement of phycocyanobilin and biliverdin as chromophores,
829
Fig. 4. Bacteriophytochrome photocycle and structure. Schematic structure of the chromophore-binding domain and simplified scheme of the photocycle of bacteriophytochromes. Assignment of proton release between I700 and Ibl1 is tentative. For further explanation: see text.
next to the phytochromobilin. But as discussed above, its mode of attachment can also vary, including noncovalent association of the chromophore with the apo-protein, as well as the rate of thermal relaxation of the Pfr form back to Pr. A very surprising observation was recently made by Verméglio and co-workers: Rps. palustris BPhyP3 (in contrast to Rps. BPhyP2, which shows conventional red- and blue-shifts), is converted by red light to a species that absorbs in the near-red (i.e., shifts to the blue) rather than to the red with respect to the most stable ground state (Giraud et al., 2005). This new bacteriophytochrome species is therefore referred to as Pnr (for: pigment absorbing in the near-red). Nevertheless, even in Rps. palustris BPhyP3, primary photochemistry is assumed to be based on regular Z → E isomerization (see below under VI.C). This wide variety in spectral shifts and relaxation rates may turn out to be a valuable asset in future studies in synthetic biology, in which one would try to construct light responsive systems with predefined properties. For this discussion, however, it implies that the details that have so far been uncovered of the (ultra)fast spectroscopy of phytochromes may apply only to a sub-group of members of this family. By far most of the information available has been obtained in
830
J. Hendriks, M. A. van der Horst, T. K. Chua, M. Á. Pérez, L. J. van Wilderen, M. T. A. Alexandre, M.-L. Groot, J. T. M. Kennis and K. J. Hellingwerf
experiments performed with the bacteriophytochrome Cph1 from Synechocystis PCC6803. The photocycle of this bacteriophytochrome shows striking similarities with the main one in plant phytochromes and it can be produced heterologously in Escherichia coli, even as a holoprotein (Landgraf et al., 2001). Therefore, most of the discussion below will be centered on Cph1 and its homologs. The photochemistry of, and transient intermediates involved in the photocycle of Cph1 have been studied with both optical absorption and vibrational spectroscopy, and both in a time-resolved mode and by making use of low temperature trapping (e.g., Remberg et al., 1997). This applies in particular to the photo-transformation of Pr into Pfr; the opposite reaction has been studied less intensively, but shows large similarities. The processes can be summarized as shown in Fig. 4. With a relatively low quantum yield (on the order of 0.2), the Pr form of bacteriophytochrome is converted to a first transient ground state intermediate with relatively slow picosecond kinetics. This form is generally referred to as I700 (in time resolved studies) or lumi-R in low-temperature trapping experiments (see Mizutani et al. (1994) for related studies in oat phytochrome A). I700 is formed — after initial relaxations in the excited state — in about 15 to 20 ps (Heyne et al., 2002), a surprisingly slow rate compared to the primary photochemistry in other photoreceptors, including the photoactivation of Pfr (Heyne et al., 2002; van der Horst and Hellingwerf, 2004). Subsequently, one (in PhyA: Zhang et al., 1992; Michler and Braslavsky, 2001), or at least two intermediates (in Cph1: van Thor et al., 2001; Borucki et al., 2005a) are formed that are blue-shifted with respect to Pr. In this sense phytochromes follow the temporal order of color changes of the other two families that contain an isomerizable chromophore, the rhodopsins and the xanthopsins (Hellingwerf et al., 1996). Often, however, more complex photocycle schemes have to be proposed (like the occurrence of similar, parallel intermediates, or of equilibria between successive intermediates) to explain the details of the recorded spectral changes. In parallel to the formation and decay of the longest-living intermediate in this process, protonation changes take place in both the chromophore and the apo-protein, presumably in part to facilitate proton transfer from the C- to the D-ring of the extended linear tetrapyrrole. There is, however, also net proton release upon formation of Pfr, which results from the conformational
transition in the CBD. However, using heteronuclear NMR studies it was shown that the chromophore is protonated in both the Pr and Pfr states (Strauss et al., 2005). Recent quantum chemical calculations support the view that no proton transfer from the chromophore to the apo-protein takes place after activation (Durbeej and Eriksson, 2006). With continuous illumination of a Cph1 sample with ~630 nm light, the protein can be converted almost completely to the Pfr form. Photoactivation of Pfr with 727 nm light leads to an ultrafast re-isomerization of the bilin chromophore to form a first transient ground-state intermediate (tentatively referred to here as Ifr) with a quantum yield of 0.16 (Lamparter et al., 1997). The kinetics of this reaction is biphasic, with time constants of 0.5 and 3 ps, respectively (Heyne et al., 2002). The ultra-fast nature of this reaction — and instrument limitations — precluded observations of possible excited state processes involved in the photoactivation of Pfr. Time-resolved and temperature trapping experiments have revealed that following Ifr, also a blue-shifted transient intermediate is formed, prior to re-formation of Pr. In parallel, the state of protonation of the bilin chromophore must be adjusted. These latter reactions, however, have been characterized in little detail until now. C. Signaling-State Formation in Bacteriophytochromes Light-activated signal generation and transmission in plant phytochromes still contains many mysteries. It is not even clear whether or not these are true (auto)kinases, and if so, whether kinase activity serves signal transfer, signal adaptation or both (Kim et al., 2005). In the bacteriophytochromes the situation has been quite a bit further resolved. Several bacteriophytochromes contain a C-terminal histidine kinase domain. This domain is an integral member of one of the families of domains that constitute the so-called two-component systems. In line with this, light-regulated kinase activity has been demonstrated in several of these bacteriophytochromes (for a review see Rockwell et al., 2006), where the phosphorylated regulators often serve as transcriptional activators (see also above). The dominant mode of action seems to be Pfr-activated histidine kinase activity. Nevertheless, the variety in detail of this mechanism is impressive. In selected examples the CBD module (i.e., the combination of the PLD, GAF and PHY domains) activates and represses the kinase (or perhaps even
Chapter 41
Photoreceptor Proteins from Purple Bacteria
phosphatase) activity of the histidine kinase domain. But modulating activity is also observed on other types of output domains, like MCP domains (Zhulin, 2000; Bhaya, 2004; Fiedler et al., 2005), and domains that interact with the anti-repressor PpsR (Giraud et al., 2002); note that modulation of PP2C phosphatase and DUF domain activity has not yet been experimentally demonstrated. Furthermore, besides Pfr as the activated form, other family members have Pr as the activated form. Moreover, thermal darkreversion rates vary significantly between different members of the (cyano)bacteriophytochrome family. Which exact characteristic of the above applies to a specific (cyano)bacteriophytochrome family member has to be determined separately for each individual member. However, sequence similarity with other members of the family will often suggest a good working hypothesis. An approximate picture of the structural basis of signaling state formation in the input domains of the ‘classical’ Pr → Pfr type bacteriophytochromes can be constructed by combining the spectroscopic (and NMR) data obtained with Cph1, the use of C15=16 cis- and trans-locked chromophores (Inomata et al., 2005), the spatial structure of the CBD of the Pr form of BphP from Deinococcus radiodurans (Rockwell et al., 2006), and recent ultrafast mid-infrared spectroscopy (van Thor et al., 2007). Here light absorption alters the configuration of the cyanophycobilin chromophore from the 5Zsyn,10Zsyn,15Zanti- to the 5Zsyn,10Zsyn,15Eanti-configuration, i.e., a cis to trans isomerization of its C15=16 double bond. Additional single bond rotations have been invoked in this process (either between the A and B ring or between the C and D ring), but may be restricted by a possible clash of the D ring with surrounding amino acids like Y176, Y263 or even D207 of BphP. Rotation of the D ring due to isomerization of the C15=16 double bond is facilitated by the generally sparse packing of side chains around this ring: the residues that line the pocket of ring D are not tightly packed against this ring. Reorientation of the D ring may lead to disruption of the hydrogen bond to H290 and the van der Waals contact with e.g., Y263. This in turn could lead to a conformational change on the surface of the GAF domain in a surface patch that is composed of D207, Y263 and F203, and is conserved in many different phytochromes. Through this patch, information may be relayed to a downstream (kinase) domain. This and related models can only be rigorously tested by resolving the spatial structure of the Pfr
831
intermediate. It remains to be determined to what extent the above details will apply to other bacteriophytochromes, like those with strongly shifted spectral characteristics. It is even more uncertain what this will teach us with respect to the molecular basis of the function of the significant number of non-photochromic bacteriophytochromes and (bacterio)phytochrome-related proteins. VII. Concluding Remarks Progress in the field of the structure/function relationships in light-sensing signal input domains of photoreceptor proteins has been very impressive during the past five years, thanks in large part to the combined application of structural and spectroscopic techniques. Detailed descriptions are now available of the primary photochemistry of the chromophores and the initial dynamics of the surrounding apoprotein structures of the photosensory receptors discussed in this chapter, i.e., LOV, BLUF, PYP and PHY domains. Furthermore, in part owing to the application of a range of ‘omics’ techniques, detailed insight is also available at the physiological level about the biological processes triggered by light absorption and the signal-transduction chain leading to this response. The anti-transcriptional regulator AppA from Rba. sphaeroides serves as a good example in this respect (Hendriks et al., 2007; for a discussion of the role of AppA, see also Chapter 35, Bauer et al.). A much tougher problem to solve is obtaining insight into the middle part of these signal transduction chains, i.e., the structural basis of the transmission of signals between the separate domains in multi-domain proteins. Progress in this field hinges on our ability to crystallize these multi-domain proteins. Such progress is important if we want to quantitatively understand signaling, which is essential for analyzing how different signal-transduction chains (may) interact. Nevertheless, it is important to keep in mind that the diversity of the photosensory receptors in purple bacteria may be richer than the representatives of the four families discussed in this review. In the introduction to this review we have already emphasized the possibility that purple bacteria may contain members of the Rhodopsin and/or the Cryptochrome family. Nevertheless, more families of photoreceptors may exist and may have members in this sub-domain of the Bacteria. Proteins related to the orange carotenoid
832
J. Hendriks, M. A. van der Horst, T. K. Chua, M. Á. Pérez, L. J. van Wilderen, M. T. A. Alexandre, M.-L. Groot, J. T. M. Kennis and K. J. Hellingwerf
protein (Kirilovsky, 2007) may be candidates. More and more one realizes that photosensory receptors are abundant also in chemotrophic (micro) organisms (van der Horst et al., 2007). This fact considerably broadens the scope and impact for investigation of the detailed molecular characteristics of these receptors and calls for the further development of methodology to achieve this to the benefit of our better understanding of both types of organisms. Acknowledgments The research of M.L.G, J.T.M.K. and K.J.H. is generously supported by grants from the Life Sciences Council of the Nederlandse Organisatie voor Wetenschappelijk Onderzoek (NWO-ALW). References Alexandre MTA, Arents J, van Grondelle R, Hellingwerf KJ and Kennis JTM (2007) A base-catalyzed mechanism for dark state recovery in the Avena sativa phototropin 1 LOV 2 domain. Biochemistry 46: 3129–3137 Anderson S, Dragnea V, Masuda S, Ybe J, Moffat K and Bauer C (2005) Structure of a novel photoreceptor, the BLUF domain of AppA from Rhodobacter sphaeroides. Biochemistry 44: 7998–8005 Ataka K, Hegemann P and Heberle J (2003) Vibrational spectroscopy of an algal Phot-LOV1 domain probes the molecular changes associated with blue-light reception. Biophys J 84: 466–474 Avila-Pérez M, Hellingwerf KJ and Kort R (2006) Blue light activates the sigma(B)-dependent stress response of Bacillus subtilis via YtvA. J Bacteriol 188: 6411–6414 Baca M, Borgstahl GE, Boissinot M, Burke PM, Williams DR, Slater KA and Getzoff ED (1994) Complete chemical structure of photoactive yellow protein: Novel thioester-linked 4-hydroxycinnamyl chromophore and photocycle chemistry. Biochemistry 33: 14369–14377 Baltuška A, van Stokkum IHM, Kroon A, Monshouwer R, Hellingwerf KJ and van Grondelle R (1997) The primary events in the photoactivation of yellow protein. Chem Phys Lett 270: 263–266 Bhaya D (2004) Light matters: Phototaxis and signal transduction in unicellular cyanobacteria. Mol Microbiol 53: 745–754 Borgstahl GE, Williams DR and Getzoff ED (1995) 1.4 Å structure of photoactive yellow protein, a cytosolic photoreceptor: Unusual fold, active site, and chromophore. Biochemistry 34: 6278–6287 Borucki B, von Stetten D, Seibeck S, Lamparter T, Michael N, Mroginski MA, Otto H, Murgida DH, Heyn MP and Hildebrandt P (2005a) Light-induced proton release of phytochrome is coupled to the transient deprotonation of the tetrapyrrole chromophore. J Biol Chem 280: 34358–34364 Borucki B, Kyndt JA, Joshi CP, Otto H, Meyer TE, Cusanovich
MA and Heyn MP (2005b) Effect of salt and pH on the activation of photoactive yellow protein and gateway mutants Y98Q and Y98F. Biochemistry 44: 13650–13663 Buttani V, Losi A, Polverini E and Gartner W (2006) Blue news: NTP binding properties of the blue-light sensitive YtvA protein from Bacillus subtilis. FEBS Lett 580: 3818–3822 Changenet-Barret P, Espagne A, Charier S, Baudin JB, Jullien L, Plaza P, Hellingwerf KJ and Martin MM (2004) Early molecular events in the photoactive yellow protein: Role of the chromophore photophysics. Photochem Photobiol Sci 3: 823–829 Changenet P, Zhang H, van der Meer MJ, Hellingwerf KJ and Glasbeek M (1998) Subpicosecond fluorescence upconversion measurements of primary events in yellow proteins. Chem Phys Lett 282: 276–282 Cho H-Y, Tseng T-S, Kaiserli E, Sullivan S, Christie JM and Briggs WR (2007) Physiological roles of the LOV domains of phototropin 1 and phototropin 2 in Arabidopsis. Plant Physiol 143: 517–529 Chosrowjan H, Mataga N, Nakashima N, Imamoto Y and Tokunaga F (1997) Femtosecond-picosecond fluorescence studies on excited state dynamics of photoactive yellow protein from Ectothiorhodospira halophila. Chem Phys Lett 270: 267–272 Chosrowjan H, Mataga N, Shibata Y, Imamoto Y and Tokunaga F (1998) Environmental effects on the femtosecond-picosecond fluorescence dynamics of photoactive yellow protein: Chromophores in aqueous solutions and in protein nanospaces modified by site-directed mutagenesis. J Phys Chem B 102: 7695–7698 Christie JM and Briggs WR (2005) Blue light sensing and signaling by the phototropins. In: Briggs WR, Spudich JL (eds) Handbook of Photosensory Receptors, pp 277–303. Wiley-VCH Verlag GmbH & Co., Weinheim Christie JM, Swartz TE, Bogomolni RA and Briggs WR (2002) Phototropin LOV domains exhibit distinct roles in regulating photoreceptor function. Plant J 32: 205–219 Crosson S (2005) LOV-domain structure, dynamics, and diversity. In: Briggs WR, Spudich JL (eds) Handbook of Photosensory Receptors, pp 323–336. Wiley-VCH Verlag GmbH & Co., Weinheim Crosson S and Moffat K (2001) Structure of a flavin-binding plant photoreceptor domain: Insights into light-mediated signal transduction. Proc Natl Acad Sci USA 98: 2995–3000 Crosson S and Moffat K (2002) Photoexcited structure of a plant photoreceptor domain reveals a light-driven molecular switch. Plant Cell 14: 1067–1075 Crosson S, Rajagopal S and Moffat K (2003) The LOV domain family: Photoresponsive signaling modules coupled to diverse output domains. Biochemistry 42: 2–10 Cusanovich MA and Meyer TE (2003) Photoactive yellow protein: A prototypic PAS domain sensory protein and development of a common signaling mechanism. Biochemistry 42: 4759–4770 Davis SJ, Vener AV and Vierstra RD (1999) Bacteriophytochromes: Phytochrome-like photoreceptors from nonphotosynthetic eubacteria. Science 286: 2517–2520 Demchuk E, Genick UK, Woo TT, Getzoff ED and Bashford D (2000) Protonation states and pH titration in the photocycle of photoactive yellow protein. Biochemistry 39: 1100–1113 Devanathan S, Pacheco A, Ujj L, Cusanovich M, Tollin G, Lin S and Woodbury N (1999) Femtosecond spectroscopic observations of initial intermediates in the photocycle of the photoactive yellow protein from Ectothiorhodospira halophila. Biophys J
Chapter 41
Photoreceptor Proteins from Purple Bacteria
77: 1017–1023 Dunlap JC and Loros JJ (2005) Neurospora photoreceptors In: Briggs WR, Spudich JL (eds) Handbook of Photosensory Receptors, pp. 371–389. Wiley-VCH Verlag GmbH & Co., Weinheim Durbeej B and Eriksson LA (2006) Protein-bound chromophores astaxanthin and phytochromobilin: Excited state quantum chemical studies. Phys Chem Chem Phys 8: 4053–4071 Eitoku T, Nakasone Y, Matsuoka D, Tokutomi S and Terazima M (2005) Conformational dynamics of phototropin 2 LOV2 domain with the linker upon photoexcitation. J Am Chem Soc 127: 13238–13244 Fedorov R, Schlichting I, Hartmann E, Domratcheva T, Fuhrmann M and Hegemann P (2003) Crystal structures and molecular mechanism of a light-induced signaling switch: The Phot-LOV1 domain from Chlamydomonas reinhardtii. Biophys J 84: 2474–2482 Fiedler B, Borner T and Wilde A (2005) Phototaxis in the cyanobacterium Synechocystis sp PCC 6803: Role of different photoreceptors. Photochem Photobiol 81: 1481–1488 Freddolino PL, Dittrich M and Schulten K (2006) Dynamic switching mechanisms in LOV1 and LOV2 domains of plant phototropins. Biophys J 91: 3630–3639 Fukushima Y, Okajima K, Shibata Y, Ikeuchi M and Itoh S (2005) Primary intermediate in the photocycle of a blue-light sensory blue-light FAD-protein, Tll0078, of Thermosynechococcus elongatus BP-1. Biochemistry 44: 5149–5158 Gauden M, Yeremenko S, Laan W, van Stokkum IHM, Ihalainen JA, van Grondelle R, Hellingwerf KJ and Kennis JTM (2005) Photocycle of the flavin-binding photoreceptor AppA, a bacterial transcriptional antirepressor of photosynthesis genes. Biochemistry 44: 3653–3662 Gauden M, van Stokkum IHM, Key JM, Luhrs DC, Van Grondelle R, Hegemann P and Kennis JTM (2006) Hydrogen-bond switching through a radical pair mechanism in a flavin-binding photoreceptor. Proc Natl Acad Sci USA 103: 10895–10900 Gauden M, Grinstead JS, Laan W, van Stokkum HM, Avila-Pérez M, Toh KC, Boelens R, Kaptein R, van Grondelle R, Hellingwerf KJ and Kennis JTM (2007) On the role of aromatic side chains in the photoactivation of BLUF domains. Biochemistry 46: 7405–7415 Genick UK, Borgstahl GE, Ng K, Ren Z, Pradervand C, Burke PM, Srajer V, Teng TY, Schildkamp W, McRee DE, Moffat K and Getzoff ED (1997a) Structure of a protein photocycle intermediate by millisecond time-resolved crystallography. Science 275: 1471–1475 Genick UK, Devanathan S, Meyer TE, Canestrelli IL, Williams E, Cusanovich MA, Tollin G and Getzoff ED (1997b) Active site mutants implicate key residues for control of color and light cycle kinetics of photoactive yellow protein. Biochemistry 36: 8–14 Gensch T, Gradinaru CC, van Stokkum IHM, Hendriks J, Hellingwerf KJ and van Grondelle R (2002) The primary photoreaction of photoactive yellow protein (PYP): Anisotropy changes and excitation wavelength dependence. Chem Phys Lett 356: 347–354 Getzoff ED, Gutwin KN and Genick UK (2003) Anticipatory active-site motions and chromophore distortion prime photoreceptor PYP for light activation. Nat Struct Biol 10: 663–668 Giraud E, Fardoux L, Fourrier N, Hannibal L, Genty B, Bouyer P, Dreyfus B and Verméglio A (2002) Bacteriophytochrome
833
controls photosystem synthesis in anoxygenic bacteria. Nature 417: 202–205 Giraud E, Zappa S, Vuillet L, Adriano JM, Hannibal L, Fardoux J, Berthomieu C, Bouyer P, Pignol D and Verméglio A (2005) A new type of bacteriophytochrome acts in tandem with a classical bacteriophytochrome to control the antennae synthesis in Rhodopseudomonas palustris. J Biol Chem 280: 32389–32397 Gomelsky M and Klug G (2002) BLUF: A novel FAD-binding domain involved in sensory transduction in microorganisms. Trends Biochem Sci 27: 497–500 Grinstead JS, Hsu ST, Laan W, Bonvin AM, Hellingwerf KJ, Boelens R and Kaptein R (2006a) The solution structure of the AppA BLUF domain: Insight into the mechanism of lightinduced signaling. ChemBioChem 7: 187–193 Grinstead JS, Avila-Pérez M, Hellingwerf KJ, Boelens R and Kaptein R (2006b) Light-induced flipping of a conserved glutamine sidechain and its orientation in the AppA BLUF domain. J Am Chem Soc 128: 15066–15067 Groot ML, van Wilderen L, Larsen DS, van der Horst MA, van Stokkum IHM, Hellingwerf KJ and van Grondelle R (2003) Initial steps of signal generation in photoactive yellow protein revealed with femtosecond mid-infrared spectroscopy. Biochemistry 42: 10054–10059 Guo HM, Kottke T, Hegemann P and Dick B (2005) The Phot LOV2 domain and its interaction with LOV1. Biophys J 89: 402–412 Haker A, Hendriks J, Gensch T, Hellingwerf K and Crielaard W (2000) Isolation, reconstitution and functional characterisation of the Rhodobacter sphaeroides photoactive yellow protein. FEBS Lett 486: 52–56 Haker A, Hendriks J, van Stokkum IHM, Heberle J, Hellingwerf KJ, Crielaard W and Gensch T (2003) The two photocycles of photoactive yellow protein from Rhodobacter sphaeroides. J Biol Chem 278: 8442–8451 Han YC, Braatsch S, Osterloh L and Klug G (2004) A eukaryotic BLUF domain mediates light-dependent gene expression in the purple bacterium Rhodobacter sphaeroides 2.4.1. Proc Natl Acad Sci USA 101: 12306–12311 Han YC, Meyer MHF, Keusgen M and Klug G (2007) A haem cofactor is required for redox and light signalling by the AppA protein of Rhodobacter sphaeroides. Mol Microbiol 64: 1090–1104 Harper SM, Neil LC and Gardner KH (2003) Structural basis of a phototropin light switch. Science 301: 1541–1544 Harper SM, Christie JM and Gardner KH (2004) Disruption of the LOV-J alpha helix interaction activates phototropin kinase activity. Biochemistry 43: 16184–16192 Hasegawa K, Masuda S and Ono T (2006) Light induced structural changes of a full-length protein and its BLUF domain in YcgF(Blrp), a blue-light sensing protein that uses FAD (BLUF). Biochemistry 45: 3785–3793 Hellingwerf KJ (2005) Bacterial observations: A rudimentary form of intelligence? Trends Microbiol 13: 152–158 Hellingwerf KJ, Hoff WD and Crielaard W (1996) Photobiology of microorganisms: How photosensors catch a photon to initialize signalling. Mol Microbiol 21: 683–693 Hellingwerf KJ, Hendriks J and Gensch T (2003) Photoactive Yellow Protein, a new type of photoreceptor protein: Will this ‘yellow lab’ bring us where we want to go? J Phys Chem A 107: 1082–1094
834
J. Hendriks, M. A. van der Horst, T. K. Chua, M. Á. Pérez, L. J. van Wilderen, M. T. A. Alexandre, M.-L. Groot, J. T. M. Kennis and K. J. Hellingwerf
Hendriks J, Hoff WD, Crielaard W and Hellingwerf KJ (1999a) Protonation/deprotonation reactions triggered by photoactivation of photoactive yellow protein from Ectothiorhodospira halophila. J Biol Chem 274: 17655–17660 Hendriks J, van Stokkum IH, Crielaard W and Hellingwerf KJ (1999b) Kinetics of and intermediates in a photocycle branching reaction of the photoactive yellow protein from Ectothiorhodospira halophila. FEBS Lett 458: 252–256 Hendriks J, van Stokkum IHM and Hellingwerf KJ (2003) Deuterium isotope effects in the photocycle transitions of the photoactive yellow protein. Biophys J 84: 1180–1191 Hendriks J, Avila-Pérez M, Bruggeman F and Hellingwerf KJ (2007) Photosensory proteins as a tool in synthetic biology: Bridging computational biophysics and systems biology. In: Hansmann UHE, Meinke J, Mohanty S and Zimmermann O (eds) From Computational Biophysics to Systems Biology (CBSB07), Vol 36, pp 149–153. John von Neumann Institute for Computing, Juelich Heyne K, Herbst J, Stehlik D, Esteban B, Lamparter T, Hughes J and Diller R (2002) Ultrafast dynamics of phytochrome from the cyanobacterium Synechocystis, reconstituted with phycocyanobilin and phycoerythrobilin. Biophys J 82: 1004–1016 Heyne K, Mohammed OF, Usman A, Dreyer J, Nibbering ETJ and Cusanovich MA (2005) Structural evolution of the chromophore in the primary stages of trans/cis isomerization in photoactive yellow protein. J Am Chem Soc 127: 18100–18106 Hoff WD, Dux P, Hard K, Devreese B, Nugteren-Roodzant IM, Crielaard W, Boelens R, Kaptein R, van Beeumen J and Hellingwerf KJ (1994) Thiol ester-linked p-coumaric acid as a new photoactive prosthetic group in a protein with rhodopsin-like photochemistry. Biochemistry 33: 13959–13962 Hoff WD, Van Stokkum IHM, Gural J and Hellingwerf KJ (1997) Comparison of acid denaturation and light activation in the eubacterial blue-light receptor photoactive yellow protein. Biochim Biophys Acta-Bioenerg 1322: 151–162 Holzer W, Penzkofer A, Fuhrmann M and Hegemann P (2002) Spectroscopic characterization of flavin mononucleotide bound to the LOV1 domain of Phot1 from Chlamydomonas reinhardtii. Photochem Photobiol 75: 479–487 Holzer W, Penzkofer A and Hegemann P (2005) Absorption and emission spectroscopic characterisation of the LOV2-His domain of phot from Chlamydomonas reinhardtii. Chem Phys 308: 79–91 Huang KY and Beck CF (2003) Photoropin is the blue-light receptor that controls multiple steps in the sexual life cycle of the green alga Chlamydomonas reinhardtii. Proc Natl Acad Sci USA 100: 6269–6274 Ihee H, Rajagopal S, Srajer V, Pahl R, Anderson S, Schmidt M, Schotte F, Anfinrud PA, Wulff M and Moffat K (2005) Visualizing reaction pathways in photoactive yellow protein from nanoseconds to seconds. Proc Natl Acad Sci USA 102: 7145–7150 Imamoto Y, Kataoka M, Tokunaga F, Asahi T and Masuhara H (2001) Primary photoreaction of photoactive yellow protein studied by subpicosecond-nanosecond spectroscopy. Biochemistry 40: 6047–6052 Inomata K, Hammam MAS, Kinoshita H, Murata Y, Khawn H, Noack S, Michael N and Lamparter T (2005) Sterically locked synthetic bilin derivatives and phytochrome Agp1 from Agrobacterium tumefaciens form photoinsensitive Pr- and Pfr-like adducts. J Biol Chem 280: 24491–24497
Iseki M, Matsunaga S, Murakami A, Ohno K, Shiga K, Yoshida K, Sugai M, Takahashi T, Hori T and Watanabe M (2002) A blue-light-activated adenylyl cyclase mediates photoavoidance in Euglena gracilis. Nature 415: 1047–1051 Ito S, Murakami A, Sato K, Nishina Y, Shiga K, Takahashi T, Higashi S, Iseki M and Watanabe M (2005) Photocycle features of heterologously expressed and assembled eukaryotic flavinbinding BLUF domains of photoactivated adenylyl cyclase (PAC), a blue-light receptor in Euglena gracilis. Photochem Photobiol Sci 4: 762–769 Iwata T, Nozaki D, Tokutomi S, Kagawa T, Wada M and Kandori H (2003) Light-induced structural changes in the LOV2 domain of Adiantum phytochrome3 studied by low-temperature FTIR and UV-visible spectroscopy. Biochemistry 42: 8183–8191 Iwata T, Nozaki D, Tokutomi S and Kandori H (2005) Comparative investigation of the LOV1 and LOV2 domains in Adiantum phytochrome3. Biochemistry 44: 7427–7434 Jenal U and Malone J (2006) Mechanisms of Cyclic-di-GMP Signaling in Bacteria. Annu Rev Genet 40: 385–407 Jiang ZY, Swem LR, Rushing BG, Devanathan S, Tollin G and Bauer CE (1999) Bacterial photoreceptor with similarity to photoactive yellow protein and plant phytochromes. Science 285: 406–409 Joshi CP, Borucki B, Otto H, Meyer TE, Cusanovich MA and Heyn MP (2005) Photoreversal kinetics of the I-1 and I-2 intermediates in the photocycle of photoactive yellow protein by double flash experiments with variable time delay. Biochemistry 44: 656–665 Jung A, Domratcheva T, Tarutina M, Wu Q, Ko WH, Shoeman RL, Gomelsky M, Gardner KH and Schlichting L (2005) Structure of a bacterial BLUF photoreceptor: Insights into blue light-mediated signal transduction. Proc Natl Acad Sci USA 102: 12350–12355 Jung A, Reinstein J, Domratcheva T, Shoeman RL and Schlichting I (2006) Crystal structures of the AppA BLUF domain photoreceptor provide insights into blue light-mediated signal transduction. J Mol Biol 362: 717–732 Kaneko T and Tabata S (1997) Complete genome structure of the unicellular cyanobacterium Synechocystis sp. PCC6803. Plant Cell Physiol 38: 1171–1176 Karniol B, Wagner JR, Walker JM and Vierstra RD (2005) Phylogenetic analysis of the phytochrome superfamily reveals distinct microbial subfamilies of photoreceptors. Biochem J 392: 103–116 Kasahara M, Swartz TE, Olney MA, Onodera A, Mochizuki N, Fukuzawa H, Asamizu E, Tabata S, Kanegae H, Takano M, Christie JM, Nagatani A and Briggs WR (2002) Photochemical properties of the flavin mononucleotide-binding domains of the phototropins from Arabidopsis, rice, and Chlamydomonas reinhardtii. Plant Physiol 129: 762–773 Kay CWM, Schleicher E, Kuppig A, Hofner H, Rüdiger W, Schleicher M, Fischer M, Bacher A, Weber S and Richter G (2003) Blue light perception in plants — Detection and characterization of a light-induced neutral flavin radical in a C450A mutant of phototropin. J Biol Chem 278: 10973–10982 Kennis JTM and Alexandre MTA (2006) Mechanisms of light activation in flavin-binding photoreceptors. In: Silva E and Edwards AM (eds) Flavins: Photochemistry and Photobiology, Comprehensive Series in Photochemical and Photobiological Sciences Vol 6, pp 287–319. RSC Publishing, Cambridge Kennis JTM, van Stokkum IHM, Crosson S, Gauden M, Moffat K
Chapter 41
Photoreceptor Proteins from Purple Bacteria
and van Grondelle R (2004) The LOV2 domain of phototropin: A reversible photochromic switch. J Am Chem Soc 126: 4512–4513 Kim JI, Park JE, Zarate X and Song PS (2005) Phytochrome phosphorylation in plant light signaling. Photochem Photobiol Sci 4: 681–687 Kirilovsky D (2007) Photoprotection in cyanobacteria: The orange carotenoid protein (OCP)-related non-photochemical-quenching mechanism. Photosynth Res 93: 7–16 Kita A, Okajima K, Morimoto Y, Ikeuchi M and Miki K (2005) Structure of a cyanobacterial BLUF protein, Tll0078, containing a novel FAD-binding blue light sensor domain. J Mol Biol 349: 1–9 Koh M, Van Driessche G, Samyn B, Hoff WD, Meyer TE, Cusanovich MA and Van Beeumen JJ (1996) Sequence evidence for strong conservation of the photoactive yellow proteins from the halophilic phototrophic bacteria Chromatium salexigens and Rhodospirillum salexigens. Biochemistry 35: 2526–2534 Kottke T, Heberle J, Hehn D, Dick B and Hegemann P (2003a) Phot-LOV1: Photocycle of a blue-light receptor domain from the green alga Chlamydomonas reinhardtii. Biophys J 84: 1192–1201 Kottke T, Dick B, Fedorov R, Schlichting I, Deutzmann R and Hegemann P (2003b) Irreversible photoreduction of flavin in a mutated Phot-LOV1 domain. Biochemistry 42: 9854–9862 Kraft BJ, Masuda S, Kikuchi J, Dragnea V, Tollin G, Zaleski JM and Bauer CE (2003) Spectroscopic and mutational analysis of the blue-light photoreceptor AppA: A novel photocycle involving flavin stacking with an aromatic amino acid. Biochemistry 42: 6726–6734 Kyndt JA, Meyer TE and Cusanovich MA (2004a) Photoactive yellow protein, bacteriophytochrome, and sensory rhodopsin in purple phototrophic bacteria. Photochem Photobiol Sci 3: 519–530 Kyndt JA, Hurley JK, Devreese B, Meyer TE, Cusanovich MA, Tollin G and Van Beeumen JJ (2004b) Rhodobacter capsulatus photoactive yellow protein: Genetic context, spectral and kinetics characterization, and mutagenesis. Biochemistry 43: 1809–1820 Kyndt JA, Fitch JC, Meyer TE and Cusanovich MA (2005) Thermochromatium tepidum photoactive yellow protein/bacteriophytochrome/diguanylate cyclase: Characterization of the PYP domain. Biochemistry 44: 4755–4764 Laan W, van der Horst MA, van Stokkum IH and Hellingwerf KJ (2003) Initial characterization of the primary photochemistry of AppA, a blue-light-using flavin adenine dinucleotide-domain containing transcriptional antirepressor protein from Rhodobacter sphaeroides: A key role for reversible intramolecular proton transfer from the flavin adenine dinucleotide chromophore to a conserved tyrosine? Photochem Photobiol 78: 290–297 Laan W (2005) Signal sensing and transduction in the blue-light photoreceptor AppA and the cyanobacterial phytochrome Cph1. PhD thesis, University of Amsterdam Laan W, Bednarz T, Heberle J and Hellingwerf KJ (2004) Chromophore composition of a heterologously expressed BLUFdomain. Photochem Photobiol Sci 3: 1011–1016 Laan W, Gauden M, Yeremenko S, van Grondelle R, Kennis JTM and Hellingwerf KJ (2006) On the mechanism of activation of the BLUF domain of AppA. Biochemistry 45: 51–60 Lamparter T, Mittmann F, Gartner W, Borner T, Hartmann E and
835
Hughes J (1997) Characterization of recombinant phytochrome from the cyanobacterium Synechocystis. Proc Natl Acad Sci USA 94: 11792–11797 Lamparter T, Michael N, Mittmann F and Esteban B (2002) Phytochrome from Agrobacterium tumefaciens has unusual spectral properties and reveals an N-terminal chromophore attachment site. Proc Natl Acad Sci USA 99: 11628–11633 Landgraf FT, Forreiter C, Pico AH, Lamparter T and Hughes J (2001) Recombinant holophytochrome in Escherichia coli. FEBS Lett 508: 459–462 Larsen DS and van Grondelle R (2005) Initial photoinduced dynamics of the photoactive yellow protein. Chem Phys Chem 6: 828–837 Larsen DS, van Stokkum IHM, Vengris M, van der Horst MA, de Weerd FL, Hellingwerf KJ and van Grondelle R (2004a) Incoherent manipulation of the photoactive yellow protein photocycle with dispersed pump-dump-probe spectroscopy. Biophys J 87: 1858–1872 Larsen DS, Vengris M, van Stokkum IHM, van der Horst MA, de Weerd FL, Hellingwerf KJ and van Grondelle R (2004b) Photoisomerization and photoionization of the photoactive yellow protein chromophore in solution. Biophys J 86: 2538–2550 Lee IR, Lee W and Zewail AH (2006) Primary steps of the photoactive yellow protein: Isolated chromophore dynamics and protein directed function. Proc Natl Acad Sci USA 103: 258–262 Losi A (2006) Flavin-based photoreceptors in bacteria. In: Silva E and Edwards AM (eds) Flavins: Photochemistry and Photobiology, Comprehensive Series in Photochemical and Photobiological Sciences Vol 6, pp 217–269. RSC Publishing, Cambridge Losi A and Braslavsky SE (2003) The time-resolved thermodynamics of the chromophore-protein interactions in biological photosensors as derived from photothermal measurements. Phys Chem Chem Phys 5: 2739–2750 Losi A, Quest B and Gartner W (2003) Listening to the blue: The time-resolved thermodynamics of the bacterial blue-light receptor YtvA and its isolated LOV domain. Photochem Photobiol Sci 2: 759–766 Masuda S and Bauer CE (2002) AppA is a blue light photoreceptor that antirepresses photosynthesis gene expression in Rhodobacter sphaeroides. Cell 110: 613–623 Masuda S, Hasegawa K, Ishii A and Ono T (2004) Light-induced structural changes in a putative blue-light receptor with a novel FAD binding fold sensor of blue-light using FAD (BLUF); Slr1694 of Synechocystis sp PCC6803. Biochemistry 43: 5304–5313 Masuda S, Hasegawa K and Ono T (2005a) Adenosine diphosphate moiety does not participate in structural changes for the signaling state in the sensor of blue-light using FAD domain of AppA. FEBS Lett 579: 4329–4332 Masuda S, Hasegawa K and Ono TA (2005b) Tryptophan at position 104 is involved in transforming light signal into changes of beta-sheet structure for the signaling state in the BLUF domain of AppA. Plant Cell Physiol 46: 1894–1901 Meyer TE (1985) Isolation and characterization of soluble cytochromes, ferredoxins and other chromophoric proteins from the halophilic phototrophic bacterium Ectothiorhodospira halophila. Biochim Biophys Acta 806: 175–183 Meyer TE, Yakali E, Cusanovich MA and Tollin G (1987) Properties of a water-soluble, yellow protein isolated from a halophilic
836
J. Hendriks, M. A. van der Horst, T. K. Chua, M. Á. Pérez, L. J. van Wilderen, M. T. A. Alexandre, M.-L. Groot, J. T. M. Kennis and K. J. Hellingwerf
phototrophic bacterium that has photochemical activity analogous to sensory rhodopsin. Biochemistry 26: 418–423 Meyer TE, Fitch JC, Bartsch RG, Tollin G and Cusanovich MA (1990) Soluble cytochromes and a photoactive yellow protein isolated from the moderately halophilic purple phototrophic bacterium, Rhodospirillum salexigens. Biochim Biophys Acta 1016: 364–370 Michler I and Braslavsky SE (2001) Time-resolved thermodynamic analysis of the oat phytochrome A phototransformation. A photothermal beam deflection study. Photochem Photobiol 74: 624–635 Mizutani Y, Tokutomi S and Kitagawa T (1994) Resonance Raman-spectra of the intermediates in phototransformation of large phytochrome — deprotonation of the chromophore in the bleached intermediate. Biochemistry 33: 153–158 Nozaki D, Iwata T, Ishikawa T, Todo T, Tokutomi S and Kandori H (2004) Role of Gln1029 in the photoactivation processes of the LOV2 domain in Adiantum phytochrome3. Biochemistry 43: 8373–8379 Nozue K, Kanegae T, Imaizumi T, Fukuda S, Okamoto H, Yeh KC, Lagarias JC and Wada M (1998) A phytochrome from the fern Adiantum with features of the putative photoreceptor NPH1. Proc Natl Acad Sci USA 95: 15826–15830 Ntefidou M, Iseki M, Watanabe M, Lebert M and Hader DP (2003) Photoactivated adenylyl cyclase controls phototaxis in the flagellate Euglena gracilis. Plant Physiol 133: 1517–1521 Okajima K, Yoshihara S, Fukushima Y, Geng XX, Katayama M, Higashi S, Watanabe M, Sato S, Tabata S, Shibata Y, Itoh S and Ikeuchi M (2005) Biochemical and functional characterization of BLUF-type flavin-binding proteins of two species of cyanobacteria. J Biochem 137: 741–750 Okajima K, Fukushima Y, Suzuki H, Kita A, Ochiai Y, Katayama M, Shibata Y, Miki K, Noguchi T, Itoh S and Ikeuchi M (2006) Fate determination of the flavin photoreceptions in the cyanobacterial blue light receptor TePixD (T110078). J Mol Biol 363: 10–18 Otto H, Lamparter T, Borucki B, Hughes J and Heyn MP (2003) Dimerization and inter-chromophore distance of Cph1 phytochrome from Synechocystis, as monitored by fluorescence homo and hetero energy transfer. Biochemistry 42: 5885–5895 Perman B, Srajer V, Ren Z, Teng T, Pradervand C, Ursby T, Bourgeois D, Schotte F, Wulff M, Kort R, Hellingwerf K and Moffat K (1998) Energy transduction on the nanosecond time scale: early structural events in a xanthopsin photocycle. Science 279: 1946–1950 Quail PH, Boylan MT, Parks BM, Short TW, Xu Y and Wagner D (1995) Phytochromes — photosensory perception and signaltransduction. Science 268: 675–680 Rajagopal S and Moffat K (2003) Crystal structure of a photoactive yellow protein from a sensor histidine kinase: Conformational variability and signal transduction. Proc Natl Acad Sci USA 100: 8037–8037 Rajagopal S, Key JM, Purcell EB, Boerema DJ and Moffat K (2004) Purification and initial characterization of a putative blue light-regulated phosphodiesterase from Escherichia coli. Photochem Photobiol 80: 542–547 Remberg A, Lindner I, Lamparter T, Hughes J, Kneip C, Hildebrandt P, Braslavsky SE, Gartner W and Schaffner K (1997) Raman spectroscopic and light-induced kinetic characterization of a recombinant phytochrome of the cyanobacterium Synechocystis. Biochemistry 36: 13389–13395
Rockwell NC, Su YS and Lagarias JC (2006) Phytochrome structure and signaling mechanisms. Annu Rev Plant Biol 57: 837–858 Romling U and Amikam D (2006) Cyclic di-GMP as a second messenger. Curr Op Microbiol 9: 218–228 Romling U, Gomelsky M and Galperin MY (2005) C-di-GMP: The dawning of a novel bacterial signalling system. Mol Microbiol 57: 629–639 Ryan WL, Gordon DJ and Levy DH (2002) Gas-phase photochemistry of the photoactive yellow protein chromophore trans-p-coumaric acid. J Am Chem Soc 124: 6194–6201 Schleicher E, Kowalczyk RM, Kay CWM, Hegemann P, Bacher A, Fischer M, Bittl R, Richter G and Weber S (2004) On the reaction mechanism of adduct formation in LOV domains of the plant blue-light receptor phototropin. J Am Chem Soc 126: 11067–11076 Schuttrigkeit TA, Kompa CK, Salomon M, Rüdiger W and MichelBeyerle ME (2003) Primary photophysics of the FMN binding LOV2 domain of the plant blue light receptor phototropin of Avena sativa. Chem Phys 294: 501–508 Shimizu N, Kamikubo H, Mihara K, Imamoto Y and Kataoka M (2002) Effect of organic anions on the photoreaction of photoactive yellow protein. J Biochem (Tokyo) 132: 257–263 Shimizu N, Imamoto Y, Harigai M, Kamikubo H, Yamazaki Y and Kataoka M (2006) pH-dependent equilibrium between long lived near-UV intermediates of photoactive yellow protein. J Biol Chem 281: 4318–4325 Strauss HM, Hughes J and Schmieder P (2005) Heteronuclear solution-state NMR studies of the chromophore in cyanobacterial phytochrome Cph1. Biochemistry 44: 8244–8250 Suetsugu N and Wada M (2005) Photoreceptor Gene Families in Lower Plants. In: Briggs WR, Spudich JL (eds) Handbook of Photosensory Receptors, pp 349–369. Wiley-VCH Verlag GmbH & Co., Weinheim Swartz TE and Bogomolni RE (2005) LOV-Domain Photochemistry. In: Briggs WR, Spudich JL (eds) Handbook of Photosensory Receptors, pp 305–323. Wiley-VCH Verlag GmbH & Co., Weinheim Ujj L, Devanathan S, Meyer TE, Cusanovich MA, Tollin G and Atkinson GH (1998) New photocycle intermediates in the photoactive yellow protein from Ectothiorhodospira halophila: picosecond transient absorption spectroscopy. Biophys J 75: 406–412 Unno M, Sano R, Masuda S, Ono TA and Yamauchi S (2005) Light-induced structural changes in the active site of the BLUF domain in AppA by Raman spectroscopy. J Phys Chem B 109: 12620–12626 Unno M, Masuda S, Ono TA and Yamauchi S (2006) Orientation of a key glutamine residue in the BLUF domain from AppA revealed by mutagenesis, spectroscopy, and quantum chemical calculations. J Am Chem Soc 128: 5638–5639 van Beeumen JJ, Devreese BV, van Bun SM, Hoff WD, Hellingwerf KJ, Meyer TE, McRee DE and Cusanovich MA (1993) Primary structure of a photoactive yellow protein from the phototrophic bacterium Ectothiorhodospira halophila, with evidence for the mass and the binding site of the chromophore. Protein Sci 2: 1114–1125 van Brederode ME, Gensch T, Hoff WD, Hellingwerf KJ and Braslavsky SE (1995) Photoinduced volume change and energy storage associated with the early transformations of the photoactive yellow protein from Ectothiorhodospira halophila.
Chapter 41
Photoreceptor Proteins from Purple Bacteria
Biophys J 68: 1101–1109 van der Horst MA and Hellingwerf KJ (2004) Photoreceptor proteins, ‘star actors of modern times’: A review of the functional dynamics in the structure of representative members of six different photoreceptor families. Accounts Chem Res 37: 13–20 van der Horst MA, van Stokkum IHM, Dencher NA and Hellingwerf KJ (2005a) Controlled reduction of the humidity induces a shortcut recovery reaction in the photocycle of photoactive yellow protein. Biochemistry 44: 9160–9167 van der Horst MA, Laan W, Yeremenko S, Wende A, Palm P, Oesterhelt D and Hellingwerf KJ (2005b) From primary photochemistry to biological function in the blue-light photoreceptors PYP and AppA. Photochem Photobiol Sci 4: 688–693 van der Horst MA, Key J and Hellingwerf KJ (2007) Photosensing in chemotrophic, non-phototrophic bacteria: let there be light sensing too. Trends Microbiol 15: 554–562 van Thor JJ, Borucki B, Crielaard W, Otto H, Lamparter T, Hughes J, Hellingwerf KJ and Heyn MP (2001) Light-induced proton release and proton uptake reactions in the cyanobacterial phytochrome Cph1. Biochemistry 40: 11460–11471 van Thor JJ, Roynane KL and Towrie M (2007) Formation of the early photoproduct Lumi-R of cyanobacterial phytochrome Cph1 observed by ultrafast mid-infrared spectroscopy. J Am Chem Soc 129: 126–132
837
van Wilderen L, Van der Horst MA, van Stokkum IHM, Hellingwerf KJ, van Grondelle R and Groot ML (2006) Ultrafast infrared spectroscopy reveals a key step for successful entry into the photocycle for photoactive yellow protein. Proc Natl Acad Sci USA 103: 15050–15055 Wagner JR, Brunzelle JS, Forest KT and Vierstra RD (2005) A light-sensing knot revealed by the structure of the chromophorebinding domain of phytochrome. Nature 438: 325–331 Xie A, Kelemen L, Hendriks J, White BJ, Hellingwerf KJ and Hoff WD (2001) Formation of a new buried charge drives a large-amplitude protein quake in photoreceptor activation. Biochemistry 40: 1510–1517 Yeremenko S, van Stokkum IHM, Moffat K and Hellingwerf KJ (2006) Influence of the crystalline state on photoinduced dynamics of photoactive yellow protein studied by ultraviolet-visible transient absorption spectroscopy. Biophys J 90: 4224–4235 Zhang CF, Farrens DL, Bjorling SC, Song PS and Kliger DS (1992) Time-resolved absorption studies of native etiolated oat phytochrome. J Am Chem Soc 114: 4569–4580 Zhulin IB (2000) A novel phototaxis receptor hidden in the cyanobacterial genome. J Mol Microbiol Biotechnol 2: 491–493 Zirak P, Penzkofer A, Schiereis T, Hegemann P, Jung A and Schlichting I (2006) Photodynamics of the small BLUF protein BlrB from Rhodobacter sphaeroides. J Photochem Photobiol B-Biol 83: 180–194
Chapter 42 Foreign Gene Expression in Photosynthetic Bacteria Philip D. Laible*, Donna L. Mielke and Deborah K. Hanson Biosciences Division, Argonne National Laboratory, 9700 South Cass Avenue, Argonne, IL 60439-4803, U.S.A. Summary ............................................................................................................................................................... 840 I. Introduction .................................................................................................................................................... 840 A. Membrane Proteins: Essential, Yet Elusive ..................................................................................... 840 B. Photosynthetic Bacteria: Attractive Vehicles for Expression of Foreign Membrane Proteins .......... 840 II. Design of a Rhodobacter-Based System for the Expression of Membrane Proteins for Structural and Functional Studies .................................................................................................................................. 841 A. Vector Design Elements .................................................................................................................. 841 B. Auto-Induction of Heterologous Expression .................................................................................... 844 C. Host Design Considerations ............................................................................................................ 844 D. Host/Vector Combinations .............................................................................................................. 845 E. Detection and Quantitation of Expressed Proteins........................................................................... 846 F. Cellular Localization of Heterologously-Expressed Membrane Proteins ......................................... 846 III. Production of Foreign Membrane Proteins in Rhodobacter .......................................................................... 847 A. Comparison of Capabilities for Homologous and Heterologous Expression ................................... 847 B. Cloning and Expression of the Escherichia coli Membrane Proteome in Rhodobacter ................. 848 C. Heterologous Expression in Photosynthetic Cultures ...................................................................... 848 D. Purification of Heterologously-Expressed Membrane Proteins from the Rhodobacter Intracytoplasmic Membrane ............................................................................................................. 849 IV. Optimization and Generalization of Heterologous Expression in Rhodobacter ............................................ 850 A. Considerations for Heterologous Expression in Rhodobacter ........................................................ 850 1. Codon Usage ......................................................................................................................... 850 2. Modifications to Make Rhodobacter More User-Friendly as an Expression Host ................. 850 3. Features Influencing Transcript Stability ................................................................................ 851 4. Post-Translational Modifications ............................................................................................. 851 5. Capability for Incorporation of Selenomethionine ................................................................... 851 B. Adaptation to Higher Throughput ..................................................................................................... 851 C. Adopting Tactics from Other Systems Tailored to Membrane Protein Expression ......................... 851 D. Potential for a Cell-Free Expression System Based upon Rhodobacter ........................................ 852 V. Advantages Afforded by Rhodobacter .......................................................................................................... 852 A. Production of Natively-Folded Proteins ............................................................................................ 852 B. Utility for Proteins Whose Successful Expression Requires Multiple Gene Products ...................... 853 C. Simple Genetics Facilitate Interspecific Gene Expression and in vivo Functional Analysis............. 853 1. Metabolic Engineering............................................................................................................. 854 2. Interspecific Complementation of Functions Inherent to Rhodobacter .................................. 854 D. An Alternative Promoter for High-Level Expression ......................................................................... 854 E. Utility of a Common Bilayer .............................................................................................................. 854 F. Ability to Produce Different Membrane Samples for Structural and Functional Experiments........... 855 VI. Perspectives ................................................................................................................................................... 856 Acknowledgments ................................................................................................................................................ 856 References ........................................................................................................................................................... 856 *Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 839–860. © 2009 Springer Science + Business Media B.V.
840
Philip D. Laible, Donna L. Mielke and Deborah K. Hanson
Summary Large-scale production of membrane proteins in functional form is an arduous task. The overexpression of membrane proteins is fraught with unique challenges, and systems presently in use often fail to generate products which are compatible with subsequent solubilization, stabilization, and purification steps. To address these requirements, it is possible to exploit the unique physiology of the Rhodobacter species of photosynthetic bacteria, which produces extremely large quantities of internal membranes (invaginations of the cytoplasmic membrane) under certain growth conditions in response to changes in light intensity and/or oxygen tension. Towards this end, an expression system has been designed that coordinates synthesis of foreign membrane proteins with synthesis of new membrane into which they can be incorporated. These intracytoplasmic membrane (ICM) vesicles sequestering the newly synthesized foreign proteins are readily isolated by differential centrifugation following cell lysis. A diverse set of foreign membrane proteins – spanning a range of isoelectric points, molecular weights and predicted membrane topologies and representing several prokaryotic and eukaryotic species – has been expressed heterologously in Rhodobacter. Many target membrane proteins can be produced and purified in a semi-automated fashion at levels greater than 10 mg per liter of culture. This expression system is versatile, offering many strategies to achieve success in expression of membrane proteins. However, it is not necessarily restricted to the production of membrane proteins. For instance, Rhodobacter possesses advantages for the expression of soluble proteins that require an abundance of membrane surface (e.g., proteins only peripherally- or transiently-associated with the membrane) or proteins requiring complex redox cofactors which are native to Rhodobacter. In addition, Rhodobacter ICMs are easily isolated and are naturally enriched in target membrane proteins; thus, these membranes also facilitate biochemical studies, making this expression system valuable for a wide range of applications. I. Introduction A. Membrane Proteins: Essential, Yet Elusive The functions performed by membrane proteins are extremely important for all organisms. Despite the fact that membrane proteins represent approximately 30% of every genome and comprise more than 60% of all drug targets, only about 100 unique membrane protein structures have been determined to date (http://www.mpdb.ul.ie/; Raman et al., 2006), in contrast with unique structures representing approximately 9800 soluble protein families (Research Collaboratory for Structural Bioinformatics, 2008). A major factor influencing the paucity of membrane protein structures is that the expression levels of membrane proteins in native tissue are generally low. Abbreviations: DMSO – dimethyl sulfoxide; Escherichia – E.; FPLC – fast protein liquid chromatography; GFP – green fluorescent protein; HT – polyhistidine tag; ICM – intracytoplasmic membrane; IMAC – immobilized metal affinity chromatography; IPTG – isopropyl-β-D-thiogalactoside; LH1 – core light-harvesting antenna complex; LH2 – peripheral light-harvesting antenna complex; ORF – open reading frame; PCR – polymerase chain reaction; Rba. – Rhodobacter; RC – reaction center; SDS – sodium dodecyl sulfate; SDS-PAGE – sodium dodecyl sulfatepolyacrylamide gel electrophoresis; SeMet – selenomethionine; Tc – tetracycline
Prior to the year 2000, it is no coincidence that the majority of membrane protein structures were obtained for proteins which were purified from abundant natural sources, including many from photosynthetic bacteria (approximately 40%) (Laible et al., 2005b). While many membrane proteins have been isolated in functional form from their native host organisms, purification in such cases is highly protein-specific, is not adaptable to high-throughput methodologies, and rarely yields the amounts of pure membrane proteins that are needed for extensive biochemical studies and crystallization trials. The need for production of significant quantities of membrane proteins in functional form could be satisfied by their heterologous expression in a system that is capable of high yield. It would also be advantageous to design such a system from the start to be compatible with all subsequent steps of solubilization, stabilization and purification of the target membrane protein. B. Photosynthetic Bacteria: Attractive Vehicles for Expression of Foreign Membrane Proteins Photosynthetic bacteria are particularly capable of addressing the problem of membrane protein pro-
Chapter 42
Heterologous Expression in Rhodobacter
duction. Members of the Rhodobacter (Rba.) genus are facultative photoheterotrophs characterized by a metabolic diversity that allows them to adapt readily to a wide variety of environmental conditions. They are known to reduce nitrogen compounds, fix carbon dioxide, utilize carbon sources in an aerobic environment, grow photosynthetically under anaerobic conditions, or grow anaerobically in the dark in the presence of exogenous electron acceptors (reviewed in Imhoff, 1995). Both the mechanisms by which environmental cues are sensed and the biochemical machinery necessary to survive in a given setting are complex (Elsen et al., 2004; Roh et al., 2004). In particular, under conditions of light and/or lowered oxygen tension, the membrane surface in the organism increases many-fold as an ICM is elaborated as invaginations of the cytoplasmic membrane (Niederman et al., 1976). Concomitantly, the same environmental cues induce synthesis of the peripheral (LH2) and core (LH1) light-harvesting assemblies and the reaction center (RC), and the new ICM sequesters these complexes composed of transmembrane polypeptides and their associated hydrophobic redox and energy transfer cofactors (reviewed in Kiley and Kaplan, 1988). The inducible ICM is a particularly notable property of the Rhodobacter genus that makes these organisms extremely attractive candidates to serve as hosts for the expression of foreign membrane proteins. In Rhodobacter, the expression of the target membrane protein and formation of the new membrane into which it can be incorporated can be coordinated by placing expression of the target gene under control of a promoter(s) that drives the expression of a component of the native photosynthetic apparatus. In addition, the organisms may be cultured to high cell densities in the presence or absence of light on a variety of rich or defined minimal media, and the color of the cell culture reports that conditions leading to induction of both ICM synthesis and expression of the foreign membrane protein have been achieved. It is this ability to induce concomitant synthesis of new membrane and target membrane protein that distinguishes expression systems based on photosynthetic bacteria from other systems that were initially designed for the expression of soluble proteins and only later applied to membrane proteins. When employed for the expression of membrane proteins, the latter often fail because protein synthesis and membrane synthesis are not coupled, leading to the degradation of the expressed membrane proteins or the formation of insoluble aggregates or inclusion
841
bodies (Kiefer et al., 1999; Korepanova et al., 2005; Columbus et al., 2006). This chapter describes the features and implementation of an expression system based upon Rba. sphaeroides that is designed to overcome these limitations. II. Design of a Rhodobacter-Based System for the Expression of Membrane Proteins for Structural and Functional Studies The Rba. sphaeroides expression system is designed to take advantage of the organism’s unique inducible membrane to facilitate the production of membrane proteins in native form. To accomplish this, genes encoding the structural components of the photosynthetic machinery of the cell are replaced with the gene encoding the membrane protein of interest. By manipulating the growth conditions of the culture, it is easy to coordinate expression of the desired protein with the synthesis of the inducible membrane meant to accommodate the photosynthetic machinery. The expressed heterologous membrane protein thus has a membrane destination — the ICM — into which it can insert and assemble, thereby avoiding the often insurmountable complication of aggregation and formation of inclusion bodies. ICM vesicles enriched in heterologous membrane protein are readily isolated after cell lysis, enabling a straightforward purification of the desired membrane protein in native form for subsequent structural or functional studies. A. Vector Design Elements In designing vectors for expression of foreign genes, critical physiological characteristics of the native organism were considered. The complexes that constitute the photosynthetic apparatus differ quite dramatically in their relative abundances; genes that encode these complexes are localized to the puf, puc, and puh operons (Kiley and Kaplan, 1988). In the native ICM, the RC assembly is encompassed by the LH1 antenna complex that consists of 14–18 αβ heterodimers (Karrasch et al., 1995; Jamieson et al., 2002; Qian et al., 2003, 2005; Roszak et al., 2003). The peripheral LH2 complexes are rings made of 8–9 αβ heterodimers (McDermott et al., 1995; Koepke et al., 1996; Walz et al., 1998). Relative to the LH1 and RC complexes, the amount of the LH2 complex is variable, increasing dramatically in low
842
Philip D. Laible, Donna L. Mielke and Deborah K. Hanson
light (to up to 10–20 LH2 per LH1 and RC; Aagaard and Sistrom, 1972; Clayton, 1980; van Grondelle et al., 1983; Drews, 1985; Bahatyrova et al., 2004). Thus, if genes encoding these three complexes were replaced with foreign genes, expectations follow that replacement of the LH2 structural genes with foreign genes would lead to the highest expression level, replacement of LH1 structural genes would result in an intermediate level of expression, and replacement of RC structural genes would lead to relatively low expression of a foreign protein. The upper limits of these expectations are based on the amounts of protein that can be purified from wild-type strains expressing the native photosynthetic apparatus; in practice, RCs can be purified from chemoheterotrophic or photosynthetic cultures in yields of 10 mg l–1, and LH1 and LH2 complexes can be purified at levels exceeding 100 mg l–1. An initial set of platform vectors was constructed to take advantage of these differences in natural abundance (Fig. 1a–c), thereby allowing for high-level expression or modulation of the extent of expression of foreign proteins in cases where there is concern about toxicity, saturation of the secretory and assembly machinery, and/or formation of inclusion bodies. The pRKPLHT1 and pRKPLHT4 vectors were based on the puf operon and its promoter that responds to decreasing oxygen tension. The pRKPLHT7 vector was based on the puc operon whose promoter is activated by low light intensity and decreasing oxygen levels. In the pRKPLHT1 vector, the pufB and pufA structural genes that encode the LH1 subunits were replaced by a simple cassette carrying cloning sites followed by a C-terminal heptahistidine tag and two stop codons. In the pRKPLHT4 vector, an analogous cassette was used to replace the pufL and pufM genes that encode the L and M subunits of the reaction center. These two locations for foreign genes in the puf operon take advantage of the relative stoichiometry of the LH1 and RC polypeptides (~15:1). This difference is determined largely by differential transcript stabilities that are related to a region of RNA secondary structure elements located between the pufA and pufL genes (Fig. 1) that protects the LH1-portion of the transcript from exonuclease digestion (reviewed in Klug, 1995). Another analogous cassette was used to replace the pucB and pucA structural genes in the pRKPLHT7 vector. All vectors utilize transcription terminators present in the native operons. PCR-based cloning of foreign genes into these
expression vectors is directed by N-terminal oligonucleotides that incorporate a restriction site compatible with the vector, followed by a ribosome binding site (GGAGG, the consensus ribosome binding site of the photosynthetic gene cluster) (Naylor et al., 1999) spaced 6 –7 bases from the initiation codon. C-terminal oligonucleotides fuse a compatible restriction site in frame with the polyhistidine tag and stop codons. The same amplicon is compatible with all three platform vectors. All of these vectors are derivatives of the broadhost-range plasmid pRK404 (Ditta et al., 1985), which is maintained stably in trans in Rba. sphaeroides by selection for tetracycline resistance (1 mg l–1). This vector is relatively large (11.2 kb), and its copy number has been estimated at 4–6 per cell (Donohue and Kaplan, 1991). Smaller broad-host-range plasmids — derivatives of pBBR1 (Kovach et al., 1994, 1995) — were evaluated as expression vectors in attempts to increase both the copy number (and, thus, yield of expression) and the variety of antibiotic resistance markers that could be used in a Rhodobacter expression system. However, for reasons that are not understood, expression yields were lower, thus this strategy was not pursued further. At the outset of the process of expression vector design and construction, the sequence of pRK404 was largely unknown. Subsequently, its sequence was determined (Scott et al., 2003) in order to facilitate engineering of new generations of expression vectors. That inventory now includes vectors for either restriction enzyme-based or ligation-independent cloning (Dieckman et al., 2002) (Fig. 1f) that feature both cleavable and uncleavable N-terminal polyhistidine tags, as well as extended N- and C-terminal polyhistidine tags (10 × His, 13 × His) that enable increased availability and/or tighter binding of the tag to chromatography resin during purification by immobilized metal affinity chromatography (IMAC). Expression vectors have also been constructed which may serve to assist in membrane incorporation of foreign proteins expressed in Rhodobacter. Temporally-coordinated induction of both ICM and target protein may not always produce membrane incorporation of the foreign protein if its inherent membrane-targeting sequences are not recognized by Rba. sphaeroides. In such cases, appending peptides that are known to perform this function in Rba. sphaeroides can direct the foreign membrane protein to the ICM. To this end, the pRKPLHT1 platform vector was modified to incorporate the N-terminal
Chapter 42
Heterologous Expression in Rhodobacter
843
Fig. 1. Platform vectors for the Rhodobacter expression system, based on the broad-host-range vector pRK404, are derived from vectors described previously (Pokkuluri et al., 2002). Foreign genes are placed under control of the oxygen- and/or light-regulated puf (Ppuf) or puc (Ppuc) promoters. A region of stable RNA secondary structure (hairpin) dictates the stability of the upstream transcript. Vectors encode a C-terminal heptahistidine tag (HT) that is fused in frame with two stop codons (*). Vector d encodes an N-terminal membrane anchor/linker domain and vector e provides an N-terminal signal sequence. Restriction sites in bold type are unique. Foreign genes are inserted via the SpeI and BglII sites in vectors a-e; vector f carries a PmlI site that enables ligation-independent cloning (LIC). (a) pRKPLHT4; (b) pRKPLHT1; (c) pRKPLHT7; (d) pRKMAHT1; (e) pRKSSHT1; (f) pRKLICHT1.
membrane anchor/linker domain derived from cytochrome cy of Rba. capsulatus (Myllykallio et al., 1997) (Fig. 1d). This membrane-bound cytochrome can be expressed heterologously to high levels in Rba. sphaeroides (vide infra). In addition, another vector has been constructed to target foreign soluble
proteins or the N-terminus of foreign membrane proteins to the periplasmic space. This compartment also provides a reducing environment that can assist in protein assembly. This pRKPLHT1 derivative has been modified to include a segment that encodes the N-terminal signal sequence from Rba. sphaeroides
844
Philip D. Laible, Donna L. Mielke and Deborah K. Hanson
cytochrome c2, which is cleaved upon transport of that protein to the periplasm (MacGregor and Donohue, 1991) (Fig. 1e). After cloning steps are completed from DNA propagated in Escherichia (E.) coli hosts, expression plasmids are transferred to Rba. sphaeroides via conjugation. E. coli S17-1 (Simon et al., 1983) serves as the plasmid donor. Transconjugants are selected on minimal medium (modified Hutner’s medium; Gerhardt et al., 1994), then are purified on rich medium (YCC; Taguchi et al., 1992); plasmids are maintained with tetracycline (Tc; 1 mg l–1). B. Auto-Induction of Heterologous Expression Induction of expression of foreign genes cloned into the aforementioned platform vectors is controlled by oxygen and/or light, the same environmental cues that initiate synthesis of the ICM. Depending upon the host strain employed, cultures can be grown either photosynthetically or chemoheterotrophically. In photosynthetic cultures, anaerobiosis and/or light auto-induce the coupled synthesis. In chemoheterotrophic cultures (dark, 34 ºC, semi-aerobic), concomitant synthesis of ICM and heterologous protein is auto-induced when the oxygen tension lowers as the cell density increases. Even in chemoheterotrophic growth modes, Rhodobacter is a highly pigmented organism, and the natural pigmentation of semiaerobic Rhodobacter cultures can be exploited to indicate that the induction conditions which result in the concomitant synthesis of target membrane protein and new membrane have been achieved. Repression of the puf and puc promoters can be achieved under aerobic conditions, e.g., with rapid shaking of flasks carrying a relatively small amount of medium. This strategy can be employed when induction of ICM and heterologous protein synthesis needs to be controlled tightly, as in the case of expression of a toxic protein or incorporation of selenomethionine into induced proteins (Laible et al., 2005a). To achieve such conditions for small-scale expression screening, strains are grown under semi-aerobic, chemoheterotrophic conditions in 80 ml of YCC/Tc1 (Tc at 1 mg l–1) medium in 125 ml baffled Nephelo flasks shaking at 125 rpm. Larger-scale chemoheterotrophic cultures for protein production are grown in baffled 2.8 l Fernbach flasks containing 2 l of YCC/Tc1. When cultures are grown photosynthetically, small chemoheterotrophically-grown starter cultures are used to inoculate glass bottles that are
Table 1. Host strains of Rba. sphaeroides used in this study, differing in the complement of native proteins of the photosynthetic apparatus present in the ICM. Strain ∆∆11 PUC705-BA PUF∆LMX21 ATCC17023
RC − + − +
Native Complexes LH1 LH2 − − − + + + + +
filled completely with YCC/Tc1. Cultures are then incubated under anaerobic conditions, with stirring, at 32–34 ºC with ~ 200 W m–2 illumination from incandescent bulbs. Turbidity is monitored with a Klett-Summerson colorimeter; typically, chemoheterotrophic cultures are harvested when they reach 250–300 Klett units (2–4 days; OD600 ~2.5); photosynthetic cultures reach turbidities of > 450 Klett units (3–5 days; OD600 ~ 4). C. Host Design Considerations The expression strategy targets the inducible ICM of the Rhodobacter host for localization of the heterologously-expressed membrane proteins. Since the protein composition of the ICM can be engineered, the effect of deletion of native ICM proteins on ICM ultrastructure (morphology and volume) and on the level of incorporation of foreign membrane protein was investigated. To test the hypothesis that a partially-depleted ICM could accommodate more foreign membrane protein, four strains of Rba. sphaeroides were evaluated as hosts for heterologous expression (Table 1). These strains differ in both the nature and number of native complexes of the photosynthetic apparatus present in the ICM. They range from a true wild type to a strain that is deleted for the three transmembrane protein complexes of the photosynthetic apparatus: wild-type ATCC17023 (RC+LH1+LH2+; PS+; www.atcc.org), PUC705-BA (RC+LH1+LH2–; PS+; Lee et al., 1989), PUF∆LMX21 (RC–LH1+LH2+; PS–; Farchaus and Oesterhelt, 1989), and ∆∆11 (RC–LH1–LH2–; PS–; Pokkuluri et al., 2002). Electron microscopy studies show that in the wildtype organism the ICM appears as vesicles. The morphology of the ICM changes, however, as its protein content is manipulated. It is well known that deletion of the LH2 complex of Rba. sphaeroides yields a strain characterized by tubular membranes (Hunter et al., 1988; Kiley et al., 1988; Golecki and Heinrich,
Chapter 42
Heterologous Expression in Rhodobacter
1991; Golecki et al., 1991; Drews and Golecki, 1995; Fowler et al., 1995). Strains that synthesize the LH1 and LH2 complexes, but carry a deletion of the RC, look much like the native strain, and a strain lacking all three complexes of the photosynthetic apparatus (∆∆11) is characterized by a less structured ICM that is neither tubes nor spheres. Complementation of these deletions with genes encoding wild-type LH1/RC or LH2 complexes expressed recombinantly in trans restores the volume and many of the morphological features of the native ICM (data not shown). Similarly, spherical vesicles are restored in ∆∆11 carrying an expression plasmid producing foreign membrane proteins, and these data are corroborated by the observation that the foreign membrane protein produced in high yield is localized to the ICM (vide infra). The size of the membrane fraction in cell lysates is comparable in semi-aerobic cultures of all ∆∆11 expression strains. D. Host/Vector Combinations The ability of a combination of vector and host to produce quantities of foreign membrane proteins for structural and functional studies depends upon a number of factors. These include the ability to grow the culture to high cell density under conditions where foreign gene expression is coupled to ICM expression, leading to the incorporation of the expressed membrane protein into the ICM. To iden-
845
tify host/vector combinations where these criteria can be met adequately to produce useful quantities of foreign membrane proteins, two reporter proteins—the soluble green fluorescent protein (GFP) from Aequoria victoria (Yang et al., 1996) and the membrane-bound cytochrome cy of Rba. capsulatus (Myllykallio et al., 1997) — were cloned into the pRKPLHT1, pRKPLHT4, and pRKPLHT7 vectors and their expression levels were quantified in the four different host strains. Expression of the soluble GFP served to indicate whether chemoheterotrophic growth conditions were optimal for induction of the puf and puc promoters, while expression of the membrane-bound cytochrome cy indicated the level of membrane insertion of the expressed protein in a particular host strain. Cultures (2 l) of the expression strains were grown under chemoheterotrophic conditions, and amounts of IMAC-purified protein were determined by UV-visible spectroscopy. Figure 2 reveals the results of this analysis. The left panel shows that the synthesis of GFP was greatest from the pRKPLHT1 vector in each host, bearing out the expectation that replacement of LH1 genes by the foreign gene would result in higher yields than replacement of RC genes (pRKPLHT4) in the puf operon-based vectors. The yield of protein from the puc-based vector (pRKPLHT7) was expected to be the highest of the three platform vectors, and that expectation was borne out when yield was quantified on a per cell basis. However, oxygenation conditions
Fig. 2. Expression yields of soluble (GFP) and membrane-bound (cytochrome cy) reporter proteins from various host-vector combinations. Yields of GFP reported on success in achieving the proper culture conditions for auto-induction of expression from vectors where native genes of the photosynthetic apparatus were replaced with the foreign gene. Yields of cytochrome cy reported that membrane localization of the expressed foreign protein was maximized in host strains that were deleted for one or more of the native protein complexes of the intracytoplasmic membrane (Table 1).
846
Philip D. Laible, Donna L. Mielke and Deborah K. Hanson
that induce this promoter and produce respectable cell densities are more difficult to achieve in shake flasks, thus affecting the absolute yield of heterologous expression on a per volume basis. The right panel presents results from expression of cytochrome cy, again showing that expression is highest from the pRKPLHT1 vector. The results in this panel also underscore the fact that the wild-type host and the PUF∆LMX21 strain (retaining both light-harvesting complexes) accumulated lesser amounts of protein. Yields were highest in two deletion strains — ∆∆11 (fully deleted) and PUC705-BA (LH2 deleted) — indicating that strains with partiallydepleted membranes produce the greatest amounts of foreign membrane proteins when cultures are grown under these conditions. When all of the expression results from cytochrome cy and GFP are considered, the conclusion is that the highest yields of foreign protein are obtained with pRKPLHT1, the vector in which LH1 genes are replaced and expression is driven by the puf promoter. The host strain that produces the highest yield overall in chemoheterotrophic cultures is ∆∆11, a strain in which all three transmembrane complexes of the photosynthetic apparatus have been deleted. This vector-host combination is used below to express genes encoding hundreds of proteins, examining the versatility of Rhodobacter for expressing a wide variety of target membrane proteins. E. Detection and Quantitation of Expressed Proteins Neither native photosynthetic proteins nor overexpressed foreign membrane proteins are always clearly visible in gels stained with Coomassie Brilliant Blue. Also, membrane proteins often aggregate with heat and tend to bind sodium dodecyl sulfate (SDS) at different stoichiometries, thus they can migrate anomalously on denaturing polyacrylamide gels relative to molecular weight standards that are composed of soluble proteins. Consequently, heterologouslyexpressed membrane proteins are detected more unambiguously with immunoblotting techniques that recognize the common polyhistidine tag. To analyze expression in a large number of strains, generic screening methods were developed to determine whether a target protein is expressed successfully and incorporated into the ICM. Proteins from whole cell lysates and membrane fractions of aliquots of small-scale cultures were prepared for SDSpolyacrylamide gel electrophoresis (SDS-PAGE). Immunoblotting techniques were applied routinely
for detection and rough quantitation of expression levels in Rhodobacter of His-tagged foreign membrane proteins. Success in expressing a target membrane protein is measured by comparing the immunoblot signal from the target protein with that of a positive control protein that is known to be expressed at a level of 1 mg l–1 culture (+ control; Fig. 3). A recombinant strain carrying a platform vector lacking a cloned gene serves as the negative control (– control; Fig. 3). To be considered a ‘hit’, the signal from a target protein in the immunoblot must be equivalent to or greater than that of the positive control, indicating that it is expressing at > 1 mg pure protein per liter of culture. Purification of proteins expressed at levels below 1 mg l–1 culture is extremely cumbersome but could be pursued, depending upon the value of the target. The expression levels of many heterologous proteins rival those of native ICM proteins that can be purified at a yield of > 10 mg l–1 of culture. F. Cellular Localization of HeterologouslyExpressed Membrane Proteins The screening process includes determination of whether the foreign membrane proteins are incorporated into the ICM. This specialized membrane is contiguous with the cytoplasmic membrane, but differs from the latter in its chemical and protein composition, its morphological and physical properties, and in its kinetics of biogenesis (Parks and Niederman, 1978; Drews and Golecki, 1995; Verméglio et al., 1995). Upon cell lysis via mechanical breakage (e.g., French press, microfluidizer), the ICM invaginations break apart from the cytoplasmic membrane, becoming sealed inside-out vesicles (cytoplasmic face to the outside). Following the initial removal of cell debris, these vesicles are isolated easily by differential centrifugation. In the native organism, this fraction is rich in the integral membrane proteins that constitute the photosynthetic apparatus. In engineered expression strains, this fraction should contain the heterologous membrane protein, and its cellular localization can be tracked easily by using the polyhistidine tag. The immunoblot methods employed also report whether any His-tagged proteins are present in inclusion bodies or whether any have been cleaved by proteases in vivo. Target membrane proteins are found almost exclusively in the membranes. Thus, differential centrifugation is sufficient to afford facile (and substantial) purification of this membrane fraction from soluble proteins. In addition, ICM localization of the
Chapter 42
Heterologous Expression in Rhodobacter
847
Fig. 3. Detection and quantitation of expression levels of heterologously-expressed membrane proteins. Overexpressed proteins are not always clearly visible in Coomassie Brilliant Blue-stained SDS-PAGE gels (upper panel) in membrane extracts. Immunoblots (lower panel) are conclusive, general, and help identify anomalous mobility on gels. Expression levels of many target proteins (e.g., APC01198 and APC01217) rival or exceed levels attained by highly-expressed, native Rhodobacter ICM complexes (i.e. >10 mg protein l–1 of culture, purified). Positive control: Rba. sphaeroides membranes in which reaction center expression is limited to 1mg protein l–1 of culture. Negative control: membranes from a Rba. sphaeroides host strain carrying an empty expression vector. APC: a designator of targets for a structural genomics project.
expressed foreign protein is taken as an indicator that the protein possesses at least some degree of structural integrity that directs membrane insertion. Examination of various cell fractions for hundreds of expressed proteins has failed to produce clear evidence for the formation of inclusion bodies in any Rba. sphaeroides strain expressing foreign proteins. III. Production of Foreign Membrane Proteins in Rhodobacter A. Comparison of Capabilities for Homologous and Heterologous Expression The initial set of targets was composed of E. coli membrane proteins of unknown structure for which an ortholog could be identified in the Rba. capsulatus genome (www.ergo-light.com/ERGO), and served as a simple test of whether expression success would be
lower outside the genus of purple photosynthetic bacteria. Based on these results that indicated a relatively high success rate for expression of E. coli membrane proteins, the target set was expanded. The combination of platform vector pRKPLHT1 (puf promoter, LH1 genes replaced) and host strain ∆∆11 (RC–LH1–LH2–) has now been employed in a larger-scale effort to express target membrane proteins from several species. Rhodobacter is capable of expressing membrane proteins from a variety of other prokaryotes, as well as fungi (Aspergillus nidulans), plants (Arabidopsis thaliana), and humans. Its application to the expression of target membrane proteins across the three kingdoms is just beginning to be explored, and it has been employed in a few instances to express foreign soluble proteins (as discussed in Sections V, B and D). To date, it has been used most extensively for the production of membrane proteins that are targets of a structural genomics initiative focused upon prokaryotic pathogens (www.mcsg.anl.gov).
848
Philip D. Laible, Donna L. Mielke and Deborah K. Hanson
B. Cloning and Expression of the Escherichia coli Membrane Proteome in Rhodobacter One of the initial target organisms for this structural genomics project was E. coli. Genes representing the entire E. coli membrane proteome have now been cloned for expression in Rba. sphaeroides. Genes were selected for cloning and expression studies by bioinformatic analysis of the annotated E. coli genome (compbio.mcs.anl.gov/target; N. Maltsev, P. D. Laible, D. K. Hanson, unpublished). From the 1030 hydrophobic proteins identified, 444 were selected as representatives of unique membrane proteins encoded by this genome whose structure is unknown (i.e., they share <30% homology with any structures deposited in the Protein Data Bank). These targets — the majority of which are unannotated or ‘hypothetical’ membrane proteins—encompass a wide range of molecular weights, transmembrane topology, and isoelectric points. This set of genes has been cloned into the ∆∆11[pRKPLHT1] host-vector combination (Table 1; Fig. 1) and expression was auto-induced in chemoheterotrophic cultures. Analysis of more than 200 of the expression strains has shown that approximately 60% of the E. coli membrane proteins are expressed in Rba. sphaeroides at levels that exceed 1 mg l–1. Fig. 4 illustrates the broad size range (14 – 86 kDa, overall; 17 – 66 kDa in Fig. 4) of E. coli membrane proteins that can be expressed at or above this threshold level. In fact, many can be expressed and purified at levels of 10 – 20 mg l–1 of culture. Similarly, the isoelectric points of the ‘hits’ range from 5.9 to 11.9, and proteins having up to 14 transmembrane spans have been expressed and are localized within the Rhodobacter ICM. This dataset is large, unique, and unbiased, and will provide powerful input in effecting use of Rhodobacter as a general expression host. These expression results—both positive and negative—are in the process of being mined to discover gene, transcript, and/or protein features that influence successful expression. Of the targets that failed to be expressed, over half are currently annotated as ‘hypothetical’ proteins. Since there is no information about cofactors or protein partners that may be necessary for the assembly of these membrane proteins in functional form, their failure to be expressed may be due to the fact that they are inappropriate targets for single-gene expression strategies.
Fig. 4. Molecular weight range of E. coli membrane proteins expressed heterologously in Rhodobacter.
C. Heterologous Expression in Photosynthetic Cultures The previously-described protocols for large-scale production of membrane proteins involved dark, semi-aerobic culture of Rba. sphaeroides strains in which expression of foreign genes was driven by the oxygen-responsive puf promoter. When a photocompetent strain is employed as the expression host, the ICM is induced maximally, culture densities reach higher levels, and anaerobiosis and/or light can be used to induce expression of genes under the control of the puf and puc promoters. In either semi-aerobic or photosynthetic culture, protein synthesis is coordinate with new membrane synthesis. In semiaerobic culture, it is easy to obtain >10 mg of many purified proteins per liter of cell culture, yet others are expressed at lower levels. It is possible that photosynthetic growth could increase the amounts of protein expressed in all strains. Of particular interest are those proteins that fall into low- or non-expressing categories. The photosynthetic growth regime may also serve to elevate expression in some borderline ‘hits’ (e.g., Fig. 3, targets APC01216 and APC01218; APC, a designator of targets for a structural genom-
Chapter 42
Heterologous Expression in Rhodobacter
ics initiative) such that the amount of protein that is produced reaches useful levels, thus possibly rescuing some strains that were scored previously as non-expressors. Photosynthetic growth is not possible if expression host ∆∆11 is used, since it lacks light-harvesting antennae and a RC (Table 1). Thus, wild-type strain ATCC17023 has been used as the host in preliminary experiments to evaluate the potential for increased production of foreign proteins. In the analysis of total membrane extracts, it was clear that the light-grown culture produced more protein. At least five-fold more protein (50 mg l–1) could be purified via IMAC from membranes of the light-grown culture, and this protein was significantly purer than the protein isolated from the dark-grown culture. The samples share some common impurities, yet some of these appear to be specific to the growth mode. This culture method was not pursued in initial experiments for two reasons. The expression plasmids are selectable by tetracycline and photooxidation of this drug produces toxic products. Thus, this antibiotic had been contraindicated for light-grown cultures of Rba. sphaeroides (Davis et al., 1988; Lee et al., 1989; Varga and Kaplan, 1993). However, no phototoxicity due to the drug was observed since the culture conditions were anaerobic, while generation of toxic photoproducts requires oxygen (Martin et al., 1987). Another reason is that results from semi-aerobic cultures suggested that hosts with a partially-depleted ICM accumulated the greatest amount of foreign membrane protein. Given the high yields that have now been obtained from photosynthetic culture, this growth mode will be exploited further, especially for production of proteins driven by the light-responsive puc promoter. D. Purification of Heterologously-Expressed Membrane Proteins from the Rhodobacter Intracytoplasmic Membrane Generic, reproducible and rapid methods have been developed for solubilizing and purifying heterologously-expressed proteins from the membranes of Rba. sphaeroides. Utilizing the polyhistidine tag engineered into the expression vector, detergentsolubilized target membrane proteins can be purified readily by IMAC. This method is specific, its rapidity can facilitate purification of the target protein in its native state, and its general utility eliminates
849
Fig. 5. Semi-automated purification of an E. coli membrane protein (APC00809) expressed heterologously in Rba. sphaeroides. Analysis of purity assessed by SDS-PAGE reveals marked purity improvements with additional rounds of chromatography performed in series, starting from the isolated ICM. (Left to right) Pro Sieve Protein Markers (Cambrex), membrane fraction, affinity chromatography (5mL HiTrapTM Chelating HP Column), gel filtration chromatography (HiLoadTM 26/60 SuperdexTM 75 Prep Grade Column), and ion exchange chromatography (5mL HiTrapTM DEAE FF Ion Exchange Column).
the need to determine de novo the type of chromatography which will be successful for each protein. By combining affinity chromatography sequentially with gel filtration and/or ion exchange steps, highly purified heterologously-expressed target proteins can be recovered rapidly from ICMs (Fig. 5). Routinely, a two-step process of IMAC followed by gel filtration chromatography yields both quantities (~1–10 mg) and purities (>90%) of proteins that are sufficient for studies of structure and function. Purity levels of target membrane proteins were demonstrated on overloaded SDS-polyacrylamide gels (Fig. 6). For those proteins lacking activity assays, structural and functional integrity was suggested by the presence of strong signals in circular dichroism spectra that are lost under denaturing conditions (e.g., heat and/or guanidine-HCl). These protocols have been used successfully with a wide variety of detergents (zwitterionic, charged, nonionic, etc.). Yields have reached up to 10–20 mg of pure protein per liter of culture.
850
Philip D. Laible, Donna L. Mielke and Deborah K. Hanson
Fig. 6. SDS-PAGE analysis of E. coli membrane proteins expressed heterologously in Rhodobacter. Purified, concentrated target membrane proteins were overloaded (25 µg/lane) to assess purity.
IV. Optimization and Generalization of Heterologous Expression in Rhodobacter A. Considerations for Heterologous Expression in Rhodobacter Many aspects of heterologous gene expression in Rhodobacter are unique and special attention must be paid to them to optimize potential for success. These considerations are likely to be most important for target genes derived from organisms that are phylogenetically distant from photosynthetic bacteria. A few of the most important are outlined below. 1. Codon Usage A frequent consideration for the likelihood of success in expressing any protein heterologously is the commonality of the codon usage in the source and host organisms. The G+C content of Rhodobacter species is 68%; thus, codons ending with G or C are used preferentially in the ORFs of these organisms. Some AT-rich codons (e.g., TTA, ATA, CTA, TCA, GTA, TGT, AGA) are used less than once per thousand in coding sequences (http://www.kazusa. or.jp/codon/). How will the presence of rare codons affect the yield of protein expressed in Rhodobacter? To address this question, a silent mutation (ATT→ATA) was introduced within the first 20 amino acids of GFP to determine the effect of the rare isoleucine codon AUA on its expression level. Surprisingly, expression
levels of the mutant GFP rivaled or surpassed that of the wild-type GFP. This strongly suggests that rare codons (at least rare isoleucine codons) will not be problematic for expression of foreign genes that contain such codons in Rhodobacter. In addition, preliminary sequence analysis of target genes that have been expressed heterologously in Rhodobacter suggests that this organism’s codon preference is not a serious limitation. Targets expressed successfully were just as likely to have codons which are ‘rare’ in Rhodobacter as were targets expressed unsuccessfully. It must be noted, however, that highly expressed genes of the photosynthetic apparatus do not contain codons in the above list. In the event that rare codons present obstacles to expression, host strains can be engineered further to incorporate multiple copies of genes encoding these relatively rare tRNAs. 2. Modifications to Make Rhodobacter More User-Friendly as an Expression Host To make the Rhodobacter expression system more accommodating to foreign genes, it could benefit from engineering to remove enzymatic activities that interfere with uptake and maintenance of exogenous DNA, and the accumulation and efficient purification of expressed foreign protein. The method of choice for introduction of foreign DNA into Rhodobacter strains is not transformation, but conjugal mating. The single-stranded mechanism of delivery in the latter circumvents endogenous restriction/modification enzymes that cleave exogenous double-stranded DNA, but the need for purification of Rba. sphaeroides transconjugants from the E. coli donor strain adds several days to the gene transfer procedure. Vectors lacking sites for endogenous restriction endonucleases have been used successfully to transform Rba. sphaeroides (Fornari and Kaplan, 1982). Thus, inactivation of components of the Rba. sphaeroides restriction/modification system would enable much higher frequencies for transformation of this organism by foreign DNA. In addition, construction of recA host strains will eliminate the potential for homologous recombination between foreign gene sequences and their homologs in the Rhodobacter host genome. The Rba. sphaeroides genome encodes a Lon protease (http://merops.sanger.ac.uk/cgi-bin/speccar ds?sp=sp001916&type=P). Typically, this enzyme is inactivated in expression hosts of E. coli, e.g., BL21; this strain also lacks the OmpT protease. The Lon
Chapter 42
Heterologous Expression in Rhodobacter
protease and others detailed on the above website are candidates for inactivation in strains of Rba. sphaeroides that will be used as expression hosts, as are host proteins that copurify with target proteins during affinity chromatography. 3. Features Influencing Transcript Stability When expression of a heterologous membrane protein gene in Rba. sphaeroides fails, it is unknown whether the failure is due to proteolysis of misfolded protein or mRNA instability. Thus, experiments have been initiated to determine the fate of proteins whose expression has failed. For the Rba. sphaeroides expression strains, quantitative reverse transcriptase-PCR protocols have been developed to measure mRNA message levels that will correlate mRNA abundance (e.g., stability) with protein expression levels, allowing the identification and eventual incorporation of sequence elements that confer transcript stability. 4. Post-Translational Modifications The variety of post-translational modifications that is characteristic of eukaryotic hosts will not occur in prokaryotic cells. If the protein is a subject for structure determination, incomplete in vivo post-translational modifications often thwart crystallogenesis by introducing heterogeneity into the sample. Production of these membrane proteins in Rhodobacter will result in a more homogeneous preparation that may have a greater propensity to crystallize. However, some post-translational modifications may be required for functional integrity of a particular target membrane protein, therefore in vitro methods for introducing the modifications would need to be employed or developed for such proteins expressed heterologously in this organism. The extent of the modifications could be monitored by two-dimensional SDS-PAGE or mass spectrometry. 5. Capability for Incorporation of Selenomethionine Today’s method of choice for the determination of de novo X-ray protein structures involves the use of multiple- or single anomalous-wavelength dispersion (MAD/SAD) techniques. The incorporation of a selenium atom into the protein crystal is one way to provide an anomalous scatterer as a key to solving the crystal structure by these techniques. Selenium
851
may be incorporated into a protein sequence by use of the amino acid analog selenomethionine (SeMet) during biosynthesis of the polypeptide chain. Methods have been devised for the quantitative biosynthetic incorporation of SeMet into induced ICM proteins of Rba. sphaeroides and Rba. capsulatus, thus facilitating the determination of their structures by efficient X-ray diffraction techniques at beamlines of third generation synchrotrons (Laible et al., 2005a). B. Adaptation to Higher Throughput In the later stages of cloning and expression of the E. coli membrane proteome, several steps in the process were adapted to higher-throughput methodologies, utilizing the 96-well format and liquid-handling robots. PCR-based amplifications of target genes were tailored to this format, as was processing of amplicons for ligation-independent cloning, E. coli transformation and growth of transformants, plasmid preparations and screening by restriction enzyme digestion, and conjugation with host strains of Rba. sphaeroides. Standard protocols have been developed whereby multiple proteins expressed in Rhodobacter can be purified in semi-automated fashion by sequential steps of affinity, size exclusion, and/or ion exchange chromatography using a modified FPLC (Laible et al., 2004). Greater than 100 mg of a single heterologously-expressed target protein have been purified in a given FPLC run. These methods augment the potential utility of photosynthetic bacteria and their inducible membranes for the production of foreign membrane proteins for structural and functional studies. C. Adopting Tactics from Other Systems Tailored to Membrane Protein Expression A proliferation of membranes (intracellular or organellar) in the host organism is a theme common to several other membrane protein expression systems that are in use or in development. The system outlined here for Rhodobacter may benefit by adapting some strategies being used in optimization of these other systems. Among the alternatives for membrane protein expression are the C41 and C43 strains of E. coli that possess additional membranes (Miroux and Walker, 1996; Arechaga et al., 2000); Halobacterium salinarum and its inducible purple membrane (Turner et al., 1999); Tetrahymena thermophila (Gaertig et al., 1999) and cell wall-deficient L-forms of Proteus
852
Philip D. Laible, Donna L. Mielke and Deborah K. Hanson
mirabilis, E. coli, Bacillus subtilis, and Streptomyces hygroscopicus (Gumpert and Hoischen, 1998; Hoischen et al., 2002) that are used for surface display of expressed membrane proteins; and engineered strains of methylotrophic bacteria (Nguyen et al., 1998), the yeast Hansenula polymorpha (van Dijk et al., 2000), and photosynthetic bacteria (Graichen et al., 1999; De Smet et al., 2001; Collins and Cheng, 2004) which feature inducible ICMs. These systems utilize a variety of promoters and induction conditions. In addition, they employ different means of targeting expressed proteins to cellular compartments, providing samples of utility for various studies of structure and function.
Muller, 1990; Troschel et al., 1992; Wieseler and Muller, 1993; Garcia et al., 1994; Helde et al., 1997). These proteins were localized to Rhodobacter ICM vesicles if (and only if) the vesicles were added cotranslationally; that is, the ICM must be present during protein synthesis for the efficient membrane incorporation. Using insights gained from work with the cell-based Rhodobacter expression system, it should be possible to determine the components that are necessary to define a robust cell-free system employing Rhodobacter ICMs that can be applied generically for the heterologous production of a wide variety of membrane proteins.
D. Potential for a Cell-Free Expression System Based upon Rhodobacter
V. Advantages Afforded by Rhodobacter
Production of soluble proteins has benefited from recent advances in the quality and diversity of the methods and cell extracts that are available for in vitro translation systems (Spirin, 2004). Commercial extracts are now obtainable from a several sources (E. coli: Busso et al., 2003, 2004; Kuruma et al., 2005; wheat germ: Endo and Sawasaki, 2006; and rabbit reticulocytes: Pelham and Jackson, 1976). Yet other organisms have been employed as the source of extracts, as described in the literature (Staphylococcus: Schimz et al., 1995; and yeast: Wang, 2006). They are relatively simple and time-saving, can be used to optimize expression of a particular target protein and produce mutant proteins, can use PCR amplicons instead of plasmid constructs, can produce proteins toxic to cell-based systems, and can label proteins or incorporate non-native amino acids. However, quality-controlled extracts are expensive and their efficiency depends upon the concerted activities of multiple components, many of which are short-lived. Cell-free systems have been used only recently to generate membrane protein samples. Success has been achieved by the addition of membrane components or mimetics (detergents or E. coli lipids; Klammt et al., 2004). With the implementation of many cell-free systems based on prokaryotic extracts, it seems logical to attempt production of membrane proteins with a Rhodobacter extract. In fact, there are reports in the literature where Rhodobacter extracts have been used successfully to produce functional native membrane proteins (Chory and Kaplan, 1982; Chory et al., 1985; Hoger et al., 1986; Troschel and
Data presented herein indicate that Rhodobacter can assume an important role in serving as an alternative host for the overexpression of proteins — soluble, membrane-bound, or multisubunit complexes— whose expression has proven to be problematic in other systems for various reasons. A number of strategies can be developed to approach the problem of expressing a particular target protein, taking full advantage of the organism’s metabolic diversity to control the timing and rate of induction, as well as assembly and compartmentalization of the expressed protein. The versatility afforded by this ensemble of vectors, hosts, and culture modes increases the probability of success. A. Production of Natively-Folded Proteins A major advantage for expression of membrane proteins in Rhodobacter is that expressed proteins are localized to the ICM. This attribute suggests strongly that the expressed proteins assume a structure that directs proper insertion into the membrane. In the cases studied extensively, ICM localization has been linked to functional integrity. Thus, the working hypothesis is that this relationship holds for other membrane proteins whose functions cannot be assayed readily. Detection of expressed proteins with the anti-polyhistidine antibody has failed to produce clear evidence for formation of inclusion bodies in Rhodobacter. This is in sharp contrast to other prokaryotic expression systems — such as E. coli expression systems based on T7 polymerase (Kiefer et al., 1999; Korepanova et al., 2005; Columbus et al., 2006) — where
Chapter 42
Heterologous Expression in Rhodobacter
high-level overexpression often results in aggregation and precipitation of incompletely-folded polypeptides as inclusion bodies. The kinetics of induction and semi-aerobic (or photosynthetic) growth rate are correspondingly slower in Rhodobacter and may shift the equilibrium towards the production of the folded, functional state of the target proteins, whether they are soluble or membrane-bound. If the target protein requires membrane association—either static or transient—for its integrity, Rhodobacter possesses a greatly increased membrane surface area with which it can interact. B. Utility for Proteins Whose Successful Expression Requires Multiple Gene Products Rhodobacter has a demonstrated utility for expressing proteins requiring complex metal-containing cofactors and those that require coupled expression of several genes in order to form a mature multisubunit enzyme. A few such expression systems have been described; in these systems, induction of expression is not coupled to induction of the ICM. Expression of the soluble flavocytochrome c-sulfide dehydrogenase from Allochromatium vinosum and the membranebound homolog from Ectothiorhodospira vacuolata was achieved in both Rba. sphaeroides and Rba. capsulatus using a Rubisco promoter from Allochromatium vinosum (De Smet et al., 2001). Heterologous expression failed in E. coli but succeeded in Rhodobacter because of the endogenous heme biosynthesis and cytochrome maturation machinery in the latter host organisms (Chapter 21, Sanders et al.). Rba. sphaeroides has also been used as a host to express soluble methylamine dehydrogenase, an enzyme whose biosynthesis requires the expression of several genes in addition to the structural genes that encode the multi-subunit complex (Graichen et al., 1999). These ‘maturation’ proteins appear to include a membrane protein and a heme-containing peroxidase (Chistoserdov et al., 1994; van der Palen et al., 1997). When the complete operon encoding all of these proteins was placed under the control of the host’s oxygen-inducible coxII promoter, active periplasmically-localized methylamine dehydrogenase was produced by Rba. sphaeroides in aerobicallygrown cultures. In addition, Rba. capsulatus and the dor promoter from its dimethyl sulfoxide (DMSO) reductase operon have been used successfully to express soluble sulfite:cytochrome c oxidoreductase
853
from Starkeya novella (Kappler and McEwan, 2002). Production of the functional heterodimeric enzyme in photosynthetically-grown cultures required two different redox cofactors (molybdenum pterin and a c-type heme), translocation of the two subunits to the periplasm by two different secretory pathways, and proper assembly of protein subunits and cofactors in that compartment. In this case, the dor promoter was chosen because of its strength and stringent regulation by DMSO, molybdate, and anaerobic growth conditions. Alternatively, if ICM synthesis would benefit the functional expression of the membrane protein complex, polycistronic DNA segments can be inserted easily into existing vectors where synthesis is driven by either the puf or puc promoters (Fig. 1a–c). In this case, stoichiometric amounts of proteins would be produced from the operon (unless inherent regulatory elements dictating otherwise are present and are recognized by Rhodobacter). Another vector in this series has been constructed which allows for dual expression of two or more foreign genes — in different relative stoichiometries — by replacement of both the LH1 and RC coding regions (a combination of vectors in Fig. 1a and 1b). Two considerations drive the latter strategy: (1) the function of many multisubunit protein complexes often requires a subunit ratio other than 1:1 and (2) maturation proteins would be required in catalytic amounts, whereas overexpression of structural subunit(s) of the target complex would be desired. C. Simple Genetics Facilitate Interspecific Gene Expression and in vivo Functional Analysis Rhodobacter species are among the few photosynthetic bacteria for which genetic systems have been established. Originally, these systems were employed for structure-function studies of native and variant forms of endogenous proteins. Because they are more metabolically diverse than E. coli, Rhodobacter species serve now, in addition, as attractive vehicles for in vivo structure-function studies of proteins derived from more complex organisms that are not as amenable to rapid genetic manipulation. Expression of these interspecific genes either adds to the repertoire of metabolic processes in Rhodobacter or complements its extensive inherent capabilities.
854
Philip D. Laible, Donna L. Mielke and Deborah K. Hanson
1. Metabolic Engineering Efficient gene manipulation has enabled the coupling of photosynthetic energy conversion to metabolic activities that are not intrinsic to Rhodobacter. In one such instance, a cellulase gene from Cellulomonas fimi was expressed in Rba. capsulatus from the puf promoter and a small, but significant, level of extracellular cellulase activity was observed (Johnson et al., 1986). Since the metabolic capabilities of photosynthetic bacteria include efficient conversion of carbon sources to hydrogen and carbon dioxide, further manipulation of this expression strain has the potential to produce an organism in which photosynthetic hydrogen production is coupled to the use of complex carbohydrates, such as cellulosic agricultural byproducts, as carbon sources (Johnson et al., 1986). 2. Interspecific Complementation of Functions Inherent to Rhodobacter The ease in genetic manipulation of Rhodobacter has led to its utility in investigating the function of proteins and protein complexes by extensive mutational analysis. Quite simply, this involves the construction of a strain that is deleted for the protein of interest followed by its complementation (in cis or in trans) by wild-type or variant forms of that protein. This complementation can be broadened to investigate the function of homologs or suspected homologs from more distantly related organisms. For example, there are several instances where deletions of the photosynthetic apparatus of Rba. sphaeroides or Rba. capsulatus have been complemented by genes from other purple bacteria (Zilsel et al., 1989; Fowler et al., 1995; Fowler and Hunter, 1996; Fulcher et al., 1998; Katsiou et al., 1998; Aklujkar et al., 2006). When coupled with known differences in amino acid sequence and subunit composition, the changes in energy- and electron-transfer efficiency or growth phenotype observed in these hybrid photosynthetic units has led to a better understanding of the interactions that are important in the native supermolecular complexes. This complementation extends well beyond that of genes from other photosynthetic bacteria. Deletion strains of Rhodobacter serve as null backgrounds that can be complemented by enzymes derived from organisms for which there is no genetic system that can be employed to determine structure-function relation-
ships. For example, the NADPH:protochlorophyllide oxidoreductase from pea (Pisum sativum) has been expressed in Rba. capsulatus from the IPTG-inducible tac promoter. The expressed pea enzyme rescued this activity in protochlorophyllide-reduction mutant host strains of Rba. capsulatus (Wilks and Timko, 1995; Lebedev and Timko, 2002). In a second example, the carotenoid content of Rba. sphaeroides has been modified by heterologous expression of genes of the carotenoid biosynthetic pathway of Erwinia herbicola (Hunter et al., 1994; Garcia-Asua et al., 2002). In one such strain, redirection of the native pathway resulted in the synthesis and incorporation of lycopene (instead of the native neurosporene) into the LH2 complex. This foreign cofactor was shown to function effectively in energy transfer to the native bacteriochlorophyll molecules bound within this hybrid complex (Garcia-Asua et al., 2002). The ability to construct such expression strains facilitates rapid evaluation of the effects of multiple amino acid or cofactor substitutions on functions such as catalytic activity, energy transfer, or the binding of substrates, cofactors and protein ligands that aid in assembly, regulation, or stabilization of active complexes. D. An Alternative Promoter for High-Level Expression A T7-based overexpression system has been constructed for Rba. capsulatus to exploit the high levels of expression that are afforded by this promoter/polymerase combination. This IPTG-inducible system is being used to address the problem of producing enzymes (e.g., hydrogenase) that require the coupled expression of many genes — not only those that encode structural subunits, but also genes that regulate enzyme assembly and cofactor biosynthesis (Drepper et al., 2005). This strategy is geared towards high-level expression of multiple genes from large bacterial operons, taking advantage of the inability of T7 RNA polymerase to recognize the numerous transcription terminators that often interrupt such operons. Protein synthesis in this instance is not coupled to ICM biosynthesis. E. Utility of a Common Bilayer Heterologous production of membrane proteins has the potential to offer the advantage of a common lipid bilayer from which the target protein is solubilized as
Chapter 42
Heterologous Expression in Rhodobacter
a first step in its purification. Discovery of conditions that optimize solubilization of membrane proteins and stabilization of targets in detergent micelles prior to purification is a time-consuming, labor-intensive and expensive process, especially for membrane proteins extracted from their native organisms. The parameter space that must be searched in this process is quite large. Anecdotal schemes abound, but no precedent exists for a single ‘magic bullet’ set of generic conditions (detergents, temperatures, incubation time, protein/surfactant ratios, etc.) under which proteins can be extracted in high yield and functional form from native hosts. Heterologous production of membrane proteins has the potential to offer significant advantages in this regard. When membrane proteins are expressed in the Rhodobacter ICM, solubilization proceeds from a common lipid bilayer of defined chemical composition (Benning, 1998; Chapter 7, Tamot and Benning). By screening a large number of detergents against a small number of target proteins, one can optimize the first step of solubilization (dismantling the bilayer) to determine which types of detergents are best at integrating, penetrating, and ultimately destroying the lipid bilayer of the ICM. The same set of detergents should work well for any membrane protein that is localized to the ICM and can be used in a generic approach. Subsequent testing for functionality will determine how robust micelles of this ‘dismantling detergent set’ are in stabilizing membrane proteins. Preliminary evidence suggests that they might be quite good, as the Rhodobacter ICM can be broken down with detergents that are considered to be fairly ‘gentle’ (e.g., Deriphat 160; Kirmaier et al., 2003). The development of these generic approaches is aided by decades of work on photosynthetic complexes and it implements detergents that have worked well in extracting complexes of the photosynthetic apparatus that are characterized by markedly different stabilities outside the lipid bilayer. F. Ability to Produce Different Membrane Samples for Structural and Functional Experiments It is possible to produce many types of membrane preparations from Rhodobacter cells, thus the design of particular structural or functional experiments can benefit from selection of different cell lysis and membrane fractionation procedures in order to take full advantage of the characteristics of the ICM bi-
855
layer housing heterologously-expressed membrane proteins. The literature contains protocols for the preparation of inside-out membrane vesicles and for the production of spheroplasts (Takemoto and Bachmann, 1979). In the former, the cytoplasmic face of the ICM proteins is exposed and periplasmic proteins are captured in the interior of the vesicles. Alternatively, in spheroplasts, the periplasmic surfaces of ICM-bound proteins are exposed, while interior cellular components are inaccessible to reagents that remain outside the spheroplasts. Likewise, other methods can produce membrane sheets that expose both membrane surfaces to the milieu (Siebert et al., 2004). In addition, many studies have illustrated the dynamic organization of components of the photosynthetic membrane in response to various growth conditions, implying the existence of specialized zones that sequester particular components (Niederman et al., 1979; Kaufmann et al., 1982; Drews, 1992; Drews and Golecki, 1995; Laible et al., 2005a). In preliminary experiments with Rba. sphaeroides strains expressing either native or foreign membrane proteins in trans, multiple different subpopulations of ICM fragments have been separated following cell lysis, differential ultracentrifugation, and sucrose density gradient centrifugation. The protein content of those membrane fractions was analyzed by SDS-PAGE; an example is shown in Fig. 7. In these cases, one particular membrane band in the density gradient was highly enriched (>50%) in the target membrane protein, whereas the other gradient fractions were nearly devoid of it. These ICM subpopulations represent a substantial purification of the target membrane protein, even though it remains in an undisturbed bilayer environment. Thus, the availability of these subpopulations could circumvent the need for extraction of the target protein from the bilayer and its solubilization in detergents. These fractions have enormous, untapped utility in functional studies, biochemical assays, and in the production of ordered arrays for structural analysis. They could also serve as starting material in a subtractive strategy for the selection of affinity reagents to exposed periplasmic or cytoplasmic surfaces of the target membrane protein—the subtractive subset could be obtained simply by producing ICM subpopulations from the corresponding engineered control strain of Rba. sphaeroides carrying an empty expression vector. Membrane fragments or vesicles harboring the expressed target membrane protein
856
Philip D. Laible, Donna L. Mielke and Deborah K. Hanson promise, especially for the production of membrane proteins for structural biology studies and the structural genomics initiatives which are ramping up into higher-production modes. Rhodobacter has the potential to serve the ‘membrane protein world’ in the same pivotal role as E. coli serves the ‘soluble protein world.’ It is not unreasonable to project that, in the very near future, cells from the genus Rhodobacter will be used as factories for the production of membrane proteins for a wide variety of studies and applications. Acknowledgments
Fig. 7. Rhodobacter ICM fractions in which the target membrane proteins are predominant were separated by sucrose density gradient centrifugation (right, arrows) and their purities were assayed using SDS-PAGE (left). The second lane displays an ICM fraction derived from a strain expressing reaction centers at ~15 mg l–1 culture. The multisubunit RCs are identified as three dark bands at ≤ 31 kDa. Summation of these three bands and comparison to the total protein suggest enrichment between ~30–40 %. The other two gel lanes display membrane fractions isolated from two different Rhodobacter strains, each expressing heterologously an E. coli membrane protein (APC00809, 20 kDa; APC00821, 40 kDa). The target protein is > 50% of the total in those membrane fractions.
could be immobilized with a defined orientation via tags expressed on its exposed surfaces, facilitating ‘sidedness’ in selection schemes or functional assays. VI. Perspectives The Rhodobacter expression system described herein has the potential to reshape the thinking of the scientific community concerning its ability to produce and study membrane proteins efficiently. This system puts to task, in a novel way, this photosynthetic species and the membranes it naturally produces in order to harness this machinery for the efficient production of foreign membrane proteins. This species is also an exceptionally capable host for production of problematic soluble proteins that require complex redox cofactors or association with the membrane, or whose successful expression requires more than a single gene expression strategy. This ensemble of expression strategies offers great
The authors have benefited greatly from the contributions of laboratory members (past and present) and colleagues at Argonne National Laboratory for various aspects of the development and evaluation of the Rhodobacter expression system, including David Mets, Matt Keller, Adam Crawford, Katie Olson, Yuri Poluektov, Dan Silverman, Mike Bellisario, Marc Wander, Chris Kors, Greg Tira, Heather Scott, Sam Hofman, Hema Joshi, Zach Morris, Adam Reynolds, Becky Koranda, James Bautista, Terrence Wong, Aaron Hata, Vicky Landorf, Shaina Bengtson, Natalia Maltsev, and Elsie Quaite-Randall. They also thank Hewson Swift for expertise in and facilities for electron microscopy, Frank Collart and Mark Donnelly for valuable discussions relating to vector design, and Lynda Henry, Brad Hales, and Ake Danielsson of GE Healthcare for suggestions regarding adaptation of ÄKTA-FPLCs for semi-automated purification of membrane proteins. Funding for these efforts was provided by the National Institutes of Health (R01 GM61887, P50 GM62414, R01 GM71318, and P01 GM75913) and the United States Department of Energy, under contract W-31-109-ENG-38. References Aagaard J and Sistrom WR (1972) Control of synthesis of reaction center bacteriochlorophyll in photosynthetic bacteria. Photochem Photobiol 15: 209–225 Aklujkar M, Prince RC and Beatty JT (2006) The photosynthetic deficiency due to puhC gene deletion in Rhodobacter capsulatus suggests a PuhC protein-dependent process of RC/LH1/PufX complex reorganization. Arch Biochem Biophys 454: 59–71 Arechaga I, Miroux B, Karrasch S, Huijbregts R, de Kruijff B, Runswick MJ and Walker JE (2000) Characterisation of new intracellular membranes in Escherichia coli accompanying large scale over-production of the b subunit of F1F0 ATP synthase.
Chapter 42
Heterologous Expression in Rhodobacter
FEBS Lett 482: 215–219 Bahatyrova S, Frese RN, Siebert CA, Olsen JD, Van Der Werf KO, Van Grondelle R, Niederman RA, Bullough PA, Otto C and Hunter CN (2004) The native architecture of a photosynthetic membrane. Nature 430: 1058–1062 Benning C (1998) Membrane lipids in anoxygenic photosynthetic bacteria. In: Siegenthaler P-A and Murata N (eds) Lipids in Photosynthesis: Structure, Function, and Genetics (Advances in Photosynthesis and Respiration, Vol 6), pp 83–101. Kluwer Academic Publishers, Dordrecht Busso D, Kim R and Kim SH (2003) Expression of soluble recombinant proteins in a cell-free system using a 96-well format. J Biochem Biophys Methods 55: 233–240 Busso D, Kim R and Kim SH (2004) Using an Escherichia coli cell-free extract to screen for soluble expression of recombinant proteins. J Struct Funct Genomics 5: 69–74 Chistoserdov AY, Chistoserdova LV, McIntire WS and Lidstrom ME (1994) Genetic organization of the mau gene cluster in Methylobacterium extorquens AM1: Complete nucleotide sequence and generation and characteristics of mau mutants. J Bacteriol 176: 4052–4065 Chory J and Kaplan S (1982) The in vitro transcription-translation of DNA and RNA templates by extracts of Rhodopseudomonas sphaeroides. Optimization and comparison of template specificity with Escherichia coli extracts and in vivo synthesis. J Biol Chem 257: 15110–15121 Chory J, Muller ED and Kaplan S (1985) DNA-directed in vitro synthesis and assembly of the form II D-ribulose-1,5-bisphosphate carboxylase/oxygenase from Rhodopseudomonas sphaeroides. J Bacteriol 161: 307–313 Clayton RK (1980) Photosynthesis: Physical Mechanisms and Chemical Patterns. Cambridge University Press, Cambridge Collins MLP and Cheng Y (2004) Host/vector system for expression of membrane proteins. Office UP.USANo. 6,680,179 Columbus L, Lipfert J, Klock H, Millett I, Doniach S and Lesley SA (2006) Expression, purification, and characterization of Thermotoga maritima membrane proteins for structure determination. Protein Sci 15: 961–975 Davis J, Donohue TJ and Kaplan S (1988) Construction, characterization, and complementation of a Puf– mutant of Rhodobacter sphaeroides. Journal of Bacteriology 170: 320–329 De Smet L, Kostanjevecki V, Guisez Y and Van Beeumen J (2001) A novel system for heterologous expression of flavocytochrome c in phototrophic bacteria using the Allochromatium vinosum rbcA promoter. Arch Microbiol 176: 19–28 Dieckman L, Gu M, Stols L, Donnelly MI and Collart FR (2002) High throughput methods for gene cloning and expression. Protein Expr Purif 25: 1–7 Ditta G, Schmidhauser T, Yakobsen E, Lu P, Liang X-W, Finlay DR, Guiney D and Helinski DR (1985) Plasmids related to the broad host range vector, pRK290, useful for gene cloning and for monitoring gene expression. Plasmid 13: 149–153 Donohue TJ and Kaplan S (1991) Genetic techniques in Rhodospirillaceae. In: Methods in Enzymology, pp 459–485. Academic Press, New York Drepper T, Arvani S, Rosenau F, Wilhelm S and Jaeger KE (2005) High-level transcription of large gene regions: A novel T(7) RNA-polymerase-based system for expression of functional hydrogenases in the phototrophic bacterium Rhodobacter capsulatus. Biochem Soc Trans 33: 56–58 Drews G (1985) Structure and functional organization of light-
857
harvesting complexes and photochemical reaction centers in membranes of phototrophic bacteria. Microbiol Rev 49: 59–70 Drews G (1992) Intracytoplasmic membranes in bacterial cells: organization, function, and biosynthesis. In: Mohan S, Dow C and Cole JA (eds) Prokaryotic Structure and Function: A New Perspective, pp 249–274. Cambridge University Press, Cambridge Drews G and Golecki JR (1995) Structure, molecular organization, and biosynthesis of membranes of purple bacteria. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 231–257. Kluwer Academic Publishers, Dordrecht Elsen S, Swem LR, Swem DL and Bauer CE (2004) RegB/RegA, a highly conserved redox-responding global two-component regulatory system. Microbiol Mol Biol Rev 68: 263–279 Endo Y and Sawasaki T (2006) Cell-free expression systems for eukaryotic protein production. Curr Opin Biotechnol 17: 373–380 Farchaus JW and Oesterhelt D (1989) A Rhodobacter sphaeroides puf L, M and X deletion mutant and its complementation in trans with a 5.3 kb puf operon shuttle fragment. EMBO Journal 8: 47–54 Fornari CS and Kaplan S (1982) Genetic transformation of Rhodopseudomonas sphaeroides by plasmid DNA. J Bacteriol 152: 89–97 Fowler GJ and Hunter CN (1996) The synthesis and assembly of functional high and low light LH2 antenna complexes from Rhodopseudomonas palustris in Rhodobacter sphaeroides. J Biol Chem 271: 13356–13361 Fowler GJ, Gardiner AT, Mackenzie RC, Barratt SJ, Simmons AE, Westerhuis WH, Cogdell RJ and Hunter CN (1995) Heterologous expression of genes encoding bacterial light-harvesting complexes in Rhodobacter sphaeroides. J Biol Chem 270: 23875–23882 Fulcher TK, Beatty JT and Jones MR (1998) Demonstration of the key role played by the PufX protein in the functional and structural organization of native and hybrid bacterial photosynthetic core complexes. J Bacteriol 180: 642–646 Gaertig J, Gao Y, Tishgarten T, Clark TG and Dickerson HW (1999) Surface display of a parasite antigen in the ciliate Tetrahymena thermophila. Nat Biotechnol 17: 462–465 Garcia AF, Meryandini A, Brand M, Tadros MH and Drews G (1994) Phosphorylation of the alpha-polypeptides and beta-polypeptides of the light-harvesting complex I (B870) of Rhodobacter capsulatus in an in vitro translation system. FEMS Microbiol Lett 124: 87–91 Garcia-Asua G, Cogdell RJ and Hunter CN (2002) Functional assembly of the foreign carotenoid lycopene into the photosynthetic apparatus of Rhodobacter sphaeroides, achieved by replacement of the native 3-step phytoene desaturase with its 4-step counterpart from Erwinia herbicola. Mol Microbiol 44: 233–244 Gerhardt P, Murray RGE, Wood W and Krieg NR (1994) Methods for General and Molecular Bacteriology. American Society for Microbiology, Washington, DC Golecki JR and Heinrich UR (1991) Ultrastructural and electron spectroscopic analyses of cyanobacteria and bacteria. J Microsc 162: 147–154 Golecki JR, Ventura S and Oelze J (1991) The architecture of
858
Philip D. Laible, Donna L. Mielke and Deborah K. Hanson
unusual membrane tubes in the B800-B850 light-harvesting bacteriochlorophyll-deficient mutant 19 of Rhodobacter sphaeroides. FEMS Microbiol Lett 77: 335–340 Graichen ME, Jones LH, Sharma BV, van Spanning RJ, Hosler JP and Davidson VL (1999) Heterologous expression of correctly assembled methylamine dehydrogenase in Rhodobacter sphaeroides. J Bacteriol 181: 4216–4222 Gumpert J and Hoischen C (1998) Use of cell wall-less bacteria (L-forms) for efficient expression and secretion of heterologous gene products. Curr Opin Biotechnol 9: 506–509 Helde R, Wiesler B, Wachter E, Neubuser A, Hoffschulte HK, Hengelage T, Schimz KL, Stuart RA and Muller M (1997) Comparative characterization of SecA from the alpha-subclass purple bacterium Rhodobacter capsulatus and Escherichia coli reveals differences in membrane and precursor specificity. J Bacteriol 179: 4003–4012 Hoger JH, Chory J and Kaplan S (1986) In vitro biosynthesis and membrane association of photosynthetic reaction center subunits from Rhodopseudomonas sphaeroides. J Bacteriol 165: 942–950 Hoischen C, Fritsche C, Gumpert J, Westermann M, Gura K and Fahnert B (2002) Novel bacterial membrane surface display system using cell wall-less L-forms of Proteus mirabilis and Escherichia coli. Appl Environ Microbiol 68: 525–531 Hunter CN, Pennoyer JD, Sturgis JN, Farrelly D and Niederman RA (1988) Oligomerization states and associations of lightharvesting pigment protein complexes of Rhodobacter sphaeroides as analyzed by lithium dodecylsulfate polyacrylamide-gel electrophoresis. Biochemistry 27: 3459–3467 Hunter CN, Hundle BS, Hearst JE, Lang HP, Gardiner AT, Takaichi S and Cogdell RJ (1994) Introduction of new carotenoids into the bacterial photosynthetic apparatus by combining the carotenoid biosynthetic pathways of Erwinia herbicola and Rhodobacter sphaeroides. J Bacteriol 176: 3692–3697 Imhoff JF (1995) Taxonomy and physiology of phototrophic purple bacteria and green sulfur bacteria. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 1–15. Kluwer Academic Publishers, Dordrecht Jamieson SJ, Wang P, Qian P, Kirkland JY, Conroy MJ, Hunter CN and Bullough PA (2002) Projection structure of the photosynthetic reaction centre-antenna complex of Rhodospirillum rubrum at 8.5 Å resolution. EMBO J 21: 3927–3935 Johnson JA, Wong WK and Beatty JT (1986) Expression of cellulase genes in Rhodobacter capsulatus by use of plasmid expression vectors. J Bacteriol 167: 604–610 Kappler U and McEwan AG (2002) A system for the heterologous expression of complex redox proteins in Rhodobacter capsulatus: characterisation of recombinant sulphite:cytochrome c oxidoreductase from Starkeya novella. FEBS Lett 529: 208–214 Karrasch S, Bullough PA and Ghosh R (1995) The 8.5 angstrom projection map of the light-harvesting complex I from Rhodospirillum rubrum reveals a ring composed of 16 subunits. EMBO Journal 14: 631–638 Katsiou E, Sturgis JN, Robert B and Tadros MH (1998) Heterologous expression of genes encoding bacterial light-harvesting complex II in Rhodobacter capsulatus and Rhodovulum sulfidophilum. Microbiological Research 153: 189–204 Kaufmann N, Horst-Helwig R, Golecki JR, Garcia AF and Drews G (1982) Differentiation of the membrane system in
cells of Rhodopseudomonas capsulata after transition from chemotrophic to phototrophic growth conditions. Archives of Microbiology 131: 313–322 Kiefer H, Maier K and Vogel R (1999) Refolding of G-protein-coupled receptors from inclusion bodies produced in Escherichia coli. Biochemical Society Transactions 27: 908–912 Kiley PJ and Kaplan S (1988) Molecular genetics of photosynthetic membrane biosynthesis in Rhodobacter sphaeroides. Microbiological Reviews 52: 50–69 Kiley PJ, Varga A and Kaplan S (1988) Physiological and structural analysis of light-harvesting mutants of Rhodobacter sphaeroides. J Bacteriol 170: 1103–1115 Kirmaier C, Laible PD, Hindin E, Hanson DK and Holten D (2003) Detergent effects on primary charge separation in wild-type and mutant Rhodobacter capsulatus reaction centers. Chem Phys 294: 305–318 Klammt C, Lohr F, Schafer B, Haase W, Dotsch V, Ruterjans H, Glaubitz C and Bernhard F (2004) High level cell-free expression and specific labeling of integral membrane proteins. Eur J Biochem 271: 568–580 Klug G (1995) Post-transcriptional control of photosynthesis gene expression. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 1235–1244. Kluwer Academic Publishers, Dordrecht Koepke J, Hu X, Muenke C, Schulten K and Michel H (1996) The crystal structure of the light-harvesting complex II (B800-850) from Rhodospirillum molischianum. Structure 4: 581–597 Korepanova A, Gao FP, Hua Y, Qin H, Nakamoto RK and Cross TA (2005) Cloning and expression of multiple integral membrane proteins from Mycobacterium tuberculosis in Escherichia coli. Protein Sci 14: 148–158 Kovach ME, Phillips RW, Elzer PH, Roop RM and Peterson KM (1994) pBBR1MCS: Broad host range cloning vector. Biotechniques 16: 800–802 Kovach ME, Elzer PH, Hill DS, Robertson GT, Farris MA, Roop II RM and Peterson KM (1995) Four new derivatives of the broad-host-range cloning vector pBBR1MCS, carrying different antibiotic-resistance cassettes. Gene 166: 175–176 Kuruma Y, Nishiyama K, Shimizu Y, Muller M and Ueda T (2005) Development of a minimal cell-free translation system for the synthesis of presecretory and integral membrane proteins. Biotechnol Prog 21: 1243–1251 Laible PD, Scott HN, Henry L and Hanson DK (2004) Towards higher-throughput membrane protein production for structure genomics initiatives. J Struct Func Genom 5: 167–172 Laible PD, Hata AN, Crawford AE and Hanson DK (2005a) Incorporation of selenomethionine into induced intracytoplasmic membrane proteins of Rhodobacter species. J Struct Func Genom 6: 95–102 Laible PD, Mielke DL and Hanson DK (2005b) Membrane protein production: A bacterial ‘factory’ in Rhodobacter. Screening 6: 30–32 Lebedev N and Timko MP (2002) POR structural domains important for the enzyme activity in R. capsulatus complementation system. Photosynth Res 74: 153–163 Lee JK, Kiley PJ and Kaplan S (1989) Post-transcriptional control of puc operon expression of B800-850 light-harvesting complex formation in Rhodobacter sphaeroides. J. Bacteriol. 171: 3391–3405 MacGregor BJ and Donohue TJ (1991) Evidence for two pro-
Chapter 42
Heterologous Expression in Rhodobacter
moters for the cytochrome c2 gene (cycA) of Rhodobacter sphaeroides. J Bacteriol 173: 3949–3957 Martin JP, Jr., Colina K and Logsdon N (1987) Role of oxygen radicals in the phototoxicity of tetracyclines toward Escherichia coli B. J Bacteriol 169: 2516–2522 McDermott G, Prince SM, Freer AA, Hawthornthwaite-Lawless AM, Papiz MZ, Cogdell RJ and Isaacs NW (1995) Crystal structure of an integral membrane light-harvesting complex from photosynthetic bacteria. Nature 374: 517–521 Miroux B and Walker JE (1996) Over-production of proteins in Escherichia coli: Mutant hosts that allow synthesis of some membrane proteins and globular proteins at high levels. J Molec Biol 260: 289–298 Myllykallio H, Jenney FE, Jr., Moomaw CR, Slaughter CA and Daldal F (1997) Cytochrome cy of Rhodobacter capsulatus is attached to the cytoplasmic membrane by an uncleaved signal sequence-like anchor. J Bacteriol 179: 2623–2631 Naylor G, Addlesee H, Gibson L and Hunter CN (1999) The photosynthesis gene cluster of Rhodobacter sphaeroides. Photosynth Res 62: 121–139 Nguyen HH, Elliott SJ, Yip JH and Chan SI (1998) The particulate methane monooxygenase from Methylococcus capsulatus (Bath) is a novel copper-containing three-subunit enzyme. Isolation and characterization. J Biol Chem 273: 7957–7966 Niederman RA, Mallon DE and Langan JJ (1976) Membranes of Rhodopseudomonas sphaeroides. IV. Assembly of chromatophores in low-aeration cell suspensions. Biochim Biophys Acta 440: 429–447 Niederman RA, Mallon DE and Parks LC (1979) Membranes of Rhodopseudomonas sphaeroides VI. Isolation of a fraction enriched in newly synthesized bacteriochlorophyll a-protein complexes. Biochim Biophys Acta 555: 210–220 Parks LC and Niederman RA (1978) Membranes of Rhodopseudomonas sphaeroides. V. Identification of bacteriochlorophyll alpha-depleted cytoplasmic membrane in phototrophically grown cells. Biochim Biophys Acta 511: 70–82 Pelham HR and Jackson RJ (1976) An efficient mRNA-dependent translation system from reticulocyte lysates. Eur J Biochem 67: 247–256 Pokkuluri PR, Laible PD, Deng YL, Wong TN, Hanson DK and Schiffer M (2002) The structure of a mutant photosynthetic reaction center shows unexpected changes in main chain orientations and quinone position. Biochemistry 41: 5998–6007 Qian P, Addlesee HA, Ruban AV, Wang P, Bullough PA and Hunter CN (2003) A reaction center-light-harvesting 1 complex (RCLH1) from a Rhodospirillum rubrum mutant with altered esterifying pigments: Characterization by optical spectroscopy and cryo-electron microscopy. J Biol Chem 278: 23678–23685 Qian P, Hunter CN and Bullough PA (2005) The 8.5 Å projection structure of the core RC-LH1-PufX dimer of Rhodobacter sphaeroides. J Mol Biol 349: 948–960 Raman P, Cherezov V and Caffrey M (2006) The membrane protein data bank. Cell Mol Life Sci 63: 36–51 Research Collaboratory for Structural Bioinformatics (2008) RCSB Protein Data Bank. http://www.rcsb.org/pdb, February 5, 2008 Roh JH, Smith WE and Kaplan S (2004) Effects of oxygen and light intensity on the transcriptome expression in Rhodobacter sphaeroides 2.4.1. J Biol Chem 279: 9146–9155 Roszak AW, Howard TD, Southall J, Gardiner AT, Law CJ, Isaacs NW and Cogdell RJ (2003) Crystal structure of the RC-LH1
859
core complex from Rhodopseudomonas palustris. Science 302: 1969–1972 Schimz KL, Decker G, Frings E, Meens J, Klein M and Muller M (1995) A cell-free protein translocation system prepared entirely from a gram-positive organism. FEBS Lett 362: 29–33 Scott HN, Laible PD and Hanson DK (2003) Sequences of versatile broad-host-range vectors of the RK2 family. Plasmid 50: 74–79 Siebert CA, Qian P, Fotiadis D, Engel A, Hunter CN and Bullough PA (2004) Molecular architecture of photosynthetic membranes in Rhodobacter sphaeroides: The role of PufX. EMBO J 23: 690–700 Simon R, Priefer U and Puhler A (1983) A broad host range mobilization system for in vivo genetic engineering: Transposon mutagenesis in gram negative bacteria. Bio/Technology 1: 37–45 Spirin AS (2004) High-throughput cell-free systems for synthesis of functionally active proteins. Trends Biotechnol 22: 538–545 Taguchi AKW, Stocker JW, Alden RG, Causgrove TP, Peloquin JM, Boxer SG and Woodbury NW (1992) Biochemical characterization and electron-transfer reactions of sym1, a Rhodobacter capsulatus symmetry mutant which affects the initial electron donor. Biochemistry 31: 10345–10355 Takemoto J and Bachmann RC (1979) Orientation of chromatophores and spheroplast-derived membrane vesicles of Rhodopseudomonas sphaeroides: analysis by localization of enzyme activities. Arch Biochem Biophys 195: 526–534 Troschel D and Muller M (1990) Development of a cell-free system to study the membrane assembly of photosynthetic proteins of Rhodobacter capsulatus. J Cell Biol 111: 87–94 Troschel D, Eckhardt S, Hoffschulte HK and Muller m (1992) Cell-free synthesis and membrane integration of the reaction center subunit H from Rhodobacter capsulatus. FEMS Microbiol Lett 91: 129–133 Turner GJ, Reusch R, Winter-Vann AM, Martinez L and Betlach MC (1999) Heterologous gene expression in an membraneprotein-specific system. Prot Expres Purif 17: 312–323 van der Palen CJ, Reijnders WN, de Vries S, Duine JA and van Spanning RJ (1997) MauE and MauD proteins are essential in methylamine metabolism of Paracoccus denitrificans. Antonie Van Leeuwenhoek 72: 219–228 van Dijk R, Faber KN, Kiel JA, Veenhuis M and van der Klei I (2000) The methylotrophic yeast Hansenula polymorpha: A versatile cell factory. Enzyme Microb Technol 26: 793–800 van Grondelle R, Hunter CN, Bakker JGC and Kramer HJM (1983) Size and structure of antenna complexes of photosynthetic bacteria as studied by singlet-singlet quenching of the bacteriochlorophyll fluorescence yield. Biochim Biophys Acta 723: 30–36 Varga AR and Kaplan S (1993) Synthesis and stability of reaction center polypeptides and implications for reaction center assembly in Rhodobacter sphaeroides. J Biol Chem 268: 19842–19850 Verméglio A, Joliot P and Joliot A (1995) Organization of electron transfer components and supercomplexes. In: Blankenship RE, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 279–295. Kluwer Academic Publishers, Dordrecht Walz T, Jamieson SJ, Bowers CM, Bullough PA and Hunter CN (1998) Projection structures of three photosynthetic complexes
860
Philip D. Laible, Donna L. Mielke and Deborah K. Hanson
from Rhodobacter sphaeroides: LH2 at 6 Å, LH1 and RC-LH1 at 25 Å. J Mol Biol 282: 833–845 Wang Z (2006) Controlled expression of recombinant genes and preparation of cell-free extracts in yeast. Methods Mol Biol 313: 317–331 Wieseler B and Muller M (1993) Translocation of precytochrome c2 into intracytoplasmic membrane vesicles of Rhodobacter capsulatus requires a peripheral membrane protein. Mol Microbiol 7: 167–176 Wilks HM and Timko MP (1995) A light-dependent complemen-
tation system for analysis of NADPH:protochlorophyllide oxidoreductase: Identification and mutagenesis of two conserved residues that are essential for enzyme activity. Proc Natl Acad Sci USA 92: 724–728 Yang F, Moss LG and Phillips GN, Jr. (1996) The molecular structure of green fluorescent protein. Nat Biotechnol 14: 1246–1251 Zilsel J, Lilburn TG and Beatty JT (1989) Formation of functional inter-species hybrid photosynthetic complexes in Rhodobacter capsulatus. FEBS Lett 253: 247–252
Chapter 43 Assembly of Bacterial Light-Harvesting Complexes on Solid Substrates Kouji Iida1,2, Takehisa Dewa2 and Mamoru Nango2* 1
Nagoya Municipal Industrial Research Institute, Rokuban 3-4-41, Atsuta-ku, Nagoya 456-0058, Japan; 2Department of Applied Chemistry, Nagoya Institute of Technology, Gokiso-cho, Showa-ku, Nagoya 466-8555, Japan
Summary .............................................................................................................................................................. 861 I. Introduction..................................................................................................................................................... 862 II. Atomic Force Microscopy Imaging of Reassociated Bacterial Light-Harvesting Complex 1 on a Mica Substrate ............................................................................................................................................... 864 A. Reassociation of Light-Harvesting 1 Complexes.............................................................................. 864 B. Near Infra-red Absorption and Fluorescence Spectra of Light-Harvesting 1 Complexes on a Mica Substrate ................................................................................................................................. 864 C. Atomic Force Microscopy Images of Light-Harvesting 1 Complexes in the Absence and Presence of Carotenoids on a Mica Substrate ................................................................................ 864 III. Conductivity of the Bacterial Reaction Center on Chemically Modified Gold Substrates Using Conductive Atomic Force Microscopy ............................................................................................................ 865 A. Chemical Modification of Gold Substrates Using a Self-assembled Monolayer of Mercaptopyridines and Reaction Center Adsorption onto the Substrate ......................................... 865 1. Preparation of Chemically Modified Au (111).......................................................................... 865 2. Preparation of Reaction Centers on Chemically Modified Au (111) ........................................ 866 B. Current-voltage Curves of Reaction Centers Adsorbed on Chemically Modified Au (111) .............. 867 1. Atomic Force Microscopy Imaging and Conductive Atomic Force Microscopy Current-voltage Measurements............................................................................................... 867 2. Current-voltage Curves Measured Using a Probe Modified with 2-Mercaptopyridine ............ 868 IV. Molecular Assembly of Photosynthetic Antenna Core Complex on an Amino-terminated Indium Tin Oxide Electrode ........................................................................................................................................ 869 A. Stability of Various Components of the Photosynthetic Unit: Light-Harvesting 1, Reaction Center, Native Reaction Center-Light-Harvesting 1, and Admixed Reaction CenterLight-Harvesting 1 Complexes ......................................................................................................... 869 B. Photoinduced Current from Adsorbed Photosynthetic Unit Components ........................................ 870 V. Concluding Remarks ..................................................................................................................................... 873 Acknowledgments ................................................................................................................................................. 873 References ............................................................................................................................................................ 873
Summary Light-harvesting (LH) and reaction center (RC) complexes were adsorbed onto mica and gold substrates, in order to assemble an artificial antenna complex on a solid support, with the eventual aim of developing useful nanodevices. The normal near-infra-red (IR) absorption spectra of photosynthetic complexes were retained indicating stable assembly on these solid substrates. Atomic Force Microscopy (AFM) of reassociated LH1 complexes, under air, showed the expected ring-like structure. *Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 861–875. © 2009 Springer Science + Business Media B.V.
862
Kouji Iida, Takehisa Dewa and Mamoru Nango
RCs adsorbed onto a self-assembled monolayer (SAM)-modified Au(111) substrate were sandwiched between a gold-coated probe and the SAM, and conductive atomic force microscopy (CAFM) was used to measure the conductivity of the RCs in vacuo, in order to derive the fundamental physical parameters that relate the electric functionality of the RCs to the properties of their redox carriers. Various compounds were prepared as SAM materials to investigate the stability and morphology of RCs on the substrate using near-IR absorbance spectroscopy and AFM, respectively. A clear rectification of this current was observed for the modification of the Au(111) substrate with the π-conjugated thiol 2-mercaptopyridine (2MP), indicating that 2MP was effective in both promoting the specific orientation of the RCs on the gold electrode and enhancing the current. The 2MP substrate is therefore a suitable chemical modifier for Au(111) with which to measure CAFM of single RCs. Reaction Center-Light-Harvesting 1 (RC-LH1) core complexes isolated from Rhodospirillum rubrum were successfully self-assembled on an indium tin oxide electrode modified with 3-aminopropyltriethoxysilane. Near IR absorbance, fluorescence and IR spectra indicated that these RC-LH1 complexes were stable on the electrode. Efficient energy transfer and photocurrent responses of these RC-LH1 complexes were observed upon illumination of the LH1 complex at 880nm. These RC-LH1 complexes are applicable to molecular devices such as photocurrent generators.
I. Introduction Solar energy absorbed by light-harvesting (LH) complexes is rapidly transferred to the reaction centers (RCs); together, these complexes form the so-called photosynthetic unit (PSU). The detailed mechanisms of energy transfer and trapping involving several types of membrane-bound pigment-protein complexes have been gradually revealed (Ke, 2001; Chapter 13, van Grondelle and Novoderezhkin). In purple bacteria light energy absorbed by LH2 (McDermott et al., 1995; Chapter 8, Gabrielsen et al.) is rapidly funneled into a core RC-LH1 complex (Roszak et al., 2003; Fotiadis et al., 2004; Cogdell et al., 2004; Chapter 9, Bullough et al.), where subsequent charge separation takes place. Analyses by X-ray crystallography (McDermott et al., 1995; Roszak et al., 2003; Cogdell et al., 2004, 2005) and by electron cryo-microscopy of 2-dimensional (2-D) crystals (Karrasch et al., 1995; Qian et al., 2003; 2005) have revealed the structures of these components of the PSU. The crystal structure of RC-LH1 core complex of Rhodopseudomonas (Rps.) Abbreviations: 2-D – 2-dimensional; 2MP – 2-mercaptopyridine; AFM – atomic force microscopy; APS – 3-aminopropyltriethoxysilane; BChl – bacteriochlorophyll; BPhe – bacteriopheophytin; CAFM – conductive atomic force microscopy; IR – infra-red; ITO – indium tin oxide; LDAO – N, N´-dimethyldodecylamine N-oxide; LH1 – light-harvesting 1 complex; LH2 – peripheral light-harvesting complex 2; MV – methyl viologen; PSU – photosynthetic unit; Rba. – Rhodobacter; RC – reaction center; RC-LH1 – reaction center-light-harvesting 1 core complex; Rsp. – Rhodospirillum; SAM – self assembly monolayer; β-OG – n-octyl-β-D-glucopyranoside
palustris has showed that the RC is surrounded by the ellipsoidal 15-membered LH1 complex (Roszak et al., 2003). The structure of the similar RC-LH1 core complex derived from Rhodospirillum (Rsp.) rubrum was demonstrated by cryo-electron microscopy (Jamieson et al., 2002; Qian et al., 2003). The nanoscale apparatus present in photosynthetic membranes has been optimized for low-power output and operation with a high quantum efficiency. In molecular electronics, great advances have been made in the integration and miniaturization of organic devices using components of PSUs. Recently an organic device using nanoassembled films has been demonstrated (Matsui et al., 2003, 2004, 2005; Nagata et al., 2003). Similar advances in bio-nanotechnology require systems for assembling arrays of membrane proteins, such as those found in the bacterial photosynthetic membrane, in both nanoassembled films and on electrodes (Iida et al., 2000a,b, 2001, 2005; Ogawa et al., 2002, 2004; Das et al., 2004; Dewa et al., 2005, 2006). Atomic force microscopy (AFM) has been used to image such arrays comprising the PSU in native membranes prepared from several species of photosynthetic bacteria, as well as artificial arrays reconstituted from purified complexes (Scheuring et al., 2001, 2003, 2004a,b; Stamouli et al., 2003; Bahatyrova et al., 2004; Fotiadis et al., 2004). This topic is reviewed in detail elsewhere in this volume (Chapter 47, Scheuring). As an example, AFM topographs have been recorded for the RC-LH1 complex in native photosynthetic membranes from Rps. viridis (Scheuring et al., 2003) and Rsp. photometricum (Scheuring et al., 2004a,b). Fotiadis et al. (2004) recorded high resolution topographs of the
Chapter 43
Assembly of Antenna Complexes on Solid Substrates
RC-LH1 complex from Rsp. rubrum reconstituted into 2-D crystals using lipids from Escherichia coli. Bahatyrova et al. (2004a,b) showed that the LH1 complexes of a mutant lacking the RC form circular, elliptical, and even polygonal ring shapes as well as arcs and open rings, and that the LH1 complexes are positioned to function as an energy collection hub from the LH2 complex to RC in membranes from the wild-type bacterium. These PSUs are an in vivo ensemble of nanoscale components which collectively are able to harvest, transfer and trap energy (Chapter 15, Şener and Schulten). The 2-D arrays of complexes that comprise the PSUs are intriguing when one considers how they function and interplay in such arrays and also how it might be possible to artificially fabricate them for developing nano-biomaterial assemblies (Bahatyrova et al., 2004b; Das et al., 2004; Reynolds et al., 2007; Escalante et al., 2008). The RC is a useful membrane protein for these types of studies because the stability and the spectroscopic properties of its pigments can be used to check that the RCs remain in a native state when assembled into nanofilms or onto electrodes. Currents produced by RC films have already been reported (Lee et al., 1997; Das et al., 2004; Stamouli et al., 2004). Scanning probe microscopy, including AFM, possesses the potential of not only imaging but also measuring a single biomolecule (Lee et al., 1997; Stamouli, 2004). Recently scanning probe analyses, including scanning tunneling microscopy and conductive-AFM (CAFM), have been shown to be very useful because they allow the study of the electronic properties of individual molecules in nanoassembled films on conductive metal substrates without the requirement for averaging inherent in most other methods. In particular, CAFM allows the contact resistance between the sample and tip to be controlled during current-voltage measurements. In a previous publication, Stamouli et al. (2004) reported the electron conduction of RCs in lipid bilayers on highly orientated pyrolytic graphite using CAFM with a Pt/Ir coated probe in air. It is desirable to establish a methodology by which fragile PSU components are put onto a solid substrate. In this area of research there are two fundamental issues. Firstly, how to order proteins on the electrodes with respect to their sidedness, and secondly how to make a good electrical contact between the protein and the electrode. The adsorption of functional PSU proteins on substrates is very difficult because they often denature and their inherent electron transfer capability is then compromised. Chemical modification of the electrodes is commonly used to solve this problem (Chen et al., 2002). In addition, these
863
chemical modifications also play a critical role in making a good electrical contact between the PSU protein and the electrode (Das et al., 2004). However, there are rather few reports of studies looking for better chemical modifiers of electrodes (Sagara et al., 1990; Taniguchi et al., 2003; Ogawa et al., 2004; Mikayama et al., 2006). This knowledge and methodology would help to achieve the aim of using PSU complexes in a nanodevice. We have recently reported that RC-LH1 core complexes isolated from Rsp. rubrum and Rps. palustris can be assembled onto a cationically-modified transparent indium tin oxide (ITO) electrode, which exhibits photoinduced current generation (Ogawa et al., 2004). Our current understanding of energy transfer and charge separation reactions in the LH2 and RC-LH1 complexes has enabled the first step to be taken towards generating artificial systems that convert light energy into a usable electrical current. Previous attempts to produce an artificial, energy-converting electrode system used either LH1 complexes (Nagata et al., 2003) or RCs immobilized on the electrodes. Alegria and Dutton (1991), Yasuda et al. (1992) and Fang et al. (1995) have all reported devices using Langmuir-Blodgett films containing bacterial RCs. Until now, there have only been a few attempts to immobilize intact core complexes, consisting of both the RC and LH1 components, onto an electrode (Das et al., 2004; Ogawa et al., 2004; Suemori et al., 2006, 2007). We have recently developed a procedure to create a self-assembled monolayer (SAM) of reconstituted LH1 complexes on a transparent indium tin oxide (ITO) electrode modified with 3-aminopropyltriethoxysilane (APSITO) between the electrode surface and the anionic LH1 polypeptides at pH 8.0 (Ogawa et al., 2002). The near infra-red (IR) absorption spectrum showed that the LH1 complex was stable when immobilized onto these electrodes. This study was extended using native RC-LH1 complexes isolated from Rsp. rubrum and Rps. palustris, which were successfully assembled onto an APS-ITO electrode; efficient energy transfer and photocurrent responses could be observed upon illumination at 880 nm (Ogawa et al., 2004). This chapter reports: (1) the morphology and stability of the Rsp. rubrum LH1 complex on a mica substrate, studied by near-IR absorption and fluorescence spectroscopy, as well as by AFM; (2) the topography and conductivity of Rba. sphaeroides RCs measured with CAFM on SAM-modified Au(111) substrates; (3) the electrochemical behavior of Rsp. rubrum complexes on a modified ITO electrode to investigate
864
Kouji Iida, Takehisa Dewa and Mamoru Nango
assembly on the electrode and to develop photocurrent generation system for these assemblies—RC-LH1, LH1 and RC complexes, and an admixed complex of LH1 and RC. We discuss the potential usage and architecture of PSU assemblies on modified electrodes in terms of artificial photosynthesis and progress towards developing useful nanodevices.
indicated stable and efficient energy transfer from carotenoid to BChl a (Iida et al., 2005). In previous work LH2 complexes were spin coated together with polyvinylalcohol on LiF (Van Oijen et al., 1999). In this case, these authors were undertaking a single molecule emission spectroscopy study of the LH2 complex; they also noted a correspondence between fluorescence properties of the LH complexes in micelles and on the LiF surface, which indicated that no denaturation had taken place.
II. Atomic Force Microscopy Imaging of Reassociated Bacterial Light-Harvesting Complex 1 on a Mica Substrate A. Reassociation of Light-Harvesting 1 Complexes It is now generally accepted that the LH1 subunits of Rsp. rubrum can be reassociated to give holo LH1 complexes where the bacteriochlorophyll (BChl) a Qy absorption band is red shifted to 870 nm, the same absorption maximum seen for native LH1 complexes (Miller et al., 1987; Visschers et al., 1991; see also Chapter 10, Loach and Parkes-Loach). The LH1 subunit complexes are reassociated to form the holo LH1 complexes by dilution of the β-octyl glucoside (β-OG) concentration below the critical micelle concentration (20 mM) as shown in Fig. 1. The BChl a molecules then become excitonically aggregated and the hydrophobic interactions between the subunit LH1 complexes play an important role in the refolding process. B. Near Infra-red Absorption and Fluorescence Spectra of Light-Harvesting 1 Complexes on a Mica Substrate LH1 complexes were spin coated onto mica and rinsed with water. Near IR absorption spectra were recorded; the BChl a Qy band was at 870 nm in the absence of carotenoids, which is consistent with that seen for the complex in aqueous solution at 25 °C. Similar results were obtained for the reassociated LH1 complex in the presence of carotenoids. These results indicate that this complex was not denatured by binding to the mica surface; the LH1 complex on the mica was stable for 24 hours in a dark room at 4 °C. It has also been shown that a Langmuir-Blodgett film can be used to deposit a membrane layer of LH1 complexes on a glass substrate (Iida et al., 2000a). The functionality of LH1 bound to mica was also supported by fluorescence emission spectra, which
C. Atomic Force Microscopy Images of LightHarvesting 1 Complexes in the Absence and Presence of Carotenoids on a Mica Substrate Figure 1 shows AFM images of LH1 complexes in the absence of both RC and carotenoids on mica, recorded using AC (tapping) mode and a standard silicon probe with either a 7.5 N/m and 110 µm length (MicroMasch NSC 12) or 40 N/m and 125 µm length (Seiko Instruments Inc., SII ). The AFM topographs of the reassociated LH1 complexes, under air, showed the expected ring-like structure observed in the 300 × 300 nm2 image (Iida et al., 2005); both individual and fused rings could be seen. The average diameters of the ring-like structures were approximately 30 nm (outers) and 8 nm (inner) (n=24), respectively. The height of the rings was measured to be 0.2–0.6 nm above background. These inner diameters are within the size for LH1 rings predicted from X-ray and cryoelectron microscopy studies (Karrasch et al., 1995; Roszak et al., 2003). However, the outer diameter was too large since it was expected to be 12 nm, on the basis of cryo-electron microscopy studies of the LH1 and RC-LH1 complexes (Karrasch et al., 1995; Jamieson et al., 2002). In the present study, the LH1 complexes were transferred to the mica from β-OG detergent solution. Thus, the large outer diameter could arise from a belt of β-OG molecules surrounding the LH1 ring, adsorbed as approximately 5 layers (since the length of the β-OG molecules is approximately 2 nm). Although it is difficult to define the subunits at this level of resolution, broken rings were observed, which indicate the presence of an incompletely refolded LH1 complexes. Bahatyrova et al. (2004b) also imaged broken rings of Rba. sphaeroides LH1 using AFM. Interestingly, a 2-D aggregated patch of LH1 complexes was seen in the presence of carotenoids, implying that the carotenoids may play a crucial role in the formation of such patches in the absence of lipid (Iida et al., 2005).
Chapter 43
Assembly of Antenna Complexes on Solid Substrates
865
Fig. 1. Reassociation of LH1 complexes from Rhodospirillum rubrum, deposition of a monolayer of the LH1 complexes on a mica substrate, and AFM topographs of the monolayers. Photosynthetic membranes from Rsp. rubrum were solubilized in buffer containing 30 mM β-OG to dissociate the RC-LH1 complexes. The LH1 subunit complexes were reassociated to form LH1 complexes in β-OG aqueous solution. The LH1 complexes were spin coated onto freshly cleaved mica and directly observed by using AFM. The area shown is 300 × 300 nm2, and the zoomed area 153 × 153 nm2. A typical height profile is shown, which corresponds to the white bar on the 300 × 300 nm2 area, with A, A’, B, B’ denoting the ~30 nm outer diameters of the ring-like structures, as well as the heights of the rings, measured to be 0.2–0.6 nm above background.
It is important to avoid denaturation of immobilized LH complexes on the various substrates and to establish that they retain efficient energy transfer so that these complexes can then be used for non-linear optics or photo-sensitizers in a photonic nanodevice. There are many polar amino acids such as glutamic acids and asparagines in the predicted extrinsic N-terminal region of the LH1α and β polypeptides from Rsp. rubrum. This suggests that these N-terminal regions may be weakly adsorbed onto the mica surface as a result of hydrophilic interactions. Thus, LH1 complexes bound to mica will be useful for studying energy transfer between complexes, or from carotenoids to BChl a within each complex, as well as for applying this knowledge of antenna function to energy transfer and non-linear optics devices. III. Conductivity of the Bacterial Reaction Center on Chemically Modified Gold Substrates Using Conductive Atomic Force Microscopy In these experiments Au(111) was used as coating
metal for both the substrate and the probe to try to obtain electrodes with work functions similar to those shown in Fig. 2 (Mikayama et al., 2006). The aim was to try and derive the fundamental physical parameters that relate the electric functionality of the RCs to the properties of their redox carriers. A. Chemical Modification of Gold Substrates Using a Self-assembled Monolayer of Mercaptopyridines and Reaction Center Adsorption onto the Substrate 1. Preparation of Chemically Modified Au (111) A 100 nm thickness of gold was epitaxially deposited onto prebaked mica with an evaporation rate of 0.3–0.4 nm s–1.The Au(111) substrates were then immersed in a 1 mM solution (in distilled ethanol) of one of the following: mercaptoacetic acid (HSCH2COOH, MAC), 2-mercaptoethanol (HSCH2CH2OH, 2ME), 4-mercaptopyridine (C5H5NS, 4MP), and 2-mercaptopyridine (C5H5NS, 2MP) (Fig. 2a) for 2 hours.
866
Kouji Iida, Takehisa Dewa and Mamoru Nango
Fig. 2. (a) The chemical structures of modifiers, MAC, 2ME, 4MP and 2MP. (b) Schematic diagram showing CAFM applied with RCs sandwiched between the Au(111) substrate and the cantilever. (c) Illustration of pigments oriented in Rba. sphaeroides RCs, from X-ray crystallography data and showing the light-induced electron transfer pathway (dashed line).
2. Preparation of Reaction Centers on Chemically Modified Au (111) RC complexes were isolated from the Rba. sphaeroides mutant PUC705BA as previously described (Okamura, 1974). 100 µl of a 1 µM RC solution in 0.8% β-OG (in 10mM Tris-HCl buffer, pH 8) was cast by syringe onto each SAM-modified Au substrate (2 cm × 1 cm) for the near-IR absorbance measurements. The chemically-modified Au(111) substrates for the AFM measurements were incubated with a 1µM RC solution in 5 ml for 10 minutes in the dark. The surface roughness of the RC films was 2.87 nm, which was smaller than the size of a single RC complex (Chang et al., 1991). Dense grains with various diameters of 5–20 nm, which corresponded to aggregates of up to 10 RCs, were observed in all the topographs of the RCs adsorbed onto all the different SAM-modified Au(111) substrates (Mikayama et al., 2006). Figure 3 shows the near-IR absorbance spectra of the RCs in solution and on the chemically-modified Au (111) substrates in the presence of the mild reducing agent, 20 mM L-ascorbic acid. The well-known RC absorbance peaks at 760 nm and 800 nm were observed for the RCs cast on the 2ME-, 2MP-, 4MPand MAC-modified Au (111) substrates, as seen with the RCs in solution. These absorbance bands come from the bacteriopheophytin (BPhe) and accessory BChl a molecules in the RCs, respectively (Ke, 2001). Interestingly, the 860 nm peak, which is due to the BChl special pair in the RC, was not observed in near-IR spectra of RCs on the SAM modified Au(111)
Fig. 3. Near-IR spectra of RCs in solution and immobilized on MAC, 2ME, 4MP and 2MP-modified Au substrates. Each spectrum is shown with an offset.
substrates before these were dipped in the reducing agent. Clearly placing the RCs onto the gold substrate oxidizes them. The Qy band of BChl a of monomeric state is 777 nm (Ke, 2001). Figure 3 shows that the decreasing absorbance at 860 nm was not accompanied by an increase in 777 nm absorbance for RCs bound to the 2ME-, 2MP-, and 4MP-modified Au (111) substrates. Thus, the presence of these three near-IR absorbance peaks clearly indicates that the RCs were native on the gold substrates. However, the RC absorbance is partly affected on MAC-modified Au(111), implying that the carboxylic acid surface might denature the RC. In the absence of the SAMs which modify the Au(111) surface the RCs were not adsorbed.
Chapter 43
Assembly of Antenna Complexes on Solid Substrates
B. Current-voltage Curves of Reaction Centers Adsorbed on Chemically Modified Au (111) 1. Atomic Force Microscopy Imaging and Conductive Atomic Force Microscopy Current-voltage Measurements The AFM imaging and CAFM measurements were carried out under vacuum (5 × 10–5 Pa) at room temperature (Mikayama et al., 2006). AFM images were recorded with an SPI-3800 controller and SPA-300HV unit (SII Nanotechnology, Japan). SiN cantilevers, with a spring constant of 0.09 N m–1, were used. The Au/Cr-coated SiN cantilevers (Olympus) had a spring constant of 0.025 N m–1 and a typical tip radius of 40 nm. The current was monitored with a highly sensitive current-voltage amplifier connected to the conductive tip. We looked for areas of densely packed single RCs by searching for 5–20 nm diameter grains by AFM imaging as shown in Fig. 4a. When such an area was found, the conductivity of each grain was measured. A ramp bias voltage, with a typical sweep time of 100 ms, was applied between the conductive tip and the lower Au(111) substrate. A typical applied cantilever force was 1 nN. Each CAFM current-voltage curve was obtained
867
by averaging 40 individual scans. Current-voltage measurements with Au-coated probes were determined in vacuo at room temperature on grains with diameters of 5–20 nm i.e. single RCs (Fig. 4a, inset). The current-voltage curves were obtained by averaging 40 individual current-voltage curves taken from different grains selected at random. Before the acquisition of every current-voltage curve the applied force of the AFM probe was increased in order to measure a detectable current. It was noted, however, that increasing this applied force from 1 nN to 5 nN, between the tip and the RCs, resulted in an unsteady current. The effective current path mediated by the four redox components in the RCs may be, therefore, sensitive to protein deformation. Zhao et al. (2004) have observed such a pressure dependent current with a metalloprotein that includes azurin as a redox component. Figure 4b shows the measured current as a function of the applied voltage between the substrate and tip. The adsorbed RCs displayed different current-voltage characteristics, depending on the type of chemical modifiers. Table 1 summarizes the data with measured currents at ±1 V and gives the rectification ratios. The current and the rectification ratio were the lowest with MAC-modified Au(111); MAC forms a negatively charged surface and this results in
Fig. 4. a) AFM topograph (500 × 500 nm2) of RCs on a 2MP-modified Au(111) substrate imaged with a Au coated cantilever. The insert shows a highresolution topograph with a 50 nm scan. The grains were measured by CAFM. b) Four I-V graphs of RCs each with differently chemically-modified surfaces; substrate/SAM/RC/probe: Au/2ME/RC/Au, Au/MAC/RC/Au, Au/2MP/RC/Au and Au/4MP/RC/Au.
Kouji Iida, Takehisa Dewa and Mamoru Nango
868
Table 1. Current and rectification ratio of RCs at ±1 V on various SAMAu(111) substrates SAM compound
I (pA) –1 (V) down
a
Rectification ratio, % +1 (V) upa
|up|/(|down| + |up|)a
MAC
–10
10
50.0
2ME
–20
120
85.7
2MP
–80
570
87.7
–8.0
25
75.8
–150
1000
87.0
4MP b
2MP-2MP a
RC anchored perpendicular down and up (up as in Fig. 1b) to Au(111). b Using a 2MP-modified Au-coated Si cantilever.
the denaturation of the RCs, as shown by the near-IR spectrum (Fig. 3). In this case the current-voltage curve was symmetric at about V= 0. The current of 570 pA at 1 V from RCs mediated by 2MP was the best and about twenty times higher than with the RCs mediated by 4MP. Current-voltage graphs of films with RCs oriented on 2ME, 2MP, and 4MP-modified Au(111) substrates showed an asymmetric conductance between the positive and negative bias voltages. This contrasts with the symmetric current-voltage graphs measured for the RCs on a MAC-modified substrate. In the former cases the current increases on reverse bias and remains small in forward bias. Current rectification by organic molecules has been reported in the literature for zwitterionic molecules (Martin et al., 1993), phythalocyanine molecules(Page et al., 1999) and photosynthetic RCs from plants (Lee et al., 1997). In the case of Photosystem I from plants the current-rectifying current-voltage curve, obtained with scanning tunneling microscopy, has been attributed to the orientation of the Photosystem I complexes on the surface (Stamouli et al., 2003) and has, therefore, been used to indicate the orientation with which the complex preferentially adsorbs on the surface. An alternative type of electron transfer is via electronic states existing in the molecule sandwiched between the two electrodes. These could be localized or delocalized over the entire molecule. Since the X-ray crystal structure of the RC has been described, this information can be used to suggest a possible molecular electron transfer pathway (Deisenhofer et al., 1985; Chang, 1991). In this study, the direction of the rectification in the cases of the Au(111) electrode
modified with 2ME, 2MP, and 4MP SAMs suggests that the RCs from Rba. sphaeroides are preferentially adsorbed at RC-H subunit and that electrons from the gold coated probe transfers to the chemically modified Au(111) substrate by the pathway as shown in Fig. 2c (and see below). Two possible reasons for the enhanced current with 2MP are (a) better orientation of the RCs and (b) lowering of the electron injection barrier. For efficient electron injection into proteins, the work function of the electrode has to match the energy level of the lowest unoccupied molecular orbital or the highest occupied molecular orbital of the conducting species. Tuning of the metal work function with the SAM enables the current density in organic solar cells to be increased (De Boer et al., 2005), as demonstrated by several studies using of π-conjugated thiols to decrease the injection barrier. If the current-voltage measurements are performed on RCs sandwiched between a 2MP-modified Au(111) substrate and a 2MP-modified Au-coated probe, then in principle the effect of lowering the carrier injection barrier can be distinguished from the effect of orientation of the RCs (Sumi, 1997). 2. Current-voltage Curves Measured Using a Probe Modified with 2-Mercaptopyridine Figure 5a shows AFM topographs of RC complexes adsorbed on a 2MP-modified Au(111) substrate, measured in vacuo (5 × 10–4 Pa) and obtained with a 2MP-modified Au-coated Si probe (Mikayama et al., 2006). The resolution is the same as using an unmodified Au-coated Si probe. These results indicate that
Chapter 43
Assembly of Antenna Complexes on Solid Substrates
869
Fig. 5. a) AFM topograph (1000 × 1000 nm2) of RCs on a 2MP-modified Au(111) substrate with a 2MP-modified gold coated cantilever. The configuration is the same as for Fig. 2. b) I-V curves of RCs: upper curve – RCs sandwiched between a 2MP-modified Au(111) substrate and a 2MP-modified Au-coated probe; lower curve – RCs measured on a 2MP-modified Au(111) substrate, but measured using a probe coated only with Au.
the tip radius is about 30 nm. There was, therefore, no increase in the contact area between probe and sample surface due to the chemical modification. Au-coated probes with tip radii of 10–20 nm have often been used in previous CAFM measurements of monolayers (Davis et al., 2004; Zhao et al., 2004). The tip radius of the chemically-modified probes used in our study are, therefore, only a little larger than those used by other groups. Stable current-voltage curves were obtained using a 2MP-modified Au-coated probe prepared with a dipping time of 2 hours. Using chemically-modified probes prepared with shorter dipping times (<1 hour) the AFM images were similar but the CAFM current through RCs became quite unstable. The current-voltage curve of RCs sandwiched between a 2MP-modified Au(111) substrate and a 2MPmodified Au-coated probe is shown in Fig. 5b and compared with the current-voltage curve obtained using an Au-coated probe alone. The current value of over 1 nA at 1 V with the 2MP-modified probe was twice as high as with the Au-coated one. This is a striking enhancement of the current, where the current rectification behavior was very similar to that seen using the unmodified Au-coated probe and the 2MP-modified Au(111) surface, implying that 2MP has an important role in lowering the electron injection barrier between RCs and the Au-coated electrode.
IV. Molecular Assembly of Photosynthetic Antenna Core Complex on an Aminoterminated Indium Tin Oxide Electrode A. Stability of Various Components of the Photosynthetic Unit: Light-Harvesting 1, Reaction Center, Native Reaction CenterLight-Harvesting 1, and Admixed Reaction Center-Light-Harvesting 1 Complexes Figure 6 shows the absorption spectra of various components of the Rsp. rubrum PSU in solution and assembled onto an APS-ITO electrode. RCs were purified from photosynthetic membranes of Rsp. rubrum using N,N´-Dimethyldodecylamine Noxide (LDAO) and ion-exchange chromatography on diethylaminoethyl cellulose (Noël et al., 1972). LH1 complexes were purified by the method of Picorel et al. (1983) using LDAO, Triton X-100, and ion-exchange chromatography. The RC-LH1 core complex, purified using LDAO and ion-exchange chromatography had an A880:A280 absorption ratio of 2.0, indicating a very pure preparation. The basic methods for preparation of these purified complexes assembled on APS-ITO electrodes have been reported previously (Ogawa et al., 2002, 2004). LH1, RC, and RC-LH1 core complexes were assembled on a piranha treated (H2SO4/H2O2) APS-ITO electrode by immers-
870
Kouji Iida, Takehisa Dewa and Mamoru Nango
Fig. 6. Near-IR absorption spectra of various components of the Rsp. rubrum PSU in solution and assembled onto an APS-ITO electrode. In each panel the solid lines correspond to spectra acquired in a β-OG micellar solution, and the dotted lines correspond to spectra of complexes on the electrode. (a) LH1-type complexes. (b) RC-LH1 complexes. (c) admixed LH1 and RC complexes. (d) The absorbance spectrum of RCs in solution.
ing it in the solution containing the photosynthetic complexes for 12 hours at 4 ˚C. An alternative method of preparing an electrode coated with components of the PSU employed a suspension of proteoliposomes composed of photosynthetic complexes and 1,2dimyristoyl-sn-glycero-3-phosphocholine (DMPC). The proteoliposomes were deposited onto a clean ITO electrode, followed by evaporation of the water in vacuo for a few hours, which resulted in a thin film of complexes on the electrode (Nagata et al., 2003). It is apparent from the data in Fig. 6 that the spectra of the LH1 complex (Fig. 6a) and the RC-LH1 core complex (Fig. 6b) on the assembled electrodes were identical to those of the purified, solubilized complexes, indicating that these complexes can be assembled on the electrode without denaturation. Fig. 6c shows the absorption spectra of admixed LH1 and RC complexes, which simulate the composition of the natural PSU. The solution (solid line) consists of a 1:1 mixture of LH1 and RC complexes; the spectral profile is therefore quite similar to that of the RC-LH1 core complex (Fig. 6b), which is composed of RC and LH1 complexes in their natural 1:1 stoichiometry, as found in the Rsp. rubrum PSU. Again, the spectrum of the admixed, reconstituted PSUs is similar to that in solution, indicating the likelihood that LH1 and RC complexes are assembled onto the electrode together. The absorption spectrum of RCs in solution
is shown in Fig. 6d; when RCs were assembled on the APS-ITO electrode, the absorption bands could not be detected, which might be due to denaturation on the electrode. Further, in the absence of the APS material on the ITO surface the PSU components were not adsorbed. In contrast, the surface coverage of complexes on the on the APS-ITO electrode was roughly estimated to be 50–60%, indicating a monolayer assembly. B. Photoinduced Current from Adsorbed Photosynthetic Unit Components Figure 7 shows that PSU components assembled on APS-ITO electrodes are capable of photoinduced current generation (Ogawa et al., 2004; Suemori et al., 2006; Suemori et al., 2007). Photocurrents were measured at –0.2V (vs. Ag/AgCl) in a homemade cell that contained three electrodes; a PSU-assembled APS-ITO electrode as a working electrode (1 cm2), an Ag/AgCl (saturated KCl) reference electrode, and a platinum flake as a counter electrode. The solution consisted of 0.1 M phosphate buffer (pH 7.0), containing 0.1M NaClO4 and 5 mM methyl viologen (MV). The electrode was irradiated periodically with 880 nm light (30 s on-off cycles as indicated by arrows in Fig. 7a). As expected, the LH1-assembled electrode
Chapter 43
Assembly of Antenna Complexes on Solid Substrates
871
Fig. 7. a) Photocurrent responses of PSU components assembled onto an APS-ITO electrode. The traces are labeled according to the complexes examined:. The assembled electrodes were irradiated with light at 880 nm periodically (30 s on – 30 s off cycles), b) absorption spectrum of RC-LH1 complexes on the ITO electrode (solid line) and the action spectrum for the photocurrent upon irradiation at 880 nm (dashed line).
did not respond to the incident light, but the RC-LH1 core complex and admixed RC and LH1 complexes did exhibit a photocurrent response. The absorption and action spectra for the RC-LH1 complex were shown in Fig. 7b). The RC-assembled electrode did not show a well-defined response, which could be due to denaturation of RCs on the surface, as mentioned above (Matsumoto et al., 1999; Suemori et al., 2006, 2007). It is interesting to note that the admixed RC/LH1 electrode exhibits a photocurrent activity comparable to that from the RC-LH1 core complex. One possibility is that the RC is stabilized by the LH1 complex, which absorbs the light energy and transfers it to RC in similar way to the intact RCLH1 core complex. A schematic illustration of the possible mechanism is depicted in Fig. 8. The present system was optimized for an electrode potential of –0.2 V; thus, it generates cathodic current by irradiation (Fig. 7a). The current-voltage curves of the PSUs measured here clearly demonstrate that these proteins do conduct electrons. There are various models to account for electron transfer in organic systems. Since the X-ray crystal structure of the RC has been described, this information can be used to suggest a possible molecular electron transfer pathway (Chang et al., 1991). In this study, the direction of the electron transfer on the modified electrode suggests that the RCs from Rsp. rubrum are adsorbed onto the electrode in the preferential orientation described below. The RC cofactors, namely the BChl dimer special pair (P) the monomeric BChls, BPhes and a quinone (QA), are the necessary electron transfer components (Fig. 8a;
Chapter 16, Jones). Direct excitation with light or, indirectly by resonance energy transfer from an antenna complex, first promotes an electron in P to an excited singlet state. Electrons are then transferred along the pigments associated with the L-subunit of the RC i.e. from P to BChl (0.47 nm transfer distance), to BPhe (0.38 nm transfer distance) and finally to QA (0.9 nm transfer distance). In this study, it is not clear where the link to MV takes place; it could be via BChl or BPhe since the redox potential of MV is much more negative than that of QA. The oxidized P might then be reduced by electron transfer from the electrode. A potential diagram of the present system is shown in Fig. 8b (Imahori, 2000; Suemori et al., 2006; Suemori et al., 2007). The photoexcited special pair (P*) transfers an electron to MV in solution and subsequently the electrode transfers an electron to reduce P+. The AFM images suggest that the P side of RC-LH1 complexes contacts the mica surface. Fotiadis et al. (2004) used AFM to show that the H subunit of the RC-LH1 complex protrudes from the membrane by 4 nm. This suggests that the C-terminal (periplasmic) side of the LH1 complex is adsorbed onto the mica, rather than the cytoplasmic face of the RC-H subunit, as depicted in Fig. 2c. This adsorption is mediated by hydrophilic interactions between the polar amino acids at the LH1 C-terminus and the amino-terminated ITO electrode. These results suggest that the C-terminus of the LH1 complex was oriented to the ITO substrate and the RC-H subunit faces the aqueous phase. Thus the rate determining step of the
872
Kouji Iida, Takehisa Dewa and Mamoru Nango
Fig. 8. a) Schematic diagram of RC-LH1 core complexes assembled onto an APS-ITO electrode together with a possible mechanism of electron flow. Direct excitation with light or, indirectly, by resonance energy transfer from the LH1 initiates electron flow between the electrode and the MV++ electron acceptor via the RC complex, catalytically under bias voltage, which generates a cathodic photocurrent as shown in Fig. 7a. Subsequently the electrode transfers an electron to reduce P+. b) Energy diagram for cathodic photocurrent generation by the RC-LH1 core complex in the present system.
photoinduced electron transfer might be between the electrode and P in the RC. In contrast negligible photocurrent was observed for LH1 or RC complexes on the electrode. When the LH1 complex of Rsp. rubrum alone was immobilized on the electrode, the observed photocurrent was mainly generated by light absorbed at 770 nm i.e. from monomeric BChl a (Nagata et al., 2003). This indicates that the core complex was well organized on the electrode and the photocurrents were driven by light that was initially absorbed by the LH components. Thus it is important to immobilize the RC-LH1 complex on various substrates without them being denatured and with a defined orientation relative to the electrode, as shown in Fig. 8. There is still considerable controversy about the density distribution at interfaces, especially in terms
of essential factors that determine the barrier height (Davis et al., 2004). If the direction of the current of the adsorbed PSUs could be controlled by their orientation, while keeping the same electrode geometry, then the proposed presence of an inherent current arising from the RCs would be verified (Akiyama et al., 2000). This, however, is a challenging problem. It is difficult to estimate the molecular ratio of admixed LH1 and RC complexes assembled on the electrode. To clarify this issue, an ITO electrode was examined whose surface was covered with a cast film consisting of LH1 and RC complexes incorporated into DMPC bilayers (Nagata et al., 2003). The advantage of this system is that the RC:LH1 ratio is controllable. The electrode coated with an LH1 film did not exhibit a photocurrent response. This
Chapter 43
Assembly of Antenna Complexes on Solid Substrates
is consistent with the result of LH1 complexes assembled on the APS-ITO electrode (Fig. 7a). As the RC:LH1 ratio rose to 1:1 the photocurrent response significantly increased., and above this ratio the photocurrent decreased (RC:LH1 = 2; Suemori et al., 2007). These results indicate that the photocurrent depends on the RC:LH1 ratio, but the reason for such dependence is not clear at the moment. Nevertheless, the demonstration that the admixed RC and LH1 complexes function as a photoinduced current generator as effectively as the RC-LH1 core complex is a new finding which reveals the potential of the methodology for assembling complexes on electrodes and for molecular assembly using wide variety of components. An LH1-type complex reassociated from subunit components (Miller et al., 1987) can be assembled on a substrate and an artificial LH1-type complex reconstituted from LH1α and LH1β polypeptides and zinc-substituted BChl a (Wakao et al., 1996; Ogawa et al., 2002; Nagata et al., 2003) can also be assembled on an APS-ITO electrode. These artificial LH1-type complexes on such substrates exhibited identical spectral profiles to those in solution. Therefore, not only the components of the native PSU, but also artificial photosynthetic components are applicable to molecular devices such as photocurrent generators. V. Concluding Remarks AFM imaging of reassociated LH1 complexes on mica, CAFM measurement of RC complexes on a gold substrate, and photocurrent responses of RC-LH1 complexes on an APS-ITO electrode upon near-IR illumination have been achieved. AFM revealed two-dimensionally self-organized LH1 complexes on a mica substrate, showing ring-like structures interpreted as individual LH1 molecules. The Qy dipole moments of the LH1 BChls are horizontal to the mica surface and efficient energy transfer from carotenoid to BChl could be detected. These results demonstrate that the self-organization of the refolded LH1 complexes on mica can be a powerful tool to better understand the supramolecular structure of LH1 complexes. Using CAFM, the electrical conduction properties of RCs sandwiched between a Au-coated probe and a Au(111) substrate in vacuo could be measured to derive the fundamental physical parameters that relate the electric functionality of the RCs. Modification of
873
the Au(111) substrate with the π-conjugated thiol, 2MP, was effective in both promoting the specific orientation of the RCs on the electrode and enhancing the current. The 2MP substrate is therefore a better chemical modifier for Au(111) with which to measure CAFM of single RCs. Further, molecular assemblies of various combinations of PSU components, LH1, RC and RC-LH1 complexes, were successfully deposited onto an APS-ITO electrode and the function of photoinduced current generation was evaluated upon excitation of the LH1 complex. These results provide useful methodologies to better understand the suprastrucure of PSUs as well as to gain knowledge of building an artificial fabrication of PSUs on solid substrates, with the eventual aim of constructing useful nanodevices. Acknowledgments M.N. thanks Professor Richard J. Cogdell and Dr. Alastair Gardiner, University of Glasgow, U.K., and Professors P. A. Loach and P. Parkes-Loach, Northwestern University, U.S.A. for the kind gifts of Rsp. rubrum and Rba. sphaeroides (PUC705-BA) and for helpful discussions on molecular assembly. The present work was partially supported by a Grant-in-Aid for Scientific Research on No.13480186, 15033236, 15655061, 18550150 from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of the Japanese Government and AOARD-06-4084. References Akiyama R, Matsumoto T and Kawai T (2000) Capacitance of a molecular overlayer on the silicon surface measured by scanning tunneling microscopy. Phys Rev B 62: 2034–2038 Alegria G and Dutton PL (1991) Langmuir-Blodgett monolayer films of bacterial photosynthetic membranes and isolated reaction centers: Preparation, spectrophotometric and electrochemical characterization. Biochim Biophys Acta 1057: 239–257 Bahatyrova S, Frese RN, Siebert CA, Olsen JD, van der Werf K, van Grondelle R, Niederman RA, Bullough PA, Otto C and Hunter CN (2004a) The native architecture of a photosynthetic membrane. Nature 430: 1058–1062 Bahatyrova S, Frese R, van der Werf KO, Otto C, Hunter CN, and Olsen JD (2004b) Flexibility and size heterogeneity of the LH1 light harvesting complex revealed by atomic force microscopy: Functional significance for bacterial photosynthesis. J Biol Chem 279: 21327–21333 Chang CH, El-Kabbani O, Tiede D, Norris J and Schiffer M (1991) Structure of the membrane-bound protein photosynthetic
874
Kouji Iida, Takehisa Dewa and Mamoru Nango
reaction center from Rhodobacter sphaeroides. Biochemistry 30: 5352–5360 Chen X, Ferrigno R, Yang J and Whitesides GM (2002) Redox properties of cytochrome c adsorbed on self-assembled monolayers: A probe for protein conformation and orientation. Langmuir 18: 7009–7015 Cogdell RJ, Gardiner AT, Roszak AW, Law CJ, Southall J and Isaacs NW (2004) Rings, ellipses and horseshoes: How purple bacteria harvest solar energy. Photosynth Res 81: 207–214 Das R, Kiley PJ, Segal M, Norville J, Yu AA, Wang L, Trammell SA, Reddick LE, Kumar R, Stellacci F, Lebedev N, Schnur J, Bruce BD, Zhang S and Baldo M (2004) Integration of photosynthetic protein molecular complexes in solid-state electronic devices. Nano Lett 4: 1079–1083 Davis JJ, Wrathmell CL, and Fletcher J (2004) The tunnelling conductance of molecularly ordered metalloprotein arrays. J Mol Recognit 17: 167–173 De Boer B, Hadipour A, Mandoc MM, van Woudenbergh T and Blom PWM (2005) Tuning of metal work functions with selfassembled monolayers. Adv Mater 17: 621–625 Deisenhofer J, Epp O, Miki K, Huber R and Michel H (1985) Structure of the protein subunits in the photosynthetic reaction centre of Rhodopseudomonas viridis at 3Å resolution. Nature 318: 618–624 Dewa T, Yamada T, Ogawa M, Sugimoto M, Mizuno T, Yoshida K, Nakao Y, Kondo M, Iida K, Yamashita K, Tanaka T and Nango M (2005) Design and expression of cysteine-bearing hydrophobic polypeptides and their self-assembling properties with bacteriochlorophyll a derivatives as a mimic of bacterial photosynthetic antenna complexes. Effect of steric confinement and orientation of the polypeptides on the pigment/polypeptide assembly process. Biochemistry 44: 5129–5139 Dewa T, Sugiura R, Suemori Y, Sugimoto M, Takeuchi T, Hiro A, Iida K, Gardiner AT, Cogdell RJ and Nango M (2006) Lateral organization of a membrane protein in a supported binary lipid domain: Direct observation of the organization of bacterial lightharvesting complex 2 by total internal reflection fluorescence microscopy. Langmuir 22: 5412–5418 Escalante M, Maury P, Bruinink C, van der Werf K, Olsen JD, Timney JA, Hunter CN, Huskens J, Subramaniam V and Otto C (2008) Directed assembly of functional light harvesting antenna complexes onto chemically patterned surfaces. Nanotechnology 19: 25101 Fang JY, Gaul DF, Chumanov G and Cotton TM (1995) Characterization of photosynthetic reaction centers from Rhodobacter sphaeroides at the air-water interface and in Langmuir-Blodgett films. Langmuir 11: 4366–4370 Fotiadis D, Qian P, Philippsen A, Bullough PA, Engel A and Hunter CN (2004) Structural analysis of the reaction center light-harvesting complex I photosynthetic core complex of Rhodospirillum rubrum using atomic force microscopy. J Biol Chem 279: 2063–2068 Iida K, Kashiwada A and Nango M (2000a) Construction of Langmuir-Blodgett films from light-harvesting complex I isolated from photosynthetic bacteria. Colloids Surf A 169: 199–208 Iida K, Ohya N, Kashiwada A, Mimuro M and Nango M (2000b) Characterization of the light-harvesting polypeptide/bacteriochlorophyll a complex isolated from photosynthetic bacteria by the linear dichroism spectra. Bull Chem Soc Jpn 73: 221–229 Iida K, Kiriyama H, Fukai A, Konings WN and Nango M (2001)
Two-dimensional self-organization of the light-harvesting polypeptides/BChl a complex into a thermostable liposomal membrane. Langmuir 17: 2821–2827 Iida K, Inagaki J, Shinohara K, Suemori Y, Ogawa M, Dewa T and Nango M (2005) Near-IR absorption and fluorescence spectra and AFM observation of the light-harvesting 1 complex on a mica substrate refolded from the subunit light-harvesting 1 complexes of photosynthetic bacteria Rhodospirillum rubrum Langmuir 21: 3069–3075 Imahori H, Yamada H, Nishimura Y, Yamazaki and Sakata Y (2000) Vectorial multistep electron transfer at the gold electrodes modified with self-assembled monolayers of ferrocene-porphyrinfullerene triads J Phys Chem B 104: 2099–2108 Jamieson SJ, Wang P, Qian P, Kirkland JY, Conroy MJ, Hunter CN and Bullough PA, (2002) Projection structure of the photosynthetic reaction centre-antenna complex of Rhodospirillum rubrum at 8.5 Å resolution. EMBO J 21: 3927–3935 Karrasch S, Bullough P and Ghosh R (1995) The 8.5 Å projection map of the light- harvesting complex 1 from Rhodospirillum rubrum reveals a ring composed of 16 subunits. EMBO J 14: 631–638 Ke B (2001) Photosynthesis, Photobiochemistry and Photobiophysics (Advances in Photosynthesis and Respiration, Vol 10). Kluwer Academic Publishers, Dordrecht Lee I, Lee JW and Greenbaum E (1997) Biomolecular electronics: Vectorial arrays of photosynthetic reaction centers. Phys Rev Lett 79: 3294–3297 Martin AS, Sambles JR and Ashwell G (1993) Molecular rectification: Dipole reversal in a cationic donor–(p-bridge)–acceptor dye. Phys Rev Lett 70: 218–221 Matsumoto K, Nomura K, Tohnai Y, Fujioka S, Wada M and Erabi T (1999) Immobilization of photosynthetic reaction center complexes onto a hydroquinonethiol-modified gold electrode. Bull Chem Soc Jpn 72: 2169–2175 McDermott G, Prince SM, Freer AA, Hawthornthwaite-Lawless AM, Papiz MZ, Cogdell RJ and Isaacs NM (1995) Crystal structure of an integral membrane light-harvesting complex from photosynthetic bacteria. Nature 374: 517–521 Mikayama T, Iida K, Suemori Y, Miyashita T and Nango M (2006) Electron transfer mediated by photosynthetic reaction center proteins between two chemical-modified metal electrodes. Mol Cryst Liq Cryst 445: 291–296 Miller JF, Hinchigeri SB, Parkes-Loach PS, Callahan PM, Sprinkle JR, Riccobono JR and Loach PA (1987) Isolation and characterization of a subunit form of the light-harvesting complex of Rhodospirillum rubrum. Biochemistry 26: 5055–5062 Nagata M, Yoshimura Y, Inagaki J, Suemori Y, Iida K, Ohtsuka T and Nango M (2003) Construction and photocurrent of lightharvesting polypeptides/zinc bacteriochlorophyll a complex in lipid bilayers. Chem Lett 32: 852–853 Noël H, van der Rest M, Gingras G (1972) Isolation and partial characterization of P870 reaction center complex from wild type Rhodospirillum rubrum. Biochim Biophys Acta 275: 219–230 Ogawa M, Kanda R, Dewa T, Iida K and Nango M (2002) Molecular assembly of light-harvesting antenna complex on ITO electrode. Chem Lett 31: 466–467 Ogawa M, Shinohara K, Nakamura Y, Suemori Y, Nagata M, Iida K, Gardiner AT, Cogdell RJ and Nango M (2004) Self-assembled monolayer of light-harvesting 1 and reaction center (LH1-RC) complexes isolated from Rhodospirillum rubrum on an amino-
Chapter 43
Assembly of Antenna Complexes on Solid Substrates
terminated ITO electrode. Chem Lett 33: 772–773 Okamura MY, Steiner LA and Feher G (1974) Characterization of reaction centers from photosynthetic bacteria. I. Subunit structure of the protein mediating the primary photochemistry in Rhodopseudomonas spheroides R-26. Biochemistry 13: 1394–1403 Page CC, Chen X, Moser CC and Dutton PL (1999) Natural engineering principles of electron tunnelling in biological oxidation-reduction. Nature 402: 47–52 Picorel R, Belanger G and Gingras G (1983) Antenna holochrome B880 of Rhodospirillum rubrum S1. Pigment, phospholipid, and polypeptide composition. Biochemistry 22: 2491–2497. Qian P, Addlesee HA, Ruban AV, Wang P, Bullough PA and Hunter CN (2003) A reaction center-light-harvesting 1 complex (RCLH1) from a Rhodospirillum rubrum mutant with altered esterifying pigments: characterization by optical spectroscopy and cryo-electron microscopy. J Biol Chem 278: 23678–23685 Qian P, Hunter CN and Bullough PA (2005) The 8.5 Å projection structure of the core RC-LH1-PufX dimer of Rhodobacter sphaeroides J Mol Biol 349: 948–960 Reynolds N, Janusz S, Escalante-Marun M, Timney J, Ducker RE,Olsen JD, Otto C, Subramaniam V, Leggett GJ and Hunter CN (2007) Directed formation of micro- and nanoscale patterns of functional light harvesting LH2 complexes. J Am Chem Soc 129: 14625–14632 Roszak AW, TD Howard, Southall J, Gardiner A, Law CJ, Isaacs NW and Cogdell RJ (2003) Crystal structure of the RC-LH1 core complex from Rhodopseudomonas palustris. Science 302: 1969–1972 Sagara T, Niwa K, Sone A, Hinnen C and Niki K (1990) Redox reaction mechanism of cytochrome c at modified gold electrodes. Langmuir 6: 254–262 Scheuring S, Reiss-Husson F, Engel A, Rigaud J-L and Ranck J-L (2001) High-resolution AFM topographs of Rubrivivax gelatinosus light-harvesting complex LH2. EMBO J 20: 3029–3035 Scheuring S, Rigaud JL and Sturgis JN (2004a) Variable LH2 stoichiometry and core clustering in native membranes of Rhodospirillum photometricum. EMBO J 23: 4127–4133 Scheuring S, Sturgis JN, Prima V, Bernadac A, Lévy D and Rigaud JL (2004b) Nanodissection and high-resolution imaging of the Rhodopseudomonas viridis photosynthetic core complex in native membranes by AFM. Proc Natl Acad Sci USA 101: 11293–11297 Stamouli A, Kafi S, Klein DCG, Oosterkamp TH, Frenken JWM, Cogdell RJ and Aartsma TJ (2003) The ring structure and organization of light harvesting 2 complexes in a reconstituted
875
lipid bilayer, resolved by atomic force microscopy. Biophys J 84: 2483–2491 Stamouli A, Frenken JWN, Oosterkamp TH, Cogdell RJ, and Aartsma TJ (2004) The electron conduction of photosynthetic protein complexes embedded in a membrane. FEBS Letters 560: 109–114 Suemori Y, Nagata M, Nakamura Y, Nakagawa K, Okuda A, Inagaki J, Shinohara K, Ogawa M, Iida K, Dewa T, Yamashita K, Gardiner A, Cogdell RJ and Nango M (2006) Self-assembled monolayer of light-harvesting core complexes of photosynthetic bacteria on an amino-terminated ITO electrode. Photosynthesis Res 90: 17–21 Suemori Y, Fujii K, Ogawa M, Nakamura Y, Shinohara K, Nakagawa K, Nagata M, Iida K, Dewa T, Yamashita K and Nango M (2007) Molecular assembly of artificial photosynthetic antenna core complex on an amino-terminated ITO electrode. Colloids and Surfaces B: Biointerfaces 56: 182–187 Sumi H (1997) Phonon-progressional V-I characteristics of STM processes mediated by redox molecular adsorbates. Chem Phys 222: 269–280 Taniguchi I, Ishimoto H, Miyagawa K, Iwai M, Nagai H, Hanazono H, Taira K, Kubo A, Nishikawa A, Nishiyama K, Dursun Z, Hareau GPJ and Tazaki M (2003) Surface functions of 2-mercaptopyridine, 2-mercaptopyrazine and 2-mercaptoquinoxaline modified Au(111) electrodes for direct rapid electron transfer of cytochrome c. Electrochem Commun 5: 857 Van Oijen A, Ketelaars M, Köhler J, Aartsma TJ and Schmidt, J. (1999) Unraveling the electronic structure of individual photosynthetic pigment-protein complexes, Science 285: 400–402 Visschers RW, Chang MC, van Mourik F, Parkes-Loach PS, Heller BA, Loach PA and van Grondelle R. (1991) Fluorescence polarization and low-temperature absorption spectroscopy of a subunit form of light-harvesting complex I from purple photosynthetic bacteria. Biochemistry 30: 5734–5742. Wakao N, Yokoi N, Isoyama N, Hiraishi A, Shimada K, Kobayashi M, Kise H, Iwaki M, Itoh S, Takaichi S and Sakurai Y (1996) Discovery of natural photosynthesis using Zn-containing bacteriochlorophyll in an aerobic bacterium Acidiphilium rubrum. Plant Cell Physiol 37: 889–893 Yasuda Y, Hirata Y, Sugino H, Kumei M, Hara M, Miyake J and Fujihira M (1992) Langmuir-Blodgett films of reaction centers of Rhodopseudomonas viridis: Photoelectric characteristics. Thin Solid Films 210/211: 733–735 Zhao J, Davis JJ, Sansom MSP and Hung A (2004) Exploring the electronic and mechanical properties of protein using conducting atomic force microscopy. J Am Chem Soc 126: 5601–5609
Chapter 44 Optical Spectroscopy of Individual Light-Harvesting Complexes from Purple Bacteria Jürgen Köhler* Experimental Physics IV and Bayreuth Institute for Macromolecular Research (BIMF), University of Bayreuth, 95440 Bayreuth, Germany
Summary ............................................................................................................................................................... 877 I. Introduction..................................................................................................................................................... 877 II. The Experimental Setup ................................................................................................................................. 879 III. Energy Transfer, Excitons, Strong and Weak Coupling ................................................................................. 880 IV. Spectroscopy of Single Light-Harvesting 2 Antenna Complexes ................................................................... 882 A. Overview .......................................................................................................................................... 882 B. The B800 Band ............................................................................................................................... 882 1. Intra- and Intercomplex Heterogeneity .................................................................................... 882 2. Chromophore-Protein Interactions in the B800 Assembly ...................................................... 884 C. The B850 Band ................................................................................................................................ 885 V. Energy Transfer in a Single Photosynthetic Unit ............................................................................................ 889 Acknowledgements ............................................................................................................................................... 891 References ............................................................................................................................................................ 891
Summary The primary reactions of purple bacterial photosynthesis take place within two well characterized pigmentprotein complexes, the core reaction center-light-harvesting 1 (RC-LH1) complex and the more peripheral light-harvesting 2 (LH2) complexes. These antenna complexes serve to absorb incident solar radiation and to transfer it to the RCs, where it is used to ‘power’ the photosynthetic redox reaction. This review describes how the use of single-molecule spectroscopy has contributed to a more detailed understanding of the molecular mechanisms involved in the energy transfer processes.
I. Introduction Since the beginnings of single-molecule spectroscopy (Moerner and Kador, 1989; Orrit et al., 1990) this technique has undergone breathtaking progress by combining two established fields: labeling biomolecules with fluorescent markers and fluorescence microscopy. These developments allowed the extension of singlemolecule experiments to ambient conditions initiating, in turn, a tremendous rush in the life sciences (Perkins et al., 1994; Funatsu et al., 1995; Nishizaka et al., 1995; Schnitzer and Block, 1997; Lu et al., 1998; Weiss, *Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 877–894. © 2009 Springer Science + Business Media B.V.
878 1999; Harms et al., 2001; van Oijen et al., 2003; Kinosita et al., 2004; Lee et al., 2006). The intriguing feature of this technique is that it enables one to elucidate information that is commonly washed out by ensemble averaging. Besides the possibility of circumventing spatial inhomogeneities, it allows for the observation of dynamical processes which are usually obscured by the lack of synchronization within an ensemble. A single molecule that undergoes a time-dependent development between different states is at any time in a distinct, well defined state and the whole sequence of steps can be studied. In particular, this allows for the identification of shortlived intermediate states that might be essential for understanding the processes under study but which would be completely masked otherwise. Usually, in optical single-molecule experiments, the molecule is detected via its laser-induced fluorescence and the prerequisites that have to be fulfilled in order to detect the signal from an individual molecule have been detailed in several reviews on this subject (Moerner and Basché, 1993; Basché et al., 1997; Plakhotnik et al., 1997; Köhler, 1999; Moerner and Orrit, 1999; Rigler et al., 2001; Moerner and Fromm, 2003; Zander et al., 2003; Kulzer and Orrit, 2004; Hofmann et al., 2007). Single-molecule techniques have been employed to investigate pigment-protein complexes from purple bacteria for nearly ten years (Bopp, 1997). In most purple bacteria, the photosynthetic unit (PSU) present in the membrane contains two types of antenna complexes, the LH1 complex, which is closely associated with the RC forming the core RC-LH1 complex, and the peripheral LH2 complex. Depending on the growth conditions some species feature another peripheral light-harvesting complex, LH3, which is a spectroscopic variant of LH2. In recent years highresolution structures of most of the building blocks of the PSU have become available (Chapter 8, Gabrielsen et al.). Remarkably, all peripheral light-harvesting complexes form circular oligomers and the basic building block is a protein heterodimer (αβ), which binds three bacteriochlorophyll (BChl) a pigments and one carotenoid molecule (McDermott et al., 1995; Koepke et al., 1996; McLuskey et al., 2001; Papiz et Abbreviations: BChl – bacteriochlorophyll; (EM)CCD – (electron multiplying) charge coupled device; FWHM – full width half maximum; LH – light-harvesting; PSB – phonon side band; PSU – photosynthetic unit; Rba. – Rhodobacter; RC – reaction center; Rps. – Rhodopseudomonas; Rsp. – Rhodospirillum; ZPL – zero phonon line
Jürgen Köhler al, 2003; Chapter 8, Gabrielsen et al.). The total LH2 complex consists either of nine (Rhodopseudomonas (Rps.) acidophila, Rhodobacter (Rba.) sphaeroides) or eight (Rhodospirillum (Rsp.) molischianum) such αβ-polypeptide heterodimers where the chromophores are arranged in two pigment pools labeled B800 and B850, according to their room-temperature absorption maxima in the near infrared. The B800 assembly comprises nine (eight for Rsp. molischianum) well-separated BChl a molecules which have their bacteriochlorin planes aligned nearly perpendicularly to the symmetry axis whereas the B850 assembly comprises eighteen (sixteen for Rsp. molischianum) BChl a molecules in close contact oriented with the planes of the molecules parallel to the symmetry axis, resembling the blades of a turbine. In contrast to LH2, the structures of RC-LH1 complexes show a strong dependence on the species of the purple bacterium (Karrasch et al., 1995; Roszak et al., 2003; Siebert et al., 2004; Qian et al., 2005; Chapter 9, Bullough et al.). More details about the structures and the mutual arrangement of the antenna complexes are presented elsewhere in this book. The first single-molecule studies on pigmentprotein complexes from purple bacteria were conducted under ambient conditions. Fluctuations of the emission intensity as well as fluctuations of the polarization state of the emitted light were observed (Bopp et al., 1997, 1999). However, under ambient conditions photobleaching of the probe molecules usually limits the observation time to some tens of seconds. Since the main cause of photobleaching is photochemical reactions in the electronically excited state in the presence of oxygen, these processes play a negligible role at cryogenic temperatures simply due to the lack of (mobile) oxygen. Low-temperature experiments have revealed valuable information about the character of the electronically excited states of the antenna complexes (van Oijen et al., 1998, 1999a,b, 2000; Tietz et al., 1999, 2000; Ketelaars et al., 2001, 2002; Matsushita et al., 2001; Gerken et al., 2003a,b; de Ruijter et al., 2004; Köhler and Aartsma, 2006; Hofmann et al., 2004a). The robustness of the light-harvesting process in purple bacteria has been demonstrated by the observation of the excitation energy transfer within a single self-aggregated photosynthetic unit in a non-membrane environment under cryogenic conditions (Hofmann et al., 2003a). More recent studies have uncovered details of the electronic coupling between the BChl a chromophores in LH2 and the electron-phonon coupling between these chro-
Chapter 44
Spectroscopy of Single LH-Complexes
mophores and the protein backbone (Hofmann et al., 2003b, 2005b). Changing the point of view and considering the weakly coupled B800 BChl a molecules as local probes to monitor their local environments led to insights about the organization of the energy landscape within the binding pocket (Hofmann et al., 2003c, 2004b). Reviews of single-molecule work on bacterial light-harvesting complexes can be found in Wrachtrup et al. (2003), Köhler and Aartsma (2006) and Cogdell et al. (2006). Often scientists from the Life Science community worry about the relevance of low-temperature spectroscopy on biomolecules. However, processes that involve only the movements of electrons are as active at cryogenic temperatures as at room temperature because electrons are very light particles. Hence, cryogenic conditions provide a very reasonable approximation of the electronic structure of biomolecules, i.e., of the spatial distribution of the electronic wavefunctions and the relaxation pathways between different states. One of the most important aspects of low-temperature spectroscopy is the improved spectral resolution resulting from a tremendous decrease of the linewidth of the optical transitions. This provides a sensitive tool to map fluctuations in the vicinity of the probe molecule via the associated spectral fluctuations. At room temperature these variations would be completely masked because in this temperature regime the much broader homogeneous linewidth of the absorptions by far exceeds the spectral changes. Clearly, at low temperatures the observed rates of the fluctuations do not reflect the natural dynamics. But given the longer relaxation times for all other processes, including conformational rearrangements, it shifts the timescale of these fluctuations into a regime that is experimentally accessible. Therefore, in the opinion of the author, low-temperature studies form a reliable basis for the discussion of the electronic properties of biomolecules under ambient conditions. Here we will focus on some examples from our own work to illustrate the value of low-temperature single molecule experiments on antenna complexes from purple bacteria. II. The Experimental Setup For our experiments we use a versatile home-built microscope that can be operated in confocal or wide-field mode under liquid helium conditions, as
879
shown schematically in Fig.1a. (A very similar setup operating in the visible rather than the near infrared spectral range has been described in detail in Lang et al., 2006). The sample is held in a cryogenic insert which also accommodates the microscope objective of numerical aperture (NA = 0.85). The insert is based on a mechanical construction that has been described previously (van der Meer et al., 1995) and allows in situ adjustments of the objective and the sample with respect to each other along two mutually orthogonal directions (focus and sample position) with submicrometer precision. In confocal mode the excitation light is focused to a spot size of 470 nm (full width half maximum, FWHM @ λ = 800 nm) whereas in wide-field mode an additional lens is introduced in the excitation path to defocus the light to an area on the sample of about 50 × 50 µm2. Switching between the two modes is accomplished with the aid of some flip mounts for moving some of the optical elements into and out of the optical path. Usually the selection of an individual pigment-protein complex for spectroscopy is done in two steps. First, a wide-field image of the sample is recorded. Therefore the excitation wavelength is tuned to the maximum of the absorption band and, after passing suitable bandpass filters to suppress residual laser light, the red-shifted fluorescence is detected by a charge coupled device (CCD) camera. The 3-D representation of a fluorescence image for LH2 from Rsp. molischianum is shown in Fig.1b where each peak corresponds to a diffraction-limited image of an individual LH2 complex, providing information about the spatial location of the individual complexes. Second, a spatially well-isolated complex is selected from the wide-field image and the microscope is switched to the confocal mode such that the excitation volume exactly coincides with the spatial position of the selected complex. The fluorescence collected by the same optics as before is directed onto a confocally placed avalanche photodiode. In this mode an excitation volume of about 1 µm3 is achieved and superior background suppression allows the recording of fluorescence-excitation spectra with high signal-to-background ratios, see Fig. 1c. Alternatively the fluorescence can be detected in wide-field mode by an electron-multiplying chargecoupled device camera (EMCCD, DV 465, Andor). Fluorescence-excitation spectra from individual LH2 complexes are obtained by scanning the laser while acquiring about 300 frames on the EMCCD, Fig. 2a. Consequently, the frame number corresponds to the
880
Jürgen Köhler
Fig.1. a) Schematic sketch of the arrangement of the low-temperature microscope. b) Three-dimensional representation of a fluorescence image of about 40 × 40 µm2. Each peak corresponds to the diffraction-limited image of an individual LH2 complex. c) Fluorescenceexcitation spectrum of an individual LH2 complex from Rsp. molischianum. Adapted from Hofmann et al. (2004b).
excitation wavelength and by integrating the total intensity of the fluorescence image of an individual complex on the CCD as a function of the read-out frame number a fluorescence-excitation spectrum can be obtained, Fig.2b. This allows for registering simultaneously the fluorescence-excitation spectra of typically 20–40 complexes imaged on the EMCCD. The spectral bandwidth of the excitation laser is 0.07 nm (1 cm–1). However, the nominal spectral resolution is determined by the mutual relationship of the scan speed of the laser and the read-out time of the EMCCD. This amounts to 0.1 nm of spectral resolution which restricts this acquisition scheme to spectral features that are sufficiently broad. With respect to the confocal operation mode, the parallel detection scheme accelerates the experimental work tremendously.
III. Energy Transfer, Excitons, Strong and Weak Coupling It has been established that the spatial structure of photosynthetic complexes, especially the mutual orientation of the pigments, determines to a large extent their spectroscopic features and excited-state dynamics (van Amerongen et al., 2000). A good starting point for the description of the electronically excited states of a molecular aggregate is provided by the formalism of Frenkel-excitons (Frenkel, 1931a,b; Knox, 1963; Davydov, 1971). As a first approximation each molecule can be described as a two-level system consisting of a ground state and a single electronically excited state separated by excitation energy E0. A state |n〉 represents molecule n in the electronically excited state and all other molecules 1,2, ..., n – 1,
Chapter 44
Spectroscopy of Single LH-Complexes
881
such as a linear chain or a ring one finds 1
k =
Fig.2. a) Sequence of EMCCD images of a sample region of 100 × 100 µm2. Each frame has been obtained with an illumination time of 500 ms and a read-out time of 120 ms. The bright dots correspond to diffraction-limited images of individual LH2 complexes. During the acquisition of the sequence the wavelength of the excitation is scanned. The straight line serves as a guide for the eye for the emission of a particular LH2 complex as imaged on the CCD camera. b) Fluorescence-excitation spectrum of a single LH2 complex, reconstructed from the CCD- images shown in a) and converted to wavelengths. (no. stands for number) Adapted from Hofmann et al. (2004a).
n + 1, ..., N in the ground state. The states |1〉 to |N〉 feature the same excitation energy, E0, which is localized on an individual molecule. The energy transfer originates from the electronic interaction V between the pigments and the electronically excited states and can be described by the Hamiltonian N
H = ∑ E0 n n + n=1
1 N ∑ ∑V n m 2 n=1 m≠ n nm
(1)
where Vnm denotes the matrix element 〈n|V|m〉 of the interaction between molecules in excited states located on molecule ‘n’ and ‘m.’ The eigenfunctions of this Hamiltonian are given by linear superpositions of the localized wavefunctions, so-called Frenkel excitons and for a one-dimensional regular arrangement
N
N
∑e
i 2 πk
n N
n
k = 0, ..., N–1
(2)
n=1
For this ideal system where all molecules are equivalent, the exciton wavefunctions have equal amplitudes on each pigment and the excitation is fully delocalized over all N pigments. The initial degeneracy of the localized excited states is lifted by the interaction and results in a manifold of energy levels — the exciton band. However, various deviations from this situation tend to localize the excitation energy on a smaller part of the aggregate. For example, local variations in the transition energies of the individual molecules represented by a random shift δn with respect to the average transition energy E0 reflect slight differences in their local environment. Alternatively, any structural deviation from perfect symmetry results in a variation of the mutual interactions, represented by ∆Vnm . These factors can be combined into the following Hamiltonian (Fidder et al., 1991): N
N
H = ∑ ( E0 + δ n ) n n + ∑ ∑ (Vnm + ∆Vnm ) n m n =1
n =1 m ≠ n
(3) Since the disorder in the energies δn affects the diagonal elements of the Hamiltonian this is commonly referred to as diagonal disorder. For the same reason the variations in the interaction strength are referred to as off-diagonal disorder. The resulting energy levels of the total system are determined by the strength of the coupling and the site-energies of the system. Commonly, one distinguishes two limiting cases. In the regime of weak coupling, |V⁄ δn| « 1, the interaction between the transition dipoles is much smaller than the difference in site energies of the pigments. Here the description of the excitations in terms of the localized states |n〉 is a good approximation and the transfer of energy between the pigments is visualized as a diffusive hopping process (incoherent energy transfer). In the other extreme, the strong coupling limit, |V⁄ δn| » 1, the interaction is much larger than the difference in site energies of the pigments. In this case Frenkel-exciton states |k〉 provide a good starting point for the description of the excited states and the transfer of excitation energy occurs in a wavelike
Jürgen Köhler
882 manner (coherent energy transfer). If the interaction strength and the site-energy differences are similar in magnitude, the energy transfer is expected to be intermediate between the extremes of incoherent (hopping) and coherent (wavelike) energy transfer. A detailed discussion of the exciton delocalization in the context of pigment protein complexes from purple bacteria can be found in Dahlbom et al. (2001). IV. Spectroscopy of Single LightHarvesting 2 Antenna Complexes A. Overview The fluorescence-excitation spectra of several individual LH2 complexes are shown in Fig.3 (van Oijen et al., 1999b; Ketelaars et al., 2001). The upper trace shows, for comparison, the fluorescence-excitation spectrum taken from a bulk sample (dashed line) together with the spectrum that results from the summation of the spectra of nineteen individual LH2 complexes (solid line). The two spectra are in excellent agreement and both feature two broad structureless bands around 800 and 860 nm corresponding to the absorptions of the B800 and B850 pigments of the complex. By measuring the fluorescence-excitation spectra of the individual complexes, remarkable features become visible which are obscured in the ensemble average. In particular, a striking difference between the B800 and B850 bands becomes evident: the spectra around 800 nm show a distribution of narrow absorption bands, whereas in the B850 spectral region 2–3 broad bands are present. The differences between the two absorption bands can be understood by considering the ratio V/δn between the intermolecular interaction strength V and the spread in transition energies for the BChl a molecules in the two ring assemblies. From the crystal structure the magnitude of the dipolar coupling strength between the pigments of the B800 ring can be estimated to be about 24 cm–1 (Sauer et al., 1996). The intermolecular interactions between the B850 BChl a molecules are more difficult to calculate because the closest distance between the pigments is rather small compared to the size of the molecules. Several theoretical approaches have been employed (Sauer et al. 1996; Alden et al., 1997; Krueger et al., 1998; Scholes et al., 1999; for a summary, see Cogdell et al., 2006) and these have resulted in values for the strength of the pigmentpigment interaction between 250 cm–1 and 400 cm–1.
Fig. 3. Fluorescence-excitation spectra of individual LH2 complexes of Rps. acidophila. The top traces show the comparison between an ensemble spectrum (dashed line) and the sum of spectra recorded from nineteen individual complexes (solid line). The lower traces display spectra from single LH2 complexes. Each individual spectrum has been averaged over all possible excitation polarizations. All spectra were measured at 1.2 K at 20 W/cm2 with LH2 dissolved in a PVA-buffer solution. Adapted from van Oijen et al. (1999b), Ketelaars et al. (2001).
A crude measure for the magnitude of the diagonal disorder δn can be obtained from the inhomogeneous linewidths of the B800 and B850 absorption bands suggesting that V/δn differs by about an order of magnitude between the two ring assemblies. This puts forward the idea that for the electronically excited states of the B850 BChl a molecules collective excitations (Frenkel excitons) play an important role (strong coupling limit); whereas the excitations of the B800 BChl a molecules can be described in first approximation as being localized on an individual BChl a molecule (weak coupling limit). B. The B800 Band 1. Intra- and Intercomplex Heterogeneity Comparison of the B800 fluorescence-excitation spectra of several individual LH2 complexes (Fig. 4) reveals significant variations in the spectral distribu-
Chapter 44
Spectroscopy of Single LH-Complexes
883
∑ I (i) ⋅ ν(i) ν= ∑ I (i) i
(4)
i
where I(i) denotes the fluorescence intensity at datapoint i, ν(i) the spectral position corresponding to datapoint i, and the sum runs over all datapoints of the spectrum and uses the distribution of spectral means as a measure for the intercomplex heterogeneity. The intracomplex heterogeneity, or diagonal disorder, was obtained by calculating the standard deviations σν of the intensity distributions in the individual spectra (5)
σ ν = [ν 2 − ν 2 ]1/2
where ν 2 is given by
ν2
∑ I (i) ⋅ ⎡⎣ ν(i) ⎤⎦ = ∑ I (i) i
2
(6)
i
Fig.4. Comparison of fluorescence-excitation spectra in the B800 spectral region for LH2 from Rps. acidophila. The upper trace shows for comparison an ensemble spectrum. The lower traces show spectra from three individual LH2 complexes averaged over all possible excitation polarizations. All spectra were measured at 1.2 K at 10 W/cm2 with LH2 dissolved in a PVA-buffer solution. Adapted from van Oijen et al. (2000).
tion of the narrow absorption bands. These bands are spread over the whole spectral region covered by the inhomogeneously broadened ensemble spectrum which directly reflects the differences in the site energies of the individual B800 BChl a molecules. Consequently, single-molecule spectroscopy offers the unique opportunity to measure the diagonal disorder (intracomplex disorder) of the B800 BChl a molecules directly and, by comparison of several complexes, to evaluate statistical information about the variation of the transition energies between different complexes (intercomplex disorder). (van Oijen et al., 2000). These authors introduced the spectral mean, ν–, of the fluorescence-excitation spectrum of a single LH2 complex:
A histogram of ν–, obtained from the fluorescenceexcitation spectra of 46 complexes, is shown in Fig. 5a and has a width of ~120 cm–1. For the intracomplex distribution, Fig.5b, one finds σν centered at a value of ~55 cm–1 which corresponds to a full width at half maximum (FWHM) of 130 cm–1 for the distribution of site energies. Clearly, an ensemble spectrum (i.e., the conventional absorption spectrum) reflects the convolution of both contributions to the overall heterogeneity. From these data a total inhomogeneous linewidth Γinhom of about 180cm–1 is calculated for the B800 band, which is in good agreement with results from the bulk spectra of LH2 of Rps. acidophila taken at 1.2 K (van Oijen et al., 2000). Obviously, the approximation of taking the inhomogeneous linewidth of the ensemble as a measure for the diagonal disorder is an oversimplification. The key parameter for the electronic coupling is the magnitude of the intermolecular interaction strength V with respect to the energy mismatch δ between adjacent B800 BChl a molecules rather than its relative size with respect to the width of the distribution of site energies ∆Eintra, as shown schematically in Fig. 5c. From the histogram for the intracomplex disorder (Fig.5b) the width of a Gaussian distribution of site energies can be estimated to be about 120 cm–1 which yields
884
Jürgen Köhler
Fig.5. a) Distribution of the spectral mean for 46 LH2 complexes featuring the amount of intercomplex heterogeneity. b) Distribution of standard deviations for the spread of absorption lines in the individual fluorescence-excitation spectra for the same 46 LH2 complexes corresponding to the intracomplex heterogeneity. Adapted from van Oijen et al. (2000). c) Schematic representation of the variations in site energy of the B800 BChl a molecules that play a role in the description of the electronically excited states. For more details see text.
V/δ ≈ 0.2 which falls clearly into the regime of weak to intermediate coupling. In a study that employed the mutual orientation of the B800 transition-dipole moments to map out the electronic coupling between the B800 chromophores of individual LH2 complexes from Rsp. molischianum, Hofmann et al. (2003b) found that for this assembly the ratio V/δ varies between 0.2–2.5. These numbers suggest a slight delocalization of the B800 excited states over 2–3 neighboring BChl a molecules with a concomitant redistribution of oscillator strength. These findings are corroborated by a theoretical study that appeared recently (Cheng and Silbey, 2006). It is worth noting that if a slight delocalization in the B800 excitations is effective the histogram in Fig. 5b is influenced by the
delocalization as well and does not reflect exclusively the diagonal disorder. 2. Chromophore-Protein Interactions in the B800 Assembly Typically, the spectrum of the first electronically excited state of an organic molecule embedded in a matrix gives rise to a homogeneously broadened zerophonon line (ZPL) accompanied by a relatively broad phonon side band (PSB) (Rebane, 1970). The ZPL results from the pure electronic transition, whereas the PSB corresponds to an electronic transition in combination with the simultaneous excitation of a vibration of the host due to linear electron-phonon
Chapter 44
Spectroscopy of Single LH-Complexes
coupling. In order to distinguish the intramolecular vibrations of the probe molecules from those of the host in which they are embedded, the latter are commonly referred to as phonons. The distribution of the intensity between the ZPL and the PSB, i.e. the profile of the electronic spectrum, is determined by the electron-phonon coupling strength, and provides information about the strength of the interaction of the probe molecule with its local surroundings, here between a B800 Bchl a molecule and its protein environment. The experimental problem that arises when attempting to measure such a profile of the electronic spectrum of an individual B800 absorption is that the spectra are subjected to temporal fluctuations (spectral diffusion) due to structural variations in the local environment of the B800 pigments; this results in temporal averaging of the signal and the subtle details of the PSB are, therefore, washed out. In order to extract the information about the electron-phonon coupling from the B800 fluorescence-excitation spectra, Hofmann et al. (2005) recorded thousands of B800 spectra from the same LH2 complex in rapid succession and employed a multivariate statistical analysis pattern recognition approach. Such algorithms have been used for the comparison of amino acid sequences of proteins from different species or for the reconstruction of the three dimensional structure of large biological macromolecules from two-dimensional projections obtained by single particle cryo-electron microscopy (van Heel et al., 2000). The basic idea of this approach is to consider an n-dimensional vector space where n corresponds to the number of data points in a spectrum. Then, each individual spectrum can be represented by a point in this vector space (the components are given by the intensity of the respective data points) and all consecutively recorded spectra form a data cloud with pronounced elongations along distinct directions. Now, the first task is to find these main directions within the data set (i.e., a eigenvector/eigenvalue or principal components analysis) and to describe the raw spectra of the individual sweeps as linear combinations of the 50–100 most significant eigenvectors — in this case ‘eigenspectra’ — of the set. It is worth noting that the eigenvectors do not need to have any resemblance with the real spectra as they simply form an orthogonal basis for the data space. This reduction of dimensionality decreases the amount of data significantly thereby facilitating its interpretation. The next step of the multivariate statistical analysis is the pattern recognition part,
885
i.e., finding similar looking spectra. The goal of this classification is to group close-lying elements into compact classes. Such a so-called optimal partition can always be defined for a given number of classes; it is the partition in which the interclass variance (between classes) is maximal and the total intraclass variance (within classes) is minimal (in a mathematical least-square sense). For this, an automatic hierarchical classification algorithm was used in which similar spectra (classes) are merged to form larger classes until a predetermined number of classes has been reached. In simple words, the classes can be envisaged as sub-clouds of points in the n-dimensional vector space, whereby the data points within each sub-cloud should be as close together as possible, while the distance between the sub-clouds should be as large as possible. For more details about the algorithm the reader is referred to van Heel et al. (1989, 2000) and Borland and van Heel (1990). The statistical analysis of such a stack of B800 spectra from an individual LH2 complex from Rsp. molischianum indeed uncovered a narrow peak accompanied by a weak broad shoulder on its highenergy side as a recurrent motif for the individual B800 absorptions. Examples are shown in Fig.6. In order to analyze the spectral profile in more detail the low-energy wing of the ZPL has been fitted by a Lorentzian that was subtracted from the data to obtain the PSB as shown in the insets of Fig. 6. The values obtained for the linewidth of the ZPL cover the range between 3 –10 cm–1 (FWHM) and are in agreement with the results of other experiments (de Caro et al., 1994; Wu et al., 1996). The different PSBs show clear variations with respect to integrated intensity, shape, width, and center frequency. Within the Born-Oppenheimer approximation the electronphonon coupling determines the Debye-Waller factor, α. From these data Hofmann et al. (2005) obtained values for the Debye-Waller factor between 0.4 and 0.9 reflecting a rather weak electron-phonon coupling. Their findings are summarized in Table 1 together with the data obtained for the average values from ensemble experiments (Small, 1995; Wu et al., 1996), which are in agreement with the averages based on the single-molecule data. C. The B850 Band The closest distance between the B850 pigments is less than 1 nm, i.e., rather small with respect to the
Jürgen Köhler
886
Fig.6. Recurrent motifs in the spectra from individual B800 absorptions after multivariate statistical analysis. For better comparison the peak positions of the sharp spectral features have been set arbitrarily to zero and the vertical scale has been normalized. The insets show an expanded view of the broad shoulders on the high-energy side of the sharp peak after subtraction of a Lorentzian that has been fitted to the low-energy wing of the sharp line. The obtained Debye-Waller factor α is given in each panel. Adapted from Hofmann et al. (2005b).
Table 1. Electron-phonon coupling in the B800 band. ΓZPL denotes the linewidth (FWHM) of a Lorentzian curve fitted to the ZPL, ΓPSB and ωm denote the linewidth (FWHM) and the center frequency of the PSB, respectively, and α is the Debye-Waller factor. For each parameter the range of values observed (upper row) as well as its average value and standard deviation (lower row) is given. The last two rows refer to ensemble averaged data from the literature. ΓZPL (cm–1) (Hofmann et al., 2005b) 4.2–9.6 5.6±1.4 (Wu et al., 1996) (Small, 1995)
ΓPSB (cm–1) 25–62
ωm (cm–1) 26–56
α 0.43–0.90
44±12
34±8
0.64±0.15
20–30
0.74
20
0.6
5 30–40
size of the molecules. Therefore a strong excitonic coupling between the molecules has to be considered to understand the optical spectra. In order to calculate the intermolecular interaction strength, V, several theoretical approaches have been employed (Sauer et al., 1996; Alden et al., 1997; Krueger et al., 1998; Scholes et al., 1999) resulting in values of V ≈ 250 cm–1 – 400 cm–1 which is about an order of magnitude larger than the interaction strength in the B800 ring. These figures suggest that the electronic excitations in the B850 are strongly delocalized among the pigments. The details of the exciton manifold of the B850 assembly of LH2 has been the subject of numerous studies (Sauer et al., 1996;
Alden et al., 1997; Krueger et al., 1998; Scholes et al., 1999; Mostovoy and Knoester, 2000; Dahlbom et al., 2001; Jang et al., 2001; Matsushita et al., 2001; Sumi, 2001; Didraga and Knoester, 2002; Hu et al., 2002). The exciton states are separated in lower (as) and upper (s) branches. This reflects the interactions within a dimer and is commonly known as Davydov splitting (Robinson, 1970; Davydov, 1971). Each branch consists of eight pairwise degenerate states (for LH2 from Rps. acidophila), labeled by kj = ±1, ±2, ±3, ±4, (j = as, s) and one nondegenerate state kj = 0. As a result of the strong interaction within the B850 ring the initial degeneracy of the excited states of the 18 B850 Bchl a molecules is lifted and the
Chapter 44
Spectroscopy of Single LH-Complexes
energy of the 18 exciton eigenstates is spread over a band with a width of about 1000 cm–1. Both the magnitude and the mutual arrangement of the individual transition-dipole moments have a crucial influence on the resulting selection rules for the optical transitions. Due to the circular arrangement of the pigments in LH2 only the exciton states kj = 0, ±1 have a non-vanishing transition-dipole moment. The kj = 0 transitions can be excited with light polarized parallel to the C9-symmetry axis of the ring whereas transitions to the kj = ±1 states can be excited with light of mutually orthogonal polarization within the plane of the ring. Since the individual transition-dipole moments of the B850 pigments are oriented mainly in the plane of the ring rather little of the total oscillator strength is associated with the kj = 0 states. Due to the head-to-tail arrangement of the transition-dipole moments within an individual dimer nearly all the oscillator strength is concentrated in the lower exciton manifold (j = as) which results in a strong electronic transition from the kas = ±1 states, seen in vivo as the strong near infrared (IR) absorption band at approximately 860 nm. The upper exciton components ks = ±1 carry less than 3% of the total oscillator strength and give rise to very weak absorptions up to about 790 nm (Sauer et al., 1996). Figure 7a shows an example of a two-dimensional representation of a stack of 50 fluorescence-excitation spectra from an individual LH2 complex where the horizontal axis corresponds to the read-out frame number of the EMCCD, i.e., to the excitation wavelength, the vertical axis to the polarization of the excitation, and the grayscale to the intensity. The spectrum that results when the whole sequence of 50 scans is averaged is shown in Fig. 7b. It features two maxima at about 856 nm and 863 nm. The maximum intensity of these peaks can be excited at mutually orthogonal polarization for the incident radiation as depicted in Fig. 7c. Apparently, the transition-dipole moments of these two transitions are perpendicular with respect to each other. Based on this observation and the fact that these bands are by far the most intense ones in the absorption spectra, these transitions have been attributed to the kas = ± 1 exciton states (van Oijen et al., 1999b; Hofmann et al. 2004a). The observed line widths of the bands reflect the ultrafast relaxation (~ 100 fs) to the kas = 0 exciton state, consistent with time-resolved data (Vulto et al., 1999). However, for the circular arrangement of the pigments, the exciton model predicts that the kas = ± 1 states are degenerate
887
Fig.7. a) Two-dimensional representation of a stack of 50 consecutively recorded fluorescence-excitation spectra. The horizontal axis corresponds to EMCCD read-out frame number or equivalently to the wavelength of the excitation. Each spectrum has been extracted for an individual LH2 complex from a sequence of EMCCD images. Between two spectra the polarization of the incident laser light has been rotated by 18°. The vertical axis corresponds to polarization and the intensity is given by the grey scale. b) Fluorescence-excitation spectrum that results from averaging the whole stack of 50 individual spectra. c) Average of 15 scans for mutually orthogonal polarization of the excitation. (no. stands for number). From Hofmann et al. (2004a).
and have the same intensity. Random diagonal disorder, caused by stochastic variations in the protein environment, introduces the term δn into Eq. 3. The main effects on the exciton manifold are a mixing of the different exciton levels, a modification of the energy separation of the exciton levels and lifting their pairwise degeneracy, and a redistribution of oscillator strength to nearby states, including the kas = 0 state. The random off-diagonal disorder described by the term ∆Vnm in Eq. 3, is caused by the variations in the coupling between the Bchl a pigments and originates from fluctuations in the orientations and positions of the individual transition-dipole moments. The influence of these various types of disorder on the B850
888 exciton states has been analyzed intensively. Since the effects of random off-diagonal disorder on the exciton dynamics are experimentally indistinguishable from those of random diagonal disorder (Fidder et al., 1991) many studies have focused on the latter type of disorder (Sauer et al., 1996; Alden et al., 1997; Jang et al., 2001). More recently, studies have also included combinations of random diagonal and correlated off-diagonal disorder (Dempster et al., 2001) as well as combinations of random on- and off-diagonal disorder and correlated on- and off-diagonal disorder on the B850 exciton states (Jang et al., 2001). Predictions based on these models differ with respect to subtle details in the spectra such as the extent and distribution of the energy separation between the kas = ±1 states and the intensity ratio of the respective transitions. In order to discriminate between the different models Hofmann et al. (2004a) applied widefield fluorescence-excitation spectroscopy in combination with a highly sensitive electron-multiplying charge coupled device (EMCCD) camera, which led to a
Jürgen Köhler massive parallelization of the experimental scheme. From these experimental data the energetic splitting of the kas = ±1 states ∆Eblue,red, the ratio of the integrated intensities Iblue/Ired, and the relative orientation of the two transition-dipole moments ∆αblue,red, were obtained, where blue (red) refers to the energetically higher (lower) absorption band(s). The distributions of these parameters are shown in the histograms in Fig. 8 together with numerical simulations that will be discussed below (van Oijen et al., 1999b; Hofmann et al., 2004a). As can be seen from Fig. 8 these parameters vary from complex to complex. The energetic separation ∆Eblue,red of the kas = ±1 states is centered at 126 cm–1 and has a width of 101 cm–1 (FWHM). The center value of the integrated intensity ratio Iblue/Ired is 0.73 (FWHM: 0.54), and the mutual orientation of the transition-dipole moments ∆αblue,red is 91° (FWHM: 19°). The experimental data have been compared with results from numerical simulations for three different models: A) random diagonal disorder (energetic disorder), B) random diagonal disorder together with correlated off-diagonal disorder (structural disorder),
Fig.8. Comparison of the experimental distributions (histograms) for the energetic separation ∆Eblue,red (top), the intensity ratio Iblue/Ired (center), and the relative orientation of the transition dipole moments ∆αblue,red (bottom) of the two broad transitions in the B850 spectra with numerical simulations (dots) for a) random diagonal disorder, b) random diagonal and correlated off-diagonal disorder, and c) random and correlated diagonal disorder. The experimental data refer to the left vertical scale and the simulations refer to the right vertical scale. From Hofmann et al. (2004a).
Chapter 44
Spectroscopy of Single LH-Complexes
and C) random- and correlated diagonal disorder. It has been found that introducing only random diagonal disorder, model A), could not explain the observations even if higher exciton states were included in the calculation (Fig. 8a). Instead, a correlated off-diagonal disorder of C2 symmetry, model B), was introduced as a regular modulation of the interaction strength in the Hamiltonian to explain the observed large splitting of the kas = ±1 states (van Oijen et al., 1999a,b; Ketelaars et al., 2001; Matsushita et al., 2001). The dominant effect of such a modulation is a coupling between exciton states that differ in their quantum numbers by ∆k = ±2. Since C2 is the lowest symmetry component of an ellipse the origin of this modulation could be visualized as an elliptical deformation of the complexes with a deformation amplitude δr/r0 = 7–8% where r0 is the radius of the unperturbed ring and the long and short axes of the ellipse deviate from r0 by ±δr. Such a structural deformation will lead to spatial displacements of the BChl a molecules and consequently to a modulation of the interaction strengths. Based on fluorescence-polarization experiments, other groups came to similar conclusions (Bopp et al., 1999; Tietz et al., 2000). The experiments by Bopp et al. (1999) indicate, moreover, that the structural deformation undergoes fluctuations on a timescale of seconds under ambient conditions. In combination with random diagonal disorder the model explained the splitting of the kas = ±1 states and the distribution of this parameter as well as the mutual orthogonal transition-dipole moments of the two kas = ±1 transitions, but it failed to explain the observed intensity ratio for these two transitions. However, it is reasonable to assume that a possible structural deformation would also affect the conformation of the protein residues in the binding pocket of the chromophores. Since it is already known for the B800 absorptions that tiny relative changes of distance in the binding pocket are sufficient to result in large spectral shifts (Zazubovich et al., 2002) it is obvious to ask whether a moderate modulation of the site energies would be compatible with the experimental data as well. If the modulation in site energies follows an imposed structural deformation of the LH2 complex a less pronounced deviation from circular symmetry (which might not be resolved by X-ray crystallography) could be sufficient to account for the observations. Indeed a model that takes into account random and correlated diagonal disorder, model C), is in good agreement with all measured experimental distributions, Fig. 8c (Hofmann et al., 2004a).
889
Finally it is worth noting that theoretical analysis of the data presented in van Oijen et al. (1999b) and which is consistent with the data presented above has shown that both inter- and intracomplex disorder of the BChl a site energies are effective in the B850 manifold as well (Mostovoy and Knoester, 2000). From ensemble experiments it is hard to distinguish between these two contributions of disorder, yet only the intracomplex type of disorder is of relevance for the delocalization of the excitation energy. V. Energy Transfer in a Single Photosynthetic Unit In Fig. 9a, a series of wide-field images from a sample of RC-LH1 complexes from Rba. sphaeroides is shown as a function of the excitation wavelength. Each bright dot corresponds to the diffraction-limited image of an individual core complex. In order to obtain these images the excitation wavelength has been switched between 800 nm, 835 nm and 870 nm while for all images the detection wavelength has been fixed to 917 nm, i.e., the maximum of the LH1 emission. Exciting the sample at 870 nm the wide-field fluorescence image (Fig. 9a, right), shows several individual RC-LH1 complexes from Rba. sphaeroides. After changing the excitation to 800 or to 835 nm (Fig. 9a, left and center, respectively), only one of the features detected previously is still able to emit light at 917 nm. In order to investigate this finding in more detail we have recorded the fluorescence-excitation spectrum of this particular feature (Fig. 9b). It consists of several narrow absorptions at 800 nm and two broad bands at about 830 nm and 870 nm which feature further substructure. Our interpretation of this finding is that the fluorescence-excitation spectrum, Fig. 9b, corresponds to the superposition of the respective spectra from an individual LH2 and an individual RC-LH1 complex. Hence we assign the spectral features around 800 and 835 nm to absorptions from LH2 and the band at 870 nm to an absorption of RC-LH1 and interpret this observation as excitation energy transfer within a single PSU. In order to substantiate this conjecture we recorded the relative fluorescence intensity as a function of the polarization of the emission for excitation wavelengths at 800 nm, 835 nm and 870 nm as indicated by the arrows in part b) of Fig. 9. Within experimental accuracy no difference in the orientation of the emitting transition-dipole moment is observable, Fig. 9c; this provides strong evidence
890
Jürgen Köhler
Fig.9. Fluorescence-imaging and fluorescence-excitation spectroscopy of a single photosynthetic unit from Rba. sphaeroides. The fluorescence was detected at 917 nm. a) Wide-field fluorescence images (20 × 25 µm2) from LH1-RC dissolved in a polymer film at 1.4 K. From left to right the excitation wavelengths were 800, 835 and 870 nm respectively at an intensity of 250 W/cm2. b) Fluorescenceexcitation spectrum of the feature encircled in a). The signal is given in counts per second (cps), the excitation intensity was 20 W/cm2. c) Intensity of the LH1-RC fluorescence as a function of the polarization of the emission for the feature encircled in a). For better comparison the three traces have been normalized and correspond to excitation wavelengths of 800 nm (full line), 835 nm (dotted line) and 870 nm (dashed line) as indicated by the arrows in a). From Hofmann et al. (2003a).
that the emission at 917 nm always stems from the same electronic state irrespective of the excitation wavelength, which supports our assignment. The trace that corresponds to an excitation wavelength of 800 nm (full line) shows a significant decrease in intensity beyond a polarization angle of 270°. This can be easily understood if one considers that the pigment-protein complexes are susceptible to light-induced spectral fluctuations resulting in slight changes of the spectral positions of the individual absorptions during the experiment. This is of minor influence for the broad B850 and B870 spectral features but the relatively narrow absorption lines of the B800 band might get shifted out of resonance with the excitation laser during data acquisition. From these results we conclude that we indeed have observed the excitation-energy transfer in a single
photosynthetic unit, as sketched in Fig. 10. Generally one might have expected a stronger quenching of the LH1 emission by the RC. But especially at low temperatures the lowest energy level of a significant fraction of the LH1 complexes is thought to be shifted to the red with respect to the energy level of the primary donor (P) of the RC, thus enhancing fluorescence from LH1 because of reduced trapping efficiency (Valkunas et al., 1992; Somsen et al., 1994). Moreover, at cryogenic temperatures, light-excitation induces charge separation only between the primary electron donor P and the primary quinone acceptor QA; electron transfer to the secondary quinone acceptor QB is, in fact, blocked below 150 K (Kleinfeld et al., 1984; Xu and Gunner, 2001). As a consequence, once QA has been reduced following the first photon, no further photochemistry and concomitant quench-
Chapter 44
Spectroscopy of Single LH-Complexes
ing of the fluorescence can occur. From the relative intensity of the LH2 bands with respect to the LH1 band in Fig. 9b we have to conclude that the transfer of excitation energy is very efficient even in a non-membrane environment. Due to chromatic aberrations in our low-temperature microscope the effective excitation intensity (photon energy / time·area) is lower at 800 nm as compared with the intensity at 870 nm. As a consequence of this, the efficiency of the LH2-LH1 energy transfer is even underestimated by the above mentioned criterion. In summary, the data indicate clearly an energy transfer from the peripheral LH2 to RC-LH1 complex in a single supramolecular LH2-LH1-RC aggregate. The observation of narrow lines in the B800 region of the fluorescence excitation spectrum suggests that one or, at most two, LH2 molecules are attached to the RC-LH1 complex. A larger number of LH2 complexes would result in a significant change of the spectral shape of the B800 absorption due to ‘ensemble’ averaging. The detection of the fluorescence feature presented in Fig. 9, reflecting energy transfer within an individual PSU, was a relatively rare event that could be observed only for a few percent of the studied complexes. Since the concentration of LH2 complexes in our LH1-RC preparations is very low, i.e., below the detection limit of a conventional ensemble absorption spectrum (Francia et al., 1999), it is reasonable to argue that the LH2 present in the preparation is all bound to RC-LH1 complexes. This suggests that the interaction between the peripheral and core antenna complexes is strong enough to withstand some detergent exposure during the purification procedure. Previously, excitation-energy transfer from LH2 to RC-LH1 has been observed at a reduced rate for assemblies reconstituted into liposomes (Moskalenko et al., 1992). Our work shows the formation of functional PSUs even in a non-membrane environment. These findings are supported by further experimental results. For instance mixing of two independent solutions which contained either isolated LH2 or isolated RC-LH1 complexes resulted in PSU fluorescence-excitation spectra similar to those shown in Fig. 9b. These results show that the single-molecule approach allows the investigation of energy transfer and protein-protein interactions in spontaneously reconstituted PSUs.
891
Fig.10. Schematic sketch of the supramolecular arrangement of a photosynthetic unit adapted from Kühlbrandt (1995), Hofmann et al. (2003a). The arrows indicate the excitation and emission (wavy arrows) as well as the energy transfer (full arrows) pathways. The figures refer to the wavelengths of the absorption and emission maxima of the ensemble spectrum.
Acknowledgements I thank T. J. Aartsma, R. J. Cogdell, F. Francia, M. van Heel, H. Michel, D. Oesterhelt, and G. Venturoli for fruitful collaborations and A. van Oijen, M. Ketelaars, M. Matsushita and C. Hofmann who performed the described experimental work. Financial support from the Volkswagen Foundation is gratefully acknowledged. References Alden RG, Johnson E, Nagarajan V, Parson WW, Law CJ and Cogdell RJ (1997) Calculations of spectroscopic properties of the LH2 bacteriochlorophyll-protein antenna complex from Rhodopseudomonas acidophila. J Phys Chem B 101: 4667– 4680 Basché T, Moerner WE, Orrit M and Wild UP (1997) Single Molecule Optical Detection Imaging and Spectroscopy. VCH, Weinheim Bopp M, Jia Y, Li L, Cogdell RJ and Hochstrasser RM (1997) Fluorescence and photobleaching dynamics of single light harvesting complexes. Proc Natl Acad Sci USA 94: 10630–10635 Bopp M, Sytnik A, Howard TD, Cogdell RJ and Hochstrasser RM (1999) The dynamics of structural deformations of immobilized single light-harvesting complexes. Proc Natl Acad Sci USA 96: 11271–11276 Borland L and van Heel M (1990) Classification of image data in conjugate representation spaces. J Opt Soc Am A 7: 601–610 Cheng YC and Silbey RJ (2006) Coherence in the B800 ring of purple bacteria LH2. Phys Rev Lett 9602: 028103-1– 028103-4 Cogdell RJ, Gall A and Köhler J (2006) The architecture and function of the light-harvesting apparatus of purple bacteria: From single molecules to in vivo membranes. Quart Rev Biophys 39: 227–324
892 Dahlbom M, Pullerits T, Mukamel S and Sundström V (2001) Exciton delocalization in the B850 light-harvesting complex: Comparison of different measures. J Phys Chem B 105: 5515–5524 Davydov AS (1971) Theory of Molecular Excitons. Plenum, New York De Caro C, Visschers RW, van Grondelle R and Völker S (1994) Inter- and intraband energy transfer in LH2-antenna complexes of purple bacteria. A Fluorescence line-narrowing and holeburning study. J Phys Chem 98: 10584–10590 Dempster SE, Seogijoo J and Silbey RJ (2001) Single molecule spectroscopy of disordered circular aggregates: A perturbation analysis. J Chem Phys 114: 10015–10023 de Ruijter WPF, Oellerich S, Segura J-M, Lawless AM, Papiz M and Aartsma TJ (2004) Observation of the energy-level structure of the low-light adapted B800 LH4 complex by single-molecule spectroscopy. Biophys J 87: 3413–3420 Didraga C and Knoester J (2002) Exchange narrowing in circular and cylindrical molecular aggregates: degenerate versus nondegenerate states. Chem Phys 275: 307–318 Fidder H, Knoester J and Wiersma DA (1991) Optical properties of disordered molecular aggregates: A numerical study. J Chem Phys 95: 7880–7890 Francia F, Wang J, Venturoli G, Melandri BA, Barz WP and Oesterhelt D (1999) The reaction center-LH1 antenna complex of Rhodobacter sphaeroides contains one PufX molecule which is involved in dimerization of this complex. Biochemistry 38: 6834–6845 Frenkel J (1931a) On the transformation of light into heat in solids I. The Physical Review 37: 17–44 Frenkel J (1931b) On the transformation of light into heat in solids II. The Physical Review 37: 1276–1294 Funatsu T, Harada Y, Tokunaga M, Saito K and Yanagida T (1995) Imaging of single fluorescent molecules and individual ATP turnovers by single myosin molecules in aqueous solution. Nature 374: 555–559 Gerken U, Jelezko F, Goetze B, Branschädel M, Tietz C, Ghosh R and Wrachtrup J (2003a) Membrane environment reduces the accessible conformational space available to an integral membrane protein. J Phys Chem B 107: 338–343 Gerken U, Lupo D, Tietz C, Wrachtrup J and Ghosh R (2003b) Circular symmetry of the light-harvesting 1 complex from Rhodospirillum rubrum is not perturbed by interaction with the reaction center. Biochemistry 42: 10354–10360 Harms GS, Cognet L, Lommerse PHM, Blab GA and Schmidt Th (2001) Autofluorescent proteins in single-molecule research: Applications to live cell imaging microscopy. Biophys J 80: 2396–2408 Hofmann C, Francia F, Venturoli G, Oesterhelt D and Köhler J (2003a) Energy transfer in a single self-aggregated photosynthetic unit. FEBS Lett 546: 345–348 Hofmann C, Ketelaars M, Matsushita M, Michel H, Aartsma TJ and Köhler J (2003b) Single-molecule study of the electronic couplings in a circular array of molecules: light-harvesting-2 complex from Rhodospirillum molischianum. Phys Rev Lett 90: 013001–013004 Hofmann C, Aartsma TJ, Michel H and Köhler J (2003c) Direct observation of tiers in the energy landscape of a chromoprotein: A single-molecule study. Proc Natl Acad Sci USA 100: 15534–15538 Hofmann C, Aartsma TJ and Köhler J (2004a) Energetic disorder
Jürgen Köhler and the B850-exciton states of individual light-harvesting 2 complexes from Rhodopseudomonas acidophila. Chem Phys Lett 395: 373–378 Hofmann C, Aartsma TJ, Michel H and Köhler J (2004b) Spectral dynamics in the B800 band of LH2 from Rhodospirillum molischianum: A single-molecule study. New J Phys 6: 1–15 Hofmann C, Michel H, van Heel M and Köhler J (2005) Multivariate analysis of single-molecule spectra: Surpassing spectral diffusion. Phys Rev Lett 9419: 195501–195504 Hofmann C, Kulzer F, Zondervan R, Köhler J and Orrit M (2007) Single biomolecules at cryogenic temperatures: From structure to dynamics. In: Rigler R and Vogel H (eds), Single Molecules and Nanotechnology. Springer, Berlin Hu X, Ritz T, Damjanovic A, Autenrieth F and Schulten K (2002) Photosynthetic apparatus of purple bacteria. Quart Rev Biophys 35: 1–62 Jang S, Dempster SE and Silbey RJ (2001) Characterization of the static disorder in the B850 band of LH2. J Phys Chem B 105: 6655–6665 Karrasch S, Bullough PA and Ghosh R (1995) The 8.5 Å projection map of the light-harvesting complex I from Rhodospirillum rubrum reveals a ring composed of 16 subunits. The EMBO Journal 14: 631–638 Ketelaars M, van Oijen AM, Matsushita M, Köhler J, Schmidt J and Aartsma TJ (2001) Spectroscopy on the B850 band of individual light-harvesting 2 complexes of Rhodopseudomonas acidophila I. Experiments and Monte Carlo simulations. Biophys J 80: 1591–1603 Ketelaars M, Hofmann C, Köhler J, Howard TD, Cogdell RJ, Schmidt J and Aartsma TJ (2002) Spectroscopy on individual light-harvesting 1 complexes of Rhodopseudomonas acidophila. Biophys J 83: 1701–1715 Kinosita Jr K, Adachi K and Itoh H (2004) Rotation of F1-ATPAse: How an ATP-driven molecular machine may work. Annu Rev Biophys Biomol Struct 33: 245–268 Kleinfeld D, Okamura MY and Feher G (1984) Electron-transfer kinetics in photosynthetic reaction centers cooled to cryogenic temperatures in the charge-separated state: Evidence for lightinduced structural changes. Biochemistry 23: 5780–5786 Knox RS (1963) The Electronic Structure. Theory of Excitons. Academic Press, New York Koepke JHX, Muenke C, Schulten K and Michel H (1996) The crystal structure of the light-harvesting complex II (B800-850) from Rhodospirillum molischianum. Structure 4: 581–597 Köhler J (1999) Magnetic resonance of a single molecular spin. Phys Rep 310: 261–339 Köhler J and Aartsma TJ (2006) Single molecule spectroscopy of pigment protein complexes from purple bacteria. In: Grimm B, Porra RJ, Rüdiger W and Scheer H (eds) Chlorophylls and Bacteriochlorophylls: Biochemistry, Biophysics, Functions and Applications (Advances in Photosynthesis and Respiration, Vol 25), pp 309–321. Springer, Dordrecht Krueger BP, Scholes GD and Fleming GR (1998) Calculation of couplings and energy-transfer pathways between the pigments of LH2 by the ab initio transition density cube method. J Phys Chem 102: 5378–5386 Kühlbrandt W (1995) Many wheels make light work. Nature 374: 497–498 Kulzer F and Orrit M (2004) Single-molecule optics. Annu Rev Phys Chem 55: 585–611 Lang E, Baier J and Köhler J (2006) Epifluorescence confocal
Chapter 44
Spectroscopy of Single LH-Complexes
and total internal reflection microscopy for single molecule experiments: A quantitative comparison. J Microscopy 222: 118–123 Lee JB, Hite RK, Hamdan SM, Xie XS, Richardson CC and van Oijen AM (2006) DNA Primase acts as a molecular brake in DNA replication. Nature 439: 621–624 Lu HP, Xun L and Xie S (1998) Single molecule enzymatic dynamics. Science 282: 1877–1882 Matsushita M, Ketelaars M, van Oijen AM, Koehler J, Aartsma TJ and Schmidt J (2001) Spectroscopy on the B850 band of individual light-harvesting 2 complexes of Rhodopseudomonas acidophila II. Exciton states of an elliptically deformed ring aggregate. Biophys J 80: 1604–1614 McDermott G, Prince SM, Freer AA, Hawthornthwaite-Lawless AM, Papiz AZ, Cogdell RJ and Isaacs NW (1995) Crystal structure of an integral membrane light-harvesting complex from photosynthetic bacteria. Nature 374: 517–521 McLuskey K, Prince SM, Cogdell RJ and Isaacs NW (2001) The crystallographic structure of the B800-820 LH3 light-harvesting complex from the purple bacteria Rhodopseudomonas acidophila strain 7050. Biochemistry 40: 8783–8789 Moerner WE and Basché Th (1993) Optical spectroscopy of single impurity molecules in solids. Angew Chem Int Ed Engl 32: 457–476 Moerner WE and Fromm DP (2003) Methods of single-molecule fluorescence spectroscopy and microscopy. Rev Sci Instr 74: 3597–3619 Moerner WE and Kador L (1989) Optical detection and spectroscopy of single molecules in a solid. Phys Rev Lett 62: 2535–2538 Moerner WE and Orrit M (1999) Illuminating single molecules in condensed matter. Science 283: 1670–1676 Mostovoy MW and Knoester J (2000) Statistics of optical spectra from single-ring aggregates and its application to LH2. J Phys Chem B 104: 12355–12364 Moskalenko AA, Toropygina O, Godik VI, Timpmann K and Freiberg A (1992) Investigation of spatial relationships and energy transfer between complexes B800-850 and B890-RC from Chromatium minutissimum reconstituted into liposomes. FEBS Lett 308: 133–136 Nishizaka T, Miyata H, Yoshikawa H, Ishiwata S and Kinosita K (1995) Unbinding force of a single motor molecule of muscle measured using optical tweezers. Nature 377: 251–254 Orrit M and Bernard J (1990) Single pentacene molecules detected by fluorescence excitation in a p-terphenyl crystal. Phys Rev Lett 65: 2716–2719 Papiz MZ, Prince SM, Howard T, Cogdell RJ and Isaacs NW (2003) The structure and thermal motion of the B800-850 LH2 complex at 2.0 Å resolution and 100 K: New structural features and functionally relevant motions. J Mol Biol 226: 1523–1538 Perkins TT, Quake SR, Smith DE and Chu S (1994) Relaxation of a single DNA molecule observed by optical microscopy. Science 264: 822–826 Plakhotnik T, Donley EA and Wild UP (1997) Single molecule spectroscopy. Ann Rev Phys Chem 48: 181–212 Qian P, Hunter CN and Bullough PA (2005) The 8.5Å projection structure of the core RC-LH1-PufX dimer of Rhodobacter sphaeroides. J Mol Biol 349: 948–960 Rebane KK (1970) Impurity Spectra in Solids. Plenum Press, New York
893
Rigler R, Orrit M and Basché Th (2001) Single Molecule Spectroscopy. Springer Berlin Robinson GW (1970) Electronic and vibrational excitons in molecular crystals. Annu Rev Phys Chem 21: 429–474 Roszak AW, Howard TD, Southall J, Gardiner AT, Law CJ, Isaacs NW and Cogdell RJ (2003) Crystal structure of the RC-LH1 core complex from Rhodopseudomonas palustris. Science 302: 1969–1972 Sauer K, Cogdell RJ, Prince SM, Freer A, Isaacs NW and Scheer H (1996) Structure-based calculations of the optical spectra of the LH2 bacteriochlorophyll-protein complex from Rhodopseudomonas acidophila. Photochem Photobio 64: 564–576 Schnitzer MJ and Block SM (1997) Kinesin hydrolyses one ATP per 8 nm step. Nature 388: 386–390 Scholes GD, Gould IR, Cogdell RJ and Fleming GR (1999) Ab initio molecular orbital calculations of electronic couplings in the LH2 bacterial light-harvesting complex of Rps. acidophila. J Phys Chem B 103: 2543–2553 Siebert CA, Qian P, Fotiadis D, Engel A, Hunter CN and Bullough PA (2004) Molecular architecture of photosynthetic membranes in Rhodobacter sphaeroides: The role of PufX. EMBO J 23: 690–700 Small GJ (1995) On the validity of the standard model for primary charge separation in the bacterial reaction center. Chem Phys 197: 239–257 Somsen OJ, van Mourik F, van Grondelle R and Valkunas L (1994) Energy migration and trapping in a spectrally and spatially inhomogeneous light-harvesting antenna. Biophys J 66: 1580–1596 Sumi H (2001) Bacterial photosynthesis begins with quantummechanical coherence. Chem Rec 1: 480–493 Tietz C, Cheklov O, Draebenstedt A, Schuster J and Wrachtrup J (1999) Spectroscopy on single light harvesting complexes at low temperature. J Phys Chem B 103: 6328–6333 Tietz C, Gerken U, Jelezko F and Wrachtrup J (2000) Polarization measurements on single pigment protein complexes. Single Molecules 1: 67–72 Valkunas L, van Mourik F and van Grondelle R (1992) On the role of spectral and spatial antenna inhomogeneity in the process of excitation energy trapping in photosynthesis. J Photochem Photobiol B 15: 159–170 van Amerongen H, Valkunas L and van Grondelle R (2000) Photosynthetic Excitons. World Scientific, Singapore van der Meer H, Disselhorst JAJM, Köhler J, Brouwer ACJ, Groenen EJJ and Schmidt J (1995) An insert for single molecule magnetic resonance spectroscopy in an external magnetic field. J Rev Sci Instrum 66: 4853–4856 van Heel M (1989) Classification of very large electron microscopical image data sets. Optik 82: 114–126 van Heel M, Gowen B, Matadeen R, Orlova EV, Finn R, Pape T, Kohen D, Stark H, Schmidt R, Schatz M and Patwardhan A (2000) Single-particle electron cryo-microscopy: Towards atomic resolution. Quart Rev Biophys 33: 307–369 van Oijen AM, Ketelaars M, Köhler J, Aartsma TJ and Schmidt J (1998) Spectroscopy of single light-harvesting complexes from purple photosynthetic bacteria at 1.2 K. J Phys Chem B 102: 9363–9366 van Oijen AM, Ketelaars M, Köhler J, Aartsma TJ and Schmidt J (1999a) Spectroscopy of individual LH2 complexes of Rhodopseudomonas acidophila: Localized excitations in the B800 band. Chem Phys 247: 53–60
894 van Oijen AM, Ketelaars M, Köhler J, Aartsma TJ and Schmidt J (1999b) Unraveling the electronic structure of individual photosynthetic pigment-protein complexes. Science 285: 400–402 van Oijen AM, Ketelaars M, Köhler J, Aartsma TJ and Schmidt J (2000) Spectroscopy of individual LH2 complexes of Rhodopseudomonas acidophila: Diagonal disorder, sample heterogeneity, spectral diffusion, and energy transfer in the B800 band. Biophys J 78: 1570–1577 van Oijen AM, Blainey PC, Crampton DJ, Richardson CC, Ellenberger T and Xie XS (2003) Single-molecule kinetics of λ exonuclease reveal base dependence and dynamic disorder. Science 301: 1235–1238 Vulto SIE, Kennis JTM, Streltsov AM, Amesz J and Aartsma TJ (1999) Energy relaxation within the B850 absorption band of the isolated light-harvesting complex LH2 from Rhodopseudomonas acidophila at low temperature. J Phys Chem B 103: 878–883 Weiss S (1999) Fluorescence spectroscopy of single biomolecules. Science 283: 1676–1683 Wrachtrup J, Aartsma TJ, Köhler J, Ketelaars M, van Oijen AM,
Jürgen Köhler Matsushita M, Schmidt J, Tietz C and Jelezko F (2003) Spectroscopy of individual photosynthetic pigment-protein complexes. In: Zander C, Enderlein J and Keller RA (eds) Single Molecule Detection in Solution. Methods and Application, pp 185–229. Wiley-VCH, Berlin Wu HM, Savikhin S, Reddy NRS, Jankowiak R, Cogdell RJ, Struve WS and Small GJ (1996) Femtosecond and hole-burning studies of B800’s excitation energy relaxation dynamics in the LH2 antenna complex of Rhodopseudomonas acidophila (Strain 10050). J Phys Chem 100: 12022–12033 Xu Q and Gunner MR (2001) Trapping conformational intermediate states in the reaction center protein from photosynthetic bacteria. Biochemistry 40: 3232–3241 Zander C, Enderlein J and Keller RA (eds) (2003) Single Molecule Detection in Solution. Methods and Application. Wiley-VCH, Berlin Zazubovich V, Matsuzaki S, Johnson TW, Hayes JM, Chitnis PR and Small GJ (2002) Red antenna states of photosystem I from cyanobacterium Synechococcus elongatus: A spectral hole burning study. Chem Phys 275: 47–59
Chapter 45 De novo Designed Bacteriochlorophyll-Binding Helix-Bundle Proteins Wolfgang Haehnel Institut für Biologie - Biochemie der Pflanzen, Universität, Schänzlestr. 1, D-79104 Freiburg, Germany
Dror Noy Plant Sciences Department, Weizmann Institute of Science, Rehovot 76100, Israel
Hugo Scheer* Department Biologie I – Botanik, Menzinger Str. 67, D-80638 München, Germany
Summary ............................................................................................................................................................... 895 I. Introduction .................................................................................................................................................... 896 II. Chlorophyll Structures and Interactions with Natural Proteins ....................................................................... 897 III. Challenges in Designing de novo Chlorophyll- and Bacteriochlorophyll-binding Proteins ............................ 899 IV. Modular Organized Chlorophyll Proteins Based on Branched Four-helix Bundle Proteins ........................... 901 A. Choice of Central Metal .................................................................................................................... 901 B. Protein Design and Modeling ........................................................................................................... 903 V. Incorporating Chlorophylls and Bacteriochlorophylls into Self-assembling Protein Maquettes ...................... 904 VI. From Water-soluble to Amphiphilic Chlorophyll- and Bacteriochlorophyll-protein Maquettes ........................ 905 Acknowledgments ................................................................................................................................................. 907 References ............................................................................................................................................................ 907
Summary The construction of small synthetic proteins that bind only one or a small number of (bacterio)chlorophylls ((B)Chls) is a powerful approach to understand the concepts and guidelines of protein-(B)Chl interactions and assembly; the systems can be extended, in a stepwise fashion, by adding other cofactors or by oligomerization. We review the unique aspects of the de novo design and construction of (B)Chl-binding proteins, and describe recent progress and challenges in designing new BChl-protein platforms for delineating general rules and guidelines of (B)Chl-protein assembly, structure, and function. The relevant aspects of chlorophyll (Chl) and BChl chemical structures are outlined, as well as the modes of interactions with natural proteins. Two distinct strategies are then described for designing de novo water-soluble (B)Chl-binding proteins. The first strategy is based on the covalent assembly of modular four-helix bundle proteins, the second follows the original non*Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 895–912. © 2009 Springer Science + Business Media B.V.
896
Wolfgang Haehnel, Dror Noy and Hugo Scheer
covalent heme-binding protein maquette design, which relies on self-assembly of amphiphilic helices that is primarily driven by the hydrophobic effect. Finally, we demonstrate the extension of the latter to designing transmembrane-like versions of (B)Chl-protein maquettes. I. Introduction Natural (bacterio)chlororophyll ((B)Chl) proteins are a diverse group of generally membrane bound photosynthetic complexes that are unusually rich in cofactors, most of which are Chls or BChls, in combination with carotenoids, lipids, and, in reaction centers (RCs), also hemes, quinones and metal ions. These cofactors can make up as much as 35% of the weight of the holoproteins, which results in a complex network of interactions among the various components. Uniquely, interactions of Chls with the protein, neighboring Chls, and other cofactors significantly contribute to the assembly, stability and function of these photosynthetic complexes. Many molecular details of these interactions are currently being revealed by combining an extensive database of biochemical and spectroscopic information with high resolution structural information from crystal structures of a variety of Chl and BChl proteins. Understanding the underlying concepts and guidelines of protein-Chl/BChl assembly still remains a challenge, however, because of the system’s complexity. One approach to critically test this understanding is the construction of small synthetic proteins that bind only one or a small number of (B)Chls and then extend the system, in a stepwise fashion, by adding other cofactors or by oligomerization. This approach was pioneered by the protein maquette concept introduced by the Dutton and DeGrado groups in 1994: simple and robust complexes of de novo designed proteins and cofactors are used as flexible, minimal working scaffolds to study a selected function abstracted from highly complex proteins (Robertson et al., 1994). This was complemented by the modular design of branched proteins applied by Haehnel’s group (Rau and Haehnel, 1996). Both concepts have since been applied to a variety of cofactors such as mono- and dinuclear Abbreviations: AP – amphiphilic; BChl – bacteriochlorophyll; BChlide – bacteriochlorophyllide; BPhe – bacteriopheophytin; BPheide – bacteriopheophorbide; CD – circular dichroism; Chl – chlorophyll; Chlide – chlorophyllide; HP – hydrophilic; LP – lipophilic; [M]-(B)Chlide – (B)Chlide in which the central metal has been replaced by metal ‘M’; MOP – modular organized protein; NIR – near infrared; Phe – pheophytin; RC – reaction center; UV – ultraviolet; Vis – visible
metal centers (Klemba et al., 1995; Dieckmann et al., 1997; Rau et al., 1998; Willner et al., 1999; De Jonge et al., 1999; Schnepf et al., 2001; Summa et al., 2002; Marsh and DeGrado, 2002; Maglio et al., 2003; Ghosh et al., 2005), flavins (Sharp et al., 1998b), iron-sulfur clusters (Gibney et al., 1996; Mulholland et al., 1998, 1999; Kennedy and Gibney, 2002; Nanda et al., 2005), ATP (Butterfield and Waters, 2003), quinones (Li et al., 2006), various forms of heme (Rau and Haehnel, 1998; Katz et al., 1998; Sharp et al., 1998a; Shifman et al., 1998, 2000; Fahnenschmidt et al., 1999; Willner et al., 1999; Fahnenschmidt et al., 2000; Gibney et al., 2000; Rau et al., 2000; Grosset et al., 2001; Huang et al., 2004; Ghirlanda et al., 2004), and non-natural porphyrin derivatives (Cochran et al., 2005). A comprehensive review of the maquette approach including details of design, structure and function of other protein maquettes, and their contribution to understanding the underlying properties of specific enzymes can be found in Koder and Dutton (2006); the orthogonal synthesis of branched, modular organized proteins (MOP) and their orientated attachment to solid phases is reviewed in Haehnel (2004). Here, we review the unique aspects of the de novo design and construction of Chl- and BChl-binding proteins, and describe recent progress and challenges in designing new BChl-protein platforms for delineating general rules and guidelines of (B)Chlprotein assembly, structure, and function. We start by presenting the relevant aspects of Chl and BChl chemical structures and the modes of interactions with natural proteins. We follow by describing two distinct strategies for designing de novo water-soluble (B)Chl-binding proteins. The first strategy is based on the covalent assembly of modular four-helix bundle proteins, whereas the second follows the original non-covalent heme-binding protein maquette design, which relies on self-assembly of amphiphilic helices that is primarily driven by the hydrophobic effect. Finally, we demonstrate the extension of the latter to designing transmembrane-like versions of (B)Chl protein maquettes.
Chapter 45
Synthetic Bacteriochlorophyll-Binding Proteins
897
Fig. 1. Three types of cyclic tetrapyrroles, porphyrin, chlorin and bacteriochlorin, exemplified by heme, Chl(ide) a and BChl(ide) a, respectively. Pigments are shown with two extra ligands to the central metal, one above (Lβ) and one below the macrocyle (Lα)
II. Chlorophyll Structures and Interactions with Natural Proteins Chls and BChls are cyclic tetrapyrroles carrying a characteristic isocyclic five-membered ring, which is biosynthetically derived from the C-13 propionic acid side-chain of the common precursor, protoporphyrin IX. They generally contain a central magnesium ion (Mg2+), and the C-17 propionic acid side chain is usually esterified with a C-15 (BChls c, d, e) or C-20 isoprenoid alcohol (Chls a, b, d; BChls a, b, g). Chemically, the Chls can be characterized by the degree of unsaturation of the macrocycle (Fig. 1). The fully unsaturated porphyrin system is present in the ctype chlorophyllides (Chlides)† of chromophyte algae and some prokaryotes; this is also the conjugation system of heme, the Fe-complex of protoporphyrin IX. The 17,18-dihydroporphyrin (= chlorin) system is present in the Chls a, b and d of oxygenic organisms, and also in the BChls c, d and e of green anoxygenic bacteria, the 7,8,17,18-tetrahydroporphyrin (= bacteriochlorin) system is present in the BChls a, b and g of anoxygenic bacteria. The three Chl types have distinct spectroscopic properties: The c-type Chlides absorb moderately in the 620 nm region and strongly † The ending ‘ide’ denotes an unesterified C-173 acid group; most c-type cholorophylls are thereby chlorophyllides.
in the 400–450 nm region, the chlorin-type Chls are characterized by two intense absorptions at the edges of the visible spectral region (Vis) at 400–470 and 640–700 nm. The bacteriochlorin-type Chls, on which this review is focused, have intense absorption bands (ε ~105 M–1cm–1) in the near ultraviolet (Bx, By) and the near infrared (Qy). The fourth band in the visible spectral region (Qx) has only moderate intensity (ε ~104 M–1cm–1), but is of considerable diagnostic importance as a reporter of the ligation state and charge density of the central metal (Evans and Katz, 1975; Hartwich et al., 1998; Noy et al., 1998, 2000; Yerushalmi et al., 2006). Chls and BChls are rather large and relatively hydrophobic cofactors. With a surface area of about 750 Å2, they have more than twice the surface of ATP, and even the macrocycle alone (~ 450 Å2) has a larger surface than the latter. The most polar feature is the central, coordinatively unsaturated Mg (Katz et al., 1978), followed by carbonyl groups (and a hydroxyl group in BChls c, d, e) at the periphery of the tetrapyrrole moiety, while the esterifying alcohols are hydrophobic throughout. Chls are therefore amphiphilic molecules, which readily aggregate in polar as well as in apolar environments (Katz et al., 1978, 1991; Scherz et al., 1991; Agostiano et al., 2002; Balaban et al., 2004; Kunieda et al., 2004). In non-polar solvents, coordination of the central Mg
898 with peripheral polar substituents, in particular the 131-C=O group (Scheer and Katz, 1975; Oba and Tamiaki, 1999), is the major intermolecular interaction. In hydrophilic and amphiphilic environments, these interactions are less significant and aggregation is more determined by π-π and hydrophobic interactions (Scheer et al., 1985). A wealth of information has been derived from X-ray structures of Chl proteins, which now cover members of several RCs, light-harvesting systems and a water-soluble Chl protein not involved in photosynthesis (Michel et al., 1986; Yeates et al., 1988; El-Kabbani et al., 1991; Tronrud and Matthews, 1993; Sturgis et al., 1995; Hofmann et al., 1996; Zouni et al., 2001; Jordan et al., 2001; Spiedel et al., 2002; Roszak et al., 2003; Gall et al., 2003; Ben-Shem et al., 2003; Biesiadka et al., 2004; Ferreira et al., 2004; Liu et al., 2004; Standfuss et al., 2005; Horigome et al., 2007). The best established interaction with the proteins is the coordination of polar protein side chains with the coordinatively unsaturated central Mg. In the structures there is always only a single extra ligand, so the Mg2+ is therefore five-coordinated (but see Fiedor, 2006). In approximately 50% of the natural binding sites this extra ligand is histidine; the remainder comprise a variety of ligands including glutamine, asparagine, cysteine (rarely) as well as backbone C=O groups, N-terminal formyl groups, and also water. This is different from the hemes, where the central Fe is often 6-coordinated (Fig. 1). Due to the asymmetric substitution of the macrocycle and the presence of one (Chls c) to five (BChl a) asymmetric centers at the periphery, the two faces of the macrocycle are inequivalent, and therefore ligation is diastereotopic; this aspect is addressed in Chapter 46 (Braun and Fiedor). The ligands L1 and L2 in Fig. 1 are shown in the β- and α-positions, respectively. More detailed information on ligation has been obtained by spectroscopic techniques. Ligation of the central metal can be monitored in all Chls by vibrational spectroscopy (Lutz and Mäntele, 1991; Robert, 1996). A particular advantage of BChls is the sensitivity of the QX-band to the ligation state (see above). In solution, the central Mg2+ of BChls is five- and six-fold coordinated in solution, so it has one or two extra ligands besides the four nitrogens of the tetrapyrrole. Six-fold coordination is frequently seen with nitrogen ligands such as imidazole; it was therefore surprising that all X-ray structures and, until recently, spectroscopic studies showed only five-coor-
Wolfgang Haehnel, Dror Noy and Hugo Scheer dinated Mg2+ in Chl proteins. Appreciable populations (≤30%) of six-fold coordinated complexes have been recognized recently, however, in BChl a-containing light-harvesting proteins from purple bacteria (Fiedor, 2006). Their function is presently unclear. Ligand dissociation is a mechanism of rapid internal conversion, and thereby energy dissipation (Musewald et al., 1999; Noy et al., 2000). This pathway is reduced, and therefore thermal losses minimized, in Chl proteins that have only a single, relatively stable ligand. Addition of a sixth ligand might then be involved in light-protection. All Chls contain a 131-C=O group and a 173-ester group; a 133-methyl ester group is present in all Chls but BChls c, d, e, and many Chls also have polar groups at C-31 and C-7. While principally all of these groups can interact with the protein, those of the isocyclic ring seem to be most crucial if judged from pigment exchange experiments (Scheer and Hartwich, 1995; Davis et al., 1996; Vavilin et al., 2003; Scheer, 2003, 2006). In particular, an important role has been established for the 131-C=O group in the assembly of bacterial light-harvesting complexes (Garcia-Martin et al., 2006; Chapter 46, Braun and Fiedor). Interactions of the large (~ 450 Å2) and largely hydrophobic macrocycle with the apoproteins are presently only poorly understood on a molecular basis. Mutations of histidines in bacterial RCs that are ligands to the central Mg2+ to large, non-polar amino acids lead, in vivo, to a replacement of the respective (B)Chl by the metal-free (bacterio)pheophytin ((B)Phe). Conversely, BChl is bound instead of BPhe, if a coordinating amino acid such as histidine is introduced at a position where it can bind to the central Mg (Coleman and Youvan, 1990; Frank et al., 1993; Chapter 16, Jones). Obviously, the binding-site has a specificity for the macrocyclic system, irrespective of the presence of the central Mg. Braun and coworkers have analyzed the environments of the nearly 100 Chls of Photosystem I (PS I) and found a preference for specific amino acid residues in certain ‘regions’ of the macrocycle (see Chapter 46, Braun and Fiedor). However, large-scale mutations were possible in bacterial light-harvesting complexes without losing the pigments. The situation is complicated by the fact that there are also contacts with neighboring cofactors, namely BChls and carotenoids. (B)Chl(B)Chl interactions have been suggested early on as a major driving force in the formation of (B)Chl proteins (Katz et al., 1978; Scherz and Parson, 1986).
Chapter 45
Synthetic Bacteriochlorophyll-Binding Proteins
Control of aggregation is probably a major function of the protein, mainly to avoid non-radiative losses of excitation energy. Distortions of the macrocycle introduce a red shift in absorption (Gudowska-Nowak et al., 1990). In isolated pigments, this is possible by steric hindrance of peripheral substituents (Woodward et al., 1960; Medforth et al., 1992; Gentemann et al., 1994; Senge et al., 1998, 2006; Senge, 2000); compare, for example, the 10 nm red shift in BChl c bearing a 20CH3 group with BChl d having an H instead (Senge and Smith, 1994). Distortions can also be introduced into isolated porphyrins by central metals that do not fit snugly into the central hole of the macrocycle (Buchler, 1975; Sparks et al., 1993, 1993). In situ, non-planarity can also be introduced by interactions with the proteins. This is seen in all highly-resolved structures of Chl-proteins; it is particularly pronounced in one of the BChl-B850 molecules of the purple bacterial LH2 where it probably contributes to the red shift of the site energy (Robert et al., 2003; Chapter 11, Robert). Even less is known about interactions of the esterifying alcohol with the protein, which exert considerable effects on the aggregation of Chls, especially in micellar systems (Scheer et al., 1985). This suggests a function of the alcohols in interactions with the environment including the apoprotein, other Chls, carotenoids and lipids. Most of the alcohols are sesqui- or diterpenes (Scheer, 2003) that are rather sticky due to the methyl groups protruding from the main chain. This appears to be important in positioning Chls, thereby ensuring the proper spacing and orientation among the pigments and other cofactors which govern energy and electron transfer in Chl-proteins. Such a function is supported by all available X-ray structures: the esterifying alcohols are generally remarkably well resolved and engaged in interactions, especially among each other and with the carotenoids (see references above). Certain conformations of the respective alcohols are even conserved in an evolutionary sense over long periods; for example, from the bacterial RC to the plant PS II-RC (Zinth, personal communication). There are also some remarkable examples of variations in the alcohols, e.g., in Rhodospirillum rubrum, where the alcohol in the RC is different from that of the antenna (Walter et al., 1979). It should also be noticed that hydrolysis of the alcohol is an early step in Chl degradation (Kräutler and Hörtensteiner,
899
2006), and that the enzyme chlorophyllase can carry out this reaction in integral Chl proteins (Schoch and Brown, 1986). Restriction of conformational motions is also a way of minimizing internal conversion. This may be another reason for the high structural definition and surprisingly few contacts of the alcohol side chain with membrane lipids: the alcohols are rarely at the surface of Chl proteins. It will be interesting in this respect to analyze in more detail the water-soluble Chl proteins (WSCP) that lack the photoprotective carotenoids (Noguchi et al., 1999; Schmidt et al., 2003; Reinbothe et al., 2004), and the Chlide c-containing light-harvesting complexes (MacPherson and Hiller, 2003). Unfortunately, there exists only a single X-ray structure for the former group (Horigome et al., 2007), and none for the latter. III. Challenges in Designing de novo Chlorophyll- and Bacteriochlorophyllbinding Proteins The structural properties of Chls and BChls described above largely define the problems of, and boundary conditions for, designing Chl- and BChl-binding proteins. Designing de novo a protein that will fold and self-assemble with organic molecules as large as (B)Chls is rather difficult because it requires a fine balance of hydrophobic and polar interactions between peptides, cofactors, and solvent with the organic cofactors unconstrained by any covalent link to the peptide backbone (Barker, 2003). Recently, Cochran et al. (2005) introduced the first completely computational de novo designed four-helix bundle protein that selectively and non-covalently binds non-biological porphyrin derivatives. This marked a significant breakthrough, but the general applicability of their computational algorithm is yet to be established. This notwithstanding, the successful design of heme-binding four-helix bundle proteins has demonstrated that significant progress can be made without resorting to explicit computational design algorithms that are both time consuming and resource intensive. So far, much progress has been made by applying simple guidelines through an iterative learning and application process. Water-soluble four-helix bundles are the simplest, best characterized, and most robust de novo designed protein templates. Their successful application for
Wolfgang Haehnel, Dror Noy and Hugo Scheer
900
i
Si Bj
Bj
c CGGG E E E E
Si
d L L L R
e W L L L
f K K K K
g L K L K
a H F H L
b E E E
iv f
iii
b
c
e
g a
ii
x4
d
v c
f
g a
d
a
e
e
c
e
a
b
d
f
g
b
b
d
a
d
f
g c
g
e
c
b f
Fig. 2. Left: Scheme for the synthesis of a combinatorial library of four-helix bundles on solid support: (i) Synthesis of a cyclic decapeptide with two pairs of Cys protected by Acm and S-But; (ii) stepwise construction of a spacer with an acid-cleavable linker onto the cellulose membrane; (iii) coupling of the template onto the derivatized membrane, sequential cleavage of the protecting groups (1. S-But, 2. Acm) and coupling of antiparallel helices Si and Bj, respectively, led to a library of immobilized four-helix bundles. Finally, metal bacteriopheophorbides (M-BPheide, symbolized by the white rectangle) were incorporated. Arrows indicate the N → C direction of the peptide chains. Right: The binary patterning approach for designing four-helix bundles. An alternating pattern of hydrophobic (light grey) and hydrophilic (dark grey) helix-forming residues (top) is mapped onto a seven-residue two-turn heptad repeat sequence of an idealized α-helix. (iv) In aqueous solutions, this sequence folds into a helix with a hydrophobic face and a hydrophilic face (middle). (v) Self-assembly into four helix bundles (bottom) is driven by the hydrophobic effect and forms a bundle with a hydrophobic core and hydrophilic interior. This bundle may be functionalized for heme, Chl, or BChl binding by proper placement of histidine residues (top, dark grey circles) within the hydrophobic core.
binding heme and other porphyrin derivatives makes them obvious candidates for binding (B)Chls. However, despite their close chemical relation to heme, native (B)Chl-proteins are significantly different from native and de novo designed heme-binding proteins. Most importantly, the former are primarily transmembrane proteins that are compatible with the water-insoluble (B)Chls. Hydrolysis of the longchain esterifying alcohol of Chls and BChls results in (bacterio)chlorophyllides ((B)Chlides) that are much more water-soluble and could be reconstituted into globular water-soluble heme-proteins such as myoglobin (Boxer et al., 1982; Pearlstein et al., 1982; Schlichter et al., 2001; Markovic et al., 2007) and hemoglobin (Moog et al., 1984). Another significant difference between Chls and heme is the coordination of the central metal atom: Fe in heme tends to form primarily six-fold coordination complexes whereas Mg and Zn in native (B)Chl proteins prefer five-fold coordination. The next sections describe two strategies for de-
signing Chl- and BChl-binding four-helix bundle proteins, shown schematically in Fig. 2. The first relies on covalent assembly by the orthogonal synthesis method introduced by Mutter and Vuilleumier (1989), where four helical segments are grafted onto cyclic oligopeptides, in order to obtain water-soluble four-helix bundles. Specific binding sites for (B)Chls and their metal-substituted analogs are built into the four-helix scaffold by a combination of rational design and combinatorial synthesis. The second strategy is inherently non-covalent and relies on analogs of the original water-soluble heme-binding protein maquettes, in which self-assembly is primarily driven by the hydrophobic effect. We demonstrate that this strategy can be extended for the design of transmembrane versions of (B)Chl-protein maquettes. These designs self-assemble in non-polar environments such as lipid membranes and detergent micelles thereby circumventing the marginal solubility and self-aggregation tendency of Chls and BChls in aqueous solutions. They are extremely valuable for
Chapter 45
Synthetic Bacteriochlorophyll-Binding Proteins
reproducing the fundamental functions of respiration and photosynthesis, charge separation and proton gradients, which requires a supporting membrane and its dielectric, ion, and solute impermeability. Other examples of trans-membrane (B)Chl maquettes are given in Chapter 10 (Loach and Parkes-Loach) and Chapter 46 (Braun and Fiedor). These represent a top-down design strategy aiming at miniaturizing and simplifying the native protein fold of small natural transmembrane (B)Chl-binding proteins such as the light harvesting complexes of purple photosynthetic bacteria. Other approaches to the synthetic (B)Chl proteins include a number of truncated or otherwise modified natural proteins (Paulsen and Kuttkat, 1993; Meadows et al., 1995; Chapter 10, Loach and Parkes-Loach; Chapter 46, Braun and Fiedor). In addition, repeat-peptides have been generated that form more or less stable complexes with (B)Chl in solution (Dudkowiak et al., 1998; Miyake et al., 1998; Dudkowiak et al., 1999; Chen et al., 2005; Nango, 2006; Paulsen, 2006) and or can be attached to surfaces (Chapter 43, Iida et al.). IV. Modular Organized Chlorophyll Proteins Based on Branched Four-helix Bundle Proteins The concept of modular design originates from work of Mutter and Vuilleumier (1989); it has been adapted for parallel screening for binding as shown in Fig. 2. Using an orthogonal design, four helices have been grafted in anti-parallel fashion on to a cyclic peptide, to which natural and synthetic cofactors including heme have been attached covalently or non-covalently (Rau and Haehnel, 1996; Rau et al., 2000, 2001). A particular advantage of the orthogonal synthesis is that relatively large proteins (≈12 kDa) can be constructed in a controlled, modular fashion from individual short (30 aa) helices and an orthogonally de-protectable cyclic peptide template. By exploiting the opposite face of the cyclic peptide template, it was furthermore possible to covalently attach the construct to solid matrices (Rau et al., 2000). Binding to a gold electrode resulted in transfer of electrons between the cofactors and the metal (Katz et al., 1998). More recently, metal centers have been created by the same concept (Schnepf et al., 2001). In a first attempt to design a totally synthetic Chl protein (Rau et al., 2001), the pigment was modified by removing the phytyl side chain which renders the
901
pigment very hydrophobic and unlikely to fit into a short four-helix bundle. Furthermore, it was bound covalently via C=O groups to the modular protein carrying a lysine with its ε-amino group modified by an aminoxy group near the position of the liganding histidine that is placed appropriately to bind the central Mg of the (B)Chl. The covalent binding allows for a tighter control of the attachment and facilitates a directed orientation of the pigment. This approach is somewhat unsatisfactory, however, because it requires suitably placed reactive C=O groups, the introduction of non-natural amino-acid side chains, and arbitrarily reduces the accessible conformational space. By a combination of computer–aided rational design and permutations of selected amino acids in the putative binding pocket, a total of 216 MOPs was designed. They were bound covalently to a solid cellulose support by a cleavable linker. This allowed a rapid screening by visual inspection (Fig. 3) and optical spectroscopy (absorption and fluorescence), as well as washing with a variety of solvents to quickly judge the binding strength, and repeated binding of different pigments after removal of the previously used one with an appropriate eluant. Methods have been devised for semi-automated spectral analysis of the protein spots on the cellulose membranes, for the representation of the data, and for correlating physical properties of the varied amino acids with binding strength and specificity. They allow for a simultaneous screening of MOP and pigments. They are expected, at the same time, to provide a very useful guide for another round of screening starting with the best complexes of the current set. For a more thorough characterization of individual MOPs, the covalent linker is cleavable by acid. This allows for mass spectrometry of spots of interest. Finally, MOPs of particular interest were synthesized without the linker and analyzed in solution by optical spectroscopy including circular dichroism, or the binding constants determined. A. Choice of Central Metal This set of MOPs was screened for binding with BChlides in which the central Mg was replaced by Fe, Ni, and Zn. The rationale for selecting Fe was to obtain information on the influence of the macrocyclic system on pigment binding. The binding of Fe-protoporphyrin IX (heme) has been studied already in considerable detail (Rabanal et al., 1996; Rau et al., 2000). As compared to the fully unsaturated porphy-
902
Wolfgang Haehnel, Dror Noy and Hugo Scheer
Fig. 3. Top:Helical net representation of shielding helices Si and binding helices Bj. The amino acids that were varied (Xk for Si and Zl for Bj) are emphasized in gray, the ligating histidine of the binding helices in black. The tables contain the amino acids of the individual helices at the varied positions. Long, solid arrows indicate the N → C direction of the peptide chains, dashed arrows the dipole moments of salt bridges. Bottom: photograph of a cellulose membrane carrying 38 modular proteins, after incubation with Ni-BPheide in buffer/ DMF and washing in buffer. Columns 2–7 correspond to identical binding helices Bj (j = 7 – 12), rows 1–6 to identical shielding helices Si (i = 1 – 6). The empty spots in columns 1 and 8 correspond to modular proteins that have been punched out for mass spectrometry; the two isolated spots in the rightmost column were not used for the analyses. See also Color Plate 16, Fig. 26.
rin macrocycle of protoheme, the BChl macrocyclic system is much less symmetric due to reduction at rings B and D, but also due to its substituents (Fig. 1). On the other hand, Fe-BChl is more rigid than heme due to the presence of the isocyclic ring E, much less hydrophilic due to the esterification at C-133, and more reactive on account of the β-ketoester system at ring E. Zn was chosen because it can replace the central Mg in almost any enzymatic reaction of (B)Chl, with the exception of the Mg chelatase (Masuda et al., 1999), and in all accessible (B)Chl binding sites without significant changes in reactivity and/or properties (Scheer and Hartwich, 1995; Scheer, 2003; Frigaard et al., 2006; Rüdiger, 2006).
Most emphasis was given to Ni because (i) it can also replace Mg in many (B)Chl binding sites (Fiedor et al., 2001; Scheer et al., 1985; Scheer and Hartwich, 1995; Lapouge et al., 1998), and (ii) it shows a rich ligation behavior including a coordination number nc = 4 (unlike Mg), which has been studied in some detail and can be readily accessed by optical spectroscopy (Noy et al., 2000; Snigula, 2004; Yerushalmi et al., 2006). Ni-tetrapyrroles have extremely short excited state lifetimes which are quite distinct from the very long ones of the Mg and Zn complexes (Musewald et al., 1999; Fiedor et al., 2001; Yerushalmi et al., 2006), but lack the complicated redox chemistry and complexation with oxygen typical for Fe.
Chapter 45
Synthetic Bacteriochlorophyll-Binding Proteins
B. Protein Design and Modeling Most of the modular proteins were modifications of a previously used symmetric backbone model designed for heme as tetrapyrrole cofactor (Rau et al., 2000). In the model, four helices are positioned almost parallel to each other with their centers forming a slightly distorted square with side lengths between 1.00 and 1.07 nm and diagonals of 1.47 nm. The proportions of the helices follow those of the mono heme-binding four-helix bundle of cytochrome b562 (Zhang et al., 1998). However, in favor of a stable binding pocket the cofactor has not been placed at the end but in the hydrophobic core of the four-helix bundle. The position of the metal center at the intersection of the diagonals was defined by placing two histidines in the middle of the binding helices. As in the previous design of a heme protein (Rau et al., 2000), nearby amino acids forming the putative binding pocket were chosen to leave optimum space for the cofactor. Modeling was originally performed with heme as cofactor, because it is the only tetrapyrrole for which a force field was available. The binding helices Bi (Fig. 2) allow for axial coordination of the central metal (M) by the histidines; the histidine planes are perpendicular to the macrocycle, and the M-NHis bond lengths range between of 0.208 and 0.203 nm. In order to allow a tight knob-into-hole packing of the residues between the helices (Harbury et al., 1993), the shielding helices Si have been shifted vertically by 0.27 nm relative to the binding helices. The propionate side chain(s) of the tetrapyrrole are oriented along the protein axis as in former models designed for heme binding, where an arginine was placed inside the four-helix bundles for charge neutralization. However, we have not used this arginine in the current model because the M-BPheides contain only one propionate side chain, which is supposed to interact with one of the lysine residues of the helices. Shielding (Sx) and binding helices (By) were selected by varying amino acid residues lining the hydrophobic putative binding pocket in the vicinity of the histidine binding to the central metal. The major criterion for the quality of a binding pocket was the tight packing of the residues, which was tested by letting the software (Insight II, Accelerys) fit water molecules into all voids within a 1 nm radius around the central metal. Substitution of alanine by the hydrophobic residues phenylalanine, leucine or valine was used to minimize the number of water molecules. After each change of a residue,
903
the model was energetically relaxed with the autorotamer function of Insight II. Sequences were considered suitable if ≤5 water molecules fitted in the 1 nm sphere around the central metal. An alternative approach, one that eventually proved superior, was based on the structure of the heme-binding bundle of helices 41, 43, 45 and 48 of cytochrome b (Cyt b) in the ubiquinol-cytochrome c oxidoreductase (chain C of PDB entry 1bcc; Zhang et al., 1998). The structure is tighter than our symmetric model, but also leaves a large gap in the protein, by which the tetrapyrrole is accessible to external water. The following changes were introduced: 1) since this portion of Cyt b is in a hydrophobic environment, all amino acids on the outside of the bundle were exchanged to form a hydrophilic exterior with salt bridges. 2) the unit contains four different helices, but only two different helices were used in the current synthetic scheme of modular proteins. A symmetrizing pre-selection was made, again by the criterion of best space filling, on the amino acids in the hydrophobic core, in order to obtain only two alternating pairs of helices. For yet another optimization, the tetrapyrrole was rotated by about 50° in the symmetrized bundle, by placing an arginine outside the hydrophobic core, such that the (charged) propionate group gained direct contact with external water. Tightly binding peptides were obtained using only the latter approaches; heme binding is clearly visible by inspection (columns 10–12 in Fig. 3), and a binding constant in the range of 10–7 M–1 has been determined for one of the peptides in potassium phosphate buffer (10 mM, pH = 7). The results of such an approach are summarized in the following. The respective complexes with the three differently metalated BChlides generally show similar binding patterns to the MOP, with an overall decrease in binding strength in the order Zn > Ni > Fe. This binding is mainly controlled by the binding helices, while the shielding helices provide only a modulating influence. The M-BPheide binding pattern is very different from that of heme; MOPs that bind heme strongly show only poor binding of the [M]-BChlides, and vice versa. Since heme and [Fe]BChlide contain the same central metal, namely Fe, this points to an important effect of the peripheral substituents on binding specificity. A large proportion of all three pigments is initially bound with a coordination number nc = 4 of the central metal, with no ligands other than the four nitrogens of the tetrapyrrole macrocycle. Ligation
904 states with nc > 4 were visible only as shoulders in the QX-region of the spectrum, in particular with Ni-BPheide. This is unlike the situation in natural BChl proteins, where nc = 4 is practically unknown for the coordinatively unsaturated central Mg, and the binding of (generally one) ‘extra’ ligand (nc = 5) is an important contribution to binding specificity and to pigment organization. Binding of Zn-, Fe- and, in particular, Ni-BPheide with nc = 4 to the MOPs is moderately strong, but there is currently no information whether it occurs at the designed site or is unspecific. The ligation situation is changed after washing the complexes with increasing amounts of dimethylfomamide. The proportion of complexes with nc = 5 or 6 is increased, pointing to an increasing contribution of the central metal to binding under more stringent conditions. This process at the same time decreases site heterogeneity, as judged by the concomitant line narrowing. Depending mainly on the binding helix Bj, complexes with [Ni]-BChl have been defined which after washing preferentially have the central metal in ligation states with nc = 4, 5 or 6. It was furthermore possible to define by their spectral properties two types of complexes with nc = 6, but with different ligands. Due to the particularly rich ligation behavior of Ni++ as central metal, the focus of the detailed analysis of the binding was on the Ni-complexes. In the symmetric MOP used, the ligation state with nc = 5 is of particular interest, because the symmetric design with two histidines positioned opposite to each other, is expected to invite a situation with nc = 6. While this is indeed observed in most cases where strong binding occurs (notably those with binding helix B11), even with this symmetric design there are three binding helices (B3, B5 and B10) which show significant contributions of complexes with the unusual nc = 5. Obviously, even in a (formally) symmetric situation, such an asymmetric ligation is possible. This indicates that the asymmetric cofactor may induce different packing of amino acid residues on the two sides of the macrocycle resulting in different ligation of the histidines. A dynamic equilibrium is also conceivable with the metal flipping through the tetrapyrrole ring. As shown for three selected MOPs with nc = 5 or 6 (type I and type II), the screening results can readily be scaled up to preparative work, and transferred to solution conditions. Ni-BPheide binds at least as well in solution as on the solid phase support, and the general spectral features are well reproduced. An
Wolfgang Haehnel, Dror Noy and Hugo Scheer upper limit of the dissociation constant of Ni-BPheide-B11S8, Kd ≤ 100 nM, indicates that this MOP is an excellent candidate for a second round of synthesis and screening, in order to increase binding strength and specificity even more. V. Incorporating Chlorophylls and Bacteriochlorophylls into Self-assembling Protein Maquettes The underlying concept of protein maquette design is the simple binary patterning approach (Fig. 2) (Ho and DeGrado, 1987; Regan and DeGrado, 1988; DeGrado et al., 1989). It assumes an idealized α-helix and maps an alternating pattern of hydrophobic and hydrophilic helix-forming residues onto a seven-residue two-turn heptad repeat sequence. This enables the convenient design of α-helices with distinct hydrophobic and hydrophilic faces leading, due to the hydrophobic effect, to self-assembly into tetrameric four-helix bundles, with a hydrophobic interior and a polar exterior. Additionally, placing specific residues within the hydrophobic core can confer a prescribed functionality to the protein. Thus, placing histidine residues as axial ligands within the four-helix bundle core resulted in high heme affinity, which provided a robust template for reproducing heme-binding motifs from the transmembrane cytochrome b domain of the cytochrome bc1 complex in a water-soluble four-helix bundle (Robertson et al., 1994; Ghirlanda et al., 2004). The structural and functional properties of the original design were fine tuned further to be more native-like. Functionally, it was possible to reproduce key elements of natural heme proteins such as the modulation of heme redox potentials including electrostatic coupling between redox centers, and between redox centers and proton transfer sites (Shifman et al., 1998, 2000). Structurally, a series of iterative redesign steps resulted in a heme-binding protein maquette that was native-like in its apo-form (Gibney et al., 1999). Then, the detailed structural information obtained from the apoprotein by NMR (Skalicky et al., 1999) and crystallographic methods (Huang et al., 2003) was utilized in the next stage of design leading to a di-heme protein complex with a native-like structure (Huang et al., 2004). Given the structural and chemical homology between heme and (B)Chls, and the successful introduction of Chl and BChl derivatives into natural heme binding proteins such as myoglobin (Wright
Chapter 45
Synthetic Bacteriochlorophyll-Binding Proteins
905
Fig. 4. Heme binding effect on helix orientation. Ligation of two histidines (N, representing an imidazole nitrogen) to the heme iron (Fe) forces them to be perpendicular to the heme plane (bold line). To accommodate this geometry, helices must undergo significant reorientation upon heme binding, exposing hydrophobic residues and burying hydrophilic ones.
and Boxer, 1981; Moog et al., 1984), it was expected that incorporating Chls and BChls into heme binding protein maquettes would be straightforward. Actually, it proved to be a significant challenge, primarily because of the poor solubility of the pigments in water. The hydrophobic nature of Chls and BChls severely limits incorporation into water-soluble proteins because of their tendency to self-aggregate in polar solvents. So far, attempts to bind native Chls, BChls, and their hydrophobic derivatives to the water-soluble protein maquettes have failed, mainly because of the poor solubility of the pigments in water. Conversely, water-soluble pigment derivatives such as chlorin e6 and [Zn]-BChlide are readily incorporated into the heme binding sites of four-helix bundle proteins, albeit with slightly lower affinity than heme (Razeghifard and Wydrzynski, 2003; Koder et al., 2006; Mennenga et al., 2006). Increasing the solubility of a pigment and preventing its self-aggregation is crucial for successful incorporation of Chl and BChl derivatives into water-soluble protein maquettes and opens the way for designing more specific (B)Chl binding sites within water soluble protein maquettes. Yet, this is only the first step towards revealing the particular protein-pigment interactions that discriminate between heme, Chl, and BChl binding in protein maquettes. An obvious target is the coordination properties of the central metal atom in native and metal-substituted (B)Chls that are distinctly different from those of heme-iron. Heme coordination by histidine residues has been the key for designing the recent HP (hydrophilic) series of maquettes which features a native-like structure upon binding two heme molecules. This design was made possible by realizing that heme ligation by two
histidine residues forces significant reorientation of the helices because the planes of the heme and the histidines’ imidazole groups must be perpendicular (Huang et al., 2004). This exposes non-polar residues and buries polar ones (Fig. 4). Thus, the binary pattern of each helix in the HP maquettes was modified in order to accommodate the new helix orientation, thereby leading to a native-like conformation of the heme-protein complex as revealed by NMR. However, when [Zn]-BChlide is incorporated into the HP7 protein maquette only a single helix acquires a native-like conformation whereas incorporating [Mn]BChlide results in a molten globule conformation of the protein (Koder et al., 2006). These observations can be rationalized by the different coordination properties of heme, and [Zn]- and [Mn]-BChlide: the first binds two histidines, the second only one, and the latter has higher affinity for carboxylate ligands and therefore does not bind any histidines. We are still missing critical details of the Chl- and BChl-protein molecular structures, and the differences between these and heme complexes. A detailed investigation of the coordination environment around (B)Chls incorporated within maquettes is therefore crucial for designing more specific maquettes and is currently underway. VI. From Water-soluble to Amphiphilic Chlorophyll- and Bacteriochlorophyllprotein Maquettes Another way of increasing (B)Chl solubility is to use non-polar solvent environments such as detergent micelles or lipid membranes. However, using detergents
906 with water-soluble four-helix bundle maquettes was shown to induce BChl binding but at the expense of losing most of the protein α-helical structure (Kashiwada et al., 2000). Similarly, Eggink and Hoober have shown that a hydrophobic 16-amino acids polypeptide based on a natural motif from plant light harvesting complexes specifically binds Chl in a detergent environment only when the peptide is unfolded (Eggink and Hoober, 2000). Therefore, it is necessary to design stable lipophilic or amphiphilic versions of the water-soluble protein maquettes. Unfortunately, our understanding of the principles that underlie the design of structure and folding of membrane proteins lags significantly behind our understanding of water-soluble proteins partly because the emergence of high-resolution structural information for native membrane proteins is relatively recent and confined to considerably fewer structural examples (Bowie, 2000). Nonetheless, as in the case of water-soluble maquettes, using simple guidelines and robust templates is an effective design strategy. Discher et al. (2005) introduced a modular design strategy that combines the water-soluble (HP) maquettes with lipophilic (LP) four-helix bundle designs to produce amphiphilic (AP) maquette scaffolds
Wolfgang Haehnel, Dror Noy and Hugo Scheer (Discher et al., 2005). The HP domain is de novo designed according to the rules of water-soluble four-helix bundle engineering (exposing charged and polar residues to the aqueous environment and secluding hydrophobic residues in the maquettes’ interior), whereas the LP module may be either a synthetic, de novo designed sequence or a motif from natural transmembrane peptides of known structure. The HP and LP blocks are connected to align the α-helical sequence according to the hydrophilicity of the residues and placement of histidine residues into the interior of the bundle. In the first members of the AP maquette family (Fig. 5), the HP domain was based on the water-soluble heme-binding maquettes. For the LP domain, AP maquette prototypes incorporated four-helix proton channel sequences such as the natural transmembrane segment of the M2 influenza proton channel (Nishimura et al., 2002), or a synthetic channel based on de novo designed sequence (Lear et al., 1988). Once the solubility barrier has been removed, hydrophobic BChl derivatives could be incorporated into heme binding AP maquettes. However, unlike heme, BChls bound only to heme binding sites within the LP domain. For example, the AP3 maquette design
Fig. 5. Summary of the amphiphilic modular design strategy leading to the AP maquette family. Each quarter of a cylinder corresponds to an α-helical peptide for which sequences are given. Hydrophilic residues are in italics, histidine ligands to heme/BChl are bold, and loop sequence italic bold.
Chapter 45
Synthetic Bacteriochlorophyll-Binding Proteins
(Fig. 5) utilized an LP domain based on the natural sequence (residues 188 to 201) from the trans-membrane D helix of the cytochrome bc1 complex which is a part of a four-helix bundle binding site of the high-potential heme (b562) (Iwata et al., 1998). This maquette was capable of binding [Zn]-BChl as well as the extremely hydrophobic [Ni]-BChl. As discussed above, the latter is an excellent spectroscopic probe for the local coordination environment of the central metal. Therefore, it was thoroughly examined using a variety of steady state and time-resolved spectroscopic techniques (Noy et al., 2005). These have shown that the pigment is confined within the fourhelix bundle core of AP3 and ligated by one or two histidines with severely restricted mobility. Acknowledgments Work of the authors described was supported by the Bundesministerium für Forschung und Technologie (BMFT) of Germany (WH, HS), the Deutsche Forschungsgemeinschaft (HS), the Volkswagen Foundation (HS), USA’s National Institute of Health (DN) and the Human Frontiers Science Program Organization (DN). References Agostiano A, Cosma P, Trotta M, Monsu-Scolaro L and Micali N (2002) Chlorophyll a behavior in aqueous solvents: Formation of nanoscale self-assembled complexes. J Phys Chem B 106: 12820–12829 Balaban TS, Linke-Schaetzel M, Bhise AD and Vanthuyne N and Roussel C (2004) Green self-assembling porphyrins and chlorins as mimics of the natural bacteriochlorophylls c, d, and e. Eur J Org Chem 2004: 3919–3930 Barker PD (2003) Designing redox metalloproteins from bottom-up and top-down perspectives. Curr Op Struct Biol 13: 490–499 Ben-Shem A, Nelson N and Frolow F (2003) Crystallization and initial X-ray diffraction studies of higher plant Photosystem I. Acta Crystallogr D Biol Crystallogr 59: 1824–1827 Biesiadka J, Loll B, Kern J, Irrgang K-D and Zouni A (2004) Crystal structure of cyanobacterial Photosystem II at 3.2 Å resolution: A closer look at the Mn cluster. Phys Chem Chem Phys 6: 4733–4736 Bowie JU (2000) Understanding membrane protein structure by design. Nature Struct Biol 7: 91–94 Boxer SG, Kuki A, Wright KA, Katz BA and Xuong N (1982) Oriented properties of the chlorophylls — electronic absorption-spectroscopy of orthorhombic pyrochlorophyllide α-apomyglobin single-crystals. Proc Natl Acad Sci USA 79:
907
1121–1125 Buchler JW (1975) Static coordination chemistry of metalloporphyrins. In: Smith KM (ed) Porphyrins and Metalloporphyrins, pp 157–232. Elsevier, Amsterdam Butterfield SM and Waters ML (2003) A designed hairpin peptide for molecular recognition of ATP in water. J Am Chem Soc 125: 9580–9581 Chen M, Eggink LL, Hoober JK and Larkum AW (2005) Influence of structure on binding of chlorophylls to peptide ligands. J Am Chem Soc 127: 2052–2053 Cochran FV, Wu SP, Wang W, Nanda V, Saven JG, Therien MJ and DeGrado WF (2005) Computational de novo design and characterization of a four-helix bundle protein that selectively binds a nonbiological cofactor. J Am Chem Soc 127: 1346–1347 Coleman WJ and Youvan DC (1990) Spectroscopic analysis of genetically modified photosynthetic reaction centers. Ann Rev Biophys (Biophys Chem.) 19: 333–367 Davis CM, Parkes-Loach PS, Cook CK, Meadows KA, Bandilla M, Scheer H and Loach PA (1996) Comparison of the structural requirements for bacteriochlorophyll binding in the core light-harvesting complexes of Rhodospirillum rubrum and Rhodobacter sphaeroides using reconstitution methodology with bacteriochlorophyll analogs. Biochemistry 35: 3072–3084 De Jonge N, Rau HK and Haehnel W (1999) Light-induced electron transfer in synthetic metalloproteins. Z Phys Chem 213: 175–180 DeGrado WF, Wasserman ZR and Lear JD (1989) Protein design, a minimalist approach. Science 243: 622–628 Dieckmann GR, McRorie DK, Tierney DL, Utschig LM, Singer CP, O’Halloran TV, Penner-Hahn JE, DeGrado WF and Pecoraro VL (1997) De novo design of mercury-binding two- and threehelical bundles. J Am Chem Soc 119: 6195–6196 Discher BM, Noy D, Strzalka J, Ye S, Moser CC, Lear JD, Blasie JK and Dutton PL (2005) Design of amphiphilic protein maquettes: controlling assembly, membrane insertion, and cofactor interactions. Curr Op Chem Biol 44: 12329–12343 Domanskii V, Rassadina V, Gus-Mayer S, Wanner G, Schoch S and Rüdiger W (2003) Characterization of two phases of chlorophyll formation during greening of etiolated barley leaves. Planta 216: 475–483 Dudkowiak A, Nakamura C, Arai T and Miyake J (1998) Interactions of chlorophyll a with synthesized peptide in aqueous solution. J Photochem Photobiol B 45: 43–50 Dudkowiak A, Kusumi T, Nakamura C and Miyake J (1999) Chlorophyll a aggregates stabilized by a synthesized peptide. J Photochem Photobiol. A 129: 51–55 Eggink LL and Hoober JK (2000) Chlorophyll binding to peptide maquettes containing a retention motif. J Biol Chem 275: 9087–9090 El-Kabbani O, Chang C-H, Tiede D, Norris J and Schiffer M (1991) Comparison of reaction centers from Rhodobacter sphaeroides and Rhodopseudomonas viridis: Overall architecture and protein-pigment interactions. Biochemistry 30: 5361–5369 Evans TA and Katz JJ (1975) Evidence for 5-and 6-coordinated magnesium in bacteriochlorophyll a from visible absorption spectroscopy. Biochim Biophys Acta 396: 414–426 Fahnenschmidt M, Bittl R, Rau HK, Haehnel W and Lubitz W (2000) Electron paramagnetic resonance and electron nuclear double resonance spectroscopy of a heme protein maquette. Chem Phys Lett 323: 329–339 Fahnenschmidt M, Rau HK, Bittl R, Haehnel W and Lubitz W
908 (1999) Characterization of a de novo designed heme protein by EPR and ENDOR spectroscopy. Chem Eur J 5: 2327–2334 Ferreira KN, Iverson TM, Maghlaoui K, Barber J and Iwata S (2004) Architecture of the photosynthetic oxygen-evolving center. Science 303: 1831–1838 Fiedor L (2006) Hexacoordination of bacteriochlorophyll in photosynthetic antenna LH1. Biochemistry 45: 1910–1918 Fiedor L, Leupold D, Teuchner K, Voigt B, Hunter CN, Scherz A and Scheer H (2000) Excitation trap approach to analyze size and pigment-pigment coupling: Reconstitution of LH1 antenna of Rhodobacter sphaeroides with Ni-substituted bacteriochlorophyll. Biochemistry 40: 3737–3747. Frank HA, Innes J, Aldema M, Neumann R and Schenck CC (1993) Triplet state EPR of reaction centers from the His(L173) → Leu(L173) mutant of Rhodobacter sphaeroides which contains a heterodimer primary donor. Photosynth. Res. 38: 99–101. Freer A, Prince S, Sauer K, Papiz M, Hawthornthwaite-Lawless A, McDermott G, Cogdell R and Isaacs NW (1996) Pigment. pigment interactions and energy transfer in the antenna complex of the photosynthetic bacterium Rhodopseudomonas acidophila. Structure 4: 449–462 Frigaard NU, Chew AGM, Maresca JA and Bryant DA (2006) Bacteriochlorophyll biosynthesis in green bacteria. In: Grimm B, Porra R, Rüdiger W and Scheer H (eds) Chlorophylls and Bacteriochlorophylls: Biochemistry, Biophysics, Functions and Applications (Advances in Photosynthesis and Respiration, Vol 25), pp 201–221. Springer, Dordrecht Gall A, Cogdell RJ and Robert B (2003) Influence of carotenoid molecules on the structure of the bacteriochlorophyll binding site in peripheral light-harvesting proteins from Rhodobacter sphaeroides. Biochemistry 42: 7252–8 Garcia-Martin A, Kwa L, vonJan M and Braun P (2006) Assembly of novel bacteriochlorophyll proteins in the native lipid environment. In: Grimm B, Porra R, Rüdiger W and Scheer H (eds) Chlorophylls and Bacteriochlorophylls: Biochemistry, Biophysics, Functions and Applications (Advances in Photosynthesis and Respiration, Vol 25), pp 387–396. Springer, Dordrecht Gentemann S, Medforth CJ, Forsyth TP, Nurco DJ, Smith KM, Fajer J and Holten D (1994) Photophysical properties of conformationally distorted metal-free porphyrins. Investigation into the deactivation mechanisms of the lowest excited singlet state. J Am Chem Soc 116: 7363–7368 Ghirlanda G, Osyczka A, Liu W, Antolovich M, Smith KM, Dutton PL, Wand AJ and DeGrado WF (2004) De novo design of a D2-symmetrical protein that reproduces the diheme four-helix bundle in cytochrome bc1. J Am Chem Soc 126: 8141–8147 Ghosh D, Lee KH, Demeler B and Pecoraro VL (2005) Linear free-energy analysis of mercury(II) and cadmium(II) binding to three-stranded coiled coils. Biochemistry 44: 10732–10740 Gibney BR, Mulholland SE, Rabanal F and Dutton PL (1996) Ferredoxin and ferredoxin-heme maquettes. Proc Natl Acad Sci USA 93: 15041–15046 Gibney BR, Rabanal F, Skalicky JJ, Wand AJ and Dutton PL (1999) Iterative protein redesign. J Am Chem Soc 121: 4952–4960 Gibney BR, Isogai Y, Rabanal F, Reddy KS, Grosset AM, Moser CC and Dutton PL (2000) Self-assembly of heme A and heme B in a designed four-helix bundle: Implications for a cytochrome c oxidase maquette. Biochemistry 39: 11041–11049 Grosset AM, Gibney BR, Rabanal F, Moser CC and Dutton PL (2001) Proof of principle in a de novo designed protein maquette: An allosterically regulated, charge-activated confor-
Wolfgang Haehnel, Dror Noy and Hugo Scheer mational switch in a tetra-alpha-helix bundle. Biochemistry 40: 5474–5487 Gudowska-Nowak E, Newton MD and Fajer J (1990) Conformational and environmental effects on bacteriochlorophyll optical spectra: Correlations of calculated spectra with structural results. J Phys Chem 94: 5795–5801 Haehnel W (2004) Chemical synthesis of TASP-arrays and their application in protein design. Mol Diversity 8: 219–229 Harbury PB, Zhang T, Kim PS and Alber T (1993) A switch between two-, three-, and four-stranded coiled coils in GCN4 leucine zipper mutants. Science 262: 1401–1407 Hartwich G, Fiedor L, Simonin I, Cmiel E, Schaefer W, Noy D, Scherz A and Scheer H (1998) Metal-substituted bacteriochlorophylls. 1. Preparation and influence of metal and coordination on spectra. J Am Chem Soc 120: 3675–3683 Ho SP and DeGrado WF (1987) Design of a 4-helix bundle protein: Synthesis of peptides which self-associate into a helical protein. J Am Chem Soc 109: 6751–6758 Hofmann E, Wrench PM, Sharples FP, Hiller RG, Welte W and Diederichs K (1996) Structural basis of light-harvesting by carotenoids: peridinin-chlorophyll-protein from Amphidinium carterae. Science 272: 1788–1791 Horigome D, Satoh H, Itoh N, Mitsunaga K, Oonishi I, Nakagawa A and Uchida A (2007) Structural mechanism and photoprotective function of water-soluble chlorophyll-binding protein. J Biol Chem 282: 76525–6531 Huang SS, Gibney BR, Stayrook SE, Dutton PL and Lewis M (2003) X-ray structure of a maquette scaffold. J Mol Biol 326: 1219–1225 Huang SS, Koder RL, Lewis M, Wand AJ and Dutton PL (2004) The HP-1 maquette: From an apoprotein to a structured hemoprotein designed to promote redox-coupled proton exchange. Proc Natl Acad Sci USA 101: 5536–5541 Hübschmann T, Borner T, Hartmann E and Lamparter T (2001) Characterization of the Cph1 holo-phytochrome from Synechocystis sp. PCC 6803. Eur J Biochem 268: 2055–2063 Ivancich A, Artz K, Williams JC, Allen JP and Mattioli TA (1998) Effects of hydrogen bonds on the redox potential and electronic structure of the bacterial primary electron donor. Biochemistry 37: 11812–11820 Iwata S, Lee JW, Okada K, Lee JK, Iwata M, Rasmussen B, Link TA, Ramaswamy S and Jap BK (1998) Complete structure of the 11-subunit bovine mitochondrial cytochrome bc1 complex. Science 281: 64–71 Jordan P, Fromme P, Witt HT, Klukas O, Saenger W and Krauss N (2001) Three-dimensional structure of cyanobacterial Photosystem I at 2.5 Å resolution. Nature 411: 909–917 Kashiwada A, Watanabe H, Tanaka T and Nango M (2000) Molecular assembly of zinc bacteriochlorophyll a by synthetic hydrophobic 1α-helix polypeptides. Chem Lett 24–25 Katz E, Heleg-Shabtai V, Willner I, Rau HK and Haehnel W (1998) Surface reconstitution of a de novo synthesized hemoprotein for bioelectronic applications. Angew Chem Int Ed 37: 3253–3256 Katz JJ, Shipman LL, Cotton TM and Janson TR (1978) Chlorophyll aggregation: coordination interactions in chlorophyll monomers, dimers and oligomers. In: Dolphin D (ed) The Porphyrins. Physical Chemistry, Part C, pp 402–458. Academic Press, New York Katz JJ, Bowman MK, Michalski TJ and Worcester DL (1991) Chlorophyll aggregation: Chlorophyll-water micelles as models
Chapter 45
Synthetic Bacteriochlorophyll-Binding Proteins
for in vivo long-wavelength chlorophyll. In: Scheer H (ed) Chlorophylls, pp 211–236. CRC-Press, Boca Raton Kennedy ML and Gibney BR (2002) Proton coupling to [4Fe-4S]2+ and [4Fe-4Se]2+ oxidation and reduction in a designed protein. J Am Chem Soc 124: 6826–6827 Klemba M, Gardner KH, Marino S, Clarke ND and Regan L (1995) Novel metal-binding proteins by design. Nature Struct Biol 2: 368–373 Koder RL and Dutton PL (2006) Intelligent design: The de novo engineering of proteins with specified functions. Dalton Transact 25: 3045–3051 Koder RL, VKG, Cerda J, Noy D, Smith KM, Wand AJ and Dutton PL (2006) Native-like structure in designed four α-helix bundles driven by buried polar interactions. J Am Chem Soc 128: 14450–14451 Kräutler B and Hörtensteiner S (2006) Chlorophyll catabolites and the biochemistry of chlorophyll breakdown. In: Grimm B, Porra R, Rüdiger W and Scheer H (eds) Chlorophylls and Bacteriochlorophylls: Biochemistry, Biophysics, Functions and Applications (Advances in Photosynthesis and Respiration, Vol 25), pp 237–260. Springer, Dordrecht Kunieda M, Mizoguchi T and Tamiaki H (2004) Diastereoselective self-aggregation of synthetic 3-(1-hydroxyethyl)-bacteriopyrochlophyll a as a novel photosynthetic antenna model absorbing near the infrared regions. Photochem Photobiol 79: 55–61 Küpper H, Dedic R, Svoboda A, Hala J and Kroneck PM (2002) Kinetics and efficiency of excitation energy transfer from chlorophylls, their heavy metal-substituted derivatives, and pheophytins to singlet oxygen. Biochim Biophys Acta 1572: 107–113 Lapouge K, Näveke A, Sturgis JN, Hartwich G, Renaud D, Simonin I, Lutz M, Scheer H and Robert B (1998) Non-bonding molecular factors influencing the stretching wavenumbers of the conjugated carbonyl groups of bacteriochlorophyll a. J Raman Spectrosc 29: 977–981 Lear JD, Wasserman ZR and Degrado WF (1988) Synthetic amphiphilic peptide models for protein ion channels. Science 240: 1177–1181 Li WW, Hellwig P, Ritter M and Haehnel W (2006) De novo design, synthesis, and characterization of quinoproteins. Chem Eur J 12: 7236–7245 Liu Z, Yan H, Wang K, Kuang T, Zhang J, Gui L, An X and Chang W (2004) Crystal structure of spinach major light-harvesting complex at 2.72 Å resolution. Nature 428: 287–292 Lutz M and Mäntele W (1991) Vibrational spectroscopy of chlorophylls. In: Scheer H (ed) Chlorophylls, pp 855–902. CRC Press, Boca Raton, FL Ma JG, Zhang J, Franco R, Jia SL, Moura I, Moura JJG, Kroneck PMH and Shelnutt JA (1998) The structural origin of nonplanar heme distortions in tetraheme ferricytochromes c3. Biochemistry 37: 12431–12442 Macpherson A and Hiller R (2003) Light-harvesting Systems in Chl c-containing Algae. In: Green B and Parson W (eds) Light-Harvesting Antennas in Photosynthesis (Advances in Photosynthesis and Respiration, Vol 13), pp 323–352. Kluwer Academic Publishers, Dordrecht Maglio O, Nastri F, Pavone V, Lombardi A and DeGrado WF (2003) Preorganization of molecular binding sites in designed diiron proteins. Proc Natl Acad Sci USA 100: 3772–3777 Markovic D, Pröll S and Scheer H (2007) Thermal unfolding of myoglobin: Influence of chromophore structure on stability.
909
Biochim Biophys Acta 1767: 897–904 Marsh EN and DeGrado WF (2002) Noncovalent self-assembly of a heterotetrameric diiron protein. Proc Natl Acad Sci USA 99: 5150–5154 Masuda T, Inoue K, Masuda M, Nagayama M, Tamaki A, Ohta H, Shimada H and Takamiya K (1999) Magnesium insertion by magnesium chelatase in the biosynthesis of zinc bacteriochlorophyll a in an aerobic acidophilic bacterium Acidiphilium rubrum. J Biol Chem 274: 33594–33600 McLuskey K, Prince SM, Cogdell RJ and Isaacs NW (1999) Crystallization and preliminary X-ray crystallographic analysis of the B800-820 light-harvesting complex from Rhodopseudomonas acidophila strain 7050. Acta Cryst D Biol Cryst 55: 885–887 Meadows KA, Iida K, Tsuda K, Recchia PA, Heller BA, Antonio B, Nango M and Loach PA (1995) Enzymatic and chemical cleavage of the core light-harvesting polypeptides of photosynthetic bacteria: Determination of the minimal polypeptide size and structure required for subunit and light-harvesting complex formation. Biochemistry 35: 1559–1574 Medforth CJ, Senge MO, Smith KM, Sparks LD and Shelnutt JA (1992) Nonplanar distortion modes for highly substituted porphyrins. J Amer Chem Soc 114: 9859–9869 Mennenga A, Gärtner W, Lubitz W and Gorner H (2006) Effects of noncovalently bound quinones on the ground and triplet states of zinc chlorins in solution and bound to de novo synthesized peptides. Phys Chem Chem Phys 8: 5444–5453 Michel H, Epp O and Deisenhofer J (1986) Pigment protein interactions in the photosynthetic reaction center from Rhodopseudomonas viridis. EMBO J 5: 2445–2451 Miyake J, Kusumi T, Dudkowiad A, Goc J and Frackowiak D (1998) The interactions between bacteriochlorophyll c and amphiphilic peptides. J Photochem Photobiol A 116: 147–151 Moog RS, Kuki A, Fayer MD and Boxer SG (1984) Excitation transport and trapping in a synthetic chlorophyllide substituted hemoglobin: Orientation of the chlorophyll S1 transition dipole. Biochemistry 23: 1564–1571 Mulholland SE, Gibney BR, Rabanal F and Dutton PL (1998) Characterization of the fundamental protein ligand requirements of [4Fe-4S]2+ clusters with sixteen amino acid maquettes. J Am Chem Soc 120: 10296–10302 Mulholland SE, Gibney BR, Rabanal F and Dutton PL (1999) Determination of nonligand amino acids critical to [4Fe4S]2+ assembly in ferredoxin maquettes. Biochemistry 38: 10442–10448 Musewald C, Hartwich G, Lossau H, Gilch P, Pöllinger-Dammer F, Scheer H and Michel-Beyerle ME (1999) Ultrafast photophysics and photochemistry of [Ni]-bacteriochlorophyll a. J Phys Chem B 103: 7055–7060 Mutter M and Vuilleumier S (1989) A chemical approach to protein design — template assembled synthetic proteins (DASP). Angew Chem Int Ed 28: 535–554 Nanda V, Rosenblatt MM, Osyczka A, Kono H, Getahun Z, Dutton PL, Saven JG and DeGrado WF (2005) De novo design of a redox-active minimal rubredoxin mimic. J Am Chem Soc 127: 5804–5805 Nango M (2006) Molecular assembly of bacteriochlorophyll complexes using synthetic light-harvesting (LH) model polypeptides. In: Grimm B, Porra R, Rüdiger W and Scheer H (eds) Chlorophylls and Bacteriochlorophylls: Biochemistry, Biophysics, Functions and Applications (Advances in Photosynthesis
910 and Respiration, Vol 25), pp 365–373. Springer, Dordrecht Nicholson DW and Thornberry NA (2003) Apoptosis. Life and death decisions. Science 299: 214–215 Nishimura K, Kim SG, Zhang L and Cross TA (2002) The closed state of a H+ channel helical bundle combining precise orientational and distance restraints from solid state NMR. Biochemistry 41: 13170–13177 Noguchi T, Kamimura Y, Inoue Y and Itoh S (1999) Photoconversion of a water-soluble chlorophyll protein from Chenopodium album: Resonance Raman and Fourier transform infrared study of protein and pigment structures. Plant Cell Physiol 40: 305–310 Noy D and Dutton PL (2006) Design of a minimal polypeptide unit for bacteriochlorophyll binding and self-assembly based on photosynthetic bacterial light-harvesting proteins. Biochemistry 45: 2103–2113 Noy D, Fiedor L, Hartwich G, Scheer H and Scherz A (1998) Metal-substituted bacteriochlorophylls. 2. Changes in redox potentials and electronic transition energies are dominated by intramolecular electrostatic interactions. J Am Chem Soc 120: 3684–3693 Noy D, Yerushalmi R, Brumfeld V, Ashur I, Scheer H, Baldridge KK and Scherz A (2000) Optical absorption and computational studies of [Ni]-Bacteriochlorophyll a. New insight into charge distribution between metal and ligands. J Am Chem Soc 122: 3937–3944 Noy D, Discher BM, Rubtsov IV, Hochstrasser RM and Dutton PL (2005) Design of amphiphilic protein maquettes: Enhancing maquette functionality through binding of extremely hydrophobic cofactors to lipophilic domains. Biochemistry 44: 12344–12354 Oba T and Tamiaki H (1999) Why do chlorosomal chlorophylls lack the C132-methoxycarbonyl moiety? An in vitro model study. Photosynth Res 61: 23–31 Olsen JD, Sturgis JN, Westerhuis WH, Fowler GJ, Hunter CN and Robert B (1997) Site-directed modification of the ligands to the bacteriochlorophylls of the light-harvesting LH1 and LH2 complexes of Rhodobacter sphaeroides. Biochemistry 36: 12625–32 Paulsen H (2006) Reconstitution and pigment exchange. In: Grimm B, Porra R, Rüdiger W and Scheer H (eds) Chlorophylls and Bacteriochlorophylls: Biochemistry, Biophysics, Functions and Applications (Advances in Photosynthesis and Respiration, Vol 25), pp 375–385. Springer, Dordrecht Paulsen H and Kuttkat A (1993) Pigment complexes of lightharvesting chlorophyll a/b binding protein are stabilized by a segment in the carboxyterminal hydrophilic domain of the protein. Photochem Photobiol 57: 139–142 Pearlstein RM, Davis RC and Ditson SL (1982) Giant circular dichroism of high molecular-weight chlorophyllide-apomyoglobin complexes. Proc Natl Acad Sci USA 79: 400–402 Rabanal F, DeGrado WF and Dutton PL (1996) Toward the synthesis of a photosynthetic reaction center maquette: A cofacial porphyrin pair assembled between two subunits of a synthetic four-helix bundle multiheme protein. J Am Chem Soc 118: 473–474 Rau HK and Haehnel W (1996) Biomimetic constructs: de-novo design of redox proteins. Ber Bunsenges Phys Chem 100: 2052–2056 Rau HK and Haehnel W (1998) Design, synthesis, and proper-
Wolfgang Haehnel, Dror Noy and Hugo Scheer ties of a novel cytochrome b model. J Am Chem Soc 120: 468–476 Rau HK, DeJonge N and Haehnel W (1998) Modular synthesis of de novo-designed metalloproteins for light-induced electron transfer. Proc Natl Acad Sci USA 95: 11526–11531 Rau HK, DeJonge N and Haehnel W (2000) Combinatorial synthesis of four-helix bundle hemoproteins for tuning of cofactor properties. Angew Chem Int Ed 39: 250–253 Rau HK, Snigula H, Struck A, Robert B, Scheer H and Haehnel W (2001) Design, synthesis and properties of synthetic chlorophyll proteins. Eur J Biochem 268: 3284–3295 Razeghifard AR and Wydrzynski T (2003) Binding of Zn-chlorin to a synthetic four-helix bundle peptide through histidine ligation. Chem Lett 42: 1024–1030 Regan L and DeGrado WF (1988) Characterization of a helical protein designed from first principles. Science 241: 976–978 Reinbothe C, Satoh H, Alcaraz J-P and Reinbothe S (2004) A novel role of water-soluble chlorophyll proteins in the transitory storage of chlorophyllide. Plant Physiol 134: 1355–1365 Robert B (1996) Resonance Raman studies in photosynthesischlorophyll and carotenoid molecules. In: Amesz J and Hoff AJ (eds) Biophysical Techniques in Photosynthesis (Advances in Photosynthesis and Respiration, Vol 3), pp 161–176. Kluwer Academic Publishers, Dordrecht Robert B, Cogdell R and Van Grondelle R (2003) The light-harvesting system of purple bacteria . In: Green B and Parson W (eds) Light-Harvesting Antennas in Photosynthesis (Advances in Photosynthesis and Respiration, Vol 13), pp 169–194. Kluwer Academic Publishers, Dordrecht Robertson DE, Farid RS, Moser CC, Urbauer JL, Mulholland SE, Pidikiti R, Lear JD, Wand AJ, DeGrado WF and Dutton PL (1994) Design and synthesis of multi-haem proteins. Nature 368: 425–431 Roszak AW, Howard TD, Southall J, Gardiner AT, Law CJ, Isaacs NW and Cogdell RJ (2003) Crystal structure of the RC-LH1 core complex from Rhodopseudomonas palustris. Science 302: 1969–1972 Rüdiger W (2006) Biosynthesis of chlorophylls a and b: The Last Steps. In: Grimm B, Porra R, Rüdiger W and Scheer H (eds) Chlorophylls and Bacteriochlorophylls: Biochemistry, Biophysics, Functions and Applications (Advances in Photosynthesis and Respiration, Vol 25) , pp 189–200. Springer, Dordrecht Scheer H (2003) The pigments. In: Green B and Parson W (eds) Green B and Parson W (eds) Light-Harvesting Antennas in Photosynthesis (Advances in Photosynthesis and Respiration, Vol 13), pp 29–81. Kluwer Academic Publishers, Dordrecht Scheer H (2006) Overview. In: Grimm B, Porra R, Rüdiger W and Scheer H (eds) Chlorophylls and Bacteriochlorophylls: Biochemistry, Biophysics, Functions and Applications (Advances in Photosynthesis and Respiration, Vol 25), pp 1–26. Springer, Dordrecht Scheer H and Hartwich G (1995) Bacterial reaction centers with modified tetrapyrrole chromophores. In: Blankenship R, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 649–663. Kluwer Academic Publishers, Dordrecht Scheer H and Katz JJ (1975) Nuclear magnetic resonance spectroscopy of porphyrins and metalloporphyrins. In: Smith KM (ed) Porphyrins and Metalloporphyrins, pp 399–524, Elsevier, New York
Chapter 45
Synthetic Bacteriochlorophyll-Binding Proteins
Scheer H, Paulke B and Gottstein J (1985) Long-wavelength absorbing forms of bacteriochlorophylls. II. Structural requirements for formation in Triton X-100 micelles and in aqueous methanol and acetone. In: Blauer G and Sund H (eds) Optical Properties and Structure of Tetrapyrroles, pp 507–521. De Gruyter, London Scherz A and Parson WW (1986) Interactions of the bacteriochlorophylls in antenna bacteriochlorophyll-protein complexes of photosynthetic bacteria. Photosynth Res 9: 21–32 Scherz A, Rosenbach-Belkin V, Michalski TJ and Worcester DL (1991) Chlorophyll aggregates in aqueous solutions. In: Scheer H (ed) Chlorophylls, pp 237–268. CRC-Press, Boca Raton Schlichter J, Friedrich J, Parbel M and Scheer H (2001) Influence of isotopic substitution on the conformational dynamics of frozen proteins. J Chem Phys 114: 9638–9644 Schmidt K, Fufezan C, Krieger-Liszkay A, Satoh H and Paulsen H (2003) Recombinant water-soluble chlorophyll protein from Brassica oleracea var. Botrys binds various chlorophyll derivatives. Biochemistry 42: 7427–7433 Schnepf R, Horth P, Bill E, Wieghardt K, Hildebrandt P and Haehnel W (2001) De novo design and characterization of copper centers in synthetic four-helix-bundle proteins. J Am Chem Soc 123: 2186–95 Schoch S and Brown J (1986) The action of chlorophyllase on chlorophyll-protein complexes. J Plant Physiol 126: 483–494 Senge MO (2000) Highly substituted porphyrins. In: Kadish KM, Smith KM, and Guilard R (eds) The Porphyrin Handbook, pp 239–347. Academic Press, San Diego Senge MO and Smith KM (1994) Structure and conformation of photosynthetic pigments and related compounds. 7. On the conformation of the methyl ester of (20-methyl-phytochlorinato) nickel(II) — A bacteriochlorophyll c model compound. Photochem Photobiol 60: 139–142 Senge MO, Kalisch WW and Runge S (1998) Conformationally distorted chlorins via diimide reduction of nonplanar porphyrins. Tetrahedron 54: 3781–3798 Senge MO, Wiehe A and Ryppa C (2006) Synthesis, reactivity and structure of chlorophylls. In: Grimm B, Porra R, Rüdiger W and Scheer H (eds) Chlorophylls and Bacteriochlorophylls: Biochemistry, Biophysics, Functions and Applications (Advances in Photosynthesis and Respiration, Vol 25), pp 27–37. Springer, Dordrecht Sharp RE, Diers JR, Bocian DF and Dutton PL (1998a) Differential binding of iron(III) and zinc(II) protoporphyrin IX to synthetic four-helix bundles. J Am Chem Soc 120: 7103–7104 Sharp RE, Moser CC, Rabanal F and Dutton PL (1998b) Design, synthesis, and characterization of a photoactivatable flavocytochrome molecular maquette. Proc Natl Acad Sci USA 95: 10465–10470 Shifman JM, Moser CC, Kalsbeck WA, Bocian DF and Dutton PL (1998) Functionalized de novo designed proteins: Mechanism of proton coupling to oxidation/reduction in heme protein maquettes. Biochemistry 37: 16815–16827 Shifman JM, Gibney BR, Sharp RE and Dutton PL (2000) Heme redox potential control in de novo designed four-α-helix bundle proteins. Biochemistry 39: 14813–14821 Skalicky JJ, Gibney BR, Rabanal F, Bieber Urbauer RJ, Dutton PL and Wand AJ (1999) Solution Structure of a Designed Four-α-Helix Bundle Maquette Scaffold. J Am Chem Soc 121: 4941–4951
911
Snigula H (2004) (Bacterio)Chlorophyll-Modifikationen zur Einlagerung in synthetische Peptide. Dissertation, LudwigMaximilians-Universität , München Sparks LD, Medforth CJ, Park M-S, Chamberlain JR, Ondrias MR, Senge MO, Smith KM and Shelnutt JA (1993) Metal dependence of the nonplanar distortion of octaalkyltetra-phenylporphyrins. J Amer Chem Soc 115: 581–591 Spiedel D, Roszak AW, McKendrick K, McAuley KE, Fyfe PK, Nabedryk E, Breton J, Robert B, Cogdell RJ, Isaacs NW and Jones MR (2002) Tuning of the optical and electrochemical properties of the primary donor bacteriochlorophylls in the reaction centre from Rhodobacter sphaeroides: Spectroscopy and structure. Biochim Biophys Acta 1554: 75–93 Standfuss J, Terwisscha van Scheftinga AC, Lamborghini M and Kühlbrandt W (2005) Mechanisms of photoprotection and nonphotochemical quenching in pea light-harvesting complex at 2.5 Å resolution. EMBO J 24 : 919–928 Sturgis JN, Jirsakova V, Reiss-Husson F, Cogdell RJ and Robert B (1995) Structure and properties of the bacteriochlorophyll binding site in peripheral light-harvesting complexes of purple bacteria. Biochemistry 34: 517–23 Summa CM, Rosenblatt MM, Hong JK, Lear JD and DeGrado WF (2002) Computational de novo design, and characterization of an A2B2 diiron protein. J Mol Biol 321: 923–938 Tronrud DE and Matthews BW (1993) Refinement of the structure of a water-soluble antenna complex from green photosynthetic bacteria by incorporation of the chemically determined amino acid sequence. In: Deisenhofer J and Norris JR (eds) The Photosynthetic Reaction Center, pp 13–22. Academic Press, New York Vavilin D, Xu H, Lin S and Vermaas W (2003) Energy and electron transfer in Photosystem II of a chlorophyll b-containing Synechocystis sp. PCC 6803 mutant. Biochemistry 42: 1731–1746 Walter E, Schreiber J, Zass E and Eschenmoser A (1979) Bakteriochlorophyll aGG und Bakteriophäophytin aP in den photosynthetischen Reaktionszentren von Rhodospirillum rubrum G9. Helv Chim Acta 62: 899–920 Willner I, Heleg-Shabtai V, Katz E, Rau HK and Haehnel W (1999) Integration of a reconstituted de novo synthesized hemoprotein and native metalloproteins with electrode supports for bioelectronic and bioelectrocatalytic applications. J Am Chem Soc 121: 6455–6468 Woodward RB, Ayer WA, Beaton JM, Bickelhaupt F, Bonnett R, Buchschacher P, Closs GL, Dutler H, Hannah J, Hauck FP, Ito S, Langemann A, LeGoff E, Leimgruber W, Lwowski W, Sauer J, Valenta Z and Voltz H (1960) The total synthesis of chlorophyll. J Amer Chem Soc 82: 3800–3802 Wright KA and Boxer SG (1981) Solution properties and synthetic chlorophyllide-apomyoglobin and bacteriochlorophyllide-apomyoglobin complexes. Biochemistry 20: 7546–7556 Yeates TO, Komiya H, Chirino A, Rees DC, Allen JP and Feher G (1988) Structure of the reaction center from Rhodobacter sphaeroides R-26 and 2.4.1: Protein-cofactor (bacteriochlorophyll, bacteriopheophytin and carotenoid) interactions. Proc Natl Acad Sci USA 85: 7993–7997 Yerushalmi R, Ashur I and Scherz A (2006) Metal-substituted bacteriochlorophylls: novel molecular tools. In: Grimm B, Porra R, Rüdiger W and Scheer H (eds) Chlorophylls and Bacteriochlorophylls: Biochemistry, Biophysics, Functions and
912 Applications (Advances in Photosynthesis and Respiration, Vol 25), pp 495–506. Springer, Dordrecht Zhang Z, Huang L, Shulmeister VM, Chi Y-I, Kim KK, Hung L-W, Crofts AR, Berry EA and Kim S-H (1998) Electron transfer by domain movement in cytochrome bc1. Nature 392: 677–684
Wolfgang Haehnel, Dror Noy and Hugo Scheer Zouni A, Witt HT, Kern J, Fromme P, Krauss N, Saenger W and Orth P (2001) Crystal structure of Photosystem II from Synechococcus elongatus at 3.8 Å resolution. Nature 409: 739–743
Chapter 46 Design and Assembly of Functional Light-Harvesting Complexes Paula Braun* LM University of Munich, Department I, Botany, Menzingerstr. 67, D-80638 Munich, Germany
Leszek Fiedor Faculty of Biochemistry, Biophysics and Biotechnology, Jagiellonian University, 30-387 Cracow, Gronostajowa 7, Poland Summary ............................................................................................................................................................... 914 I. Introduction..................................................................................................................................................... 914 II. Design of Model Light-Harvesting Proteins .................................................................................................... 916 A. Towards In silico Prediction of (Bacterio)chlorophyll and Carotenoid-binding Motifs....................... 916 1. Statistical Analyses of Chlorophyll-Binding Pockets ............................................................... 916 2. Statistical Analyses of Carotenoid-Binding Pockets ............................................................... 919 B. Model Light-Harvesting Proteins as a Tool to Study the Assembly of Light-Harvesting Systems ... 920 1. Light-Harvesting 2-like Complexes with Model Bacteriochlorophyll and Carotenoid Binding Sites ........................................................................................................................... 920 2. Assembly Motifs at the Bacteriochlorophyll/Protein Interface ................................................ 922 3. Hydrogen-Bonding at the Bacteriochlorophyll/Protein Interface ............................................. 922 4. Binding of Light-Harvesting-Active Carotenoid ...................................................................... 924 III. Assembly of Functional Light-Harvesting 1 Complexes ................................................................................. 924 A. Model Systems for the in vitro Assembly of Functional Light-Harvesting 1 Complexes .................. 924 1. Rhodobacter sphaeroides ..................................................................................................... 925 2. Rhodospirillum rubrum .......................................................................................................... 926 B. Reconstitution of the Light-Harvesting 1 Complex from B820 Subunits........................................... 926 C. Comparison of Reconstituted and Native Light-Harvesting 1 Complexes ....................................... 926 1. Rhodobacter sphaeroides ..................................................................................................... 926 2. Rhodospirillum rubrum .......................................................................................................... 927 D. Introduction of Ni-Bacteriochlorophyll a into the Light-Harvesting 1 Complex as the Excitation Trap ................................................................................................................................. 928 1. Pigment Coupling and the Size of the Light-Harvesting 1 Complex ....................................... 928 2. Exciton Delocalization in the Light-Harvesting 1 Complex ...................................................... 929 E. The Role of Carotenoids in the Light-Harvesting 1 Complex ........................................................... 929 1. Light-Harvesting 1 Complexes with Modified Carotenoids...................................................... 929 2. Effects of Carotenoids on Assembly of the Light-Harvesting 1 Complex ................................ 930 3. Identification of an Intermediate of Assembly of the Light-Harvesting 1 Complex .................. 932 F. Model for Assembly of the Light-Harvesting 1 Complex in vitro ....................................................... 933 1. Stage 1: Formation of Monomeric Subunits B780 .................................................................. 933 2. Stage 2: Formation of Dimeric/Tetrameric Subunits B820 ...................................................... 934 3. Stage 3: Oligomerization of Subunits to Larger Assemblies ................................................... 934 4. Stage 4: Assembly of Complete Antenna B880 ...................................................................... 934 IV. Conclusions and Prospects ............................................................................................................................ 935 Acknowledgments ................................................................................................................................................. 935 References ............................................................................................................................................................ 936 *Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 913–940. © 2009 Springer Science + Business Media B.V.
914
Paula Braun and Leszek Fiedor
Summary Two complementary model systems are described, which are used to study the assembly of functional light-harvesting (LH) complexes. One system is based on rational design of cofactor-binding motifs and their capacity to assembly model LH2 complexes via expression in native-like membranes. The second takes advantage of the highly reversible self-assembly of the LH1 complex in artificial membranes and provides a convenient tool for design of model complexes with modified cofactors. In essence, re-design of the cofactor binding pockets in LH2 enables exploration of the underlying principles that enable particular amino acid combinations to sustain stable and functional assembly of LH-active arrays. Cofactor-binding motifs predicted in silico are tested in the context of the LH2 complex. In this way, H-bonding at the bacteriochlorophyll (BChl)/protein interface and the presence of aromatic residues were identified as critical for assembly of BChl and carotenoid (Crt). Moreover, the volumes of particular residues in the vicinity of BChl were shown to be critical for fine-tuning the spectroscopic properties. The LH1 reconstitution system, on the other hand, provides new information on the cofactor-related determinants of formation and functioning of this LH complex. Using the excitation trap approach, the coupling between BChl and excitation delocalization over the LH1 ring could be evaluated, while, by the replacement of Crts, their contribution to the assembly was assessed and for the first time a Crt-binding intermediate of LH1 assembly was identified. A new challenge is to make the two model approaches more interchangeable, thus allowing us to compare the same factors in different LH complexes, and eventually to identify on a molecular level what renders these apparently similar complexes so different. I. Introduction Most photosynthetic pigment-protein complexes comprise integral membrane proteins. The central structural elements of these proteins are the transmembrane α-helices (TMHs) which traverse the lipid bilayer and bind the photosynthetic cofactors, (B)Chl and Crts. In native photosynthetic membranes, such pigment-TMH assemblies are formed either by oligomerization of monotopic polypeptides, as in LH systems of purple photosynthetic bacteria, or by association of helices from polytopic helix-bundle polypeptides, as in most other LH and reaction center (RC) complexes. Due to the increasing numbers of high resolution structures, the availability of genomic sequencing data and numerous model studies, the structures of certain membrane proteins are known in considerable detail, many of them with nearly atomic resolution, and thus the understanding of membrane protein folding has advanced to a considerable extent (for recent reviews see Mackinnon and von Heijne, 2006; Sachs and Engelman, 2006; Walters and DeAbbreviations: (B)Chl – (bacterio)chlorophyll; AA – amino acid; CD – circular dichroism; Crt – ccarotenoid; ET – energy transfer; H-bond – hydrogen-bond; IR – infra-red; LDAO – lauryl dimethyl amine oxide; LH – light-harvesting; PS I – Photosystem I; PS II – Photosystem II; Rba. – Rhodobacter; RC – reaction center; Rsp. – Rhodospirillum; TM – transmembrane; TMH – transmembrane helix; WT – wild type; β-OG – n-octyl β-D-glucopyranoside
Grado, 2006). In addition, it is now generally accepted that interactions with the surrounding membrane lipids are compounded and often critical for the functional assembly of complex protein networks such as the photosynthetic machinery (for recent reviews see Pali et al., 2003; Fyfe and Jones, 2005; Sperotto et al., 2006). The assembly of the photosynthetic pigmentproteins into functional units requires the correct integration and organization not only of polypeptides but also of photoactive cofactors, the (B)Chl and Crt molecules. Their proper functioning depends on the proteins which non-covalently bind and spatially position these chromophores. Importantly, the electronic properties of the cofactors are markedly modulated via interactions with amino acid residues, which is crucial in mediating energy and electron transfer. In turn, both the (B)Chl and Crt pigments have been shown to be essential for the proper folding and assembly of photosynthetic proteins (Kim et al., 1994; Davis et al., 1995; Plumley and Schmidt, 1995; Croce et al., 1999; Horn and Paulsen, 2002). Therefore, any model of the organization of photosynthetic pigment-protein assemblies in a lipid bilayer has to account for significant structural contributions from both types of cofactor. In LH complexes, the structural function of (B)Chls relies in large part on the ability of the central metal ion (normally Mg2+) to coordinately interact with amino acid residues and this has long been recognized to be critical for the assembly of BChl-proteins
Chapter 46
Design of Light-Harvesting Complexes
(Coleman and Youvan, 1990; Olsen et al., 1997). Only recently has the importance of stereochemical aspects in (B)Chl ligation been realized (Balaban et al., 2002; Balaban, 2003; Kania and Fiedor, 2006). Usually, (B)Chls in photosynthetic complexes accommodate one extra ligand (= pentacoordination of the central metal) from the surrounding protein. Due to the presence of several chiral centers in these molecules, the two faces of their macrocycles are diastereotopic. Depending on the location of the ligand residue relative to the macrocycle, coordination is either of the α- or β-type. In (B)Chl-proteins, both the α- and β-type of ligation are realized but they are unevenly distributed; for example, in the plant Photosystem I (PS I) only 15 out of the 96 Chl molecules are ligated in the β-position whereas in the bacterial peripheral antenna (LH2), two out of three BChls are β-ligated. The exact role of the (B)Chl ligated in the β-position is not understood. In LH2-type complexes, histidine is strictly required as the residue ligating the strongly excitonically coupled BChl-B850 (Olsen et al., 1997), while the B800 BChls can be ligated by other residues. Also in the core antenna, LH1, where hexacoordination of BChls has recently been shown to occur (Fiedor, 2006), other residues may replace the histidine (Coleman and Youvan, 1990). The functional binding of (B)Chls requires yet additional types of interactions which involve the peripheral substituents of the macrocycle. For instance, in RC and LH complexes hydrogen (H)-bonding to the C3 acetyl group of BChl, significantly contributes to the tuning of chromophore properties (Fowler et al., 1994; Allen et al., 1996). H-bonding to the oxo groups of the isocyclic ring appears to be widespread but its influence on the chromophore redox and spectral properties is still disputed (Mattioli et al., 1991; Spiedel et al., 2002; Witt et al., 2002). Moreover, the impact of H-bonding at the BChl/protein interface on the assembly and stability of functional BChl-proteins remains to be clarified. The importance of Crts for both oxygenic and anoxygenic types of photosynthesis became obvious in early studies of Crt biosynthesis inhibition (Griffiths et al., 1955). Crts have several different functions in the photosynthetic reactions; a very important one is that of photoprotection against excess radiative energy (Young, 1993). Additionally, in the antenna complexes, Crts, having strong absorption in the visible, contribute to the LH process in spectral regions where the absorption of light by tetrapyrroles is neg-
915
ligible (Koyama et al., 1996). The efficiency of the energy transfer (ET) from the Crt to BChl molecules varies from 30 to nearly 90%, strongly depending on the number of conjugated double bonds in the Crt (Akahane et al., 2004; Polivka and Sundström, 2004; Chapter 12, Frank and Polívka). In spite of intensive structural and spectroscopic investigations, many questions as to the molecular mechanisms involved in Crt-BChl ET remain open (Koyama et al., 1996; Ritz et al., 2000; Papagiannakis et al., 2002) and there is a need for modeling LH complexes where the factors determining the rates of ETs can be systematically varied. The presence of Crts is critically required for the assembly of some of the pigment-protein complexes; their function in protein assembly is usually that of providing structural stabilization (Cohen-Bazire and Stanier, 1958; Lang and Hunter, 1994; Davis et al., 1995; Zurdo et al., 1995; Croce et al., 1999). In this context, the hydrophobic interactions of Crts with aromatic and other nonpolar amino acid residues were conventionally considered most important. Indeed, as indicated in the amino acid sequences of TMHs and shown by structural analysis, there is clustering of aromatic residues in the vicinity of Crt molecules in LH1 and LH2 (Wang and Hu, 2002; Leonov and Arkin, 2005). In addition, Crt-(B)Chl interactions (with possible mediation via phytol side chains) and π-π stacking between Crts and aromatic amino acid residues have to be taken into account as stabilizing factors (Wang and Hu, 2002). Particularly, experimental evidence for the role of aromatic residues is still wanting. The factors which drive the assembly of (B)Chl and polypeptides into unique arrangements have been addressed in numerous studies. Several experimental approaches have previously been taken to investigate the interplay between the proteins and pigments in Chl- and BChl-proteins. Principally, these approaches focused on (i) exchange of (B)Chls with chemically modified pigments in isolated pigment-protein complexes in vitro (Struck and Scheer, 1990; Struck et al., 1990a; Lapouge et al., 2000), (ii) chemical synthesis of de novo polypeptides or of truncated versions of natural ones followed by their reconstitution into complexes with pigments in vitro (Kehoe et al., 1998; Meadows et al., 1998; Kashiwada et al., 1999; Rau et al., 2001; Chapter 10, Loach and Parkes-Loach; Chapter 45, Haehnel et al.) (iii) mutagenesis and overexpression of the polypeptide, followed by in vitro reconstitution with the cofactors (Davis et al.,
916 1997; Todd et al., 1998; Bassi et al., 1999; Heinemann and Paulsen, 1999; Remelli et al., 1999; Chapter 10, Loach and Parkes-Loach) and (iv) site-directed or combinatorial mutagenesis combined with assembly of the complexes in vivo (Fowler et al., 1992; Hu et al., 1998; Olsen et al., 2003). Here, we describe the development of two new approaches to investigate the mechanisms and factors responsible for the assembly of fully functional LH complexes. One approach relies on the design and the in vivo expression of model TMHs. These model TMHs permit systematic studies of cofactor-protein interactions in a sequence context devoid of many of the natural interactions. Importantly, this system does not rely on reconstitution with pigments in a non-native, usually aqueous system, either by use of detergent or by rendering the helices water soluble. It is thus aimed at exploration of the determinants of cofactor binding and assembly of LH systems in their native membrane environment. The other approach is focused on the methods of LH1 reconstitution that not only allow complexes to be obtained with a full complement of cofactors but in particular provide a means to replace the native cofactors, either BChl or Crt, with their modified analogs. Its main goal is to construct model LH1 complexes which contain pigments modified at specific sites or in which the polypeptide sequence is systematically varied in order to investigate the effects of these modifications on the functioning of the complexes. II. Design of Model Light-Harvesting Proteins The use of a model amino acid (AA) sequence context as opposed to a native sequence context makes it possible to directly and selectively address the protein-related factors that contribute to the assembly of distinct and functional pigment-protein sites. Some of the underlying principles such as the binding and spectral adaptation of (B)Chl by the polypeptide have been clarified by classical mutagenesis in native complexes. This includes ligation of the central Mg ion of the BChl by nucleophilic amino acid residues and H-bonding between aromatic residues and the C3 acetyl group. Recently, it has been shown that there is a significant contribution of steric effects in the control of (B)Chl ligation in solution (Kania and Fiedor, 2006) but many of the determinants respon-
Paula Braun and Leszek Fiedor sible for recognition of (B)Chl, stable binding and spatial arrangement by the protein scaffold remain unraveled. Theoretical model studies of (B)Chlproteins should prove a useful approach towards the elucidation of these principles but so far this approach is quite limited as force field calculations of (B)Chl molecules have been up till now only partly successful (Linnanto and Korppi-Tommola, 2004). Even less is known concerning the binding and recognition of Crt molecules by the polypeptides. One promising approach, however, is the use of experimental model systems. Two main lines of mutagenesis procedure are generally taken to reveal the determinative patterns of cofactor-protein interaction. One is random sampling of mutations at the BChl/protein interface as, for example, in combinatorial mutagenesis. The other is introduction of directed mutations based on a rational approach, which was initiated through the use of bioinformatics to predict Crt- and BChlbinding motifs. A. Towards In silico Prediction of (Bacterio)chlorophyll and Carotenoid-binding Motifs Prior to experimental studies on the assembly of LH-active BChl and Crt cofactors in model pigmentproteins, the prediction of BChl- and Crt-protein interaction motifs can be attempted by a number of different strategies: firstly, statistical analysis of cofactor binding pockets in existing high resolution structures, in particular PS I (Jordan et al., 2001) and Photosystem II (PS II) (Loll et al., 2005) both of which bind nearly 160 Chl and 33 carotene molecules (Braun et al., 2003; Kwa et al., 2004); secondly, alignment of AA sequences of LH2 subunits; and, thirdly, computational analysis of the AA distribution in putative (B)Chl-binding pockets of ‘non-homologous’ (B)Chl-proteins as found in protein databases (Braun et al., 2003; Kwa et al., 2004; Garcia-Martin et al., 2006b). 1. Statistical Analyses of Chlorophyll-Binding Pockets Cofactor-protein contacts within a radius of ≤ 3.5–5 Å in the structures of PS I and PS II from Thermosynechoccocus elongatus were determined for each atom of the cofactor molecules. The determination of cofactor-protein contacts was performed by a selfmade Visual Basic® plug-in program for Microsoft
Chapter 46
Design of Light-Harvesting Complexes
Office Excel 2003, consisting of several functional modules. The PDB formatted files of PS I and PS II high resolution structures were used as sources for molecular data on an atomic level, which were copied into an internal, multi-dimensional numerical array. This allowed a fast, random access to any atomic position without the slower process of addressing single cells on the spreadsheet. A user interface generated on an Excel worksheet enabled the name of the residue to be entered, the acronyms of the first and last element of the searched molecule section, as well as a choice of contact distances in the range from d(min, max) = 0.5–4 nm. The calculation was performed by three obligate steps: 1) The PDB data file was accessed, analyzed for compatibility and written into the array. 2) The distance r(n,m) from atom m of the search frame with the spatial coordinate pxyz(m) to atom n of the array with the coordinate pxyz(n) was calculated using the following formula: r ( n, m) = ( px ( n) − px ( m))2 + ( p y ( n) − p y ( m))2 + ( pz ( n) − pz ( m))2
(1) and subsequently the contact criterion was tested by p d min ≤ r(n, m) ≤ d max f c(m, a) ⊕ 1 with a ∈ A)
(2)
917
For a ‘true’ condition, the counter c(m, a) for the consecutively numbered functional atom specification a out of the atom set A was incremented by one. 3) A suitable output was chosen for the data. The structural analyses of the photosystems showed that remarkably consistent patterns of interactions exist at Chl/protein interfaces. Particularly obvious are the patterns between the binding helices and the attached Chl in PS I (Fig. 1). The atoms of the Chl macrocycles are observed chiefly to interact with the AA residues at positions –4, +3 –1 and –7 in the binding helix relative to the liganding histidine at position 0 (Fig. 1). Most noticeably, the C132 oxygens are almost exclusively located in close vicinity of the residues at position –4 while the C32 atoms are primarily located close to the residues at +3, and to a lesser extent to the residues at –1, –4 and –7 (Fig. 1). In line with these findings, previous sequence alignment and mutation ‘hot spot’ studies of the protein α-subunits of the peripheral antenna of purple bacteria indicate that only two residues besides histidine at position 0 in the TMH are strictly conserved: one of these residues is located at position –4 and one at position +4 (Zuber and Cogdell, 1995; Braun et al., 2002). Thus, the interactions between the binding helix and the substituents of Chl macrocycle are generally limited to the residues which are located at the same helix interface as the residue which ligates
Fig. 1. Map of Chl/protein contacts. Contacts are shown at a distance ≤ 4 Å between Chl atoms and AA residues of the respective binding helices in Photosystem I (Fromme et al., 2001). Residues are presented according to their C-(+) or N-terminal (–) positions relative to the liganding histidine residues at position 0. The gray shading scale corresponds to the interaction frequencies. Note, the residues at position –4, +3 and –1 have the highest number of contacts to Chl.
Paula Braun and Leszek Fiedor
918
with juxta-positioned (B)Chl/helix units. In contrast, the C131 oxo of α-ligated Chls are significantly less involved in H-bonding interactions, and if so, then primarily with residues of loops and parallel helices (Table 1). The key role of such H-bonding motifs in the assembly of (B)Chl-proteins has been shown by mutational analyses in the model protein and native LH2 sequence contexts (see below). This suggests that H-bonding to the C131 keto group, in particular, of β-ligated (B)Chl may be one of the key structural motifs for the assembly of (B)Chl-proteins. In addition to the analysis of high resolution structures, the AA distribution in putative (B)Chl binding helices has been analyzed in search for (B)Chl-protein interaction motifs. This computational sequence analysis of (B)Chl-proteins stored in protein data banks is based on the following protocol: (i) assembly of the (B)Chl-protein data sets, (ii) prediction of TMH helices by use of the prediction servers TOPPRED2, HMMTOP2 and TMHMM2 (in case of deviations MEMSAT was used as well) according to the ‘consensus-principle’ (Nilsson et al., 2000), and (iii) development of a computer program which implements the double t-test and compares the expected frequency to the observed frequency for each position and residue. The frequencies of occurrence of each AA in the (B)Chl-binding sites of a set of BChl-binding proteins (2104 TM residues) have been determined. It showed that some AA residues are found with significantly higher or lower probability than expected from random distribution at the positions identified to be critical for Chl binding (Braun
the central Mg atom. Generally, the pockets for the Chl macrocycles include residues ranging from positions at ±8, hence, within two helical turns from the ligating residue towards either the C-terminal or N-terminal side. In addition, the analyses of the Chl binding pockets indicate that H-bonding at the Chl-protein interface is widespread: nearly half of the 100 Chls in PS I are likely to be H-bonded to the surrounding polypeptide, frequently through the Cl31 carbonyl group oxygen, but the oxygen atoms of other Chl substituents may also participate in H-bonding (Liebl et al., 1996; Braun et al., 2003). The groups participating in these bonds include backbone NH groups, heteroatoms of aromatic residues, and various sidechains: the amino group of lysine, the guanidinium of arginine, the amides of glutamine and asparagine, and the hydroxyl groups of serine and threonine residues. The exact H-bonding patterns between the Chl and apoproteins, however, depend on the topology of the Chl in its binding pocket (Garcia-Martin et al., 2006a). They are clearly distinct for Chls in the α- and β-ligated states (Balaban et al., 2002; Balaban, 2003). The latter frequently employ their C131 oxo groups in H-bonds to neighboring helices and subunits (Table 1) (Garcia-Martin et al., 2006a). This may be explained by the distinct spatial positioning of the C131 groups in the two different ligation states. In the α-type ligation, they face the binding helix and in the β-type ligation, they face away from the helix. Therefore, the C131 oxo groups of β-ligated Chls are more likely to be involved in tertiary interactions
Table 1. H-bonding between Chl and protein in plant photosystems as a function of the Chl ligation state. H-bonds and contacts of the C131 oxo of β- and α-ligated Chls are listed as percentage of total contacts. H-bonding interactions are identified by use of the graphics program (Weblab Viewer 3.7). H-bonding to C131 oxo β-ligated Chl PS I PS II H-bond Contact H-bond Contact 14 14 14 14
Neighboring subunitsa Neighboring helicesb c
Binding helices
d
Loops, and parallel Helices
α-ligated Chl PS I PS II H-bond Contact H-bond Contact 0 3 0 7
30
36
43
57
8
10
4
25
0
0
0
0
6
14
7
18
21
21
0
14
25
47
7
39
Neighboring pigments, H20 etc.e n.d. 29 n.d. 14 n.d. 26 n.d 7 a H-bonds between Chl molecules which are ligated by their central Mg to one subunit and residues of another subunit (e.g., Chl is ligated to subunit A, the C131 group makes a H-bond to a residue of subunit J); b H-bonds between Chl which are ligated to one TM-helix and the residues of neighboring TM-helices (other than the binding helices within one subunit); c H-bonds between Chl and residues of the binding helices; d H-bonds between Chl and residues from loop structures (including helices parallel to the membrane); e H-bonds between Chl and neighboring Chl or carotenoids (contacts to Chl are only included if no contacts to polypeptide are present).
Chapter 46
Design of Light-Harvesting Complexes
et al., 2003; Kwa et al., 2004; Garcia-Martin et al., 2006b). For example, at position –4, small residues such as alanine and aspartate are present with high probability. Additionally, the bulky residue phenylalanine is observed frequently at position –4. At position +3, large residues like leucine, tryptophan, histidine and even lysine are present with a probability higher than expected from a random distribution. In general, the results from the structural analysis of PS I are consistent with the sequence analysis of a large set of Chl-binding proteins: at numerous positions, there is good consensus between those residues which are shown to occur with elevated frequencies by the sequence analysis and those which are actually present in the Chl binding pockets of PS I. Combining the results from the structural and sequence analyses, preliminary predictions of Chl-binding sites were put forward (Garcia-Martin et al., 2005). These preliminary predictions of interaction motifs provide a starting point for the design of BChl-binding sites. (B)Chl docking, binding and functional arrangement by the model site may thus be explored within the native membrane environment.
919
2. Statistical Analyses of Carotenoid-Binding Pockets It is evident that aromatic residues cluster around Crt molecules in membrane proteins (Allen et al., 1988; Wang and Hu, 2002). PS I and PS II bind a total of 22 (Jordan et al., 2001) and 11 β-carotenes (Loll et al., 2005), respectively, and thus constitute a sizeable data base for statistical analyses of the Crt binding pockets. Although the structure of β-carotene, specifically the β-ionylidene rings at both ends of the polyene backbone, is different from the simpler Crts present in purple bacteria, it may serve as a general model for the generally very similar polyene chain and its interactions with the protein. Analyses of Crt pockets in PS I and PS II show that there are some patterns discernible in the frequency of contacts and the AA distribution around the Crt. Most contacts between the apoproteins and Crt occur at the side chain methyl groups of the polyene chains (Fig. 2). The central carbons, 14, 14´, 15, and 15´ and the carbon atoms of the terminal β-ionylidene rings are also frequently in close contact with the surrounding
Fig. 2. Carotene/protein contacts. Contacts between the AA residues and Crt atoms are shown at a radius of < 5 Å in high resolution structures of PS I and PS II (Jordan et al., 2001; Loll et al., 2005). (a) Schematic view of β-carotene. Nomenclature is according to IUPAC. (b) Number of contacts between Crt atoms and polypeptide residues. (c) AA distribution in Crt binding pockets. Gray scale codes for the residues are as indicated.
Paula Braun and Leszek Fiedor
920 protein environment. Three groups of residues are clearly overrepresented in the close vicinity of Crts: the aromatic residues, in particular phenylalanine; the aliphatic residues, in particular leucine, and short residues, in particular alanine. On average, about 25% of all contacts between the carotene and the polypeptide are to the phenylalanine residues, which at some carbons even amount to 40 % of total contacts. Only approximately 10% of the TM residues in membrane proteins are phenylalanine residues in any event (Wallin et al., 1997) and thus the interaction frequency between the Crts and phenylalanine residues is significantly higher than would be expected from a random distribution of phenylalanine residues in the photosynthetic apoproteins. Leucine residues (~20%) are also slightly overrepresented in the close vicinity of Crts (the occurrence of leucine in TM regions is estimated to amount to ~15%), but this is considerably less pronounced compared with phenylalanine. The aromatic residues tyrosine and tryptophan are primarily found in close vicinity of the carbon atoms of the ionylidene rings i.e., those which are likely to protrude into the membrane interface. The specific contribution of aromatic residues to the binding of Crt has been further explored in model and native LH proteins (see below). B. Model Light-Harvesting Proteins as a Tool to Study the Assembly of Light-Harvesting Systems For testing specific AA motifs at the cofactor/protein interface in a model sequence context, the peripheral LH complex, LH2, has been redesigned by simplifying its chromophore binding sites. An engineered strain of Rhodobacter (Rba.) sphaeroides (Jones et al., 1992) is used which lacks the operons encoding the apoproteins of the photosynthetic apparatus, but which provides the required assembly factors, pigments and membranes. The purple nonsulfur bacterium Rba. sphaeroides is genetically well accessible and a number of deletion strains lack one or all of the photosynthetic complexes. It should be noted that recently the presence of an additional copy of pucBA genes has been found in the genome of Rba. sphaeroides, which are translated into polypeptides (Zeng et al., 2003). While the contribution of such ‘new’ LH2 β-subunits is still unresolved, it has already clearly been shown that the ‘new’ α-subunits are not found in assembled LH2 complexes (Zeng et al., 2003). The possibility that unstable LH2 mutations
in the β-subunit may be partly stabilized by copies of the second polypeptide may thus not be excluded and is, at present, under investigation (C. N. Hunter, personal communication). The model proteins can be expressed either in the absence of other BChl-proteins or in the presence of alternatively LH1, or the RC-LH1-PufX core complex (Hunter, 1995). Importantly, due to the distinct spectral properties of the B800-B850 complex, assembly of the BChl-proteins can also be assayed by these properties directly in whole cells. In the model proteins, the membrane spanning portions of the LH2 subunits, which provide the protein scaffold for the light-harvesting active pigments, have been largely simplified by replacing the native sequences with alternating alanine-leucine (AL) stretches (Fig. 3) (Braun et al., 2002). Difficulties likely to arise during expression of heterologous or artificial proteins in situ have been circumvented initially by retaining the polypeptide domains protruding out of the membrane. Although they do not participate directly in pigment binding they do appear to be critical for targeting and insertion of the helices into the membrane and the overall assembly of the complex. 1. Light-Harvesting 2-like Complexes with Model Bacteriochlorophyll and Carotenoid Binding Sites In order to re-design the cofactor binding sites of LH2 complexes, the native sequences are replaced by AL repeat sequences. In the first design version, contiguous stretches of 16 and 12 residues have been replaced from Val(–1)† to Thr(+6) in the TMH of the LH2 α-subunit and from Gly(–7) to Ala(+4) in the TMH of LH2 β-subunit, respectively (Fig. 3). In wild type (WT) LH2, the stretches include all residues that are expected to interact with the BChl-B8501 of both of the α- and β-subunits at distances of <4.5 Å, except for αTrp(+9), αTyr(+13), αTyr(+14), βThr(+7) and βTrp(+9), which all lie outside the TMH (McDermott et al., 1995; Prince et al., 1997). To retain the ligation with the central metal of the BChl, His(0) was not replaced nor was the neighboring αIle(–l) residue, because of its proximity to the Mg-coordination site. Alanine and leucine residues already occur in †
The numbering specifies the amino acid position relative to the histidine, designated His (0), which binds the central Mg of the BChl-B8501. The number following BChl indicates the maximum absorption of the red-most absorption band of BChl.
Chapter 46
Design of Light-Harvesting Complexes
921
Fig. 3. LH2 complexes with alanine-leucine (AL) model sequences (a) AA sequences of the TM helices of the α- and β-subunits. The model sequences which replace the native sequences in αAL and β AL and His(0) which binds the central Mg are underlined. WT sequences are shown above the model sequences. The asterisk indicates the mutated residues. (b) Schematic view of the elementary subunit of model LH2 complexes with αAL20/S-4 and βAL12. The model AL is shown in light gray, the BChl are shown in gray, the native sequences and the Crt are shown in black. The dotted lines indicate the ligation to the Mg of BChl by His(0). For the purposes of clarity, only the polypeptide backbones are shown.
the WT sequence (Fig. 3), therefore there is a total of eight ‘new’ residues in the 16-residue stretch and five ‘new’ residues in the 12-residue-long stretch enclosing all TMH residues of the BChl-B850 site. The model LH2 complexes consist of assemblies of either chimeric α-subunits and WT β-subunits and vice versa or of both, chimeric α-subunits and chimeric β-subunits (Fig. 3). The model LH2 complexes in which the residues of the BChl-B850 binding site are replaced by alanine or leucine residues assemble to form LH2-like complexes in the membrane. α-AL16 plus a complementary WT β-subunit as well as β-AL12 with the complementary WT α-subunits (Fig. 3) support binding and assembly of BChl into structures with a geometry producing the red-shift characteristic of BChl-B850 and BChl-B800 (Braun et al., 2002; Braun et al., 2003; Kwa et al., 2004) . There are minor, but noticeable, red-shifts from 849 to 853 nm in the Qy-transition of α-AL16 which may indicate minor alterations in the BChl-B850 geometry. The structural information responsible for the shift in the absorption of the Qy band from 770 nm (‘free’ BChl) to 850 nm is, however, clearly retained if the stretches α-Val(–7) to Thr(+6) and β-Gly(–7) to Ala(+4) (Fig. 3) are replaced by the much simplified Ala-Leu sequence .
The BChl-B850 arrangement in LH2 with α- or β-AL seems to be largely preserved as judged by the circular dichroism (CD) spectra, which serve as fingerprints for the BChl geometries (Cogdell and Scheer, 1985; van Grondelle, 1985; Braun and Scherz, 1991; Koolhaas et al., 1998). Typically, their spectra show a conservative, S-shaped CD signal in the near infra-red (near-IR) with a positive peak near 848 nm, a negative peak near 872 nm and a negative trough near 800 nm, very similar to the one observed in WT LH2 (Bandilla et al., 1998; Georgakopoulou et al., 2002). The fitness of the model TMH to support a proper LH function of the complex is demonstrated by ET within the pigment assembly. The similarity of the excitation spectra with the absorption spectra shows that efficient ET takes place from BChl-B800 to BChl-B850. In particular, there is a pronounced 800 nm excitation band when detecting fluorescence emission at 900 nm. Moreover, the LH functions of Crts are also reproduced in the model complexes; the presence of characteristic bands in the range between 450 and 550 nm in the excitation spectra indicates that Crt-to-B850 ET occurs with high efficiency (Braun et al., 2002, 2003; Kwa et al., 2004). Thus, the model sequences support the assembly of pigments and protein to function as a LH unit. The residues in this region of the TMH, notably also at
922 the BChl-protein interface, do not seem to be critical for the functional specification of the BChl-B850 array. It should be noted that tuning of the BChl-B850 spectral properties (see below) is principally possible by altering the surrounding residues in the TMH (Hu et al., 1998). Perhaps, the minimal requirements for the assembly of functional BChl-B850 in LH2 are in large part fulfilled already by the histidine ligand, which is strictly required (Coleman and Youvan, 1990; Olsen et al., 1997), together with a few key residues, such as the aromatic residues anchored in the bilayer interface which have been shown to form H-bonds with BChl-B850 (Fowler et al., 1992; Fowler et al., 1994). In view of that, it may not be surprising that the additional extension of the alanine-leucine (AL) model sequence by three residues towards the N-terminus of the TMH in LH2-α (α-AL20/S-4, Fig. 3) does not impair BChl-B850 assembly. There is, however, significant loss of the Crt and the BChl-B800, indicating the importance of this region for Crt and B800pigment binding (see below). Notably, the extension of the AL sequence by only one residue towards the N-terminus of β-AL12 results in the loss of LH2 complex from the membrane. This illustrates that the residues of the β-subunits, in particular those outside of the BChl-B850 binding site, essentially contribute to the assembly of the LH complex contrary to the residues of the α-subunits. In summary, the native interactions at the BChl-B850/protein interface have been significantly altered and/or eliminated in the model LH2 αAL16 without significantly affecting LH function. 2. Assembly Motifs at the Bacteriochlorophyll/ Protein Interface To test the impact on BChl binding, residues at critical positions at the BChl/protein interface (see above) were systematically mutated in the model and native sequence contexts. In particular, the residues at the positions –4 and –1 which have been predicted to be frequently involved in contacts to the (B)Chl (Fig. 1) have been replaced by almost all 20 amino acids. It is shown by scanning permutagenesis that the stable assembly and particularly the spectral properties of LH2 are dependent on the side chain volume of the residue –4. In the model sequence context, LH2like complex assembly is possible only if alanine, serine or threonine is present at position –4 while all other residues tested at this position obstructed the assembly. In the WT-sequence context, only a few
Paula Braun and Leszek Fiedor of the bulky residues (phenylalanine, glutamic acid, tyrosine) obstruct LH2 assembly; however, nearly all residues, except for serine (as in WT) and the very similar alanine residue, introduce alterations in the spectral properties of BChl-B850. As a general trend, it is observed that the more voluminous the residue at position –4 the larger the blue-shift of the maximum absorption of the BChl-B850 red-most transition. Interestingly, the replacement of serine with glycine (no side chain) results in a minor red-shift of this transition. Thus, an exact fit of the residue at position –4 at the BChl/protein interface appears to be critical for maintaining the optimal BChl configuration. The mutations at the –1 position appear much less critical, as assembly of LH2 is retained with distinctively different residues at this position. However, both the stability and spectral properties of LH2 complex are also significantly affected by the type of residue at position –1 (Braun, unpublished). In this case, there is no correlation observed between the side chain volume and BChl-B850 spectral shifts. 3. Hydrogen-Bonding at the Bacteriochlorophyll/Protein Interface The effects of single residues at the BChl-protein interface are amplified in the model sequence context. Thus this system enables straightforward recognition of residues and motifs that contribute to the assembly of such systems, and may serve for extrapolation to cognate, native systems. The thermal stability of the LH2 model αAL16 complex appears, in spite of the functional retention, significantly reduced (Kwa et al., 2004). Simultaneous replacement of the native TMH of both the LH2 α- and β-subunits with the model AL (see Fig. 1) abolishes the assembly of LH2 in the membrane. To assess the contribution of previously predicted, potentially stabilizing, interaction motifs single residues are inserted at the model protein/BChl interface of αAL16/βAL12 to ‘rescue’ the assembly of LH2-like complex. This enables the identification and in-depth analysis of local BChlprotein interactions in the model sequence context. By use of this ‘rescue-mutagenesis’ approach we have thus identified H-bonding as a key assembly motif in the model BChl-proteins. The implementation of a strong H-bond at the BChl/TMHAL interface, most likely between the re-introduced serine at position –4 in α-AL(α-AL16/S-4) and the C131 oxo group of β-BChl-B850, has been confirmed by resonance Raman spectroscopy and protein
Chapter 46
Design of Light-Harvesting Complexes
923
Fig. 4. H-bonding drives assembly of LH2 complex with model BChl binding site. (a) ‘In situ’ near-IR absorption spectra of LH2 WT and of model LH2 with αAL16/βAL12 and with αAL16/S-4/βAL12. (b) Model of H-bonds at the BChl-B850/TMH interface in LH2 from Rba. sphaeroides. In the high resolution structure of Rps. acidophila residues 18–37 of the α-subunit have been replaced with residues 18–37 of the α-subunit of Rba. sphaeroides. Energy minimisation has been carried out on the replaced stretch and the BChl-B850 carbomethoxy groups (software WebLab ViewerProTM 3.7). The BChl-B850 molecules and α-serine (–4) are depicted in detail. The HO of serine and the O of the C-131, C133 and C173 carbonyl groups are shown as spheres. Putative H-bonds to the C-131 carbonyl group of β-BChl-B850 and the C-173 carbonyl group of the α-BChl-B850, are indicated by dotted lines.
modeling (Garcia-Martin et al., 2006a) (Fig. 4). This H-bond restores assembly of the model LH2 with both αAL16 and βAL12. The BChl arrangment in α-AL16/S-4 is very similar to that in WT LH2 as confirmed by absorption, CD, fluorescence excitation and resonance Raman spectroscopy (Kwa et al., 2004). In the model LH2 αAL16/S-4, re-insertion of serine at position –4 clearly elevated the structural stability of LH2 in the membrane. Accordingly, replacement of the serine by glycine results in significant destabilization of LH2 WT (Garcia-Martin et al., 2006a). Curiously, replacement of serine by alanine only slightly affects the thermal stability in WT LH2 (Braun et al., 2003). Furthermore, most LH2 α-subunits in related purple bacteria species possess an alanine residue at –4. One possible explanation for this apparent discrepancy may be that alanine (–4) favorably interacts with the 131 group either by weak H-bonding interactions with Cβ-H of the methyl side chain (Jiang and Lai, 2002) or by close packing interactions within the native sequence context. These interactions may be prevented in the model sequence context due to the slight reorganization of the pigments as reflected in minor red-shifts of the red-most transition (Fig. 3). Nevertheless, in the model sequence context, a single intra-membrane H-bond at the BChl/TMH interface
converts the highly unstable de novo BChl protein complex into a complex with almost native-like stability and expression levels in the membrane. Both BChl-B850 molecules within the protomer unit of LH2 (Chapter 8, Gabrielsen et al.) are β-ligated by the histidine residues of the juxta-positioned TMH of the α- and β-subunits. Mutation of the H-bond donor to the C131 carbonyl of one of the BChl-B850 molecules results in the significant impairment of LH2 assembly. This demonstrates the significance of this H-bond for the assembly of LH2 and further supports the notion that (B)Chl ligated in the β-position may have a pronounced structural impact on (B)Chlproteins (Garcia-Martin et al., 2006a). The extent to which the H-bond increases the inherent stability of the complex has not yet been determined. Using this system, however, the contribution of a single H-bond between a (B)Chl and its binding polypeptide can now be directly assessed and quantified within its native membrane environment. H-bonds have previously been identified in antenna and RC complexes (Fowler et al., 1994; Gall et al., 1997; Olsen et al., 1997; Sturgis and Robert, 1997). Hitherto, H-bonding between (B)Chl and a polypeptide has been reported to participate primarily in modulation of the spectral and redox potential properties of the (B)Chls. By use
924
Paula Braun and Leszek Fiedor
Fig. 5. Phenylalanine promotes binding of light-harvesting active Crt to LH2 complex. Fluorescence excitation spectra (a) of model LH2 with αAL20/S-4 (----), αAL20/S-4 βF(–8)G WT (—) and carotenoid-less LH2 from Rba. sphaeroides R26.1 (···), and (b) of double mutant LH2 αF(–12)L βF(–8)G (– · –) and LH2 WT (—).
of model and native LH complexes, the contribution of such intra-membrane H-bonding at BChl/protein interfaces to the stability and assembly of these complexes has been demonstrated. 4. Binding of Light-Harvesting-Active Carotenoid The mode of Crt binding to a polypeptide and the way in which this binding specifies its function is, as yet, largely unknown. Mutational studies of residues potentially contributing to Crt binding have been inconclusive. For example, systematic mutations of the aromatic residues surrounding the Crt in the bacterial RC did not dramatically affect the binding or the optical properties of these pigments (Vershinin, 1999). A phenylalanine residue has been shown to act as a gatekeeper of the Crt binding site and to determine the direction in which the Crt enters the site (Roszak et al., 2004). Statistical analysis of the plant photosystems shows that aromatic residues make up a significant part of the Crt binding pockets (see Fig. 2). By use of the model LH2 complex (Fig. 3) it is shown that phenylalanine residues in close vicinity of the Crt polyene backbone significantly affect the binding of Crt to the protein. Insertion/deletion of a phenylalanine near the centre of the alanine/leucine model helix, αAL20/S-4, modulates the binding of the functional Crts (Fig. 5). This effect is strongly enhanced by the removal of an additional phenylalanine located in the center of the β-TMH at position –8 of the WT β-subunit. Remarkably, the mutation of only two phenylalanine residues, αPhe(–12) and β-Phe(–8), in the WT sequence results in the loss of about 50% of functional Crt from the complex. This
indicates that π-π or hydrophobic interactions with aromatic AA residues critically contribute to the binding of Crt in these complexes. The assembly of LH2 and the binding of Crt in particular are significantly impaired by mutating the aromatic residues in their binding pockets. Resonance Raman studies isndicate that the conformations of the Crts are altered by the absence of these two phenylalanine residues, suggesting that the phenylalanine side chain is required to lock the Crt into a precise, well-defined geometry. Thus specific binding and functional assembly of Crt in model and WT LH2 complexes are reliant on the aromatic residue phenylalanine. The use of the LH complex as a model system thus provides evidence that the aromatic residue phenylalanine is a key factor for the binding of Crts to membrane proteins. III. Assembly of Functional LightHarvesting 1 Complexes A. Model Systems for the in vitro Assembly of Functional Light-Harvesting 1 Complexes The design and construction of model complexes in which their components can be independently manipulated is an alternative approach to the understanding of factors determining the assembly of functional photosynthetic pigment-protein complexes, as in other multicomponent biological systems. Exchange and reconstitution of cofactors has already proven to be very successful in elucidating structure-function relationships in photosynthetic RC complexes (Shuvalov and Duysens, 1986; Struck and Scheer, 1990b;
Chapter 46
Design of Light-Harvesting Complexes
925
Fig. 6. Absorption and circular dichroism (insets) spectra of isolated native (solid lines) and reconstituted (empty circles) LH1 complexes from Rba. sphaeroides (a) and Rsp. rubrum (b). Also absorption spectra (black dots) of the LH1 complex (Rba. sphaeroides) reconstituted with 21% Ni-BChl a and of the LH1 complex (Rsp. rubrum) reconstituted with spheroidene. The circular dichroism spectra of LH1 complexes from Rba. sphaeroides were recorded in the range between 350 and 950 nm and those from Rsp. rubrum in the region of carotenoid absorption.
Struck et al., 1990b; Scheer and Struck, 1993) as well as in antennae (Noguchi et al., 1990;Loach and Parkes-Loach, 1995; Davis et al., 1996; Frank et al., 1996; Desamero et al., 1998; Frank, 1999; Lapouge et al., 2000; Chapter 10, Loach and Parkes-Loach). However, there are not many examples of full reconstitution systems in which the antenna contains all its components, (B)Chl a, Crt and polypeptides, on a scale permitting a detailed characterization of the reconstituted complexes. Here we describe recent developments of experimental model systems in which cofactor exchange and its effects on the assembly and other properties of functional pigment-protein arrays can be studied in vitro. The model complexes are based on LH1 antennae from two species of purple photosynthetic bacteria, Rba. sphaeroides and Rhodospirillum (Rsp.) rubrum. In spite of many apparent similarities between these antennae, different approaches had to be developed to achieve reconstitution of functional LH1 complexes. 1. Rhodobacter sphaeroides The functional LH1 complex from Rba. sphaeroides was reconstituted from a crude methanol/chloroform (1:1) extract of freeze-dried LH1-only membranes
devoid of RC and LH2 complexes, obtained from the strain DD13[RKEK] of Rba. sphaeroides (Jones et al., 1992). The method is based on the high solubility of all vital components of LH1 complex, including the α and β polypeptides, in anhydrous organic solvents. This approach has already been used in characterizing these polypeptides from many species of purple bacteria (Theiler et al., 1984). Importantly, it was shown by 2-D nuclear magnetic resonance spectroscopy that these hydrophobic polypeptides not only are not denatured by organic media but even retain nearly native conformations in neat solvents (Conroy et al., 2000; Wang et al., 2005). Now this approach of using organic solvents can be applied to reconstitution of the LH1 complex with a full complement of cofactors. The reconstitution of the LH1 complex with the main absorption band at 875 nm occurs when the solvent extracted components of LH1 are solubilized in concentrated n-octyl β-D-glucopyranoside (β-OG; optimally 20% or 3.4%), followed by dilution to 1.5–1.1%. The treatment yields a complex containing both BChl a and Crts as judged from the absorption and CD spectra (Fig. 6). Generally, the reconstitution using low detergent concentrations (3.4%) results in higher yields of reconstituted complex (up to 40%).
Paula Braun and Leszek Fiedor
926 The reconstituted complex can be purified by ionexchange chromatography on a DEAE-Sepharose column, performed in the presence of β-OG. A very high ratio of the absorption intensities at 875 nm (Qy) to that at 275 nm (A875/A275), reaching 2.5, and only minor contributions by free BChl and B820 fractions, suggest that just one chromatographic step is sufficient to purify the complex.
X-100 in the case of spirilloxanthin). The electronic absorption and CD spectra of the purified, reconstituted complex are shown in Fig. 6. B. Reconstitution of the Light-Harvesting 1 Complex from B820 Subunits An alternative method of reconstruction of functional LH1 complexes from Rsp. rubrum has been developed recently, after the finding of a strong affinity of Crts to the B820 subunits (Fiedor and Scheer, 2005). Addition of Crt, as a concentrated solution in organic solvent (acetone or benzene), into the B820 system in 1% β-OG results in the immediate and quantitative formation of spectral forms with the Qy transition band shifted to 880–881 nm. The Crt absorption bands in these complexes also appear red-shifted with respect to the Crt absorption bands recorded in acetone or in 1% β-OG without LH1 proteins (Table 2). At a molar Crt-to-BChl a ratio of 1:1 applied during the reconstitution, the LH1 complexes produced from the B820 subunits are identical to the ones obtained in the LDAO-based system. However, the use of sub-stoichiometric amounts of Crts in the reconstitution system with B820 leads to the formation of intermediate complexes with lowered Crt content (see below).
2. Rhodospirillum rubrum In this case, successful reconstitution of functional LH1 relies on the use of Crt-depleted chromatophores, solubilized in lauryl dimethyl amine oxide (LDAO). The carotenoidless (‘blue’) chromatophores can be obtained either from wild type Rsp. rubrum via extensive extraction of the lyophilized photosynthetic membranes with benzene, or from carotenoidless mutants (Fiedor et al., 2004). A stable B880 complex is readily formed upon the addition of spirilloxanthin in acetone to the LDAO-solubilized membranes, provided very pure preparations of the Crt are used. Acetone has already been found useful as a carrier for most Crts in previous LH1 reconstitution studies (Davis et al., 1995). The characteristic red-shift of the Qy transition to 880 nm could be achieved only if Crt was present in the reconstitution mixture. The complex reconstitution is almost quantitative as judged by a disappearance of characteristic absorption bands of the carotenoidless subforms of LH1 (see below), and typically it is complete when the molar ratio of Crt to BChl a reaches 0.95:1. The resulting complex can be purified via ion-exchange chromatography on DEAE-cellulose in the presence of LDAO (Triton
C. Comparison of Reconstituted and Native Light-Harvesting 1 Complexes 1. Rhodobacter sphaeroides The B875 complex reconstituted from native com-
Table 2. The positions of the absorption maxima of chromophores in LH1 complex from Rsp. rubrum reconstituted with various C40 carotenoids. For comparison, the absorption maxima of isolated carotenoids in acetone are also presented. λmax [nm] Carotenoid (n) LH1 QY
Neu (9) 879
Sph (10) 882
Lyc (11) 879
Rhd (11) 881
Anv (12) 882
Spx (13) 883
Crt transitions
0→0 0→1 0→2 0→0 0→1 0→2 0→0 0→1 0→2 0→0 0→1 0→2 0→0 0→1 0→2 0→0 0→1 0→2
acetone
468
439
417
486
455
430
504
473
446
500
471
446
516
485
−
526
493
467
LH1
483
452
424
502
469
442
519
486
457
516
484
458
536
503
−
549
513
484
∆
15
13
7
16
14
12
15
13
11
16
13
12
20
18
−
23
20
17
Neu: neurosporene, Sph: spheroidene, Lyc: lycopene, Rhd: rhodopin, Anv: anhydrorhodovibrin, Spx: spirilloxanthin.
Chapter 46
Design of Light-Harvesting Complexes
927
Fig. 7. The effects of Ni-BChl a insertion into LH1 complex (from Rba. sphaeroides) on (a) the steady state emission of the complex in the near-IR region and (b) the dynamics of the ground state recovery, monitored at 870 nm.
ponents is spectroscopically almost indistinguishable from the one isolated from LH1-only membranes. In both complexes, the BChl near-IR absorption band (Qy) is red-shifted to 876 nm, and the Crts have characteristic broad, structured absorption bands located between 450 and 550 nm, with a maximum at 500 nm (Fig. 6a). The complexes also show Crt-BChl ET, as well as similar emission maxima and fluorescence yields. The CD spectra of the preparations are closely similar and have the features typical of LH1 complexes (Bolt et al., 1981), i.e., an intense S-shaped signal in the Crt range (400–600 nm) and a relatively weak, W-shaped signal in the Qy region, between 850 and 900 nm (Fig. 6a). Most importantly, the two complexes are functionally identical, as indicated by the efficiency of the Crt to BChl ET and the excited state lifetime. If excited at 500 nm, both complexes show an emission band in the near-IR, with the maximum at 897 nm (Fig. 7a). The peak positions and emission intensities are identical to the values obtained for the LH1-only membranes or native LH1 (not shown), indicating a complete reconstitution of ET from Crt to the B875 BChls. Also, the two complexes do not differ in terms of size distribution as ascertained by gel filtration and sucrose density centrifugation (Fiedor et al., 2001). The pigment composition of the isolated and reconstituted complexes has been analyzed by normalphase high performance liquid chromatography. Both the native LH1 complex and the complexes reconstituted with the native, non-modified pigments contain
two major Crts and BChl at similar ratios. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the native and reconstituted LH1 complexes confirms the presence of only two small polypeptides (α, β) in the preparations (Fiedor et al., 2001). Taken together, these results suggest the same cofactor arrangement and interactions are the same within the reconstituted and the native complexes. 2. Rhodospirillum rubrum The spectroscopic properties of LH1 reconstituted with spirilloxanthin compare very well with those of the isolated native complex (spirilloxanthin content >95%). The absorption spectrum of the reconstituted LH1 complex closely matches the spectrum of the native one, in particular in terms of reproducing the red-shifted positions of the BChl Qy transition and the Crt absorption bands (Fig. 6b). Specific binding of spirilloxanthin in the complex is further confirmed by CD spectroscopy. A broad negative band in the region of spirilloxanthin absorption (Fig. 6b) indicates specific interactions of the Crt and the chiral pigment-protein environment (Davis et al., 1995). The native antenna shows somewhat weaker CD signals in that region (Fig. 6b), perhaps due to lipid molecules remaining bound to the antenna, thereby reducing the contact of Crts with the antenna polypeptides (Cogdell and Scheer, 1985). The LH1 complex reconstituted with spirilloxanthin is stable and can be isolated by ion exchange chromatography in the presence of
928 Triton X-100. The protein analysis of this complex by sodium dodecyl sulfate-polyacrylamide gel electrophoresis confirmed the presence of only two polypeptides of low molecular masses, identical to those found in the isolated antenna (Fiedor et al., 2001). This comparison shows that the pigment-pigment and pigment-polypeptide interactions are correctly reproduced within the reconstituted complex. D. Introduction of Ni-Bacteriochlorophyll a into the Light-Harvesting 1 Complex as the Excitation Trap The possibility of BChl exchange in LH1 has been shown in a series of reconstitution experiments using BChl analogs modified at the peripheral positions of the macrocycle, and with some metallo-substituted BChls (Loach and Parkes-Loach, 1995; Davis et al., 1996). These studies, however, focused on carotenoidless complexes. The present method of LH1 reconstitution from the dried methanol/chloroform extract of LH1-only membranes provides a very convenient procedure for the introduction of modified BChl a into LH1 comprising all cofactors. The replacement of the native BChl a with Ni-BChl a can be achieved by addition of the modified pigment to the LH1 reconstitution mixture at the stage of the chloroform/methanol extract, followed by solvent removal and LH1 reformation in β-OG. High performance liquid chromatography analyses of the samples showed, besides the original Crts and BChl a, only one additional pigment, which was identified as NiBChl a. The content of Ni-BChl a in the reconstituted complex increases with the amount added externally, but it is consistently higher than its proportion in the reconstitution mixture. At low Ni-BChl a concentrations, the enrichment factor reaches 3.2, while at the highest concentrations used for reconstitution (30% Ni-BChl a), it amounts to 1.9. A similar preferential binding of Ni-BChl a versus native pigment has previously been seen in the reconstituted LH1 from Rsp. rubrum (Näveke et al., 1995). The enrichment of LH1 in Ni-BChl a is very likely due to the higher strength of ligand binding to Ni2+ in comparison to Mg2+, and perhaps due to additional stabilization of the low-spin d8 complexes (Douglas, McDaniel, and Alexander, 1994). The incorporation of Ni-BChl a into LH1 influences the ground state absorption (Fig. 6a) and CD properties (not shown) of the reconstituted LH1 complex only slightly. The Qx band of BChl a in the reconstituted LH1 does not show any shift (to the
Paula Braun and Leszek Fiedor red) upon Ni-BChl incorporation, being indicative of a pentacoordinated state of the central Ni2+ ion in the complex-bound pigment (Hartwich et al., 1998). 1. Pigment Coupling and the Size of the Light-Harvesting 1 Complex In contrast to the minor changes in electronic absorption, the incorporation of Ni-BChl a into LH1 has dramatic effects on the emission properties of the complex. The steady state emission spectra, obtained for a series of LH1 complexes containing increasing amounts of incorporated Ni-BChl a, are shown in Fig. 7. Qualitatively the spectra are identical, showing a single band centered at 897 nm. However, the emission intensities decrease sharply in the presence of even small amounts of Ni-BChl a. The incorporation of only 3.2% Ni-BChl a causes a 50% decrease in fluorescence emission, while at Ni-BChl a contents above 20% the B875 emission is practically completely quenched. The same quenching efficiencies of incorporated Ni-BChl a on LH1 emission have been observed upon excitation in the Soret region (Fiedor et al., 2001). Obviously, Ni-BChl a incorporated into LH1 acts as an extremely efficient deactivation channel (Hartwich et al., 1998; Musewald et al., 1999). The strong quenching effect of the incorporated Ni-BChl a on LH1 emission is confirmed by time-resolved emission measurements in the sub-nanosecond domain. The non-modified (native) LH1 complex shows a mono-exponential emission decay with a characteristic lifetime of 850±100 ps (Fiedor et al., 2000). In the presence of Ni-BChl a, the decay profiles can be fitted well mono-exponentially with practically the same time constant, regardless of the concentration of the excitation trap (Ni-BChl a). Importantly, the incorporation of Ni-BChl a into LH1 affects only the amplitudes of the emission signal, quantitatively in the same way as the reduction of fluorescence yield. These observations can be rationalized by a two-component system, comprising only (i) a subpopulation of complexes with native fluorescence quantum yield and a decay-time of about 850 ps, and (ii) a second sub-population of complexes, which are practically non-fluorescent on the time-scale of several tens of picoseconds. A very similar effect on the fluorescence was observed when another type of excitation trap, a chemically induced BChl a cation radical, was introduced into LH1 from Rhodobium marinum (Law and Cogdell, 1998). The pigment unit size in reconstituted LH1 sub-
Chapter 46
Design of Light-Harvesting Complexes
stituted with increasing amounts of Ni-BChl a can be determined via statistical analysis of the emission quenching data. The simulations, based on Poisson distribution of the trap and on the two-component model, resulted in a unit size of 20 pigment molecules. This value differs from the range of sizes (24–32) obtained by pigment analyses of core complexes and by electron microscopy and X-ray crystallography of a variety of LH1 complexes (Loach and Sekura, 1968; Loach et al., 1970; Stark et al., 1984; Ueda et al., 1985; Dawkins et al., 1988; Francke and Amesz, 1995; Karrasch et al., 1995; Akiyama et al., 1999; Jungas et al., 1999; Roszak et al., 2003). An even smaller number, i.e., 16 BChl per LH1 unit, has been obtained by numerical modeling of annihilation process in the isolated LH1 antenna, observed in ultrafast fluorescence depolarization studies (Bradforth et al., 1995). Even if one assumes losses as high as 10% of BChl a during the analysis of pigment composition, the upper size limit would just be brought close to the lowest number of 24 BChls per unit obtained from the biochemical studies. There are no indications that the presence of the modified pigment could be responsible for the unusually low aggregation number, nor that it results in a more heterogeneous size distribution. However, the reconstituted LH1 may be different in size from the freshly isolated complex and in particular from the one in native membrane environment, due, for example, to the loss of certain lipids or the PufX protein, while the spectral and functional features of LH1 are not affected by this difference. Another factor, which may appreciably influence the size of LH1 ring, is the presence/absence of the RC, as revealed in a recent study showing flexibility of LH1 rings (Bahatyrova et al., 2004). Therefore, the actual size of the complex may be determined by the procedure of LH1 isolation from the photosynthetic membranes, i.e., due to the inherent exposure to detergents, which may cause formation of closed rings of reduced sizes. 2. Exciton Delocalization in the LightHarvesting 1 Complex The introduction of Ni-BChl a into LH1 as an ultrafast excitation trap provided, for the first time, a possibility of a direct experimental verification of the mechanisms of the ultrafast excitation relaxation and intracomplex excitation transfer. This has been addressed in a study using pump-probe absorption spectroscopy with ~35 fs time resolution (Fiedor
929
et al., 2000). Incorporation of Ni-BChl a into LH1 results in the tremendous shortening of the ground state recovery time (Fig. 7b). The major conclusions from the femtosecond absorption measurements were that trapping occurs on a timescale comparable to or even shorter than the resolution of the apparatus (ground state recovery in 60 fs), and that it occurs already in the non-relaxed, non-fluorescent state of the antenna. Moreover, the statistical analysis shows that just one Ni-BChl a per 20 BChls in the ring is sufficient for complete excitation deactivation via the 60 fs channel. Obviously, this excludes localized excitation, monomeric as well as dimeric, and subsequent Förster hopping in LH1. These findings suggest that even on the subpicosecond time scale the excitation is likely to be delocalized over >10 BChl molecules, and presumably even over the entire LH1 unit of 20 BChls of the antenna (Fiedor et al., 2001). E. The Role of Carotenoids in the LightHarvesting 1 Complex 1. Light-Harvesting 1 Complexes with Modified Carotenoids The reconstitution assay, based on LDAO-solubilized LH1 from Rsp. rubrum, allows for an easy construction of LH1 complexes with non-native Crts. In addition to spirilloxanthin, a number of different Crts, with different functional groups and conjugation length varying from 9 to 12, were reconstituted with almost quantitative yields into the LH1 complex. The reconstituted complexes were purified by the same method, i.e., ion exchange chromatography on DEAE-cellulose, in the presence of detergent LDAO, or Triton X-100 (0.03–0.04%). The absorption spectra of complexes reconstituted with different Crts are very similar in the region of BChl a absorption, closely resembling the spectrum of the native LH1 antenna. The spectrum of LH1 reconstituted with spheroidene is shown, as an example, in Fig. 6b; the positions of the relevant absorption bands in all reconstituted complexes are listed in Table 2. In particular, the characteristic red-shift of the Qy transition is reproduced but its exact position slightly varies (by 3–4 nm), depending on the Crt. The red-most Qy positions at 883 nm are present in the complexes reconstituted with spirilloxanthin, anhydrorhodovibrin, rhodopin and spheroidene, whereas in the complexes reconstituted with lycopene and neurosporene, the Qy maximum never shifts above 879–880 nm.
930 The increase in length of the conjugation system in the Crts, resulting in narrowing of the energetic gap between the S0 state and the S3 state (new notation, Christensen, 1999; Chapter 12, Frank and Polívka), is reflected in the spectra of the model LH1 complexes in the region between 400 and 550 nm (Fig. 6, Table 2). In the complexes reconstituted with neurosporene and spheroidene (conjugation size 9 and 10, respectively), the Crt absorption bands are the most blue-shifted and therefore the BChl Qx transition can be seen as a separate band (Fig. 6b). These model LH1 complexes have shown their usefulness in the analysis of the coordination state of BChl a in LH1, which revealed in LH1 a subpopulation of BChl a binding two axial ligands (Fiedor, 2006). Absorption of light by the reconstituted complexes in the visible range is in large part determined by the type of Crt bound (see Chapter 6, Takaichi, for a review of Crt biosynthesis). Therefore, their colors closely resemble the coloration of the strains of photosynthetic bacteria, which accumulate the corresponding pigments as their major Crts, namely the G1c mutant of Rba. sphaeroides (neurosporene), wild type Rba. sphaeroides 2.4.1 (spheroidene), Rsp.
Paula Braun and Leszek Fiedor molischianum (lycopene), Rhodobium marinum (anhydrorhodovibrin) and Rsp. rubrum S1 (spirilloxanthin), respectively. 2. Effects of Carotenoids on Assembly of the Light-Harvesting 1 Complex The presence of Crts in the reconstitution system provides a convenient intrinsic probe for direct spectroscopic monitoring of assembly of the complex. The sequence of events leading to LH1 formation with increasing amounts of rhodopin was monitored by absorption spectroscopy (Fig. 8a). In the absence of Crt, the detergent (0.033% LDAO) induces a partial dissociation of LH1 and all forms of the complex (B780, B820 and B870) are present in equilibrium. With the introduction of Crt, the equilibrium between the subforms shifts towards the Crt-binding B880 complex, as indicated by the shift of the Qy band, reaching 880–882 nm at the point of saturation with Crt. It is noteworthy that, even with only a small amount of added Crt, when all subforms of LH1 are still present the Crt absorption bands already appear maximally red-shifted to almost the same positions as
Fig. 8. Effects of carotenoids on the formation of LH1 complex from Rsp. rubrum: changes in the absorption spectra of the carotenoidless reconstitution mixture (a) and of isolated B820 subunits (b) observed during the treatment with rhodopin, added as acetone solution, and (c) changes in the emission spectra during LH1 formation induced by spheroidene (left: excitation at 590 nm, right: excitation at 500 nm).
Chapter 46
Design of Light-Harvesting Complexes
in the final reconstitution stage (Fig. 8a). Crts exert practically the same effects on isolated LH1 subunits, as can be seen from the changes in the absorption spectra of the reconstitution system induced by Crt on the B820 subunits in 1% β-OG (Fig. 8b). In both reconstitution systems, a significant red-shift of the BChl a Qy transition indicates that the interactions with Crts induce a considerable change in the aggregation state of the subunits. As all Crts studied cause similar effects, irrespective of their rather varied structures (conjugation length and side groups), these interactions are not restricted to a specific type of Crt. Additionally, the red-shifts of the Crt absorption transitions are further independent indicators of the Crt-LH1 subunit interactions (Table 2). If the red-shift of Crt absorption is taken as a measure of their interactions with the environment (= αβ-BChl units), these interactions do not change throughout formation of the LH1 complex (Fig. 8) and remain similarly strong in the reconstitution mixture (solution) and in the isolated complex. The monitoring of the process by emission spectroscopy supports this conclusion. From the early stages of the Crt-induced complex formation, ET from Crt to BChl a can be observed which, however, occurs exclusively to the B880 BChls (Fig. 8c). This absence of other fluorescent forms upon excitation at 500 nm implies the appearance of only a single type of Crt-binding species under the conditions of the reconstitution, and it also provides a good verification of the assembly process. Seemingly, as soon as the Crt is added to the LH1 subunit (in a mixture or in a single form), the pigment molecules strongly interact with the LH1 subforms and cause their rapid aggregation to the B880 state. Consequently, the B780-B820-B870 bands seen in the absorption spectra and the fluorescence maxima, appearing upon excitation at 590 nm (Fig. 8), can be ascribed to the carotenoidless forms. The presence of Crts markedly shifts the equilibrium between the LH1 subunits towards B880 formation, implying that Crts have a high affinity for their binding sites and are involved in strong interactions from the very moment of being added in the reconstitution system. However, they do not distribute evenly among the available binding sites because the far-red emitting B880 complex is formed even when the ratio of the Crt to the carotenoidless αβ-BChl a subunits is still very low. The binding constant for formation of the Crt-to-αβ-BChl a unit as well as the self-aggregation constant for this initially formed unit must be
931
relatively high, pointing to a cooperative mechanism of LH1 formation induced by the Crt. The presence of Crts affects the properties of LH1 complexes on several levels. The complex formation itself, triggered by Crt added to the mixture of the antenna subunits, is the earliest indication of the Crt effect. The assembly occurs under conditions which otherwise favor a complete dissociation of the complex into the subunits, as observed previously in a similar LH1 reconstitution system (Davis et al., 1995). Another such indication is the enhanced resistance of the Crt-containing LH1 to organic media, to as much as 20% aqueous acetone at the final stage of reconstitution. However, the comparison of the relative amounts of Crt required for the formation of stable B880 and also varying stabilities of the complexes reconstituted with different Crts (reflected, for example, in the chromatographic conditions) indicates that the stabilizing effect is distinct for each individual Crt. Based on their contribution to the assembly of LH1, the six Crts used in the present study, can be divided into two groups; lycopene and neurosporene (plain hydrocarbons) give less stable complexes and thus higher concentrations are required to form B880, while spirilloxanthin, anhydrorhodovibrin, rhodopin and spheroidene form species of higher stability. Also, the latter four pigments show higher affinity for the BChl-αβ subunits, apparently due to the presence of the hydroxy (rhodopin) and methoxy side groups (anhydrorhodovibrin, spheroidene, spirilloxanthin). This observation indicates that, in addition to hydrophobic interactions (Sturgis and Robert, 1994) and more specific π-π stacking interactions (Wang et al., 2002), interactions of other types (possibly electrostatic interactions, for example) also play a role in binding of Crt to LH1 and in stabilization of the complex. Recently, the forces which stabilize B820 subunits were characterized as originating to a large extent from hydrophobic interactions between N-terminal extensions of the α and β polypeptides at the membrane interface (Parkes-Loach et al., 2004). However, the stabilizing energy cannot be large, as the B820 subunits retain the ability to interact and bind free Crts. Two modes may be envisioned for the binding of Crts to B820 subunits: (1) Crt molecules bind between some of the hydrophobic surfaces (TMHs of the α and β polypeptides) of adjacent B820 subunits, thus ‘cementing’ the structure. As in other photosynthetic complexes, here the interactions of Crt with the long-
932
Paula Braun and Leszek Fiedor
Fig. 9. Absorption spectra (a) of the native LH1 antenna (dotted line) from Rsp. rubrum and the iB873 complex (solid line), formed from isolated B820 subunits in the presence of sub-stoichiometric amounts of spirilloxanthin; (b) spectral changes occurring during the titration of iB873 with rhodopin; (c) the excitation (left) and emission (right) spectra, and (d) circular dichroism spectra of the native LH1 antenna (dotted line) and the intermediate iB873 complex (solid line) from Rsp. rubrum. The emission spectra were normalized to match the maximum intensities; the excitation spectra were normalized to the intensity of the Qx bands.
chain alcohol residue of BChl may also play some role; and/or (2) interactions of Crt molecules with α–helical stretches may cause some rearrangement of their N-terminal regions, which then enforce intersubunit interactions. 3. Identification of an Intermediate of Assembly of the Light-Harvesting 1 Complex B820 is, at least in vitro, an intermediate in the assembly of carotenoid-less LH1 (B870) from its components, namely BChl and the α and β-polypeptides, and the subunit seems indeed devoid of Crts, even if derived from Crt-containing bacteria (Moskalenko et al., 1996; Loach and Parkes-Loach, 1995). The aggregation states and association of B820 have been studied in considerable detail. Depending on the detergent:protein ratio, dimers (Arluison et al., 2002; Vegh and Robert, 2002) and higher aggregates (Pandit et al., 2003) have been observed as intermediates during the B870 formation, which showed relatively small red-shifts of the Qy transitions (845–850 nm).
However, no such intermediates have been observed in the few LH1 reconstitution studies carried out in the presence of Crts (Davis et al., 1995; Fiedor et al., 2001; Fiedor et al., 2004). The formation of a Crt-containing B880 complex by solely adding Crt to B820, without changing the detergent concentration, provides clear evidence for interactions between B820 and Crts. If, however, Crts were added to B820 in sub-stoichiometric amounts the formation of new types of complexes can be observed. The characteristic features of these complexes, named by us as iB873, are a low absorption in the Crt region and a slightly blue-shifted (873 nm) position of its Qy absorption maximum (Fig. 9a). While complexes of this type were spectroscopically detectable with several Crts, only the new complex formed with spirilloxanthin is sufficiently stable to be isolated using ion exchange chromatography. In addition to the stabilizing effects of Crts in LH1 complexes discussed above, the increased stability of iB873 formed with spirilloxanthin in particular may be due to specific interactions (e.g., hydrogen bonds) of the methoxy
Chapter 46
Design of Light-Harvesting Complexes
side groups of the Crt with the polar residues of the α and β polypeptides. The spirilloxanthin:BChl a ratio in the iB873 complex is only 40% of that in the native LH1, where the Crt content is 2 BChls:1 Crt, or 1 Crt per protomer (Picorel et al., 1983). After purification, it can be converted by adding Crt to a species with a Qy absorption maximum located at 882 nm (Fig. 9b), characteristic of the fully assembled LH1 complex from Rsp. rubrum. The iB873 complex is the first complex binding functional Crt with absorption properties intermediate between those of B820 and of fully assembled LH1, formed by B820 subunits and Crts. It has the characteristics of an intermediate species according to several criteria: (i) an intermediate position of its Qy transition between those of B820 and B880, (ii) an intermediate Crt content, and (iii) the conversion to B880. This is the first observation of an intermediate in the assembly of an LH1 antenna in the presence of Crt. Emission and CD measurements clearly indicate a close interaction of Crt with the other components, BChl a and the polypeptides, within iB873. Following the excitation of the complex into the spirilloxanthin absorption at 505 nm, a typical LH1 fluorescence is observed in the near-IR region (Fig. 9c). The position of the emission band maximum near 900 nm confirms the ET from spirilloxanthin to the iB873 BChls, and implies close contacts of Crt molecules with the BChl complement. Characteristically for spirilloxanthin (n = 13) in LH1 antennas (Noguchi et al., 1990), Crt-toB873 ET in iB873 is not particularly efficient (~30%); yet it is clearly visible in the excitation spectrum (Fig. 9c, left). Nevertheless, taking into account the reduced Crt content in iB873, the spirilloxanthin-to-BChl a ET efficiency in this complex is quite similar to that of the native LH1 (Noguchi et al., 1990; Davis et al., 1995). Strong CD signals, seen in iB873, differ from those of Crt in native LH1 not only by their sign, but also by the structure and positions of the maxima (Fig. 9d). Usually, in LH1 and LH2 complexes there is a non-coincidence of absorption and CD bands (Cogdell and Scheer, 1985). Excitonic interactions among Crts could produce such shifts, but the occurrence of direct Crt-Crt interactions has hitherto been demonstrated conclusively only in a single case, the peridinin-chlorophyll protein (Hofmann et al., 1996; Pilch and Pawlikowski, 1998). This seems not to be the case with iB873 because its CD spectrum shows quite well defined maxima in the Crt region, whose positions coincide with band progression in the absorption (Fig. 9). This points to the monomeric
933
character of Crt in iB873, with no significant interCrt interactions. Such interactions in iB873 are also unlikely from a Poisson analysis of the occupancies of the Crt binding sites in this complex, which were modeled assuming that the Crt molecules distribute randomly within the iB873 ensemble. A simulation of a 40% saturation of Crt in iB873 (assuming one Crt binding site per protomer) reveals that as much as 93% of the bound Crt molecules are likely not to have another Crt in their immediate neighborhood but rather an empty Crt binding site. Hence, almost all Crt molecules retain a monomeric character and, as shown by the efficient ET, they remain strongly coupled to BChls. If no further rearrangements take place during the conversion of iB873 to B880, the inversion of the CD signal of spirilloxanthin in the native LH1 antenna from Rsp. rubrum as well as the non-coincidence of CD and absorption maxima, result from Crt-Crt interactions rather than from Crt-BChl a interactions in LH1. F. Model for Assembly of the LightHarvesting 1 Complex in vitro The finding of an LH1 assembly intermediate and of cooperative formation of LH1 in the presence of Crts, combined with previously obtained data, can be summarized in a model of the assembly for this antenna complex. In the absence of RCs the stages of the assembly, shown schematically in Scheme 1, comply with the extended two-stage model of membrane protein folding, proposed by Popot and Engelman (1990). Integration of the role of the RC in the assembly of the core complex poses still a challenge. 1. Stage 1: Formation of Monomeric Subunits B780 LH1 α and β polypeptides spontaneously fold into transmembrane α-helices across the photosynthetic membrane (Drews, 1996). This is followed by binding of prosthetic groups as newly biosynthesized BChl a molecules are quickly coordinated to both polypeptides to form B780 monomeric subunits (Fig. 10). The driving force for this process is provided by strong interactions of His residues with the central Mg atoms of BChl molecules (Olsen et al., 1997). In addition to the coordination, BChl seems to be involved in other type of interactions with the apoprotein, probably of hydrophobic character, as
Paula Braun and Leszek Fiedor
934
Fig. 10. Stages of LH1 formation in the presence of carotenoids in vitro. See also Color Plate 16, Fig. 27.
indicated by the shift of the Qy band from 770 to 780 nm. The formation of B780 seems equivalent to the stage 1 of the two-stage model. 2. Stage 2: Formation of Dimeric/Tetrameric Subunits B820 Following assembly of monomeric B780 subunits, giving rise to dimeric (or tetrameric) B820 forms, the antenna folding proceeds to the early phase of stage 2, i.e., to the formation of higher order structures (Fig. 10). Most probably, the dimerization is caused in part by pigment-pigment interactions (π-π stacking) and in part due to pigment-polypeptide and polypeptide-polypeptide hydrophobic interactions in which the phytyl moiety would be expected to be involved (Freer et al., 1996). The proper folding of B820 is secured by complementarity of α and β polypeptide structures whereas the coordination state of BChl a does not change. The characteristic red-shift of the BChl Qy band originates from excitonic coupling occurring within the αβ(BChl a)2-heterodimers (Scherz and Parson, 1984; Chang et al., 1990). The tetrameric form of B820, with an almost parallel arrangement of four TMHs, comprises a four-helix bundle, a motif commonly found in the structure of membrane proteins. The Crts seem not to be involved at this stage of the assembly but, as shown above, they are definitely able to interact with both the B780 and B820 subunits and promptly induce stage 3.
3. Stage 3: Oligomerization of Subunits to Larger Assemblies The interactions of the B780 and B820 subunits with Crts provide a considerable driving force leading to the cooperative formation of larger assemblies. The association-driving interactions must be mainly of hydrophobic character because Crts, which are plain hydrocarbons (neurosporene and lycopene), also induce the oligomerization. Initially, stable intermediates with sub-stoichiometric Crt content, such as the iB873 complex, can be formed. The spectral shift of the Qy transition to 870 nm is partly due to the excitonic coupling between strongly overlapping BChl a molecules (or B820 subunits) within the B870 array. Also the interactions of BChl a with the apoprotein contribute somewhat to the shift. It is likely that the sizes of the B870 assemblies with similar spectral properties vary over a broad range, especially in the absence of the RC, as shown directly by atomic force microscopy (Bahatyrova et al., 2004). The formation of a full B870 ring terminates the folding of this membrane pigment-protein complex (Fig. 10). 4. Stage 4: Assembly of Complete Antenna B880 The intermediate forms (B870, iB873) are converted to the fully assembled antenna with a full complement of B880 BChl a and Crt molecules (Fig. 10). The shift from 870 to 880 nm is clearly correlated
Chapter 46
Design of Light-Harvesting Complexes
with the binding of Crt molecules into the complex. Crts stabilize the entire assembly and enhance the excitonic coupling between BChl a molecules in the B880 ring. Very likely, there is some excitonic coupling between Crts and BChls. The formation of the RC-LH1 complex may follow an assembly pathway that differs from that of LH1 alone; obviously, the core antenna must entirely embrace the RC complex, and surely, in native membranes, lipids are involved in the stabilization of the assembly. Therefore, formation of the core unit must be synchronized with biosynthesis of its polypeptides and the cofactors, and the oligomerization of these components might be induced by the RC itself. In addition, in some cases the symmetry of the entire complex in vivo can be broken by the presence of an additional short polypeptide (e.g., the PufX protein), which appears to bind BChl in vitro (Law et al., 2003; Chapter 10, Loach and Parkes-Loach) although it remains to be established if this is the case in vivo (Recchia et al., 1998; Roszak et al., 2003; see Chapter 9, Bullough et al. for discussion of PufX). IV. Conclusions and Prospects In this chapter the developments of two complementary model systems are described, aimed at studying assembly of functional LH complexes. One system is based on a rational design of cofactor-binding polypeptide motifs and their capacity to assembly model LH2 complexes via expression in native-like membranes. The second takes advantage of a highly reversible self-assembly of LH1 complex in artificial membranes and provides a convenient tool for design of model LH complexes with modified cofactors. Intriguingly, the two approaches are presently not mutually exchangeable, i.e., while genetic engineering of LH1 is well advanced reconstitution of the nine-fold LH2 complex has not been achieved so far. Therefore, the two model systems provide different and complementary information about LH complexes. In essence, re-design of the cofactor binding pockets in LH2 explores the underlying principles of how particular AA combinations sustain stable assembly and functional tuning of the LH-active pigment arrays. The LH1 reconstitution system, on the other hand, provides information about the cofactor-related determinants of LH complex formation and functioning. Using the excitation trap approach, the coupling between BChls and excitation delocalization over the LH1 ring
935
can be evaluated, while by the replacement of Crts, their contribution to the assembly was assessed and for the first time a Crt-binding intermediate of LH1 assembly was identified. The two systems neatly complement each other, having both distinct advantages and disadvantages: in vivo expression of model systems does not permit entire control over potential adjustments of the manipulated cell. The surrounding lipid membrane, however, is entirely natural, which is important considering that the lipids are principally involved in assembly and function of LH and membrane protein. Of course, in the reconstitution system, the composition of the reconstitution mixture and hence the lipid (detergent) membrane is largely controllable, but currently still far from the one in the native membrane. A new challenge is to make the two model approaches more interchangeable, thus allowing us to compare how the same factors contribute to the assembly and function of the different LH complexes and eventually to identify on a molecular level what renders these apparently similar complexes so different. In this respect, a re-design of LH1 polypeptides could prove very fruitful. On the other hand, the replacement of B850 BChls and systematic substitution of Crts poses a challenge as well. Finally, the assembly of the entire RC-LH1 core complex remains quite poorly understood. Hopefully, a combination of the two approaches to model investigations of photosynthetic proteins will provide the synergy required to understand this fascinating ensemble and pave ways to design nature-inspired artificial devices for efficient conversion of solar energy into useful forms. Acknowledgments These projects were supported by the Alexander von Humboldt Foundation and Volkswagen Stfitung (grant I/77 876) and the Deutsche Forschungsgemeinschaft, Bonn (grants: 1991 ‘In vivo Darstellung von de-novo-Bakteriochlorophyll-Proteinen’; SFB 533: ‘Licht-induzierte Proteindynamik.’ LF would also like to acknowledge the Krzyżanowski Fellowship from the Jagiellonian University and PB would like to acknowledge the ‘Stipendium für exzellenten wissenschaftlichen Nachwuchs’ from the LudwigMaximilians-University, Munich.
936 References Akahane J, Rondonuwu FS, Fiedor L, Watanabe Y, Koyama Y (2004) Dependence of singlet-energy transfer on the conjugation length of carotenoids reconstituted into the LH1 complex from Rhodospirillum rubrum G9. Chem Phys Lett 393: 184–191 Akiyama M, Nagashima KVP, Hara M, Wakao N, Tominaga K, Kise H and Kobayashi M (1999) Stoichiometries of LH1/RC determined by the molar ratio of BChl/BPhe analyzed by HPLC in seven species of purple bacteria containing LH1 only. Photomed Photobiol 21: 105–110 Allen JP, Feher G, Yeates TO, Komiya H and Rees DC (1988) Structure of the reaction center from Rhodobacter sphaeroides R26 — Protein cofactor (Quinones and Fe2+) interactions. Proc Natl Acad Sci USA 85: 8487–8491 Allen JP, Artz K, Lin X, Williams JC, Ivancich A, Albouy D, Mattioli TA, Fetsch A, Kuhn M and Lubitz W (1996) Effects of hydrogen bonding to a bacteriochlorophyll-bacteriopheophytin dimer in reaction centers from Rhodobacter sphaeroides. Biochemistry 35: 6612–6619 Arluison V, Seguin J and Robert B (2002) Biochemical characterization of the dissociated forms from the core antenna proteins from purple bacteria. Biochemistry 41: 11812–11819 Bahatyrova S, Frese RN, van der Werf KO, Otto C, Hunter CN and Olsen CN (2004) Flexibility and size heterogeneity of the LH1 light harvesting complex revealed by atomic force microscopy. J Biol Chem 279: 21327–21333 Balaban TS (2003) Hypothesis — Are syn-ligated (bacterio)chlorophyll dimers energetic traps in light-harvesting systems? FEBS Lett. 545: 21327–21333 Balaban TS, Fromme P, Holzwarth AR, Krauß N and Prokhorenko VI (2002) Relevance of diastereotopic ligation of magnesium atoms of chlorophylls in Photosystem I. Biochim Biophys Acta 1556: 197–207 Bandilla M, Ucker B, Ram M, Simonin I I, Gelhaye E, McDermott G, Cogdell RJ and Scheer H (1998) Reconstitution of the B800 bacteriochlorophylls in the peripheral light harvesting complex B800-850 of Rhodobacter sphaeroides 2.4.1 with BChl a and modified (bacterio-)chlorophylls. Biochim Biophys Acta 1364: 390–402 Bassi R, Croce R, Cugini D and Sandona D (1999) Mutational analysis of a higher plant antenna protein provides identification of chromophores bound into multiple sites. Proc Natl Acad Sci USA 96: 10056–10061 Bolt JD, Hunter CN, Niederman RA and Sauer K (1981) Linear and circular dichroism and fluorescence polarization of the B875 light-harvesting bacteriochlorophyll-protein complex from Rhodopseudomonas sphaeroides. Photochem Photobiol 653–656 Bradforth SE, Jimenez R, van Mourik F, van Grondelle R and Fleming GR (1995) Excitation transfer in the core light-harvesting complex (LH-1) of Rhodobacter sphaeroides: An ultrafast fluorescence depolarization and annihilation study. J Phys Chem 99: 16179–16191 Braun P and Scherz A (1991) Polypeptides and bacteriochlorophyll organization in the light- harvesting complex B850 of Rhodobacter sphaeroides R-26.1. Biochemistry 30: 5177–5184 Braun P, Olsen J, Strohmann B, Hunter CN and Scheer H (2002)
Paula Braun and Leszek Fiedor Assembly of light-harvesting bacteriochlorophyll in a model transmembrane helix in its natural environment. J Mol Biol 318: 1085–1095 Braun P, Vegh A, Strohmann B, Hunter N, Robert B and Scheer H (2003) Hydrogen bonding between the C131 keto group of bacteriochlorophyll and intramembrane serine residue α27 stabilizes LH2 antenna complex. Biochim Biophys Acta 1607: 19–26. Chang MC, Callahan PM, Parkes-Loach PS, Cotton TM and Loach PA (1990) Spectroscopic characterization of the light-harvesting complex of Rhodospirillum rubrum and its structural subunit. Biochemistry 29: 421–429 Christensen RL (1999) The electronic states of carotenoids. In: Frank HA, Young AJ, Britton G and Cogdell RJ (eds) The Photochemistry of Carotenoids (Advances in Photosynthesis and Respiration, Vol 8), pp 137–159. Kluwer Academic Publishers, Dordrecht Cogdell RJ and Scheer H (1985) Circular dichroism of lightharvesting complexes from purple photosynthetic bacteria. Photochem Photobiol 42: 669–678 Cohen-Bazire G, and Stanier RY (1958) Specific inhibition of carotenoid synthesis in a photosynthetic bacterium and its physiological consequences. Nature 181: 250–252 Coleman WJ and Youvan DC (1990) Spectroscopic analysis of genetically modified photosynthetic reaction centers. Annu Rev Biophys Chem 19: 333–367 Conroy MJ, Westerhuis W, Parkes-Loach PS, Loach PA, Hunter CN and Williamson MP (2000) The solution structure of the Rhodobacter sphaeroides LH1β polypeptide reveals two helical domains separated by a flexible region: structural consequences for the LH1 complex. J Mol Biol 298: 83–94. Croce R, Weiss S and Bassi R (1999) Carotenoid-binding sites of the major light-harvesting complex II of higher plants. J Biol Chem 274: 29613–29623 Davis CM, Bustamante PL and Loach PA (1995) Reconstitution of the bacterial core light-harvesting complexes of Rhodobacter sphaeroides and Rhodospirillum rubrum with isolated α- and β-polypeptides, bacteriochlorophyll a, and carotenoid. J Biol Chem 270: 5793–5804 Davis CM, Parkes-Loach PS, Cook CK, Meadows KA, Bandilla M, Scheer H and Loach PA (1996) Comparison of the structural requirements for bacteriochlorophyll binding in the core light-harvesting complexes of Rhodospirillum rubrum and Rhodobacter sphaeroides using reconstitution methodology with bacteriochlorophyll analogs. Biochemistry 35: 3072–3084 Davis CM, Bustamante PL, Todd JB, Parkes-Loach PS, McGlynn P, Olsen JD, McMaster L, Hunter CN and Loach PA (1997) Evaluation of structure-function relationships in the core light- harvesting complex of photosynthetic bacteria by reconstitution with mutant polypeptides. Biochemistry 36: 3671–3679 Dawkins DJ, Ferguson LA and Cogdell RJ (1988) The structure of the ‘core’ of the purple bacterial unit. In: Scheer H and Schneider S (eds) Photosynthetic Light Harvesting Systems pp 115–127. Walter de Gruyter, Berlin Desamero RZB, Chynwat V, van der Hoef I, Jansen FJ, Lugtenburg J, Gosztola D, Wasielewski MR, Cua A, Bocian DF and Frank HA (1998) Mechanism of energy transfer from carotenoids to bacteriochlorophyll: Light-harvesting by carotenoids having different extents of π−electron conjugation incorporated into
Chapter 46
Design of Light-Harvesting Complexes
the B850 antenna complex from the carotenoidless bacterium Rhodobacter sphaeroides R-26.1. J Phys Chem 102: 8151–8162. Douglas B, McDaniel D and Alexander J (1994) Concepts and Models of Inorganic Chemistry. John Wiley & Sons, Inc., New York Drews G (1996) Formation of the light-harvesting complex I (B870) of anoxygenic phototrophic purple bacteria. Arch Microbiol 166: 151–159 Fiedor L (2006) Hexacoordination of bacteriochlorophyll in photosynthetic antenna LH1. Biochemistry 45: 1910–1918 Fiedor L and Scheer H (2005) Trapping of an intermediate of LH1 complex assembly beyond the B820 subunit. J Biol Chem 280: 20921–20926 Fiedor L, Scheer H, Hunter CN, Tschirschwitz F, Voigt B, Ehlert J, Nibbering E, Leupold D and Elsaesser T (2000) Introduction of a 60 fs deactivation channel in the photosynthetic antenna LH1 by Ni-Bacteriopheophytin a. Chem Phys Lett 319: 145–152 Fiedor L, Leupold D, Teuchner K, Voigt B, Hunter CN, Scherz A and Scheer H (2001) Excitation trap approach to analyze size and pigment-pigment coupling: Reconstitution of LH1 antenna of Rhodobacter sphaeroides with Ni-substituted bacteriochlorophyll. Biochemistry 40: 3737–3747 Fiedor L, Akahane J and Koyama Y (2004) Carotenoid-induced cooperative formation of bacterial photosynthetic LH1 complex. Biochemistry 43: 16487–16496 Fowler GJ, Visschers RW, Grief GG, van Grondelle R and Hunter CN (1992) Genetically modified photosynthetic antenna complexes with blueshifted absorbance bands. Nature 355: 848–850 Fowler GJ, Sockalingum GD, Robert B and Hunter CN (1994) Blue shifts in bacteriochlorophyll absorbance correlate with changed hydrogen bonding patterns in light-harvesting 2 mutants of Rhodobacter sphaeroides with alterations at alpha-Tyr-44 and alpha-Tyr-45. Biochem J 299: 695–700 Francke C and Amesz J (1995) The size of the photosynthetic unit in purple bacteria. Photosynth Res 46: 347–352 Frank HA (1999) Incorporation of carotenoids into reaction center and light-harvesting pigment-protein complexes. In: Frank HA, Young AJ, Britton G and Cogdell, RJ (eds) Photochemistry of Carotenoids (Advances in Photosynthesis and Respiration, Vol 8), pp 235–244. Kluwer Academic Publishers, Dordrecht Frank HA, Chynwat V, Posteraro A, Hartwich G, Simonin I and Scheer, H (1996) Triplet state energy transfer between the primary donor and the carotenoid in Rhodobacter sphaeroides R-26.1 reaction centers exchanged with modified bacteriochlorophyll pigments and reconstituted with spheroidene. Photochem Photobiol 64: 823–831. Freer A, Prince S, Sauer K, Papiz M, Hawthornthwaite-Lawless A, McDermott G, Cogdell R and Isaacs NW (1996) Pigment-pigment interactions and energy transfer in the antenna complex of the photosynthetic bacterium Rhodopseudomonas acidophila. Structure 4: 449–462 Fromme P, Jordan P and Krauss N (2001) Structure of Photosystem I. Biochim Biophys Acta 1507: 5–31 Fyfe PK and Jones MR (2005) Lipids in and around photosynthetic reaction centres. Biochem Soc Trans 33: 924–930 Gall A, Fowler GJ, Hunter CN and Robert B (1997) Influence of the protein binding site on the absorption properties of the monomeric bacteriochlorophyll in Rhodobacter sphaeroides
937
LH2 complex. Biochemistry 36: 16282–16287 Garcia-Martin A, Kwa LG, von Jan M, Robert B and Braun P (2005) H-bonding drives assembly of model bacteriochlorophyll protein in the native membrane. In: van der Est A and Bruce D (eds) Fundamental aspects to global phenomena, pp 138–40. Allen Press, Montréal Garcia-Martin A, Kwa LG, Strohmann B, Robert B, Holzwarth AR and Braun P (2006a) Structural role of (bacterio)chlorophyll ligated in the energetically unfavourable β-position. J Biol Chem 281: 10626–10634 Garcia-Martin A, Kwa LG, von Jan M, Hunter CN and Braun P (2006b) Assembly of model bacteriochlorophyll proteins in the native lipid environment. In: Grimm B, Porra W, Rüdiger W and Scheer H (eds) Chlorophylls and Bacteriochlorophylls: Biochemistry, Functions and Applications (Advances in Photosynthesis and Respiration, Vol 25), pp 365–396. Springer, Dordrecht Georgakopoulou S, Frese RN, Johnson E, Koolhaas C, Cogdell RJ, van Grondelle R and van der Zwan G (2002) Absorption and CD spectroscopy and modeling of various LH2 complexes from purple bacteria. Biophys J 82: 2184–2197 Griffiths M, Sistrom WR, Cohen-Bazire G and Stainer RY (1955) Functions of carotenoids in photosynthesis. Nature 176: 1211–1214 Hartwich G, Fiedor L, Simonin I, Cmiel E, Schäfer W, Noy D, Scherz A and Scheer H (1998) Metal-substituted bacteriochlorophylls. 1. Preparation and influence of metal and coordination on spectra. J Am Chem Soc 120: 3675–3683 Heinemann B and Paulsen H (1999) Random mutations directed to transmembrane and loop domains of the light-harvesting chlorophyll a/b protein: Impact on pigment binding. Biochemistry 38: 14088–14093 Hofmann E, Wrench PM, Sharples FP, Hiller RG, Welte W and Diederichs K (1996) Structural basis of light harvesting by carotenoids: Peridinin-chlorophyll-protein from Amphidinium carterae. Science 272: 1788–1791 Horn R, Paulsen H (2002) Folding in vitro of light-harvesting chlorophyll a/b protein is coupled with pigment binding. J Mol Biol 318: 547–556 Hu Q, Sturgis JN, Robert B, Delagrave S, Youvan DC and Niederman RA (1998) Hydrogen bonding and circular dichroism of bacteriochlorophylls in the Rhodobacter capsulatus lightharvesting 2 complex altered by combinatorial mutagenesis. Biochemistry 37: 10006–10015 Hunter CN (1995) Rings of light. Curr Biol 5: 826–828 Jiang L, Lai L (2002) CH...O hydrogen bonds at protein-protein interfaces. J Biol Chem 277: 37732–37740 Jones MR, Fowler GJ, Gibson LC, Grief GG, Olsen JD, Crielaard W and Hunter CN (1992) Mutants of Rhodobacter sphaeroides lacking one or more pigment-protein complexes and complementation with reaction-centre, LH1, and LH2 genes. Mol Microbiol 6: 1173–1184 Jordan P, Fromme P, Witt HT, Klukas O, Saenger W and Krauss N (2001) Three-dimensional structure of cyanobacterial Photosystem I at 2.5 Å resolution. Nature 411: 909–917 Jungas C, Ranck J-L, Rigaud J-L, Joliot P and Verméglio A (1999) Supramolecular organization of the photosynthetic apparatus of Rhodobacter sphaeroides. EMBO J. 18: 534–542 Kania A and Fiedor L (2006) Steric control of bacteriochlorophyll ligation. J Am Chem Soc 128: 454–458
938 Karrasch S, Bullough PA and Ghosh R (1995) The 8.5 Å projection map of the light-harvesting complex from Rhodospirillum rubrum reveals a ring composed of 16 subunits. EMBO J 14: 631–638 Kashiwada A, Nishino N, Wang Z-H, Nozawa T, Kobayashi M and Nango M (1999) Molecular assembly of bacteriochlorophyll a and its analogues by synthetic 4α-helix polypeptides. Chem Lett 28:1301–1302 Kehoe JW, Meadows KA, Parkes-Loach PS and Loach PA (1998) Reconstitution of core light-harvesting complexes of photosynthetic bacteria using chemically synthesized polypeptides. 2. Determination of structural features that stabilize complex formation and their implications for the structure of the subunit complex. Biochemistry 37: 3418–3428 Kim J, Eichacker LA, Rüdiger W and Mullet JE (1994) Chlorophyll regulates accumulation of the plastid-encoded chlorophyll proteins P700 and D1 by increasing apoprotein stability. Plant Physiol. 104: 907–916 Koolhaas MH, Frese RN, Fowler GJ, Bibby TS, Georgakopoulou S, van der Zwan G, Hunter CN and van Grondelle R (1998) Identification of the upper exciton component of the B850 bacteriochlorophylls of the LH2 antenna complex, using a B800-free mutant of Rhodobacter sphaeroides. Biochemistry 37: 4693–4698 Koyama Y, Kuki M, Andersson PO and Gillbro T (1996) Singlet excited states and the light-harvesting function of carotenoids in bacterial photosynthesis. Photochem Photobiol 63: 243–256 Kwa LG, Garcia-Martin A, Vegh A, Strohmann B, Scheer H, Robert B and Braun P (2004) Hydrogen bonding in a model bacteriochlorophyll-binding site drives asssembly of light harvesting complex. J Biol Chem 279: 15067–15075 Lang HP and Hunter CN (1994) The relationship between carotenoid biosynthesis and the assembly of the light-harvesting LH2 complex in Rhodobacter sphaeroides. Biochem J 298: 197–205 Lapouge K, Naveke A, Robert B, Scheer H and Sturgis JN (2000) Exchanging cofactors in the core antennae from purple bacteria: Structure and properties of Zn-bacteriopheophytin-containing LH1. Biochemistry 39: 1091–1099 Law CJ and Cogdell RJ (1998) The effect of chemical oxidation on the fluorescence of the LH1 (B880) complex from the purple bacterium Rhodobium marinum. FEBS Lett. 432: 27–30 Leonov H, Arkin IT (2005) A periodicity analysis of transmembrane helices. Bioinformatics 21: 2604–2610 Liebl U, Nitschke W, and Mattioli TA. (1996) Pigment-protein interactions in the antenna-reaction center complex of Heliobacillus mobilis. Photochem Photobiol 64: 38–45 Linnanto J and Korppi-Tommola J. (2004) Structural and spectroscopic properties of Mg-bacteriochlorin and methyl bacteriochlorophyllides a, b, g, and h Studied by semiempirical, ab initio, and density functional molecular orbital methods. J Phys Chem A 108: 5872–5882 Loach PA and Parkes-Loach PS (1995) Structure-function relationships in core light-harvesting complexes (LHI) as determined by characterization of the structural subunit and by reconstitution experiments. In: Blankenship RE, Madigan MT, and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 437–471. Kluwer Academic Publishers, Dordrecht Loach PA and Sekura DL (1968) Primary photochemistry and electron transport in Rhodospirillum rubrum. Biochemistry
Paula Braun and Leszek Fiedor 7: 2642–2649 Loach PA, Sekura DL, Hadsell RM and Stemer A. (1970) Quantitative dissolution of the membrane and preparation of photoreceptor subunits from Rhodopseudomonas spheroides. Biochemistry 9: 724–733 Loll B, Kern J, Saenger W, Zouni A and Biesiadka J (2005) Towards complete cofactor arrangement in the 3.0 ! resolution structure of Photosystem II. Nature 438: 1040–1044 Mackinnon R, von Heijne G (2006) Membranes. Curr Opin Struct Biol 16: 431 Mattioli TA, Gray KA, Lutz M, Oesterhelt D and Robert B (1991) Resonance Raman characterization of Rhodobacter sphaeroides reaction centers bearing site-directed mutations at Tyrosine M210. Biochemistry 30: 1715–1721 McDermott G, Prince SM, Freer AA, Hawthornthwaite-Lawless AM, Papiz MZ, Cogdell RJ and Isaacs NW (1995) Crystal structure of an integral membrane light-harvesting complex from photosynthetic bacteria. Nature 374: 517–521 Meadows KA, Parkes-Loach PS, Kehoe JW and Loach PA (1998) Reconstitution of core light-harvesting complexes of photosynthetic bacteria using chemically synthesized polypeptides. 1. Minimal requirements for subunit formation. Biochemistry 37: 3411–3417 Moskalenko A, Toropygina O and Kuznetsova N (1996) Does the B820 subcomplex of the B880 complex retain carotenoids? Z. Naturforsch. 51: 309–318. Musewald C, Hartwich G, Lossau H, Gilch P, Pöllinger-Dammer F, Scheer H and Michel-Beyerle ME (1999) Ultrafast photophysics and photochemistry of [Ni]-bacteriochlorophyll a. J Phys Chem 103: 7055–7060 Näveke A, Scheer H, Robert B and Sturgis J (1995) Influence of metal exchange on absorption spectra of the light harvesting complex of Rhodospirillum rubrum. In: Mathis P (ed) Progress in Photosynthesis, pp 1–5. Kluwer Academic Publishers, Dordrecht Nilsson J, Persson B and von Heijne G (2000) Consensus predictions of membrane protein topology. FEBS Lett 486: 267–269 Noguchi T, Hayashi H and Tasumi M (1990) Factors controlling the efficiency of energy transfer from carotenoids to bacteriochlorophyll in purple photosynthetic bacteria. Biochim Biophys Acta 1017: 280–290 Olsen JD, Sturgis JN, Westerhuis WH, Fowler GJ, Hunter CN and Robert B (1997) Site-directed modification of the ligands to the bacteriochlorophylls of the light-harvesting LH1 and LH2 complexes of Rhodobacter sphaeroides. Biochemistry 36: 12625–12632 Olsen JD, Robert B, Siebert CA, Bullough PA and Hunter CN (2003) Role of the C-terminal extrinsic region of the alpha polypeptide of the light-harvesting 2 complex of Rhodobacter sphaeroides: A domain swap study. Biochemistry 42: 15114–23 Pali T, Garab G, Horvath LI, and Kota Z. Functional significance of the lipid-protein interface in photosynthetic membranes. Cell Molec Life Sci 60: 1591–1606 Pandit A, van Stokkum IHM, Georgakopoulu S, van der Zwan G and van Grondelle R (2003) Investigations of intermediates appearing in the reassociation of the light-harvesting 1 complex of Rhodospirillum rubrum. Photosynth Res 75: 235–248 Papagiannakis E, Kennis JTM, van Stokkum IHM, Cogdell RJ and van Grondelle R (2002) An alternative carotenoid-to-bacte-
Chapter 46
Design of Light-Harvesting Complexes
riochlorophyll energy transfer pathway in photosynthetic light harvesting. Proc Natl Acad Sci USA 99: 6017–6022 Parkes-Loach PS, Majeed AP, Law CJ and Loach PA (2004) Interactions stabilizing the structure of the core light-harvesting complex (LH1) of photosynthetic bacteria and its subunit (B820). Biochemistry 43: 7003–7016 Picorel R, Bélanger G and Gingras G (1983) Antenna holochrome B880 of Rhodospirillum rubrum S1. Pigment, phospholipid, and polypeptide composition. Biochemistry 22: 2491–2497 Pilch M and Pawlikowski M (1998) Circular dichroism (CD) study of peridinin-chlorophyll a protein (PCP) complexes from marine dinoflagellate algae. J Chem Soc, Farady Trans 94: 227–232 Plumley FG and Schmidt GW (1995) Light-harvesting chlorophyll a/b complexes: Interdependent pigment synthesis and protein assembly. Plant Cell 7: 689–704 Polívka T and Sundström V (2004) Ultrafast dynamics of carotenoid excited states — from solution to natural and artificial systems. Chem Rev 104: 2021–2071 Popot J-L and Engelman DM (1990) Membrane protein folding and oligomerization: The two-stage model. Biochemistry 29: 4031–4037 Prince SM, Papiz MZ, Freer AA, McDermott G, HawthornthwaiteLawless AM, Cogdell RJ and Isaacs NW (1997) Apoprotein structure in the LH2 complex from Rhodopseudomonas acidophila strain 10050: Modular assembly and protein pigment interactions. J Mol Biol 268: 412–23 Rau HK, Snigula H, Struck A, Robert B, Scheer H and Haehnel W (2001) Design, synthesis and properties of synthetic chlorophyll proteins. Eur J Biochem 268: 3284–95 Recchia PA, Davis CM, Lilburn TG, Beatty JT, Parkes-Loach PS, Hunter CN and Loach PA (1998) Isolation of the PufX protein from Rhodobacter capsulatus and Rhodobacter sphaeroides: Evidence for its interaction with the α-polypeptide of the core light-harvesting complex. Biochemistry 37: 11055–11063 Remelli R, Varotto C, Sandona D, Croce R and Bassi R (1999) Chlorophyll binding to monomeric light-harvesting complex. A mutation analysis of chromophore-binding residues. J Biol Chem 274: 33510–33521 Ritz T, Damjanovic A, Schulten K, Zhang J-P and Koyama Y (2000) Efficient light harvesting through carotenoids. 66: 125–144 Roszak AW, Howard TD, Southall J, Gardiner AT, Law CJ, Isaacs NW and Cogdell RJ (2003) Crystal structure of the RC-LH1 core complex from Rhodopseudomonas palustris. Science 302: 1969–1972 Roszak AW, McKendrick K, Gardiner AT, Mitchell IA, Isaacs NW, Cogdell RJ, Hashimoto H and Frank HA (2004) Protein regulation of carotenoid binding; gatekeeper and locking amino acid residues in reaction centers of Rhodobacter sphaeroides. Structure 12: 765–73 Sachs JN, Engelman DM (2006) Introduction to the membrane protein reviews: The interplay of structure, dynamics, and environment in membrane protein function. Annu Rev Biochem 75: 707–712 Scheer H and Struck A (1993) Bacterial reaction centers with modified tetrapyrrole chromophores. In: Deisenhofer J and Norris JR (eds) The Photosynthetic Reaction Center, pp 157–192. Academic Press, San Diego Scherz A and Parson WW (1984) Exciton interactions in dimers of bacteriochlorophyll and related molecules. Biochim Biophys
939
Acta 766: 666–678 Shuvalov VA and Duysens LNM. (1986) Primary electron transfer reactions in modified reaction centers from Rhodopseudomonas sphaeroides. Proc Natl Acad Sci USA 83: 1690–1694 Sperotto MM, May S and Baumgaertner A (2006) Modelling of proteins in membranes. Chem Phys Lipids 141: 2–29 Spiedel D, Jones MR and Robert B (2002) Tuning of the redox potential of the primary electron donor in reaction centres of purple bacteria: Effects of amino acid polarity and position. FEBS Lett 527: 171–175 Stark W, Kühlbrandt W, Wildhaber I, Wehrli E and Mühlethaler K (1984) The structure of the photoreceptor unit of Rhodopseudomonas viridis. EMBO J 3: 777–783. Struck A and Scheer H (1990) Modified reaction centers from Rhodobacter sphaeroides R26. 1. Exchange of monomeric bacteriochlorophyll with 132-hydroxy-bacteriochlorophyll. FEBS Lett 261: 385–388 Struck A, Beese D, Cmiel E, Fischer M, Müller A, Schäfer W and Scheer H (1990a) Modified bacterial reaction centers: 3. Chemically modified chromophores at sites BA, BB and HA, HB. In: Michel-Beyerle ME (ed) Reaction Centers of Photosynthetic Bacteria, pp 313–327. Springer, Berlin Struck A, Cmiel E, Katheder I and Scheer H (1990b) Modified reaction centers from Rhodobacter sphaeroides R26. 2: Bacteriochlorophyll with modified C-3 substituents at sites BA and BB. FEBS Lett. 268: 180–184 Sturgis JN and Robert B (1994) Thermodynamics of membrane polypeptide oligomerization in light-harvesting complexes and associated structural changes. J Mol Biol 238: 445–454 Sturgis JN and Robert B (1997) Pigment binding site and electronic properties in light-harvesting proteins of purple bacteria. J Phys Chem 101: 7227–7231 Theiler R, Suter F, Wiemken V and Zuber H (1984) Complete primary structure of the four light-harvesting polypeptides from Rhodopseudomonas sphaeroides R26.1. Experientia 40: 640 Todd JB, Parkes-Loach PS, Leykam JF and Loach PA (1998) In vitro reconstitution of the core and peripheral light-harvesting complexes of Rhodospirillum molischianum from separately isolated components. Biochemistry 37: 17458–17468 Ueda T, Morimoto Y, Sato M, Kakuno T, Yamashita J and Horio T (1985) Isolation, characterization, and comparison of a ubiquitous pigment-protein complex consiting of a reaction center and light-harvesting bacteriochlorophyll proteins present in purple photosynthetic bacteria. J Biochem 98: 1487–1498 van Grondelle R (1985) Excitation energy transfer, trapping and annihilation in photosynthetic systems. Biochim Biophys Acta 811: 147–195 Vegh AP and Robert B (2002) Spectroscopic characterisation of a tetrameric subunit form of the core antenna protein from Rhodospirillum rubrum. FEBS Lett 528: 222–226 Vershinin A (1999) Biological functions of carotenoids—diversity and evolution. Biofactors 10: 99–104 Wallin E, Tsukihara T, Yoshikawa S, von Heijne G and Elofsson A (1997) Architecture of helix bundle membrane proteins: An analysis of cytochrome c oxidase from bovine mitochondria. Protein Sci 6: 808–815 Walters RF, DeGrado WF (2006) Helix-packing motifs in membrane proteins. Proc Natl Acad Sci USA 103: 13658–13663 Wang Y, Hu X (2002) A quantum chemistry study of binding carotenoids in the bacterial light-harvesting complexes. J Am Chem Soc 124: 8445–8451
940 Wang Z-Y, Gokan K, Kobayashi M and Nozawa T (2005) Solution structures of the core light-harvesting α and β polypeptides from Rhodospirillum rubrum: implications for the pigment-protein and protein-protein interactions. J Mol Biol 347: 465–477 Wang Z-Y, Muraoka Y, Shimonaga M, Kobayashi M and Nozawa T (2002) Selective detection and assignment of the solution NMR signals of bacteriochlorophyll a in a reconstituted subunit of a light-harvesting complex. J Am Chem Soc 124: 1072–1078 Witt H, Schlodder E, Teutloff C, Niklas J, Bordignon E, Carbonera D, Kohler S, Labahn A and Lubitz W (2002) Hydrogen bonding to P700: Site-directed mutagenesis of threonine A739 of Photosystem I in Chlamydomonas reinhardtii. Biochemistry 41: 8557–8569 Young AJ (1993) Occurrence and distribution of carotenoids in photosynthetic systems. In: Young AJ and Britton G (eds) Carotenoids in Photosynthesis, pp 16–71. Chapman & Hall, London
Paula Braun and Leszek Fiedor Zeng X, Choudhary M and Kaplan S (2003) A second and unusual pucBA operon of Rhodobacter sphaeroides 2.4.1: Genetics and function of the encoded polypeptides. J Bacteriol 185: 6171–6184 Zuber H and Cogdell RJ (1995) Structure and organization of purple bacterial antenna complexes. In: Blankenship R, Madigan MT and Bauer CE (eds) Anoxygenic Photosynthetic Bacteria (Advances in Photosynthesis and Respiration, Vol 2), pp 315–348. Kluwer Academic Publishers, Dordrecht Zurdo J, Centeno MA, Odriozola JA, Fernandez-Cabrera C and Ramirez JM (1995) The structural role of the carotenoid in the bacterial light-harvesting protein 2 (LH2) of Rhodobacter capsulatus. A Fourier transform Raman spectroscopy and circular dichroism study. 46: 363–369
Chapter 47 The Supramolecular Assembly of the Photosynthetic Apparatus of Purple Bacteria Investigated by High-Resolution Atomic Force Microscopy Simon Scheuring* Institut Curie, UMR168-CNRS, 26 Rue d’Ulm, 75248 Paris, France
Summary ............................................................................................................................................................... 941 I. Introduction..................................................................................................................................................... 942 II. Atomic Force Microscopy Analysis of the Complexes of the Bacterial Photosynthetic Apparatus................. 943 A. The Photosynthetic Apparatus of Blastochloris viridis .................................................................... 943 B. The Photosynthetic Apparatus of Rhodospirillum photometricum .................................................. 944 C. The Photosynthetic Apparatus of Phaeospirillum molischianum .................................................... 945 D. The Photosynthetic Apparatus of Rhodobacter blasticus .............................................................. 946 E. The Photosynthetic Apparatus of Rhodopseudomonas palustris ................................................... 948 III. Conclusions .................................................................................................................................................... 949 IV. Feasibilities, Limitations and Outlook ............................................................................................................. 950 Acknowledgments ................................................................................................................................................. 951 References ............................................................................................................................................................ 951
Summary The atomic force microscope (AFM) has developed into a powerful tool in membrane protein research and is now a technique complementary to X-ray crystallography and electron microscopy (EM). The AFM allows high-resolution topographies of biological samples to be acquired under near-physiological conditions, i.e. in buffer solution at ambient temperature and pressure. The exceptionally high signal-to-noise ratio of the AFM allows imaging of single molecules, and it is therefore the only technique to date that can provide structural information on supramolecular membrane protein assemblies in native membranes at nanometer resolution. Hence, AFM can provide images of native membrane organization into which high-resolution membrane protein structures, obtained from X-ray or EM crystallographic studies, can be docked. This chapter will outline recent achievements in the elucidation of the complex assembly of the photosynthetic apparatus in different purple phototrophic bacteria, and discuss feasibilities and restrictions of future applications of AFM in membrane research.
*Email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 941–952. © 2009 Springer Science + Business Media B.V.
942 I. Introduction The structure determination of the different components of the photosynthetic apparatus, the lightharvesting complex 2 (LH2), the light-harvesting complex 1 (LH1), the reaction center (RC), the cytochrome bc1 complex, and the ATP-synthase, is almost complete. Briefly, X-ray crystallography structures of the RCs of Blastochloris (Blc.) viridis (Deisenhofer et al., 1984, 1985, 1995; Deisenhofer and Michel, 1989; Lancaster et al., 2000) and Rhodobacter (Rba.) sphaeroides (Allen et al., 1987) represent the first solved membrane protein structures in general, and have provided deep insights into their functional mechanisms (Chapter 16, Jones; Chapter 17, Axelrod et al.; Chapter 18, Williams and Allen; Chapter 19, Parson and Warshel; Chapter 20, Wraight and Gunner). The RC consists of three subunits L, M and H. While subunits L and M are similar and arranged with a quasi two-fold axis, each consisting of five transmembrane helices, subunit H forms mainly a cytoplasmic domain and is anchored by one trans-membrane helix within the membrane. The Blc. viridis RC is additionally topped by a nonmembranous tetraheme cytochrome (4Hcyt) on the periplasmic side. The LH2 from Rhodopseudomonas (Rps.) acidophila (McDermott et al., 1995) and from Phaeospirillum (Phs.) molischianum (Koepke et al., 1996) revealed nonameric and octameric cylinders, respectively, of two polypeptides (α and β), both of which span the membrane once. The inner wall of the LH2 cylinder is formed by the α-polypeptides and the outer wall by the β-polypeptides; three bacteriochlorophyll a (BChl a) and two carotenoid molecules are bound per α/β subunit (Chapter 8, Gabrielsen et al.). LH1 α/β heterodimers are considered structurally very similar to LH2 α/β heterodimers without the B800 BChl pigment ring (McDermott et al., 1995; Koepke et al., 1996; Hu and Schulten, 1998). They are closely associated with the RC to form the RC-LH1 core complex. To date, no atomic structure of this ensemble is available. However, medium and low resolution X-ray, EM and AFM data of different core complexes indicate structural variability of Abbreviations: 4Hcyt – tetraheme cytochrome; AFM – atomic force microscope; ATP – adenosine triphosphate; BChl – bacteriochlorophyll; Blc. – Blastochloris; EM – electron microscope; LH1 – light-harvesting complex 1; LH2 – light-harvesting complex 2; Phs. – Phaeospirillum; Rba. – Rhodobacter; RC – reaction center; Rps. – Rhodopseudomonas; Rsp. – Rhodospirillum; Rvi. – Rubrivivax
Simon Scheuring these complexes among species. The core complex was described to consist of a RC surrounded by an ellipse of 16 LH1 α/β heterodimers in Rhodospirillum (Rsp.) rubrum (Jamieson et al., 2002; Fotiadis et al., 2004), Bcl. viridis (Scheuring et al., 2003a), Rsp. photometricum (Scheuring et al., 2004a; Scheuring and Sturgis, 2005) and Phs. molischianum (Gonçalves et al., 2005a), or 15 LH1 α/β heterodimers plus a yet unidentified W-peptide in Rps. palustris (Roszak et al., 2003; Scheuring et al., 2006). In some bacteria dimeric core complexes are observed, consisting of two RCs surrounded by an S-shaped LH1 assembly constituted of two times 13 LH1 α/β heterodimers plus 2 PufX polypeptides at the dimer center in Rba. sphaeroides (Scheuring et al., 2004b), and Rba. blasticus, (Scheuring et al., 2005). A cryo-EM study provided a different model for the Rba. sphaeroides core dimer, in which two RCs are surrounded by an S-shaped LH1 assembly consisting of two 14 LH1 α/β heterodimer assemblies plus 2 PufX polypeptides located between LH1 subunit 14 and the RC at the peripheries of the ‘S’ (Qian et al., 2005). The structural diversity of core complexes most probably reflects adaptation to particular environmental conditions of the different species (Scheuring, 2006). Despite the diversity of core complex architectures, all structural data agree now that the LH1 assembly forms an ellipsis around the RC. A recent preliminary X-ray structure of the cytochrome bc1 complex from Rba. capsulatus (Berry et al., 2004) indicates structural similarities with the homologous cytochrome bc1 complex from mitochondria (Xia et al., 1997), and the cytochrome b6 f complexes from cyanobacteria (Kurisu et al., 2003) and algae (Stroebel et al., 2003). All are dimers consisting of three subunits: the cytochrome b with eight transmembrane helices, and the high-potential 2Fe-2S Rieske protein and the cytochrome c1, each with one transmembrane helix. The cytochrome b subunit contains two b-type hemes and houses the quinone binding sites (see Chapter 22, Berry et al. and Chapter 23, Kramer et al. for reviews of structural and mechanistic aspects of cytochrome bc1 complexes). Finally, homologous structures of both the F1 (Abrahams et al., 1994; Stock et al., 1999) and FO (Meier et al., 2005) parts of the ATP-synthase have been solved (see Chapter 24, Feniouk and Junge). Despite the wealth of functional and structural information available on the individual membrane proteins of the bacterial photosynthetic apparatus, their supramolecular organization in the native mem-
Chapter 47
Bacterial Photosynthetic Membranes Imaged by AFM
943
brane was unknown. Early spectroscopic studies of native membranes had provided information on the connectivity of bacterial complexes and the sizes of LH2 and LH1 energy transfer domains (Monger and Parson, 1977; Hunter et al., 1985). Modeling studies were also conducted (Papiz et al., 1996; Sundström et al., 1999; Hu et al., 2002), but no direct observation of the organization of bacterial photosynthetic complexes was possible until the advent of the AFM. In the last five years the application of the AFM has attained a maturity that makes it the key technique for experimental structural investigations of the photosynthetic apparatus in native membranes: it is now possible to acquire AFM topographs at ~10 Å lateral resolution at an exceptionally high signal-to-noise ratio. This has allowed the structural investigation of the individual complexes and their supramolecular architecture in native membranes of different species (Scheuring et al., 2003a, 2004a,c, 2005, 2006; Bahatyrova et al., 2004b; Gonçalves et al., 2005a; Scheuring and Sturgis, 2005; Chapter 14, Sturgis and Niederman). II. Atomic Force Microscopy Analysis of the Complexes of the Bacterial Photosynthetic Apparatus The first high-resolution AFM analysis of a constituent of the bacterial photosynthetic apparatus concerned the LH2 complex of Rubrivivax (Rvi.) gelatinosus. The stoichiometry and sidedness of purified and reconstituted complexes were analyzed by AFM. Interestingly, in reconstituted membranes of less extensively purified LH2 a minor fraction of LH1 complexes was co-purified and found assembled with LH2, providing first images of different protein complexes in one membrane (Scheuring et al., 2001). Subsequently, a series of AFM studies of purified and reconstituted proteins was conducted on: LH2 (Scheuring et al., 2003b; Stamouli et al., 2003; Gonçalves et al., 2005b), LH1 (Bahatyrova et al., 2004a), and RC-LH1 core complexes (Fotiadis et al., 2004; Scheuring et al., 2004b; Siebert et al., 2004). Finally native membranes were imaged, first containing only RC-LH1 core complexes (Scheuring et al., 2003a), then containing LH2 and RC-LH1 core complexes (Bahatyrova et al., 2004b; Scheuring et al., 2004a,c, 2005, 2006; Gonçalves et al., 2005a; Scheuring and Sturgis, 2005).
Fig. 1. High-resolution AFM analysis of the native photosynthetic membrane of Blc viridis (Scheuring et al., 2003a). a) Raw data AFM topograph of hexagonally packed core complexes in the native photosynthetic membrane of Blc. viridis. b) Using the AFM tip as a nano-dissector three sub-forms of the core complex could be imaged: 4Hcyt-RC-LH116 (1), RC-LH116 (2), and LH116 (3) c) Surface representation of the structural model of the complex assembly outlined in a) in the native membrane of Blc. viridis.
A. The Photosynthetic Apparatus of Blastochloris viridis Blc. viridis lacks an LH2 complex and therefore the core complexes form hexagonal lattices in chromatophores (Miller, 1982; Jacob and Miller, 1983, 1984). Later projection maps could not resolve the subunit architecture of the core complex due to the absence of higher packing order (Stark et al., 1984; Engelhardt et al., 1986; Ikeda-Yamasaki et al., 1998). The AFM features a signal-to-noise ratio sufficiently high that the core complex subunit architecture could be analyzed readily on single molecules in native membranes (Fig. 1a) (Scheuring et al., 2003a). High resolution images of native membranes revealed a lateral resolution of 10 Å, as judged from spectral signal-to-noise analysis gained through single particle averaging, and a vertical resolution of ~1 Å. In Blc. viridis, the RC consists of the L, M, H and the tetraheme cytochrome (4Hcyt) subunits (Deisenhofer et al., 1985). Sixteen LH1 subunits surround the
944 RC (Fig. 1a). Using the AFM tip as a nano-dissector (Hoh et al., 1991), the strongly protruding 4Hcyt was removed from the entire 4Hcyt-RC-LH116 complex (Fig. 1b; 1), and a RC-LH116 complex (Fig. 1b; 2) remained that could be imaged at the highest resolution. The topographs revealed that the LH1 subunits surround the RC as an ellipsis with axes having a length difference of ~10%. The orientation of the long axis of the LH116 ellipsis coincides with the long axis of the RC (Fig. 1b; 2). LH116 assemblies, from which the 4Hcyt and the RC proteins had been removed (Fig. 1b; 3), relax into closed circles with ~100 Å top diameter. The ellipticity of the LH116 in complex with the RC, compared to a circular LH116 after removal of the RC reflects flexibility of the LH116 assembly and strong and specific interactions between the core complex components. The distortion of the LH1 assembly from the energetically favorable circular form to an ellipsis in complex with the RC may favor the quinone/quinol passage through a closed LH1 fence (Jamieson et al., 2002; Scheuring et al., 2003a; Chapter 9, Bullough et al.). In native Blc. viridis membranes, elliptical core complexes are hexagonally packed. There is no higher packing order. The rotational orientation of each complex was independent from the orientations of neighboring cores. In such an arrangement, all core complexes are equal, i.e. they are structurally identical and their molecular neighborhood is identical. As a consequence, the whole membrane could be considered as one unit with equal energy trapping probabilities at all positions, and hence the chance that a RC is occupied by the primary photoreaction is equal for all RCs. Lacking LH2, excitation energy can be transferred between core complexes without crossing an energy barrier of ~2kT that would be posed by LH2 intercalating core complexes (Scheuring et al., 2004a; Chapter 15, Şener and Schulten). Although Blc. viridis lacks an LH2 complex, and the consequent enhanced absorption bandwidth, this AFM study shows that the homogeneity and organization of the Blc. viridis membrane provides an efficient network for harvesting and trapping light energy. B. The Photosynthetic Apparatus of Rhodospirillum photometricum The supramolecular organization of the photosynthetic apparatus of Rsp. photometricum, was studied in detail and yielded striking and novel findings concerning antenna heterogeneity, antenna domain
Simon Scheuring
Fig. 2. High-resolution AFM analysis of a high-light adapted native photosynthetic membrane of Rsp. photometricum (Scheuring et al., 2004a,c; Scheuring and Sturgis, 2005). a) Raw data AFM topograph revealing the complex mixture of LH2 and core complexes in high-light adapted native photosynthetic membranes of Rsp. photometricum. b) Surface representation of structural models of the complex assembly outlined in a) in the native membrane of Rsp. photometricum.
formation, and complex assembly (Figs. 2, 3) (Scheuring et al., 2004a,c; Scheuring and Sturgis, 2005). The majority of LH2 complexes assemble as nonameric rings. However, on closer examination LH2 complexes of various sizes were found within native membranes. Diameter distribution and image processing analysis showed heterogeneity of the LH2 complex stoichiometry around the general nonameric assembly (~70%), with smaller octamers (~15%) and larger decamers (~15%). This finding was qualitatively corroborated by examination of individual complexes in raw data images (Scheuring
Chapter 47
Bacterial Photosynthetic Membranes Imaged by AFM
Fig. 3. High-resolution AFM analysis of an antenna domain in a low-light adapted native photosynthetic membrane of Rsp. photometricum (Scheuring et al., 2004a,c; Scheuring and Sturgis, 2005). a) Raw data AFM topograph revealing a LH2 antenna domain in low-light adapted native photosynthetic membranes of Rsp. photometricum. b) Surface representation of structural model of the LH2 assembly outlined in a) in antenna domains in native membranes of Rsp. photometricum.
et al., 2004a). It seems probable that a heterogeneous stoichiometry is an inherent feature of LH2, as it has also been observed in Phs. molischianum (Gonçalves et al., 2005a) and Rps. palustris (Scheuring et al., 2006). In contrast to the heterogeneity found for LH2 complexes, the RC-LH1 core complexes appeared uniform in size; each monomeric RC was surrounded by a closed elliptical LH116 assembly, with long and short axes of 95 Å and 85 Å, following the long RC axis. Analysis of the distribution of the photosynthetic complexes of Rsp. photometricum showed significant clustering of both antenna complexes and core complexes. Membranes were found which contained domains densely packed with photosynthetic proteins, in addition to other regions composed of protein-free lipid bilayers. Clustering of complexes is a functional necessity, as each light-harvesting component must pass its harvested energy to a neighboring complex and eventually to the RC (Scheuring et al., 2004a). There is no fixed structural assembly of LH2 and core complexes; core complexes completely surrounded by LH2 (Fig. 2a, b; 1) and core complexes making multiple core-core contacts (Fig. 2a, b; 2) were found (Scheuring et al., 2004a; Scheuring and Sturgis, 2005). However, detailed pair correlation function analysis showed that the most frequent assembly was two core complexes separated by an intercalated LH2 (Scheuring and Sturgis, 2005). The supramolecular assembly of the photosynthetic complexes in Rsp. photometricum membranes from
945
cells grown under different light intensities were studied and compared. In membranes from low-light adapted cells, increased quantities of peripheral LH2 were found. Additional LH2 complexes were not randomly inserted into the membrane but formed para-crystalline hexagonally packed antenna domains (Fig. 3). Core complexes remained in domains in which they were locally more highly concentrated (LH2 rings / core complex = ~3.5), when compared with the average density under low-light cell growth (LH2 rings / core complex = ~7). Indeed, these domains in the low-light adapted membranes resembled the high-light adapted membranes in terms of protein composition and complex distribution (Scheuring and Sturgis, 2005). This indicated that complex assembly followed an eutectic phase behavior with an ideal LH2 rings/core complex ratio ~3.5 independent of the growth conditions, and additional LH2 being synthesized under low-light conditions were integrated in specialized antenna domains (Fig. 3a). The LH2 packing (Fig. 3b) in antenna domains was observed to be rigid, giving the possibility of excluding quinone/quinol diffusion (Scheuring and Sturgis, 2006). RCs that are grouped together, independent of the growth conditions, and formation of antenna domains under low-light conditions, prevent photodamage under high-light conditions and ensure efficient photon capture under low-light conditions (Scheuring and Sturgis, 2005). C. The Photosynthetic Apparatus of Phaeospirillum molischianum The supramolecular architecture of the photosynthetic complexes in native membranes of Phs. molischianum (Gonçalves et al., 2005a) was similar to that of Rsp. photometricum (Scheuring et al., 2004a,c; Scheuring and Sturgis, 2005). The proteins of the photosynthetic apparatus cluster assuring efficient energy transfer between the pigment molecules between neighboring complexes (Scheuring et al., 2004a). In native membranes of Phs. molischianum core complexes are monomeric with the RC completely enclosed by an elliptic LH1 assembly (Fig. 4a, b). A wide variety of protein arrangements exists. About 30% of core complexes were completely surrounded by LH2, and about 70% of the core complexes are connected to other cores. Core complex connectivity may be important for the function of the apparatus when photons are abundant to prevent photodamage. In agreement with the X-ray structure of the Phs.
946
Simon Scheuring
Fig. 4. High-resolution AFM analysis of a native photosynthetic membrane of Phs. molischianum (Gonçalves et al., 2005a). a) Raw data AFM topograph revealing the complex mixture of LH2 and core complexes in the native photosynthetic membrane of Phs. molischianum. b) Surface representation of structural model of the complex assembly outlined in a) in the native membrane of Phs. molischianum.
Fig. 5. High-resolution AFM analysis of an antenna domain in a native photosynthetic membrane of Phs. molischianum (Gonçalves et al., 2005a). a) Raw data AFM topograph revealing an LH2 antenna domain in native photosynthetic membranes of Phs. molischianum. b) Surface representation of structural model of the LH2 assembly outlined in a) in antenna domains in native membranes of Phs. molischianum.
molischianum LH2 complex (Koepke et al., 1996), mainly octameric LH2 complexes were found in the membranes. Surprisingly, octameric LH2 form hexagonal arrays in native membranes (Fig. 5a), as found for nonameric LH2 complexes in chromatophores of Rsp. photometricum (Fig. 3) (Scheuring and Sturgis, 2005). Octameric LH2 has a symmetry mismatch in a hexagonal lattice (Fig. 5b), but would be in symmetry match within a square lattice. In a hexagonal lattice the complex packing density is higher. The observation that both octameric (Phs. molischianum; Gonçalves et al., 2005a) and nonameric (Rsp. photometricum; Scheuring and Sturgis, 2005) LH2 complexes form hexagonal antenna domains, is strong evidence that high packing density is favored over specific protein-protein contacts. This suggests that non-specific effects such as hydrophobic mismatch or a solvation incompatibility between the branched phytyl chains exposed by LH2 pigments and the anisotropic linear aliphatic chains of the phospholipids drive complex assembly. Furthermore, as in membranes of Rsp. photometricum, evidence for variability of LH2 size was found in native chromatophores of Phs. molischianum (Gonçalves et al., 2005a).
sis, the small vesicles were fused by freeze thawing, leading to the formation of membranes with sizes up to 1µm. Due to the ‘mechanics’ of this fusion process, AFM topographs revealed proteins in ‘up and down’ orientations (Fig. 6a) (Scheuring et al., 2005). The LH2 complexes are nonameric rings, inserted with their 9-fold symmetry axis perpendicular to the membrane (Walz et al., 1998; Scheuring et al., 2003b; Gonçalves et al., 2005b). Importantly, high-resolution topographs of PufX-containing core complexes were acquired revealing that core complexes assemble in S-shaped dimers. Each core complex is composed of a clearly open LH1 assembly, with a protein-free gap of about 25 Å width, housing one RC (Fig. 6a). Furthermore, the LH1 assembly of each core is elliptical, with long and short axes of 100 Å and 90 Å. Compared to other core complexes, from Rsp. rubrum (Jamieson et al., 2002), Blc. viridis (Scheuring et al., 2003a), Rsp. photometricum (Scheuring et al., 2004a,c; Scheuring and Sturgis, 2005) and Phs. molischianum, (Gonçalves et al., 2005a), that all feature an elliptical LH1 assembly, the ellipticity is particularly astonishing, since one would expect larger flexibility of an open LH1 assembly. This was further strong evidence that LH1 ellipticity results from specific interactions of the LH1 subunits around the RC. An enlarged and depressed topography was found at the dimer center, in which averaging visualized two tiny protrusions, (Fig. 6b) attributed to the two PufX polypeptides from the two cores (Scheuring et al., 2005). The detailed structure of the Rhodobacter core
D. The Photosynthetic Apparatus of Rhodobacter blasticus The purified native photosynthetic membranes of wild type Rba. blasticus were uniformly small vesicles with a diameter of ~50 nm. In order to make these membranes amenable to high-resolution AFM analy-
Chapter 47
Bacterial Photosynthetic Membranes Imaged by AFM
Fig. 6. High-resolution AFM analysis of the native photosynthetic membrane of Rba. blasticus (Scheuring et al., 2005). a) Raw data AFM topograph revealing the complex mixture of LH2 and core complexes in the fused native photosynthetic membranes of Rba. blasticus. The PufX containing core complexes form stable dimers (~75%) with a stoichiometry of (LH113-RC-PufX)2, a smaller fraction (~25%) is monomeric LH113-RC-PufX complexes. b) Surface representation of structural model of the complex assembly outlined in a) in the native membrane of Rba. blasticus.
complex is currently matter of debate: two views about the precise position of the PufX polypeptides have been presented. The first, based on the high-resolution AFM topographs of Rba. blasticus (Scheuring et al., 2005), and reconstitution experiments and cryo-EM analysis of Rba. sphaeroides core complexes (Scheuring et al., 2004b), proposed a structural model of the subunit organization in which two RCs are each encircled by open ellipses of 13 LH1 α/β-heterodimers and two PufX subunits at the dimer center, holding the dimer together (Scheuring et al., 2005). The second, based on a 8.5 Å-resolution cryo-EM projection map of the Rba. sphaeroides core complex (Qian et al., 2005), proposed a structural model of the subunit organization in which two RCs are each encircled by ellipses of 14 LH1 α/β-heterodimers and two PufX subunits at the peripheries of the S-shaped LH1 assembly. Both models aim to explain the PufX-induced facilitation of quinone/quinol passage and dimerization of core
947
complexes. The first provides a straightforward explanation for dimerization of the Rhodobacter core complex since the PufX polypeptides are at the dimer interface, and therefore allow direct interaction of the PufX transmembrane domains; quinone/quinol exchange across a ~25 Å wide protein-free gap, located where the addition of a 14th LH1 subunit pair is sterically impossible, is proposed to be an indirect effect of the dimer architecture (Scheuring et al., 2005). The second model proposes a gating role for PufX in order to facilitate quinone/quinol exchange, and requires that dimerization of the core complex is mediated by long range PufX interactions between extended N-termini (Qian et al., 2005). These are the two working models to date, awaiting a high-resolution X-ray structure of the dimeric Rhodobacter core complex. Further discussion of this topic can be found in Chapter 9, Bullough et al. Although the membrane fusion process of wild type Rba. blasticus chromatophores impeded a detailed analysis of the large-scale association of the different photosynthetic complexes, regions comprising homogeneously inserted complexes were analyzed (Fig. 6b) (Scheuring et al., 2005). Except for the particular dimeric core complex architecture, the overall distribution of photosynthetic complexes in Rba. blasticus resembled the architectures found in Rsp. photometricum (Scheuring et al., 2004a,c; Scheuring and Sturgis, 2005), Phs. molischianum (Gonçalves et al., 2005a) and Rps. palustris (Scheuring et al., 2006). Core complexes assemble in a non-ordered manner with LH2 complexes, and without specific orientation with respect to each other within the membrane. A different architecture was found in Rba. sphaeroides membranes prepared by the action of sub-solubilizing amounts of detergent, which opened the normally closed vesicle structures so that they could lie flat on the mica surface (Bahatyrova et al., 2004b). In these membranes, dimeric core complexes formed rows intercalated by rows of LH2 rings. The LH1 complex was ideally positioned to function as an energy collection hub, temporarily storing it before transfer to the RC where photochemistry occurs: the elegant economy of the photosynthetic membrane was demonstrated by the close packing of the linear core complex arrays, which are often only separated by narrow ‘energy conduits’ of LH2 just two or three complexes wide. Furthermore, domains with complex mixtures were found (Bahatyrova et al., 2004b).
948 E. The Photosynthetic Apparatus of Rhodopseudomonas palustris The supramolecular architecture of the photosynthetic complexes of Rps. palustris was studied in native membranes from cells grown under different light conditions (Scheuring et al., 2006). One of the characteristics of Rps. palustris is the multiplicity of different LH2 complexes that can be formed, coded for by no fewer than five gene pairs. In high-light adapted membranes, absorbance spectra documented ~50% of standard B800-850 LH2 complexes and ~50% of the specific Rps. palustris B800 low-light LH2 (LL-LH2) complex, while in low-light adapted membranes only ~10% of the antennae were B800-850 LH2 and ~90% B800 LLLH2. The Rps. palustris LL-LH2 was shown to form an octameric ring (Hartigan et al., 2002), while all standard B800-850 LH2 complexes are nonameric in Rps. acidophila (McDermott et al., 1995; Gonçalves et al., 2005b), Rhodovulum sulfidophilum (Montoya et al., 1995), Rba. sphaeroides (Walz et al., 1998; Scheuring et al., 2003b), Rvi. gelatinosus (Scheuring et al., 2001), Rsp. photometricum (Scheuring et al., 2004a,c; Scheuring and Sturgis, 2005) and Rba. blasticus (Scheuring et al., 2005). The exception is the octameric LH2 complex of Phs. molischianum (Koepke et al., 1996; Gonçalves et al., 2005a). The diameters of the LH2 complexes in membranes from low-light adapted Rps. palustris were on average smaller than those in high-light adapted membranes. This finding was corroborated by detailed image analysis and biochemical analysis by fast protein liquid chromatogaphy-gel filtration. Image processing revealed classes with significant signals for either nine-fold or eight-fold symmetry. Fast protein liquid chromatogaphy-gel filtration of detergent solubilized LH2 complexes provided evidence of considerable size heterogeneity in the LH2 antenna of Rps. palustris with the B800 LL-LH2 on average smaller than the B800-850 LH2. Core complex topographs were found to be strongly elliptical with differences in axis length of ~18% (~99 Å vs. ~83 Å). The long ellipsis axis of the antenna assembly coincided with the long axis orientation of the RC. In Rps. palustris, the core complex consists of an elliptical assembly of 15 LH1 subunits and 1 W-helix, the putative quinone/quinol gate, around the RC (Roszak et al., 2003). In the X-ray structure the W-helix is located at the periapsis of the W-LH1 ellipsoid (Roszak et
Simon Scheuring al., 2003). In high-resolution AFM topographs of native membranes a remarkable topography depression of one LH1 subunit width was detected in the antenna assembly, attributed to the W-helix (Fig. 7a, arrowheads). However, no positional restrictions of the W-helix ‘gap’ were found. Core complexes, from which the RC was removed by the AFM tip, revealed ‘relaxed’ circular W-LH1 assemblies (Scheuring et
Fig. 7. High-resolution AFM analysis of a native photosynthetic membrane of Rps. palustris (Scheuring et al., 2006). a) Raw data AFM topograph revealing the complex mixture of LH2 and monomeric core complexes in the native photosynthetic membrane of Rps. palustris. b) Surface representation of structural model of the complex assembly outlined in a) in the native membrane of Rps. palustris.
Chapter 47
Bacterial Photosynthetic Membranes Imaged by AFM
al., 2006). At high resolution, the W-helix ‘gap’ is detected within these circular assemblies, indicating that the W-subunit is in contact with both neighboring α/β-LH1-heterodimers, and that W-LH1 ellipticity is RC-induced. The same behavior of re-circularization of the LH1 assembly upon RC removal was found in Blc. viridis (Scheuring et al., 2003a). The elliptical core complexes were randomly oriented with respect to each other in the native membranes of Rps. palustris (Fig. 7b). Within the different samples studied crystalline areas of core complexes were found (Scheuring et al., 2006). In some ways this is exactly the opposite of what has previously been observed in Rsp. photometricum and Phs. molischianum (Gonçalves et al., 2005a; Scheuring and Sturgis, 2005) where LH2 para-crystals co-habited with a non-crystalline mixture of LH2 and core complexes. However this behavior can easily be understood in the context of the same eutectic model developed previously (Scheuring and Sturgis, 2005), which found a critical LH2 ring / core complex ratio of ~3.5 in heterogeneous domains. Membranes that contained more LH2 integrated these complexes into para-crystalline antenna domains. The eutectic phase behavior of the complexes predicts that in cases of excess of RCs (i.e., LH2 ring / core complex ratio <<3.5), para-crystalline domains of core complexes must be formed. AFM topographs of high-light adapted membranes are consistent with this prediction, and a mixture of crystalline core complexes and a non-crystalline mixture of LH2 and core complexes (Scheuring et al., 2006) is observed. It is interesting to note that in the low-light membranes the LL-LH2 rings have a tendency to cluster in para-crystalline antenna domains even though fewer than 3.5 rings are present per core complex. Two explanations may apply: first, the reduced size heterogeneity in low-light membranes (~90% LLLH2) favors clustering, or, second, LL-LH2 has a greater inherent tendency to form para-crystalline antenna domains. III. Conclusions Taken together, several novel advances have emerged from high-resolution AFM analyses of native membranes of different purple phototrophic bacteria, which can be combined to form a general picture. All photosynthetic protein complexes have a strong tendency to cluster together. In cases when
949
the membrane surface exceeds the area needed to house these complexes pure lipid bilayer areas coexist with protein packed domains; the photosynthetic complexes prefer to cluster rather than to diffuse individually in the membranes (Scheuring et al., 2004a, 2006). This is in agreement with the function of the proteins involved in excitation energy transfer, for which direct contact between neighboring complexes is a functional necessity. Alterations in incident light levels induce a variability in size, relative abundance and membrane organization of different types of LH2 antenna complexes in native membranes. The larger the LH2 complex the farther its absorption towards longer wavelength (Scheuring et al., 2004a). There is no fixed assembly between LH2 and core complexes (RC-LH1). Depending on the species, core complexes can be completely surrounded by LH2, or they can make multiple contacts with other cores (Scheuring et al., 2004a; Bahatyrova et al., 2004b; Gonçalves et al., 2005a; Scheuring and Sturgis, 2005). However, the most frequent assembly pattern is two core complexes intercalated by one LH2 (Scheuring and Sturgis, 2005). This complex assembly follows eutectic phase behavior where ~3.5 LH2 complexes per core complex are assembled, and additional LH2 formed under low-light conditions segregate in specialized antenna domains. The fact that both nonameric (Rsp. photometricum; Scheuring and Sturgis, 2005) and octameric (Phs. molischianum; Gonçalves et al., 2005a) LH2 form hexagonal antenna domains is strong evidence for non-specific forces, rather than specific interactions, driving complex assembly. These antenna domains are probably excluded from quinone/quinol diffusion (Scheuring and Sturgis, 2006). In cases where fewer than 3.5 LH2 are synthesized per core complex, excess cores form hexagonal arrays (Scheuring et al., 2006). This description details the architecture of photosynthetic complexes in Rsp. photometricum (Scheuring et al., 2004a,c; Scheuring and Sturgis, 2005), Phs. molischianum (Gonçalves et al., 2005a) and Rps. palustris (Scheuring et al., 2006). In Rba. blasticus LH2 complexes mix with dimeric core complexes in a non-ordered manner (Scheuring et al., 2005), in contrast to the assembly found in Rba. sphaeroides where rows of dimeric core complexes are intercalated by rows of two LH2 (Bahatyrova et al., 2004b). Finally, different core complex architectures were found among species (Scheuring, 2006). The most confusing aspect of the studies on the native architectures of the photosynthetic mem-
950 branes of different photosynthetic bacteria species is the absence of the cytochrome bc1 complex in the membrane areas imaged by AFM, while functional measurements demand connectivity of cytochrome bc1 complexes and core complexes. Two explanations may apply: (i) membrane fragments that do not contain significant amounts of cytochrome bc1 complexes adsorb preferentially to the AFM support, or (ii) cytochrome bc1 complexes are not in proximity to many core complexes and the function-based structural concept comprising cytochrome bc1 and core supercomplexes must be reviewed. The discrepancy between the photosynthetic complex architecture in native membranes reported by AFM images and models based on functional measurements is the major issue that must be addressed now. For further discussion of these topics see Chapter 14, Sturgis and Niederman. High-resolution AFM topographs of supramolecular assemblies of membrane proteins can successfully be docked by atomic models resulting in structural models with high precision (Scheuring et al., 2007). Indeed, the spatial localization and the rotational alignment of the complexes can be defined with a precision of ~0.1 Å and ~4º respectively. This implies that atoms at complex interfaces and pigment positions are defined with a precision of ~3 Å. Structural models are shown in this chapter in figures 1c, 2b, 3b, 4b, 5b, 6b, and 7b, and can provide the structural basis for a deepened understanding of the molecular mechanisms of ensemble reactions in the bacterial photosynthetic apparatus, i.e., excitation transfer (Scheuring et al., 2007; Chapter 15, Şener and Schulten). IV. Feasibilities, Limitations and Outlook Use of the AFM is a fairly poor screening technique. First, scanners that are precise enough to be used for high-resolution image acquisition are limited in their full scan range ability to somewhat more than 100 µm square providing analysis of only a limited number of membranes per experiment. This problem may be overcome by precise adjustment of sample quantity used per experiment, in order to have as many membranes as possible adsorbed per area without incurring the risk that they stack or aggregate. Second, AFM imaging needs tightly adsorbed membranes. Adsorption is related to surface exposed charges and can be triggered by ionic strength in
Simon Scheuring the adsorption buffer solution (Müller et al., 1997). This is not always trivial, since retention of a native sample must be considered as determining factor for screening adsorption buffer conditions. Vesicular samples can be imaged, but it is technically more demanding than imaging sheet-like membranes. The smaller the vesicles are the more tricky this becomes (Gonçalves et al., 2006), probably because surface tension of small vesicles hampers flattening and adsorption of the spherical object to the support. This problem may be overcome by extensive screening of adsorption buffers, in order to find conditions where small vesicles tightly adsorb and open upon support adsorption. Another possibility is to use the tip as a nano-dissector to open vesicles mechanically (Scheuring et al., 2004c). Both approaches are invasive. Equally, it is possible to fragment small vesicles through addition of low concentrations of detergent (Bahatyrova et al., 2004b) or by freeze-thawing cycles (Scheuring et al., 2005). However, these procedures are undesirable since they may significantly alter the native long-range order of the complex assembly in the native membrane. Clearly, the major trump card of the AFM is its outstanding signal-to-noise ratio, i.e. its capability to image single molecules. This feature makes AFM to date the unique technique to visualize non-ordered supramolecular membrane protein assemblies in native membranes, where techniques that imply averaging or ensemble measurements fail. Some technical developments indicate that the AFM will strengthen its abilities to analyze membrane protein supramolecular assemblies in the near future. Using the AFM tip as a nano-manipulator or a force probe may allow the study of intra- and intermolecular forces of the complexes, providing insights to the underlying physical parameters that drive complex assembly. On the other hand, the simultaneous measurement of topography and either light or conductance properties using light or current sensitive probes may provide functional characterizations at the single molecule level. AFM tips with nano-tubes may provide improved reproducibility of high-resolution image acquisition and higher resolution (Cheung et al., 2000). Fast scanning AFMs that can acquire images at rates sufficient to track motion of individual molecules may allow the visualization of flexible and diffusing molecules (Viani et al., 2000; Ando et al., 2001).
Chapter 47
Bacterial Photosynthetic Membranes Imaged by AFM
Acknowledgments The author thanks former and present collaborators. This study was supported by the Institut Curie, the INSERM (Institut National de la Santé et la Recherche Médicale), an INSERM Avenir 2005 award, and a ‘ACI Nanosciences 2004’ grant (NR206). References Abrahams JP, Leslie AGW, Lutter R and Walker JE (1994) Structure at 2.8 Å resolution of F1-ATPase from bovine heart mitochondria. Nature 370: 621–628 Allen JP, Feher G, Yeates TO, Komiya H and Rees DC (1987) Structure of the reaction center from Rhodobacter sphaeroides R-26: The protein subunits. Proc Natl Acad Sci USA 84: 6162–6166 Ando T, Kodera N, Takai E, Maruyama D, Saito K and Toda A (2001) A high-speed atomic force microscope for studying biological macromolecules. Proc Natl Acad Sci USA 98: 12468–12472 Bahatyrova S, Frese RN, Van Der Werf KO, Otto C, Hunter CN and Olsen JD (2004a) Flexibility and size heterogeneity of the LH1 light harvesting complex revealed by atomic force microscopy: functional significance for bacterial photosynthesis. J Biol Chem 279: 21327–21333 Bahatyrova S, Frese RN, Siebert CA, Olsen JD, Van Der Werf KO, van Grondelle R, Niederman RA, Bullough PA and Hunter CN (2004b) The native architecture of a photosynthetic membrane. Nature 430: 1058–1062 Berry EA, Huang L-S, Saechao LK, Pon NG, Valkova-Valchanova M and Daldal F (2004) X-ray structure of Rhodobacter capsulatus cytochrome bc1: Comparison with its mitochondrial and chloroplast counterparts. Photosynth Res 81: 251–275 Cheung CL, Hafner JH and Lieber CM (2000) Carbon nanotube atomic force microscopy tips: Direct growth by chemical vapor deposition and application to high-resolution imaging. Proc Natl Acad Sci USA 97: 3809–3813 Deisenhofer J and Michel H (1989) Nobel Lecture: The photosynthetic reaction centre from the purple bacterium Rhodopseudomonas viridis. EMBO J 8: 2149–2170 Deisenhofer J, Epp O, Miki K, Huber R and Michel H (1984) Xray structure analysis of a membrane protein complex. Electron density map at 3 Å resolution and a model of the chromophores of the photosynthetic reaction center from Rhodopseudomonas viridis. J Mol Biol 180: 385–398 Deisenhofer J, Epp O, Miki K, Huber R and Michel H (1985) Structure of the protein subunits in the photosynthetic reaction centre of Rhodopseudomonas viridis at 3Å resolution. Nature 318: 618–624 Deisenhofer J, Epp O, Sinning I and Michel H (1995) Crystallographic refinement at 2.3Å resolution and refined model of the photosynthetic reaction centre from Rhodopseudomonas viridis. J Mol Biol 246: 429–457 Engelhardt H, Engel A and Baumeister W (1986) Stoichiometric model of the photosynthetic unit of Ectothiorhodospira halochloris. Proc Natl Acad Sci USA 83: 8972–8976 Fotiadis D, Qian P, Philippsen A, Bullough PA, Engel A and
951
Hunter CN (2004) Structural analysis of the reaction center light-harvesting complex I photosynthetic core complex of Rhodospirillum rubrum using atomic force microscopy. J Biol Chem 279: 2063–2068 Gonçalves RP, Bernadac A, Sturgis JN and Scheuring S (2005a) Architecture of the native photosynthetic apparatus of Phaeospirillum molischianum. J Struct Biol 152: 221–228 Gonçalves RP, Busselez J, Lévy D, Seguin J and Scheuring S (2005b) Membrane insertion of Rhodopseudomonas acidophila light harvesting complex 2 (LH2) investigated by high resolution AFM. J Struct Biol 149: 79–86 Gonçalves RP, Agnus G, Sens P, Houssin C, Bartenlian B and Scheuring S (2006) 2-Chamber-AFM: Probing membrane proteins separating two aqueous compartments. Nature Methods 3: 1007–1012 Hartigan N, Tharia HA, Sweeney F, Lawless AM and Papiz MZ (2002) The 7.5-Å electron density and spectroscopic properties of a novel low-light B800 LH2 from Rhodopseudomonas palustris. Biophys J 82: 963–977 Hoh JH, Lal R, John SA, Revel J-P and Arnsdorf MF (1991) Atomic force microscopy and dissection of gap junctions. Science 253: 1405–1408 Hu X and Schulten K (1998) Model for the light-harvesting complex I (B875) of Rhodobacter sphaeroides. Biophys J 75: 683–694 Hu X, Ritz T, Damjanovic A, Autenrieth F and Schulten K (2002) Photosynthetic apparatus of purple bacteria. Quart Rev Biophys 35: 1–62 Ikeda-Yamasaki I, Odahara T, Mitsuoka K, Fujiyoshi Y and Murata K (1998) Projection map of the reaction center-light harvesting 1 complex from Rhodopseudomonas viridis at 10 Å resolution. FEBS Lett 425: 505–508 Jacob JS and Miller KR (1983) Structure of a bacterial photosynthetic membrane. Arch Biochem Biophys 223: 282–290 Jacob JS and Miller KR (1984) Structure of a bacterial photosynthetic membrane: Integrity of reaction centers following proteolysis and detergent solubilization. Biochem Biophys Res Com 120: 164–171 Jamieson SJ, Wang P, Qian P, Kirkland JY, Conroy MJ, Hunter CN and Bullough PA (2002) Projection structure of the photosynthetic reaction centre-antenna complex of Rhodospirillum rubrum at 8.5Å resolution. EMBO J 21: 3927–3935 Koepke J, Hu X, Muenke C, Schulten K and Michel H (1996) The crystal structure of the light-harvesting complex II (B800-850) from Rhodospirillum molischianum. Structure 4: 581–597 Kurisu G, Zhang H, Smith JL and Cramer WA (2003) Structure of the cytochrome b6 f complex of oxygenic photosynthesis: Tuning the cavity. Science 302: 1009–1014 Lancaster CR, Bibikova M, Sabatino P, Oesterhelt D and Michel H (2000) Structural basis of the drastically increased initial electron transfer rate in the reaction center from a Rhodopseudomonas viridis mutant described at 2.00-Å resolution. J Biol Chem 275: 39364–39368 McDermott G, Prince SM, Freer AA, Hawthornthwaite-Lawless AM, Papiz MZ, Cogdell RJ and Isaacs NW (1995) Crystal structure of an integral membrane light-harvesting complex from photosynthetic bacteria. Nature 374: 517–521 Meier T, Polzer P, Diederichs K, Welte W and Dimroth P (2005) Structure of the rotor ring of F-Type Na+-ATPase from Ilyobacter tartaricus. Science 308: 659–662 Miller KR (1982) Three-dimensional structure of a photosynthetic
952 membrane. Nature 300: 53–55 Montoya G, Cyrklaff M and Sinning I (1995) Two-dimensional crystallization and preliminary structure analysis of light harvesting II (B800-850) complex from the purple bacterium Rhodovulum sulfidophilum. J Mol Biol 250: 1–10 Müller DJ, Amrein M and Engel A (1997) Adsorption of biological molecules to a solid support for scanning probe microscopy. J Struct Biol 119: 172–188 Papiz MZ, Prince SM, Hawthornthwaite-Lawless AM, McDermott G, Freer AA, Isaacs NW and Cogdell RJ (1996) A model for the photosynthetic apparatus of purple bacteria. Trends Plant Sci 1: 198–206 Qian P, Hunter CN and Bullough PA (2005) The 8.5Å projection structure of the core RC-LH1-PufX dimer of Rhodobacter sphaeroides. J Mol Biol 349: 948–960 Roszak AW, Howard TD, Southall J, Gardiner AT, Law CJ, Isaacs NW and Cogdell RJ (2003) Crystal structure of the RC-LH1 core complex from Rhodopseudomonas palustris. Science 302: 1969–1972 Scheuring S (2006) AFM studies of the supramolecular assembly of bacterial photosynthetic core-complexes. Curr Opinion Chem Biol 10: 387–393 Scheuring S and Sturgis JN (2005) Chromatic adaptation of photosynthetic membranes. Science 309: 484–487 Scheuring S and Sturgis JN (2006) Dynamics and diffusion in photosynthetic membranes from Rhodospirillum photometricum. Biophys J 91: 3707–3717 Scheuring S, Reiss-Husson F, Engel A, Rigaud J-L and Ranck J-L (2001) High resolution topographs of the Rubrivivax gelatinosus light-harvesting complex 2. EMBO J 20: 3029–3035 Scheuring S, Seguin J, Marco S, Lévy D, Robert B and Rigaud JL (2003a) Nanodissection and high-resolution imaging of the Rhodopseudomonas viridis photosynthetic core-complex in native membranes by AFM. Proc Natl Acad Sci USA 100: 1690–1693 Scheuring S, Seguin J, Marco S, Lévy D, Breyton C, Robert B and Rigaud J-L (2003b) AFM Characterization of tilt and intrinsic flexibility of Rhodobacter sphaeroides light harvesting complex 2 (LH2). J Mol Biol 325: 569–580 Scheuring S, Rigaud J-L and Sturgis JN (2004a) Variable LH2 stoichiometry and core clustering in native membranes of Rhodospirillum photometricum. EMBO J 23: 4127–4133 Scheuring S, Francia F, Busselez J, Melandri B, Rigaud J-L and Lévy D (2004b) Structural role of PufX in the dimerization of the photosynthetic core-complex of Rhodobacter sphaeroides.
Simon Scheuring J Biol Chem 279: 3620–3626 Scheuring S, Sturgis JN, Prima V, Bernadac A, Lévy D and Rigaud J-L (2004c) Watching the photosynthetic apparatus in native membranes. Proc Natl Acad Sci USA 101: 11293–11297 Scheuring S, Busselez J and Levy D (2005) Structure of the dimeric PufX-containing core complex of Rhodobacter blasticus by in situ AFM. J Biol Chem 180: 1426–1431 Scheuring S, Gonçalves RP, Prima V and Sturgis JN (2006) The photosynthetic apparatus of Rhodopseudomonas palustris: Structures and organization. J Mol Biol 358: 83–96 Scheuring S, Boudier T and Sturgis JN (2007) From high-resolution AFM topographs to atomic models of supramolecular assemblies. J Struct Biol 159: 268–276 Siebert CA, Qian P, Fotiadis D, Engel A, Hunter CN and Bullough PA (2004) Molecular architecture of photosynthetic membranes in Rhodobacter sphaeroides: The role of PufX. EMBO J 23: 690–700 Stamouli A, Kafi S, Klein DC, Oosterkamp TH, Frenken JW, Cogdell RJ and Aartsma TJ (2003) The ring structure and organization of light harvesting 2 complexes in a reconstituted lipid bilayer, resolved by atomic force microscopy. Biophys J 84: 2483–2491 Stark W, Kühlbrandt W, Wildhaber I, Wehrli E and Muhlethaler K (1984) The structure of the photoreceptor unit of Rhodopseudomonas viridis. EMBO J 3: 777–783 Stock D, Leslie AG and Walker JE (1999) Molecular architecture of the rotary motor in ATP synthase. Science 286: 1700–1705 Stroebel D, Choquet Y, Popot JL and Picot D (2003) An atypical haem in the cytochrome b6 f complex. Nature 426: 413–418 Sundström V, Pullerits T and van Grondelle R (1999) Photosynthetic light-harvesting: Reconciling dynamics and structure of purple bacterial LH2 reveals function of photosynthetic unit. J Phys Chem B 103: 2327–2346 Viani MB, Pietrasanta LI, Thompson JB, Chand A, Gebeshuber IC, Kindt JH, Richter M, Hansma HG and Hansma PK (2000) Probing protein-protein interactions in real time. Nature Struct Biol 7: 644–647 Walz T, Jamieson SJ, Bowers CM, Bullough PA and Hunter CN (1998) Projection structures of three photosynthetic complexes from Rhodobacter sphaeroides: LH2 at 6Å, LH1 and RC-LH1 at 25Å. J Mol Biol 282: 833–845 Xia D, Yu CA, Kim H, Xia JZ, Kachurin AM, Zhang L, Yu L and Deisenhofer J (1997) Crystal structure of the cytochrome bc1 complex from bovine heart mitochondria. Science 277: 60–66
Chapter 48 Protein Environments and Electron Transfer Processes Probed with High-Frequency ENDOR Oleg G. Poluektov* and Lisa M. Utschig Chemistry Division, Argonne National Laboratory, 9700 S. Cass Ave., Argonne, IL 60439 U.S.A.
Summary .............................................................................................................................................................. 953 I. Introduction..................................................................................................................................................... 954 II. Low Temperature Interquinone Electron Transfer in the Photosynthetic Reaction Center. Characterization of QB– States ......................................................................................................................... 956 A. Background. Electron Transfer Reactions in Photosynthesis .......................................................... 956 B. Interquinone Electron Transfer at Low Temperature........................................................................ 957 C. Trapping Kinetically Distinct Substates at Low Temperature ........................................................... 957 D. Probing Protein Environments with High Frequency Mims-type ENDOR ........................................ 959 1. Sensitivity to Local Structure and Quinone Rotation ............................................................... 959 2. The Local Environment of the QB– Site in Kinetically Distinct P+QB– States ............................... 961 III. Electron Transfer Pathways and Protein Response to Charge Separation ................................................... 963 A. Background ...................................................................................................................................... 963 B. High Frequency Spin-Correlated Radical Pair ENDOR of the P+QA– Charge-Separated State ........ 964 1. Experimental Results .............................................................................................................. 964 2. Sensitivity of Spin-Correlated Radical Pair ENDOR for Distant Nuclei ................................... 965 3. Interpretation of Spin-Correlated Radical Pair ENDOR Spectra ............................................. 966 4. The Role of the Exchangeable Protons in Regulation of Electron Transfer ............................ 967 C. Protein and Cofactor Relaxation Following Charge Separation ....................................................... 967 IV. Concluding Remarks ...................................................................................................................................... 968 Acknowledgments ................................................................................................................................................. 969 References ............................................................................................................................................................ 969
Summary Natural photosynthetic conversion of solar energy to chemical energy is a unique phenomenon which sustains all life on Earth. The key step of photosynthetic energy conversion involves rapid, photoinduced sequential electron transfers (ET) resulting in efficient charge separation across a biological membrane. Fundamental to fully understanding these ET events is discerning the involvement of heterogeneous polypeptide environments surrounding the redox cofactor sites. High-resolution X-ray crystal structures of reaction center (RC) proteins reveal the structure of cofactors and surrounding protein environments; however, static crystallographic protein structures do not readily yield details of how native dynamic solution protein structures fine-tune ET processes and coupled reactions, such as proton transfer. Thus, novel approaches complementary to crystallography are required to experimentally correlate structural, electrostatic, and dynamic features of localized protein environments with inherent ET reactions. In this chapter we discuss one of the advanced experimental *Author for correspondence, email: [email protected]
C. Neil Hunter, Fevzi Daldal, Marion C. Thurnauer and J. Thomas Beatty (eds): The Purple Phototrophic Bacteria, pp. 953–973. © 2009 Springer Science + Business Media B.V.
954
Oleg G. Poluektov and Lisa M. Utschig
techniques being developed to tackle this problem. Using recent examples from our laboratory, we demonstrate the potential of high-frequency (HF), time-resolved (TR) electron-nuclear double resonance (ENDOR) spectroscopy to reveal unique information about structure/function relationships in photosynthetic ET. First, we discuss light-induced protein conformational states of P+QB– trapped in purple photosynthetic bacterial RCs at low temperature and matrix ENDOR techniques for examining the local protein environment surrounding QB– for these states. Second, we describe the HF TR ENDOR study of the spin-correlated radical pair P+QA–. These experiments permit us to directly probe light-induced reorganization of the protein in response to the rapid formation of the charge-separated state. General principles obtained from the novel application of these methodological approaches to native biological systems will help in the development of biomimetic solar energy conversion assemblies. I. Introduction The photosynthetic system is an elaborate nano-scale biological machine that carries out solar energy conversion. The ultimate goal of natural photosynthetic research is to obtain fundamental knowledge of how photochemical processes at the molecular level are linked to the chemistry of macroscopic energy conversion. This knowledge is crucial for the future design and optimization of novel biomimetic and model artificial solar energy conversion systems. The key step of photosynthesis is a cascade of photoinduced ET processes between donor and acceptor molecules embedded in reaction center (RC) proteins. From numerous, unsuccessful attempts to fully mimic natural photosynthetic multistep ET with artificial systems constructed of covalently bound donors and acceptors, it is clear that the RC protein matrix plays a crucial role in photochemical charge separation and charge stabilization. Correlating details of how the reaction medium influences and controls ET reactions remains a significant experimental challenge. Protein conformational changes play an important part in the functioning of many proteins (O’Halloran, 1993; Utschig et al., 1995; Srajer et al., 1996; Genick et al., 1997; Luecke et al., 1999; Petsko and Ringe, 2000; Schlichting and Chu, 2000; Deng et al., 2001; Hummer et al., 2004). If the conformational changes are large-scale in nature, they often can be probed spectroscopically with X-ray diffraction, magnetic Abbreviations: A1 – phylloquinone acceptor; Blc. – Blastochloris; DAF – delay after flash; ENDOR – electron-nuclear double resonance; EPR – electron paramagnetic resonance; ESE – electron spin echo; ET – electron transfer; HF – high-frequency; HFI – hyperfine interactions; MW – microwave; P – primary donor; P700 – primary chlorophyll electron donor; Q – quinone; Rba. – Rhodobacter; RC – reaction center; RF – radiofrequency; SCRP – spin-correlated radical pair; TR – time-resolved; UQ – ubiquinone
resonance, or optical spectroscopic techniques. However, several attempts to discern similar local, largescale structural rearrangements in photosynthetic RCs have been unsuccessful (Mulkidjanian, 1999; Breton et al., 2002; Nabedryk et al., 2003; Remy and Gerwert, 2003; Baxter et al., 2004a,b, 2005; Breton, 2004; Breton et al., 2004; Pokkuluri et al., 2004). The reason for this remains unclear: do RC proteins not have inherent conformational substates linked to specific ET steps; or, are the available spectroscopic tools insensitive to detecting existing structural changes? Advanced spectroscopic techniques are necessary to address these questions. Optical spectroscopy is the most commonly used tool for studying ET processes in RCs. Optical experiments are often complicated because of problems with distinguishing primary optical markers from the cofactors involved in different ET pathways, making it difficult to correlate kinetics directly to specific charge separation events (Li et al., 1998). Thus, secondary electrochromic markers that monitor the charge distribution are often used in optical experiments (Verméglio and Clayton, 1977; Takahasi et al., 1992; Tiede et al., 1996). However, with this approach it is challenging to distinguish between ET and secondary processes, such as proton transfer or protein rearrangement (Tiede et al., 1996, 1998). The latest progress in photosynthesis research is undoubtedly linked to the high resolution X-ray RC crystal structures which reveal cofactor geometries and structures of the protein environments surrounding the cofactors (Deisenhofer et al., 1984; Allen et al., 1988; Chang et al., 1991; Jordan et al., 2001; Zouni et al., 2001; Ferreira et al., 2004; Loll et al., 2005). X-ray crystal structures provide information on the static protein structure, yet do not readily yield information on how inherent dynamic protein structure is related to function. Also, the possibility of radiation damage to the protein, i.e., X-ray induced
Chapter 48
Protein Environments and Electron Transfer Processes
photoelectrons trapped near or at the cofactor sites, complicates interpretation of X-ray data. In contrast to data from X-ray crystallography, electron paramagnetic resonance (EPR) spectroscopy can provide structural information concurrent with functional data for RCs in solution. Importantly, because EPR detects only radical species, radical intermediates of sequential ET photosynthetic processes can be directly monitored. One drawback of EPR spectroscopy, however, is the relatively lower time resolution of the technique compared to optical spectroscopy. TR EPR is sensitive to transient processes on the 100 ns or longer time-scale. For this reason, a number of short-lived radical intermediate species can not be detected by EPR. However, EPR spectroscopy is proving to be indispensable because of its high sensitivity to local cofactor environments and ability to acquire concomitant structural and kinetic data. Recent developments in advanced experimental EPR techniques, namely pulsed and high-frequency (HF) EPR have given further momentum to the application of EPR in the fields of biophysics and biochemistry. The main advantage of HF EPR compared to conventional X-band (9 GHz) EPR is a high absolute sensitivity and high spectral resolution (Kothe et al., 2008). The sensitivity of the EPR spectrometer increases with frequency in the range ν1/2 – ν9/2, depending upon the type of spectrometer and characteristics of the sample (Ernst et al., 1987; Weil et al., 1994). Increased sensitivity is especially important for biological applications where the amount of material is limited to small quantities. The spectral resolution can be characterized by the splitting δB of the two lines with g-factors g1 and g2: δB = (hν/βe) (1/g1+1/g2). Thus, the splitting increases linearly with the frequency. For the two most widely used high-frequency bands, W-band (95 GHz) and D-band (130 GHz), the increase in g-value resolution compared to X-band is 10- and 14-fold, respectively. When recorded at conventional microwave frequencies (~9 GHz), the signals of the RC oxidized electron donors and reduced electron acceptors overlap significantly, however, when measured at HF these signals are resolved. Detailed descriptions of HF EPR techniques and their applications can be found in recent reviews (Grinberg and Berliner, 2004; Thurnauer et al., 2004; Möbius et al., 2005). As will be described in this chapter, a combination of pulsed, TR, HF EPR and electron-nuclear double resonance (ENDOR) spectroscopies in combination with selec-
955
tive isotopic substitution provides new insights into the structure and electron-transfer properties of RC proteins. The ENDOR technique is a superior tool for studying local environments of paramagnetic centers by analyzing weak hyperfine interactions (HFI) with surrounding magnetic nuclei (Goldfarb and Arieli, 2004; Lendzian, 2005; Möbius et al., 2005). ENDOR allows the detection of nuclear magnetic resonance (NMR) spectra through EPR transitions and, consequently, combines the advantages of both magnetic resonance techniques. An important feature of the ENDOR technique is its high selectivity of NMR transitions. ENDOR is sensitive only to specific nuclei that are interacting with a specific electron spin. This feature establishes the main application of ENDOR spectroscopy as a probe of electronic structure, wave-function delocalization, and structure of the immediate (local) environment of a paramagnetic center (Dorio and Freed, 1979; Schweiger and Jeschke, 2001; Goldfarb and Arieli, 2004). Although the technique was initially developed for applications in solid state physics, particularly for the study of defects in semiconductor and dielectric materials (Feher et al., 1957; Eisinger and Feher, 1958), ENDOR soon became a crucial tool for biophysical research, such as studies of photosynthesis. (Goldfarb and Arieli, 2004; Möbius et al., 2005). For example, the structural characterization of the primary electron donor in the photosynthetic purple bacterial RC as a dimer of bacteriochlorophyll molecules was unequivocally established with ENDOR spectroscopy (Feher et al., 1973, 1975; Norris et al., 1973, 1974). In this chapter, we discuss new methodologies developed in our laboratory that utilize both specialized samples and pulsed, TR, HF ENDOR/EPR techniques for the study of light-induced protein conformational substates related to ET in the RCs from the purple bacterium Rhodobacter (Rba.) sphaeroides. First, we discuss the application of the matrix ENDOR technique to the study of low temperature interquinone ET. Second, we describe the high-frequency TR ENDOR of a spin-correlated radical pair state. We demonstrate that these experiments permit mapping of ET pathways within the protein and directly probe light-induced reorganization of the protein in response to rapid formation of the charge-separated state.
956
Oleg G. Poluektov and Lisa M. Utschig
II. Low Temperature Interquinone Electron Transfer in the Photosynthetic Reaction Center. Characterization of QB– States A. Background. Electron Transfer Reactions in Photosynthesis The photosynthetic bacterial reaction center (RC) protein is an integral membrane protein that couples light-induced sequential ET with proton-transfer reactions. High-resolution crystal structures show that the bacterial RC from Rba. sphaeroides (Allen et al., 1988; Chang et al., 1991 ; Ermler et al., 1994) consists of three 30–35 kDa protein subunits, L, H, and M, which bind nine cofactors: four bacteriochlorophylls (BChl), two bacteriopheophytins (BPh), two ubiquinones (UQ) and one internally bound non-heme iron. In the RC, ET occurs sequentially following photoexcitation of a bacteriochlorophyll dimer (P), through one set of cofactors, to a quinone molecule QA within 150–250 ps, resulting in a metastable charge-separated state, P+QA– (Parson, 1987; Feher et al., 1989). Subsequently, within about 200 µs, the electron reaches the final quinone acceptor QB which functions as a two electron, two proton acceptor following two successive turnovers of the RC photochemistry (Fig. 1) (Okamura et al., 2000). The heterogeneous kinetics, temperature trends, and pH dependencies of the QA– QB → QAQB– reaction show that this interquinone ET in isolated RCs is intimately linked to a complex conformational landscape (Mancino et al., 1984; Tiede et al., 1996; Li et al., 1998; Balabin and Onuchic, 2000; Schmid and Labahn, 2000; Xu and Gunner, 2001; Xu et al., 2002). The rate of the QA– QB → QAQB– ET at room temperature is independent of the driving force for the reaction, at least for the main ~100 µs component (Li et al., 1998) seen in isolated Rba. sphaeroides RCs, demonstrating that this step is limited by a protein conformational change (Graige et al., 1998). Low temperature kinetics also indicate that a structural change accompanies the first interquinone ET step. QA– QB → QAQB– ET is not observed at low temperatures (< 200 K) for RCs frozen in the dark. However, ET between the quinones does occur at cryogenic temperatures for Rba. sphaeroides RCs frozen under illumination (the so-called ‘Kleinfeld effect’; Kleinfeld et al., 1984). These observed alterations of reaction kinetics have been linked to trapping the RC in altered conformations induced by charge separation (Kleinfeld et al., 1984a). Similarly, RCs that lack QB exhibit a light-induced structural change. This is evidenced by the remark-
Fig. 1. The arrangement of cofactors of the RC from Rba. sphaeroides as revealed by X-ray crystallography. The protein matrix is not shown for clarity. P is a special pair of bacteriochlorophyll (BChl) molecules, BA/B BChls, HA/B bacteriopheophytins, QA/B ubiquinones situated around non-heme Fe ion. The pathways and time constants of the rapid sequential transfer steps are indicated by arrows. At low temperatures ET from QA– to QB is blocked, and the electron on QA– returns to P+.
ably prolonged lifetime of the radical pair state P+QA– formed from these RCs cooled to cryogenic temperatures under illumination compared with RCs frozen in the dark (Kleinfeld et al., 1984a). A number of spectroscopic tools were used to determine that the altered recombination kinetics for the former case were not the result of reorientation of the cofactors P+ and QA– (van den Brink et al., 1994; Bittl et al., 1995). On the other hand, the low temperature QA–QB → QAQB– ET has been proposed to be linked to two distinct sites of QB, the ‘distal’ and ‘proximal’ positions, observed in Rba. sphaeroides RC crystal structures. The position of QB– as determined from X-ray crystal structures of light-adapted RC crystals in the charge separated state P+QB– is located approximately 5 Å from the QB position in the charge-neutral state PQB, and has undergone a 180° propeller twist around the isoprene chain (Stowell et al., 1997). This significant shift in position and rotation of QB in the light-adapted structures relative to those adapted in the dark led to the proposal that the proximal and distal QB positions correspond to active and inactive conformations with respect to ET from QA– to QB and that movement between these two configurations represents the conformational gate for this reaction.(Stowell et al., 1997) Thus,
Chapter 48
Protein Environments and Electron Transfer Processes
957
for RCs frozen in the dark, it was suggested that QB is locked in the distal or ‘inactive’ conformation, whereas QB is shifted to the proximal or ‘active’ conformation for RCs frozen in the light. However, recent experimental results contradict this proposal and the correlation between the shift in QB configuration and rate-limiting conformational change has not been fully established (Kuglstatter et al., 2001; Pokkuluri et al., 2002, 2004; Xu et al., 2002; Baxter et al., 2004). In fact, Fourier transform infrared (FTIR) spectroscopic data, as well as optical measurements of structurally characterized RC mutants with QB in the proximal position argue against a large-scale, light-induced shift in the position of QB as a part of the gating mechanism for RCs in non-crystalline states (Breton et al., 2002; Xu et al., 2002; Nabedryk et al., 2003; Breton, 2004). Thus, the nature of these light-induced structural changes in RCs having both quinones bound remains a mystery. B. Interquinone Electron Transfer at Low Temperature To investigate light-induced protein conformational substates related to low-temperature QA– QB → QAQB– ET we have used HF pulsed EPR and ENDOR techniques. The g-tensor resolution of QA– and QB– at HF EPR allows us to spectroscopically distinguish these radical species, enabling the selective HF ENDOR of QB– and QA– sites to be obtained. As an example of HF EPR resolution enhancement, electron spin echo (ESE) induced field swept EPR spectra of P+QA– and P+QB– from protonated Fe-removed/Zn-replaced Rba. sphaeroides RCs substituted with deuterated UQ-10 are shown in Fig. 2. Note, the EPR signals were detected as the spin echo signal intensity variation as a function of magnetic field. This leads to an absorption type of signal lineshape, contrary to the first derivative type lineshape observed with conventional (continuous wave) EPR spectroscopic techniques. ESE signals of the transient P+QA– state generated at 20 K by laser pulse excitation were obtained for samples frozen in the dark. P+QB– states were generated by cooling the sample from room temperature to 20 K under 1 Hz, 605 nm laser excitation. Because of the high resolution of the g-tensors at HF, the resonances of QA– and QB– can be distinguished from each other and P+. This allows the formation or decay of each quinone species at low temperature to be directly monitored. In the case of X-band EPR spectroscopy, the signals from P+ and QA– completely overlap (Fig. 2 insert). Thus, unlike X-ray spectroscopy, local structural information can
Fig. 2. High frequency, D-band (130 GHz) and conventional Xband (9.5GHz) (insert) electron spin echo (ESE) induced EPR spectra of protonated Fe-removed/Zn-replaced Rba. sphaeroides RCs substituted with deuterated UQ-10 recorded at 20 K. The P+QA– signal (solid line) was recorded by laser pulse excitation for a sample frozen in the dark. P+QB– signals (dotted line) were generated by cooling RCs from room temperature to 20 K under 1 Hz, 605 nm laser excitation. Simulated spectra of P+ (dashed line) and QA– (dashed-dotted line) are shown for comparison.
be obtained with EPR magnetic resonance techniques for a known redox state of QB, i.e., the semiquinone QB– radical state. C. Trapping Kinetically Distinct Substates at Low Temperature Electron transfer kinetics can be directly measured for light-induced radical species resolved at HF. Two kinetically distinct conformational substates of P+QB– were observed for RCs frozen in the light: an ‘inactive’ P+QB– state and an ‘active’ P+QB– state. A typical sample frozen under illumination contained both a fraction of the ‘inactive’ state and a fraction of the ‘active’ state, as determined by direct monitoring of the EPR signals of P+QB–. For the inactive state, P+QB– is trapped with no charge recombination occurring at 20 K, and is detected as a static P+QB– EPR signal.
958 The active P+QB– state is observed by an increase in intensity of the P+QB– EPR signal at 20 K upon continuous 10 Hz laser excitation at 605 nm, and a decay of this fractional signal increase of the P+QB– EPR signal after the light is turned off. The EPR signals representing inactive and active P+QB– states observed after RCs were shock-frozen in liquid N2 following 3 s illumination at room temperature with red light are similar to the signal of P+QB– shown in Fig. 2. The population of RCs that are active to low temperature QA– QB → QAQB– ET is observed by the increased intensity of the P+QB– signal when the sample is continuously flashed with laser light at 20 K. Figure 3a shows the response of the EPR signal of the sample frozen in the dark at field position 46382 G, (corresponding to the gy canonical component of QB–) to single laser flash excitation (marked by stars) and continuous 10 Hz laser flashes (laser flashes stopped at the position indicated by arrows). After laser excitation is shut off, two kinetic components are observed: a fast decay component that corresponds to the decay of QA– and a slow decay component that corresponds to the decay of the active QB– state. The slow component has a multiexponential decay from which decay constants of ~10 s, ~30 s, and ~2 min can be obtained. These values are identical to the values obtained from kinetic traces observed after single flash or multiple flash excitation. These kinetics are distinct from the signal decay kinetics of ≤ 70 ms measured for QA– in RCs frozen in the ‘dark’ state (Fig. 3b ), or frozen under illumination in the presence of competitive inhibitors of QB (Fig. 3c). Notice that even RCs frozen in the dark still have a very small contribution of the multiexponential slow decay component. In contrast, a pure single exponential is observed in the decay kinetics of RCs that have QB replaced by inhibitor. This ≤70 ms single exponential decay is consistent with charge recombination of P+QA–, indicating that no forward ET occurs past QA– at low temperature. Similar kinetic behavior of the EPR signal was observed recently (Paddock et al., 2006) for native and mutant RCs that lack QA. By analogy to spectra obtained with the QA-less mutant RC, the authors concluded that, for native RCs frozen in the light, QB is reduced by inefficient ET along the B-branch. Our experiments with single shot laser excitation directly dispute this conclusion. Our results show that the efficiency of ET to QB in RCs frozen in the light is high, up to 50%, and, thus, can not be explained
Oleg G. Poluektov and Lisa M. Utschig
Fig. 3. Kinetics measured at 20 K by monitoring the EPR signal intensity at 46382 G. Representative samples frozen under different conditions that primarily have one state present are shown here. (a) Slow, distributed multiexponential decay was observed for QB– in the ‘active’ state. The sample was shock frozen in liquid nitrogen following 3 s illumination with red light at room temperature. (b) Predominantly fast ≤70 ms decay is observed in the sample frozen in the dark. (c) A single decay of ≤70 ms is observed for samples frozen under laser illumination, but where ET to QB is blocked with the QB inhibitor stigmatellin, consistent with recombination from QA–. The QB– signal does not decay for samples where QB– is trapped in the ‘inactive’ state (trace is not shown). 100 % of the inactive P+QB– state was generated by irradiation of the RC with 5 Hz, 605 nm laser excitation during the cooling process. Stars indicate the position of single laser flash. Arrows indicate the position where continuous laser flashes are shut off.
by ET through the B-branch. On the basis of our data, however, we cannot rule out the possibility of rereduction of QB by ET along the B-branch. The identical characteristics of the slow decay kinetics of QB– in the QA-less mutant and in the native RCs is striking (Utschig et al., 2005; Paddock et al., 2006). One possible explanation might be that, for native RCs frozen in the light, QB reduction occurs through the A-branch, while QB– oxidation occurs through the B-branch, as has been previously reported at low temperature (Xu and Gunner, 2001). The population of RCs in the trapped inactive P+QB– state is dependent on the light trapping procedure used during freezing of the RC. As the sample
Chapter 48
Protein Environments and Electron Transfer Processes
is warmed by raising the temperature of the cavity, the inactive P+QB– fraction recombines to the ground state PQB and light-induced activity is restored. Thus, the protein is not damaged by the methods for trapping P+QB–, but rather the inactive P+QB– state is a distinct conformational state of the protein that inhibits low temperature charge recombination. With optical experiments utilizing flash-induced optical absorbance transients to monitor the differences in P and P+ absorbances, an inactive state, where P+ remains oxidized at low temperatures, was observed (McElroy et al., 1974; Kleinfeld et al., 1984b; Xu and Gunner, 2001). This trapped P+ was thought to represent RCs frozen in a P+QAQB state where the concurrently generated reduced quinone has been oxidized by adventitious mediators (Kleinfeld et al., 1984b). Our results suggest that the trapped P+ state observed optically could in fact be the inactive P+QAQB– state we have detected with HF EPR as reduced quinone species are difficult to detect with difference optical absorption measurements. These results are in agreement with earlier reported optical work (Kleinfeld et al., 1984a; Xu and Gunner, 2001): Rba. sphaeroides RCs with bound QB frozen under illumination adopt a conformation that supports ET from QA– to QB and this reaction proceeds with a significant yield even at 20 K. Note that our experiments require the removal of the non-heme Fe2+ located between QA and QB and replacement with diamagnetic Zn2+. However, previous reports showed that Fe2+ is not essential for QA– QB → QAQB– ET in Rba. sphaeroides (Debus et al., 1986). D. Probing Protein Environments with High Frequency Mims-type ENDOR What is the nature of the light-induced conformational changes that enable low temperature QA– QB → QAQB– ET? The changes could be localized largescale rearrangements, such as movement of the QB cofactor between ‘proximal’ and ‘distal’ sites, or more global, less-pronounced readjustments of a region of the protein matrix. We have extended existing ENDOR methodologies to see if the light-induced conformational changes that control reactivity reside in the QB local environment. The Mims-type pulsed 1H ENDOR has a superior sensitivity to nuclei with small HFI. This sensitivity allows for observation of the so-called ‘matrix’ ENDOR (Mims, 1965; Poluektov et al., 2004, 2005). With matrix ENDOR, interactions with distant nuclei,
959
i.e., the protein environment, are mainly detected. Thus, this is an ideal method for looking at differences in the protein environments surrounding the quinones. The standard pulse sequence of the Mims-type TR ENDOR experiment is depicted in Fig. 4. Mimstype ENDOR is recorded as a stimulated electron spin echo (ESE) intensity variation as a function of radiofrequency (RF). A RF pulse is placed between the second and third microwave pulses. In the TR version of a Mims-type ENDOR experiment, (discussed below in Section III) the microwave pulse sequence follows a 5 ns laser pulse with variable delay after flash (DAF) time Dt (Fig. 4). The basics of pulsed ENDOR techniques are discussed in recent reviews (Schweiger and Jeschke, 2001; Goldfarb and Arieli, 2004; Möbius et al., 2005) Specialized samples are important for resolving local protein environments. Selective deuteration of the quinone, in a protonated protein environment, allows for spectral features of the quinone radicals to be distinguished in the ENDOR experiments from the ‘matrix’ surrounding the quinone radicals. The frequency domains of 1H and 2H ENDOR spectra are well separated at HF EPR. In fully protonated RCs substituted with deuterated quinone, 1H ENDOR performed on the QB– EPR line is a pure matrix ENDOR with respect to this radical, because QB– does not contain any protons. Thus, all nuclei which contribute to the 1H ENDOR spectrum are from the protein environment and any protein-bound water molecules. To apply these methods to examine quinone environments, we first ‘calibrate’ the Mims-type HF ENDOR technique and measure its sensitivity to small changes in both local protein structure and radical orientation in these specialized isotopically enriched samples. 1. Sensitivity to Local Structure and Quinone Rotation Since a large number of overlapping resonances from the protein contribute to the HF Mims-type ‘matrix’ ENDOR spectra, direct assignment of the resonances is not possible (see Fig. 5). Hence, we look for differences between spectra from samples generated under different conditions. The observed differences are then interpreted as a divergence in protein environment, as monitored by the proton environments near QB–. For example, Mims-type ENDOR was applied to look for conformational differences
Oleg G. Poluektov and Lisa M. Utschig
960
Fig. 4. Pulse sequence for the Mims-type ENDOR experiment. The Mims-type ENDOR is recorded as a stimulated ESE intensity variation as a function of radiofrequency. Radiofrequency π−pulse is applied between second and third microwave (MW) pulses with the separation time T. In the TR version of Mims-type ENDOR the first π/2 MW-pulse follows a 5 ns laser pulse with delay after the flash (DAF) time Dt.
near QB– between two different species of bacterial RCs. The three-dimensional crystal structures of RCs from Blastochloris (Blc.) viridis (Deisenhofer et al., 1995) and Rba. sphaeroides (Allen et al., 1988; ElKabbani et al., 1991; Ermler et al., 1994) are known to high resolution, and the QB binding pockets differ in their surrounding protein environments. Thus, as a test of matrix ENDOR to differentiate protein environments, the D-band Mims-type ENDOR of Fe-removed/Zn-replaced protonated RCs substituted with deuterated UQ-10 for RCs from Blc. viridis and Rba. sphaeroides were obtained (Fig. 5a,b). The difference spectrum (Fig. 5c) demonstrates that HF matrix ENDOR is sensitive to the differing details of the QB proton environments of the protein for the Rba. sphaeroides and Blc. viridis RCs . To visualize this difference and to estimate the number of protons contributing to the ENDOR spectra, the distance range of the local protein environment that HF ENDOR is sensitive to was calculated. Using a point-dipole approximation (Schweiger and Jeschke, 2001), the anisotropic HFI constants of the protons is given by the following equation: T = C ρ0π /R3 (3 cos2θ−1)
(1)
where, ρπ0 is the carbonyl oxygen π−spin density of the unpaired electron and C = ge gN βeβΝ / h = 79.2 MHz.Å3. Here, ge and gN are the electron and nuclear g-values, βe and βΝ are the electron and nuclear Bohr magnetons, and θ is the angle between the applied
magnetic field and the vector directed to the particular hydrogen atom. Using, for simplicity, an angle factor of 1, a minimum coupling of T = 200 kHz, and a spin density ρπ0 on each carbonyl oxygen of 25% (Sinnecker et al., 2004), we calculate that ENDOR spectra are sensitive to protons located within a sphere R ≈ 5 Å around each oxygen. The number of protons within these spheres was estimated by analysis of the X-ray crystallographic structures of Rba. sphaeroides (Ermler et al., 1994) and Blc. viridis (Lancaster and Michel, 1997) RCs. In our samples, QB is deuterated and does not contribute any protons. Based on the structural data, the number of amino acid residue protons within spheres of radius R = 4 Å, 5 Å, and 6 Å around carbonyl oxygens for Rba. sphaeroides RCs are 17, 37, 61, and, for Blc. viridis RCs, 15, 39, and 60, respectively. Fig. 5d depicts the proton positions within a R = 4 Å sphere for Rba. sphaeroides and Blc. viridis RCs by overlapping the quinone structures. Even though a comparable number of protons, with a similar distribution, contribute to the 1H ENDOR spectra of Rba. sphaeroides and Blc. viridis RCs, distinct differences in spectral ENDOR lineshapes are observed (Fig. 5c). These spectral differences illustrate the high sensitivity of HF Mims-type ENDOR to local protein environments. Therefore, any local differences between the QB– environments for the inactive and active P+QB– states, as reflected in different proton locations near QB–, should be detectable with HF ENDOR. In Rba. sphaeroides RCs, QA and QB are both
Chapter 48
Protein Environments and Electron Transfer Processes
961
Fig. 5. Comparison of D-band Mims-type 1H ENDOR of the QB– binding sites for Blc. viridis (a), and Rba. sphaeroides (b) RCs. The spectra were recorded at a magnetic field position of 46364 G. A difference spectrum of spectrum (a) – spectrum (b) is shown in (c). This spectrum (c) reflects differences in the protein environments for each RC as observed with HF 1H ENDOR spectroscopy. (d) Protons within a 4 Å sphere around each of the carbonyl oxygens (large black spheres) of QB, as determined from the X-ray crystallographic structures of the Rba. sphaeroides RC (1pcr) (Ermler et al., 1994) and Blc. viridis RC (2prc) (Lancaster and Michel, 1997). The structures of QB from each RC were overlapped, and only protons from amino acids in the local QB environments of the Blc. viridis RC (gray) and Rba. sphaeroides RC (pale gray) are shown.
ubiquinone-10 molecules, but crystal structures show that the protein environments of the two quinones are significantly different (Allen et al., 1988; Chang et al., 1991; Ermler et al., 1994). The two quinones differ in hydrogen-bonding contacts with the protein, and the protein environment surrounding QB is more polar than the protein environment surrounding QA. The difference in H-bonding is reflected in the shift of the gx component of the QB g-tensor to higher magnetic field as compared to the gx value of QA– (Gardiner et al., 1999). As a result of their different binding characteristics, these quinones have different redox potentials. QA accepts only one electron and no protons, whereas QB can be doubly reduced and protonated to form the hydroquinone (Wraight, 2004). The observed HF Mims-type 1H ENDOR spectrum of chemically reduced QA– is distinct from the spectrum of QB– (Fig. 6), further substantiating the sensitivity of the HF matrix ENDOR technique to different ubiquinone binding pockets. The orientational selectivity inherent to HF EPR spectra is apparent in the observed sensitivity of HF Mims-type ENDOR to quinone reorientation. Mims-type 1H ENDOR spectra were recorded at two different field positions, 3 G apart, within the photoaccumulated QB– resonance domain (see Fig. 6). This 3 G magnetic field shift corresponds to an effective rotation of the radical by ~5° and discernable differ-
ences in the HF ENDOR spectrum are detectable with this mere change in the g-tensor axis. Thus, the matrix HF ENDOR of samples with selective deuteration and protonation is sensitive to both different local proton environments surrounding the quinone and different orientations of the quinone. 2. The Local Environment of the QB– Site in Kinetically Distinct P+QB– States Having established the sensitivity of matrix HF ENDOR to quinone orientation and environment (see Section II.D.1), this technique was applied to RCs with kinetically defined states of P+QB– to determine if the conformational change that controls reactivity resides in the QB local environment. The HF 1HENDOR spectrum of QB– was obtained for the active P+QB– state wherein the low temperature interquinone ET proceeds. This spectrum was recorded under continuous 10 Hz laser flashing at 20 K for a RC sample that was frozen in the light and subsequently quenched at 170 K (Fig. 7b). The HF ENDOR spectrum of the QB– environment of the original active P+QB– state trapped at 20 K (in this case, the sample was not warmed to higher temperatures) observed under continuous laser flashing at 20 K (data not shown) was essentially identical to the spectrum shown in Fig. 7b. The 1H-ENDOR spectra of QB– obtained for
962
Oleg G. Poluektov and Lisa M. Utschig
Fig. 6. Sensitivity of D-band Mims-type 1H ENDOR to the protein environments and orientations of the quinone anions. Left: The 1H ENDOR spectra obtained at a magnetic field position of 46364 G for deuterated ubiquinone-10 in the QA (a) and the QB (b) binding sites of Rba. sphaeroides reflect the differences in the protein environments for the two distinct quinone binding sites. (c) the difference spectrum (spectrum (a) – spectrum (b)). Right: The 1H ENDOR obtained at two magnetic field positions, 3 G apart, of the QB– EPR line (indicated by arrows on the inset) which is equivalent to an effective ~5° rotation of QB–.
the inactive P+QB– conformation (Fig. 7a) vs. the active P+QB– state (Fig. 7b) produced by low temperature annealing were different. These differences result from the overlap of a small component of QA– signals. A 20% contribution from QA– was estimated from the simulation of the EPR spectrum recorded under continuous laser flashing conditions at 20 K. After taking into account this contribution (Fig. 7c) the difference spectrum (Fig. 7d) does not reveal any deviations in the proton surroundings around the QB– site in these kinetically distinct P+QB– states. Thus, no changes in the proton environment near QB– were observed with matrix ENDOR for RCs in the inactive P+QB– state compared to the active P+QB– state. Based on these data, any significant displacement or reorientation of QB– in the active vs. the inactive sites can be ruled out. For instance, a proximal vs. distal shift of QB– does not occur because the HF 1H ENDOR spectrum is sensitive to a change in H-bonding and 1 H hyperfine interactions that would accompany a 5 Å shift and 180° rotation in QB– position. Surprisingly, the proton environment surrounding QB–, as examined by HF Mims-ENDOR, did not change for different preparations of Fe-removal/Zn-replacement, for different pHs (pH 10, 8, and 5.5 at room temperature), or for samples in H2O-based vs. D2O-based buffer systems. The protein environment, as observed by the 1H matrix surrounding the semiquinone QB–, is quite stable and resistant to conformational changes
Fig. 7. D-band Mims-type 1H ENDOR of RCs from Rba. sphaeroides. No differences in protein structure near QB– were observed for the different kinetic P+QB– states of the RC: (a) 1H ENDOR spectrum of ‘inactive’ QB– site observed by cooling the sample to 20 K under 1 Hz, 605 nm laser excitation. (b) 1H ENDOR spectrum of the ‘active’ QB– site obtained for RCs frozen after exposure to red light at room temperature, then quenched at 170 K for 60 min. The ENDOR spectrum was collected under 10 Hz, 605 nm laser excitation at 20 K. (c) Spectrum (b) – 20% contribution of the 1H ENDOR spectrum of QA– (Fig. 6a) and (d) difference spectrum of spectrum (a) – spectrum (c) showing that the differences observed in (b) are the result of a ~20% contribution from QA– incurred by 10 Hz laser excitation. All spectra were obtained at a magnetic field position of 46364 G.
Chapter 48
Protein Environments and Electron Transfer Processes
in response to different conditions (Utschig et al., 2005). These results are consistent with tight binding of QB in the semiquinone state. Are any conformational intermediates of QB– observed when the sample temperature is raised above 20 K? RCs frozen in the light were warmed to different temperatures to examine potential conformational states intermediate between the neutral dark state QB, and the light-trapped active state of QB–. Different conformational states along the reaction path have been reported previously (Xu and Gunner, 2001). In these optical experiments, when the temperature of light-adapted RCs is raised above 120 K, the trapped conformation which can form P+QB– relaxes to an inactive conformation which is different from RCs frozen in the dark (Xu and Gunner, 2001). Our EPR spectra reveal a time-dependent decrease in the trapped inactive QB– signal upon warming RCs to 170 K or above. When cooled back down to 20 K, approximately 50% of the RCs were in an active intermediate state, transferring electrons between QA– and QB. The recombination kinetics of this active intermediate state differ from the kinetics of the active state trapped initially at 20 K. This observed difference in recombination kinetics suggests that the active intermediate P+QB– state is truly in a conformational substate that differs from the active P+QB– state. However, HF ENDOR spectra indicate that any structural differences between these two conformational substates are not situated in the local protein environment surrounding QB. III. Electron Transfer Pathways and Protein Response to Charge Separation A. Background As described in Section II, although HF Mims-type ENDOR is sensitive to small changes in local protein environments, no conformational changes in the protein environments around QB– were detected from kinetically defined states of P+QB–. Considering the observed striking variations in ET kinetics and inherent dynamic character of the protein, one finds it difficult to conceive that there are no protein conformational changes that accompany ET. This suggests that potential light-induced protein conformational changes that can be trapped at low temperatures are located in protein regions distant from QB. For example, sequential ET steps through the cofactors
963
proceeding QB– generation could induce structural changes. Hence, light-induced changes trapped at low temperature could be located in the protein environment near QA or even bacteriopheophytin, HA (Fig.1) (Tang et al., 1999; Utschig et al., 2005). On the other hand, solvent reorganization or global proton rearrangements have been postulated to stabilize the P+QA– state (McPherson et al., 1990; Franzen and Boxer, 1993; Breton and Nabedryk, 1998; Di Donato et al., 2004). Indeed, if the protein regulates ET via subtle changes in a global network of H-bonds, then local rearrangements of the H-bonds might be too small to be detected with the matrix HF-ENDOR technique. This could also explain the inability to detect light-induced conformational substates of the protein by X-ray diffraction techniques (Baxter et al., 2004) as this technique is not sensitive to protons. New spectroscopic approaches are needed to detect small-scale global rearrangements of the protein structure. One such approach is TR HF ENDOR of the spin-correlated radical pair (SCRP). These techniques are similar to those described above, however ENDOR is performed on the transient species which exhibit a spin-correlation phenomenon (Thurnauer and Norris, 1980; Thurnauer and Gast, 1985; Buckley et al., 1987; Closs et al., 1987; Stehlik et al., 1989). The transient charge separated state P+QA– has been extensively studied by TR-EPR methods. From analysis of the electron spin polarization exhibited by the TR-EPR spectra of P+Q A– it is well established that this transient charge separated state is a SCRP (Thurnauer and Norris, 1980; Thurnauer and Gast, 1985; Stehlik et al., 1989; Fuechsle et al., 1993; Snyder and Thurnauer, 1993; Van der Est et al., 1993; Prisner et al., 1995; Van der Est et al., 1995; Angerhofer and Bittl, 1996; Link et al., 2001). Recently we reported the first observation of TR HF ENDOR of the transient charge separated state P+QA– (Poluektov et al., 2004, 2005). The TR HF ENDOR spectra of protein nuclei (protons) surrounding deuterated QA– (i.e., matrix protons) exhibit complicated lineshapes which consist of narrow, derivative-like lines and differ considerably from the HF ENDOR spectra of the protein nuclei surrounding thermally equilibrated QA–. Narrowing of the lines, especially in the central part of the ENDOR spectrum, around the nuclear Larmor frequency, leads to the increased sensitivity of the ENDOR technique to the remote nuclei. The high resolution and orientational selectivity of HF ENDOR allows protein environments to be directly probed with a much improved resolution by spec-
964
Oleg G. Poluektov and Lisa M. Utschig
trally selecting specific nuclei in isotopically labeled samples. A theoretical analysis of these observations (Poluektov et al., 2005; Dubinskij et al., 2007) shows that the positions and amplitudes of ENDOR lines contain information on hyperfine interactions of a particular nucleus (a proton of the protein) with both electron spins of the SCRP, i.e., P+ and QA–. Thus, spin density delocalization in the protein environment between the SCRP donor and acceptor molecules can be revealed via HF ENDOR. Moreover, owing to the transient features of the technique, this spin density distribution can be followed in time, providing important information on the response of the protein environment to the creation of the charge-separated state P+QA–. Here we will discuss important features and applications of this novel SCRP ENDOR technique for studying the regulation of the ET reactions by the protein environment. B. High Frequency Spin-Correlated Radical Pair ENDOR of the P+QA– Charge-Separated State 1. Experimental Results The field-swept TR HF EPR and ENDOR spectra of the SCRP P+QA– in the Rba. sphaeroides RC are shown in Fig. 8a. The pulse sequence for Mims-type TR ENDOR is shown on Fig. 4. Spectra of chemically reduced QA are depicted in the top traces of Figs. 8a, b and c for comparison. Selective deuteration of the QA cofactor increases the intensity of the EPR signal and enables observation of pure matrix 1H ENDOR without interference with quinone protons, as described above in Section II. There are two striking differences between the SCRP ENDOR spectra recorded with the delay after laser flash (DAF) in the microsecond regime and the thermal equilibrium ENDOR spectra of QA– (Figs. 8b,c top traces). One difference is sharpening of particular ENDOR lines resulting in a considerable increase of the spectral resolution with the observed broad non-resolved line from the mostly overlapping resonances of matrix nuclei becoming substantially reduced. Another noticeable feature in the SCRP ENDOR spectrum is the asymmetric displacement of the spectrum around the 1H Larmor frequency, νH (Figs. 8b,c, bottom traces). Compare this to the symmetric placement of the lines in the thermal equilibrium ENDOR spectrum of QA– (Figs. 8b,c, top traces). Another effect is observed when ENDOR spectra are recorded with different delays between laser flash
Fig. 8. D-band ESE detected EPR (a) and Mims-type 1H ENDOR (b,c) spectra of Fe-removed/Zn-substituted photosynthetic bacterial RCs with deuterated QA, recorded at D-band EPR. (a, b, c) Spectra shown in top traces recorded for chemically reduced QA. Spectra shown in the bottom traces recorded for the SCRP with a DAF-time Dt = 2 µs. Duration of the π/2 MW pulses was 50 ns, separation between first and second MW pulses, τ = 200 ns, separation between second and third MW pulses, 23 µs, and length of the RF pulse, 20 µs. Temperature 50 K. Laser excitation λ = 550 nm. ENDOR spectra were recorded at the following values of the magnetic field: (b) 46380 G and (c) 46445 G. Arrows on spectra (b) and (c) indicate the 1H Larmor frequency, νH, at this field.
Chapter 48
Protein Environments and Electron Transfer Processes
and the first microwave (MW) pulse, DAF-times (Fig. 4). With an increase of the DAF-time from the microsecond to millisecond regime, the spectral lines shift and the spectrum becomes identical with the ENDOR spectrum of chemically reduced QA. This unusual effect can be clearly observed in the deuteron (2H) SCRP ENDOR spectra (Fig. 9). The ENDOR lines in Fig. 9 are from the deuterons of QA–, as the quinones are the only deuterated species in the protonated RC protein. The spectrum in Fig. 9a was acquired at a DAF-time of 2 µs and shows two lines, slightly asymmetric (less than 10 kHz) around the deuteron Larmor frequency (νD), and with a splitting of 1.04 MHz. At a delay time of 10 ms, (Fig. 9b) the ENDOR spectrum becomes symmetric around νD and the lines shift so that the splitting is 0.94 MHz. The same splitting is observed for the chemically reduced deuterated quinone, QA–, ENDOR, recorded at the same magnetic field position and with the same ENDOR pulse sequence (see Fig. 9c). If recalculated for proton coupling, the splitting of 0.94 MHz in the deuteron ENDOR spectrum corresponds to a proton HFI of 6.12 MHz. This value is in good agreement with the HFI for the QA– methyl protons (6.8 MHz for the same orientation of magnetic field with respect to the g-tensor) reported by Rohre et al. (1998) in their W-band ENDOR study of photoaccumulated QA– in the RC. 2. Sensitivity of Spin-Correlated Radical Pair ENDOR for Distant Nuclei In order to understand the unusual spectral features of SCRP ENDOR, a theoretical treatment has been developed for the high-field condition of the D-band (46,000 G; 130 GHz) EPR detected ENDOR. It was shown that the line shapes of the TR ENDOR spectra at high magnetic field can be rationalized by taking into account a three-spin system (two electrons and one nucleus) and the high spectral resolution of the HF EPR which prevents mixing between different spin substates. The theory developed for HF TR ENDOR of SCRP (Dubinski et al., 2002) explains that the experimentally observed derivative-type lines result from the simultaneous interaction of particular nuclei with both correlated electron spins. These lines become observable in the SCRP ENDOR spectra only when recorded at HF EPR: the high spectral separation of the EPR resonances in the radical pair and the high orientation selectivity are necessary prerequisites. The value of the HFI of both
965
Fig. 9. Mims-type 2H-ENDOR spectra of Fe-removed/Zn-substituted photosynthetic bacterial RCs with deuterated QA, recorded at D-band EPR for light-induced P+QA– radical pair and chemically reduced QA. (a,b) SCRP recorded with DAF-times Dt = 2 µs and Dt = 10 ms, respectively; (c) chemically reduced QA. νD indicates the position of deuteron Larmor frequency. Magnetic field position is 46385 G; τ=200 ns; Τ=33 µs, RF-pulse length –30 µs; temperature –50 K.
correlated electron spins with the particular nucleus can be extracted from the positions with respect to the Larmor frequency (HFI with QA–) and relative intensity (HFI with P +865) of the derivative lines. For pairs of distantly separated radicals, such as the ~29 Å between P +865QA–,the nuclei that meet the requirements for observing derivative ENDOR lines belong to the protein surrounding the radicals (matrix nuclei) and must be located between the two spins. (Poluektov et al., 2004; Poluektov et al., 2005) The information on spin density distribution is very important for elucidating mechanistic features of ET processes in photosynthetic proteins. Indeed, the ET between donor and acceptor in proteins occurs through the bonds of the protein scaffold and depends upon the overlap of wave functions of the cofactors and protein surroundings (Wuttke et al., 1992; Steffen et al., 1994; Langen et al., 1995; Babini et al., 2000). ENDOR spectroscopy provides a direct technique for measuring the delocalization of the electron wave function. The TR ENDOR of SCRP can provide information on the HFI of a particular nucleus with both donor and acceptor electron spins, and thus has the potential to locate probable ET pathways through the nuclei having maximum overlap of the donor and acceptor wave functions. According to
966
Oleg G. Poluektov and Lisa M. Utschig
theoretical analysis, (Poluektov et al., 2005; Dubinskij et al., 2007) this information is encoded in the SCRP ENDOR spectra via the positions and intensities of the derivative-type lines. 3. Interpretation of Spin-Correlated Radical Pair ENDOR Spectra As described above, the information content of the SCRP ENDOR spectra is significant. However, obtaining direct information about the through protein ET pathways from SCRP ENDOR is complicated by several factors. First, spectral analyses are difficult because of the large number of overlapping signals from the protons of the protein, complicating the assignment of each resonance line. Second, nuclei which are degenerate in the ENDOR spectrum of the thermalized state have different positions in the SCRP ENDOR spectrum due to the nuclear interaction with the second electron partner in the SCRP. Third, an ENDOR spectrum can not be recorded immediately after the laser flash because the πRF pulse has a length of tens of microseconds (in our case 25 µs is needed after the laser flash to start recording TR ENDOR), and significant spin-relaxation could occur during that time, leading to considerable distortions of the spectral shape from that predicted by the theory described above. Therefore, novel approaches for acquiring and analyzing SCRP ENDOR spectra in order to simplify spectral interpretations are discussed below. One approach that follows from the theoretical treatment (Poluektov et al., 2005) is decomposition of the TR ENDOR spectra into odd and even parts. There are two contributions to the SCRP ENDOR spectrum; an odd-type spectrum centered around the Larmor frequency and an even-type spectrum that resembles a typical stationary ENDOR spectrum. The odd and even parts of the SCRP ENDOR spectrum are shown in Fig. 10. This decomposition was accomplished by subtracting (odd) and adding (even) parts of the ENDOR spectrum with its mirror image obtained by reflection around the Larmor frequency. Figure 10a shows the same experimentally obtained SCRP ENDOR spectrum as Fig. 8b before deconvolution. Figs. 10b,c show the odd and even components of the spectrum of Fig. 10a, respectively, after deconvolution. According to theoretical analysis the odd part of the spectrum (Fig. 10b) shows only the nuclei which have interactions with both of the correlated electron spins P +865 and QA–. The lines in the odd component of the SCRP ENDOR spectrum arise from nuclei that
Fig. 10. Decomposition of the SCRP ENDOR spectrum from Fig. 8b into odd and even parts. (a) Original ENDOR spectrum, recorded with the parameters described in caption of Fig. 8. (b) Odd part of the deconvolution. (c) Even part of the deconvolution. Position of the 1H Larmor frequency is indicated by νH.
are potentially part of the ET pathway. The even parts of the spectra are due to nuclei that do not have significant HFI with a second electron. Comparison of the even part of the spectrum (Fig. 10c) with the ENDOR spectrum of the stationary state of QA– (Fig. 8b) shows common features. However, the relative intensities of individual lines, which contribute to the ENDOR spectrum, are different. Partially, this difference between the even-part of the SCRP spectrum and the thermalized QA– ENDOR spectrum could reflect structural relaxation of the protein matrix to accommodate charge separation. Thus, deconvolution of the SCRP ENDOR spectra into odd and even parts provides a method to separate out the nuclei involved directly in the through protein ET pathways from all of the nuclei interacting with the radicals independently. Nevertheless, the number of spectral lines in both the odd and even parts of the spectra is still large and assignment remains difficult. The ideal method for ENDOR line assignment is by specific isotope labeling, i.e., labeling a particular
Chapter 48
Protein Environments and Electron Transfer Processes
967
amino acid residue, cofactor molecule, or particular nucleus. Due to the high spectral resolution of the HF ENDOR technique (Goldfarb and Arieli, 2004; Möbius et al., 2005), spectra of such nuclei would be well separated in the frequency domain ENDOR spectrum. An example of using isotopic labeling to help resolve the ENDOR spectrum is described below. 4. The Role of the Exchangeable Protons in Regulation of Electron Transfer As an example of the isotopic labeling approach, fully deuterated Fe-removed/Zn-substituted RCs were prepared in a H2O based buffer solution. Fig. 11 shows 1H-ENDOR spectra recorded at the same magnetic field position as the 2H-ENDOR spectra shown in Fig. 9. The spectra in Figs. 11a,b are from QA–, chemically reduced and as a partner in the SCRP, respectively. The central parts of the spectra are more pronounced, with most of the spectral lines grouped around the Larmor frequency, i.e., from nuclei having small HFI. Because the protein is fully deuterated, only exchangeable protons from RC amino acids or protons from water molecules bound in the RC contribute to the 1H-ENDOR spectra. The small HFI indicates that the observed protons are positioned relatively far from QA. The largest HFI recorded is of the order of 4.5 MHz. Most likely, this HFI is from a proton that has exchanged for a deuteron hydrogenbonded to QA. Importantly, the odd part of the spectrum is considerably simplified compared to the total spectrum (Figs. 11b,d). Only three derivative-type lines are seen on both sides of the 1H-Larmor frequency. These signals are from the protons that have substantial spin density overlapping both the donor and acceptor electron wave functions. The observed lines must have sufficient spin densities from both correlated electrons and, therefore, exchangeable protons are somehow coupled with the function of P+QA–. Together with H2O molecules, labile protons of the protein environment can work as adjustable bridges to facilitate efficient ET in the RC (Bone and Pething, 1985; Dashdorj et al., 2004). We suggest that following ET, the structure of the protein is tuned to accommodate the new charge-separated state by first adjusting these flexible bridges. This protein structural tuning concomitantly changes the overlap between electron wave functions, thus redirecting ET pathways in a way that makes recombination less effective (Kleinfeld et al., 1984). By
Fig. 11. Mims-type 1H-ENDOR spectra of Fe-removed/Zn-substituted fully deuterated bacterial RC in H2O based buffer recorded at D-band EPR for: (a) chemically reduced QA; (b) light-induced P+Q A– radical pair with DAF-time Dt = 2 µs. (c,d) deconvolution of the spectrum (b) into even and odd parts, respectively. τ = 300 ns. All other parameters are the same as in caption of Fig. 8.
analogy with ‘conformational gating’ this mechanism can be called ‘conformation locking.’ C. Protein and Cofactor Relaxation Following Charge Separation TR ENDOR experiments provide direct spectroscopic observation of structural changes that constitute the protein’s dielectric response to P+QA– formation. Detailed spectral analysis demonstrates that the positions of the absorptive peaks (or even part) of the SCRP ENDOR spectra do not coincide with the positions of the same peaks in the thermalized spectrum. This can be clearly observed by comparing the positions of the outer most spectral peaks in Fig. 9c vs. 9a and Fig. 11a vs. 11b. For example, the separation of the external lines in Fig. 11a is 200 kHz larger than for the corresponding TR ENDOR spectrum in Fig. 11b, which was recorded with a total delay after flash time of ~25 µs. The same effect, but to a lesser extent, is observed for the central part of the spectra. Indeed, as we have discussed above, the absorptive lines (even part) of the spectra result from the ENDOR of the thermalized state, i.e., from nuclei which do not have interactions with the second electron in the SCRP. Thus, the positions of the lines in the even part of the SCRP spectrum and the chemically reduced QA spectrum should be identical, although their intensi-
968 ties might be different. Upon thermal relaxation, the positions of the absorptive lines in the TR ENDOR spectrum move to the same positions of the lines in the chemically reduced spectrum, i.e., spectrum of a protein state that has had time to adjust to the charge on QA. Thus, we believe that thermal relaxations of the positions of the peaks of the even part of the SCRP ENDOR spectrum reflect the tuning of the protein structure, via charge compensation or conformational relaxation events, following formation of the charge separated state P+QA–. The structural response following charge separation not only affects the protein environment, but also the cofactors. 2H ENDOR at the QA– EPR line was used to directly probe the quinone environment in samples where QA is deuterated and the protein is protonated. Fig. 9 shows how the SCRP ENDOR lines from methyl deuterons of QA– relax to the ENDOR spectrum of chemically reduced QA. ENDOR lines recorded right after the laser flash are asymmetric in amplitudes and positions around the 2H Larmor frequency. This asymmetry, measured as peak positions from the Larmor frequency, is less than 10 KHz. The question remains: can the SCRP effects solely explain the shifts of the ENDOR lines observed during thermal relaxation, or must spin redistribution due to structural relaxation also be taken into account? We can estimate the magnitude of the spectral shift due to SCRP effects within the theory of SCRP ENDOR. The SCRP effect in the ENDOR spectra is due to the magnetic field induced by the second, in this case P+, correlated electron on the nuclei. In the case of deuterated QA– in the protonated RC (Fig. 9), the nuclei we detect with ENDOR belong to QA–. From out-of-phase electron spin echo dipole-dipole modulation experiments (Dzuba et al., 1995), we know that the interaction field between the unpaired electrons of P+ and QA– in a SCRP is less than 6 MHz. Recalculation of this frequency for 2H gives 1.4 kHz. An effect of the same order of magnitude (less than 10 kHz) is observed in the SCRP ENDOR spectra. This effect, as observed in the asymmetry of the lines, is in agreement with our theory of SCRP ENDOR. However, the line shifts in the thermally relaxing spectra are of the order of 100 kHz, and much larger than the possible effects from spin correlation. Thus we believe that the observed shifts in the 2H ENDOR spectra of QA– are due to spin density redistribution within the quinone after the fast ET step. This redistribution might be induced by structural relaxation of the protein environment or conformational changes
Oleg G. Poluektov and Lisa M. Utschig within the quinone molecule. Similar line shift effects were observed in the study of Photosystem I (PS I) at X-band EPR. (Fursman et al., 2002) A small reduction of the positions of the + A1–, ENDOR lines of methyl protons in the SCRP P 700 where P700 is the primary chlorophyll electron donor and A1 is a phylloquinone acceptor, was observed right after ET compared to those of the photoaccumulated A1– state. After careful analysis, the authors came to the conclusion that the line shift can not be explained by SCRP effects. However, because the thermal re+ A1– SCRP ENDOR spectra could laxation of the P 700 not be studied due to the short lifetime of the pair, the authors concluded that the observed shift is due to small differences in local geometry within the quinone binding pocket that result from the different procedures for preparing A1– (photoaccumulation versus photoexcitation into the SCRP state). Based on our results, it is possible that the reported shift for PSI (Fursman et al., 2002), is due to the adjustment of the protein structure. Future HF TR ENDOR studies + A1– state will help address this issue. of the P 700 Interestingly, the X-band TR ENDOR spectra of + A1– SCRP exhibit both absorptive and emissive P 700 contributions that have a strong magnetic field-dependence. Although this phenomenon is a consequence of electron spin polarization in the SCRP, its manifestation in the ENDOR spectra is different than that observed with HF-ENDOR. The reduced separation between the doublet-doublet sublevels at X-band results in a substantial admixture of |S〉 and |T0〉 states, and this distorts the ENDOR spectra (Fursman et al., 2002). The ENDOR lines from transitions within the mixed electron manifolds are extremely anisotropic. They broaden, with the accompanying decrease in intensity, when detected with the reduced orientation selectivity of X-band EPR. The lines from the undistorted manifolds are detected without any differential effects. Note that while spectral analysis of HF ENDOR includes the nuclear spin interactions with both correlated electrons of the SCRP, the X-band ENDOR of the SCRP can be explained by considering HFI with only one of the two spin-correlated electrons (Fursman et al., 2002). IV. Concluding Remarks Dynamic, native solution protein structures are essential for fine-tuning ET processes and coupled reaction mechanisms in biological systems. The novel
Chapter 48
Protein Environments and Electron Transfer Processes
experimental approach described in this chapter has the potential to experimentally correlate structural, electrostatic, and dynamic features of localized protein environments with inherent ET reactions. The advantage of this approach is based on the high spectral resolution and orientational selectivity of HF ENDOR and TR HF ENDOR spectroscopy, which allows protein environments surrounding the cofactors involved in charge separation to be examined directly by spectrally selecting specific nuclei in isotopically labeled samples. In the first part of the review, we have discussed two light-induced kinetically distinct states of P+QB– trapped at low temperature and the HF matrix ENDOR approach to examine the local protein environments surrounding QB– for each of these states. The success of these experiments relies on the use of specialized samples with selective deuteration and protonation which enhances signal resolution and selectivity at HF (D-band) pulsed EPR. We demonstrated that HF matrix ENDOR is sensitive to different proton environments surrounding the quinone and to different orientations of the quinone. No structural changes near QB– were observed for Rba. sphaeroides RCs frozen in inactive and active P+QB– states with regard to ET. Interactions of the semiquinone QB– with the protein are rigid, suggesting quite an enforced protein environment surrounding QB– and that the conformational change that controls reactivity resides beyond the QB local environment. TR features of HF ENDOR spectroscopy were demonstrated in the second part of the review by examining the transient state of the spin-correlated radical pair P+QA–. SCRP ENDOR contains data on protein nuclei interactions with unpaired electrons of both donor and acceptor. This information will enable mapping of spin density overlap in the protein environment between the electron donor and acceptor in the SCRP. The spin density map will help to reconstruct electronic wave functions, such that most probable pathways for the ET between donor and acceptor through the protein environment can be predicted. It has been shown that the positions of the 1 H-ENDOR lines of the SCRP shift with an increase in the time after laser flash, which initiates ET. These shifts provide direct spectroscopic evidence of reorganization of the protein environment to accommodate the donor-acceptor charge-separated state P+QA–. The idea that protein structural changes accommodate the movement of charge that accompanies ET has been discussed in a number of publications (Treutlein et
969
al., 1992; Deisenhofer and Norris, 1993; Graige et al., 1998; Trissl et al., 2001; Xu and Gunner, 2001; Wraight, 2004). HF ENDOR spectroscopy gives a clear indication that in photosynthetic RC proteins structural changes are not localized around the donor or acceptor, but rather have a small-scale long-distance character, probably on the level of a global H-bond network. We believe that application of the HF ENDOR technique to the study of photo-initiated charge separation will provide new insights into the electron-transfer reactions in natural and artificial photosynthetic assemblies. Acknowledgments We would like to acknowledge our coworkers whose contributions are essential for our HF EPR research of the photosynthetic systems: M. C. Thurnauer, D. M. Tiede, A. M. Wagner, A. A. Dubinskij, S. V. Paschenko. The authors are grateful to S. K. Foreman for help in preparation of the manuscript. Work at ANL was supported by the U.S. Department of Energy, Office of Basic Energy Sciences, Division of Chemical Sciences, Geosciences, and Biosciences, under contract DE-AC02-06CH11357. References Allen JP, Feher G, Yeates TO, Komiya H and Rees DC (1988) Structure of the reaction center from Rhodobacter sphaeroides R-26: Protein cofactor (quinones and Fe2+) interactions. Proc Natl Acad Sci USA 85: 8487–8491 Angerhofer A and Bittl R (1996) Radicals and radical pairs in photosynthesis. Photochem Photobiol 63: 11–38 Babini E, Bertini I, Borsari M, Capozzi F, Luchinat C, Zhang X, Moura GLC, Kurnikov IV, Beratan DN, Ponce A, Di Bilio AJ, Winkler JR and Gray HB (2000) Bond-mediated electron tunneling in ruthenium-modified high-potential iron-sulfur protein. J Am Chem Soc 122: 4532–4533 Balabin IA and Onuchic JN (2000) Dynamically controlled protein tunneling paths in photosynthetic reaction centers. Science 290: 114–117 Baxter RHG, Ponomarenko N, Srajer V, Pahl R, Moffat K and Norris JR (2004a) Time-resolved crystallographic studies of light-induced structural changes in the photosynthetic reaction center. Proc Natl Acad Sci USA 101: 5982–5987 Baxter RHG, Seagle B-L, Ponomarenko N and Norris J (2004b) Specific radiation damage illustrates light-induced structural changes in the photosynthetic reaction center. J Am Chem Soc 126: 16728–16729 Baxter RHG, Seagle B-L, Ponomarenko N and Norris J (2005) Cryogenic structure of the photosynthetic reaction center of Blastochloris viridis in the light and dark. Acta Cryst D61:
970 605–612 Bittl R, Zech SG and Lubitz W (1995) Light-induced changes in transient EPR spectra of P+QA– In: Michel-Beyerle M-E (ed) The Reaction Center of Photosynthetic Bacteria: Structure and Dynamics, pp 333–339. Springer-Verlag. Berlin Bone S and Pething R (1985) Dielectric studies of protein hydration and hydration-induced flexibility. J Mol Biol 181: 323–326 Breton J (2004) Absence of large-scale displacement of quinone QB in bacterial photosynthetic reaction centers. Biochemistry 43: 3318–3326 Breton J and Nabedryk E (1998) Proton uptake upon quinone reduction in bacterial reaction centers: IR signature and possible participation of a highly polarizable hydrogen bond network. Photosynth Res 55: 301–307 Breton J, Buollais C, Mioskowski C, Sebban P, Baciou L and Nabedryk E (2002) Vibrational spectroscopy favors a unique QB binding site at the proximal position in wild-type reaction centers and in the Pro-L209-Tyr mutant from Rhodobacter sphaeroides. Biochemistry 41: 12921–12927 Breton J, Wakeham MC, Fyfe PK, Jones MR and Nabedryk E (2004) Characterization of the bonding interactions of QB upon photoreduction via A-branch or B-branch electron transfer in mutant reaction centers from Rhodobacter sphaeroides. Biochim Biophys Acta 1656: 127–138 Buckley CD, Hunter DA, Hore PH and McLauchlan KA (1987) ESR of spin-correlated radical pairs. Chem Phys Lett 135: 307–312 Chang CH, El-Kabbani O, Tiede DM, Norris J and Schiffer M (1991) Structure of the membrane-bound protein photosynthetic reaction center from Rhodobacter sphaeroides R-26. Biochemistry 30: 5353–5360 Closs GL, Forbes MDE and Norris JR (1987) Spin-polarized electron paramagnetic resonance spectra of radical pairs in micelles. Observation of electron spin-spin interactions. J Phys Chem 91: 3592–3599 Dashdorj N, Xu W, Martinsson P, Chitnis PR and Savikhin S (2004) Electrochromic shift of chlorophyll absorption in photosystem I from Synechocystis sp. PCC 6803: A probe of optical and dielectric properties around the secondary electron acceptor. Biophys J 86: 3121–3130 Debus RJ, Feher G and Okamura MY (1986) Iron-depleted reaction centers from Rhodopseudomonas sphaeroides R-26: Characterization and reconstitution with Fe2+, Mn2+, Co2+, Ni2+, Cu2+, and Zn2+. Biochemistry 25:2276–2287 Deisenhofer J and Norris JR (eds) (1993) Photosynthetic Reaction Centers. Academic Press, New York Deisenhofer J, Epp O, Miki K, Huber R and Michel H (1984) Xray structure analysis of a membrane protein complex: electron density map at 3 Å resolution and a model of the chromophores of the photosynthetic reaction center from Rhodopseudomonas viridis. J Mol Biol 180: 385–398 Deisenhofer J, Epp O, Sinning I and Michel H (1995) Crystallographic refinement at 2.3 angstrom resolution and refined model of the photosynthetic reaction center from Rhodopseudomonas viridis. J Mol Biol 246: 429–457 Deng H, Zhadin N and Callender R (2001) Dynamics of protein ligand binding on multiple time scales: NADH binding to lactate dehydrogenase. Biochemistry 40: 3767–3773 Di Donato M, Peluso A and Villani G (2004) Electron transfer between quinones in photosynthetic reaction centers. J Phys Chem B 108: 3068–3077 Dorio MM and Freed JH (eds) (1979). Multiple Electron Reso-
Oleg G. Poluektov and Lisa M. Utschig nance Spectroscopy. Plenum Press, New York Dubinski AA, Perekhodtsev GD, Poluektov OG, Rajh T and Thurnauer MC (2002) Analytical Treatment of EPR Spectra of Weakly Coupled Spin-Correlated Radical Pairs in Disordered Solids: Application to the Charge-Separated State in TiO2 Nanoparticles. J Phys Chem A 106: 938–944 Dubinskij AA, Utschig LM and Poluektov OG (2006) Spin-correlated radical pairs: The differential effect in high-field ENDOR spectra. Appl Magn Reson 30: 269–286 Dzuba SA, Gast P and Hoff AJ (1995) ESEEM study of spin-spin interactions in spin-polarized P+QA– pairs in the photosynthetic purple bacterium Rhodobacter sphaeroides R26. Chem Phys Lett 236: 595–602 Eisinger J and Feher G (1958) Hfs anomaly of Sb121 and Sb123 determined by the electron nuclear double resonance technique. Phys Rev 109: 1172–1183 El-Kabbani O, Chang C-H, Tiede DM, Norris J and Schiffer M (1991) Comparison of reaction centers from Rhodobacter sphaeroides and Rhodopseudomonas viridis: Overall architecture and protein-pigment interactions. Biochemistry 30: 5361–5369 Ermler U, Fritzsch G, Buchanan S and Michel H (1994) Structure of the photosynthetic reaction-center from Rhodobacter sphaeroides at 2.65 Å resolution- cofactors and protein-cofactor interactions. Structure 2: 925–936 Ernst RR, Bodenhausen G and Wokaun A (1987) Principles of Nuclear Magnetic Resonance in One and Two Dimensions. Clarendon Oxford, UK Feher G, Fuller CS and Gere EA (1957) Spin and magnetic moment of P32 by the electron nuclear double-resonance technique. Phys Rev 107: 1462–1464 Feher G, Hoff AJ, Isaacson R and McElroy JD (1973) Biophys Soc Abstr 13: 61a Feher G, Hoff AJ, Isaacson R and Ackerson LC (1975) ENDOR experiments on chlorophyll and bacteriochlorophyll in vitro and in the photosynthetic unit. Ann NY Acad Sci USA 244: 239–259 Feher G, Allen JP, Okamura MY and Rees DC (1989) Structure and function of bacterial photosynthetic reaction centers. Nature 339: 111–116 Ferreira KN, Iverson TM, Maghlaoui K, Barber J and Iwata S (2004) Architecture of the photosynthetic oxygen-evolving center. Science 303: 1831–1838 Franzen S and Boxer SG (1993) Temperature-dependence of the electric field modulation of electron transfer rates — charge recombination in photosynthetic reaction centers. J Phys Chem 97: 6304–6318 Füchsle G, Bittl R, Van der Est A, Lubitz W and Stehlik D (1993) Transient EPR spectroscopy of the charge separated state P+Q– in photosynthetic reaction centers. Comparison of Zn-substituted Rhodobacter sphaeroides R-26 and Photosystem I. Biochim Biophys Acta 1142: 23–25 Fursman CE, Teutloff C and Bittl R (2002) Pulsed ENDOR studies of short-lived spin-correlated radical pairs in photosynthetic reaction centers. J Phys Chem B 106: 9679–9686 Gardiner AT, Zech SG, MacMillan F, Kass H, Bittl R, Schlodder E, Lendzian F and Lubitz W (1999) Electron paramagnetic resonance studies of zinc-substituted reaction centers from Rhodopseudomonas viridis. Biochemistry 38: 11773–11787 Genick UK, Borgstahl GEO, Ng K, Ren Z, Pradervand C, Burke PM, Srajer V, Teng T, Schildkamp W, McRee DE, Moffat K and Getzoff ED (1997) Structure of a protein photocycle
Chapter 48
Protein Environments and Electron Transfer Processes
intermediate by millisecond time-resolved crystallography. Science 275: 1471–1475 Goldfarb D and Arieli D (2004) Spin distribution and the location of protons in paramagnetic proteins. Ann Rev Biophys Biomol Struct 33: 441–468 Graige MS, Feher G and Okamura MY (1998) Conformational gating of the electron transfer reaction QA– QB → QAQB– in bacterial reaction centers of Rhodobacter sphaeroides determined by a driving force assay. Proc Natl Acad Sci USA 95: 11679–11684 Grinberg OY and Berliner LJ (eds) (2004). Very High Frequency (VHF) ESR/EPR. Biological Magnetic Resonance. Kluwer Academic, Plenum Publishers, New York Hummer G, Schotte F and Anfinrud PA (2004) Unveiling functional protein motions with picosecond X-ray crystallography and molecular dynamics simulations. Proc Natl Acad Sci USA 101: 15330–15334 Jordan P, Fromme P, Witt HT, Klukas O, Saenger W and Kraub N (2001) Three-dimensional structure of cyanobacterial photosystem I at 2.5 Å resolution. Nature 411: 909–917 Kleinfeld D, Okamura MY and Feher G (1984a) Electron-transfer kinetics in photosynthetic reaction centers cooled to cryogenic temperatures in the charge-separated state: Evidence for lightinduced structural changes. Biochemistry 23: 5780–5786 Kleinfeld D, Okamura MY and Feher G (1984b) Electron transfer in reaction centers of Rhodopseudomonas sphaeroides: I. Determination of the charge recombination pathway of D+QAQB– and free energy and kinetic relations between QA– QB and QAQB–. Biochim Biophys Acta 766: 126–140 Kothe G, Norris JR, Poluektov OG and Thurnauer MC (2008) High-time resolution electron paramagnetic resonance-study of quantum beat oscillations observed in photosynthetic reaction center proteins. In: Aartsma TJ and Matysik J (eds) Biophysical Techniques in Photosynthesis II (Advances in Photosynthesis and Respiration, Vol 26), pp 305–323 . Springer, Dordrecht Kuglstatter A, Ermler U, Michel H, Baciou L and Fritzsch G (2001) X-ray structure analyses of photosynthetic reaction center variants from Rhodobacter sphaeroides: Structural changes induced by point mutations at position L209 modulate electron and proton transfer. Biochemistry 40: 4253–4260 Lancaster CRD and Michel H (1997) The coupling of lightinduced electron transfer and proton uptake as derived from crystal structures of reaction centres from Rhodopseudomonas viridis modified at the binding site of the secondary quinone, QB. Structure 5: 1339–1359 Langen R, Chang IJ, Germanas JP, Richards JH, Winkler JR and Gray HB (1995) Electron tunneling in proteins: Coupling through a beta strand. Science 268: 1733–1735 Lendzian F (2005) Structure and interactions of amino acid radicals in class I ribonucleotide reductase studied by ENDOR and high-field EPR spectroscopy. Biochim Biophys Acta 1707: 67–90 Li J, Gilroy D, Tiede DM and Gunner MR (1998) Kinetic phases in the electron transfer from P+QA– QB to P+QAQB– and the associated processes in Rhodobacter sphaeroides R-26 reaction centers. Biochemistry 37: 2818–2829 Link G, Berthold T, Bechtold M, Weidner J-U, Ohmes E, Tang J, Poluektov O, Utschig LM, Schlesselman SL, Thurnauer MC and Kothe G (2001) Structure of the P+700 A–1 radical pair intermediate in Photosystem I by high time resolution multifrequency electron paramagnetic resonance-analysis of quantum beat oscillations. J Am Chem Soc 123: 4211–4222
971
Loll B, Kern J, Saenger W, Zouni A and Biesiadka J (2005) Towards complete cofactor arrangement in the 3.0 Å structure of photosystem II. Nature 438: 1040–1044 Luecke H, Schobert B, Richter HT, Cartailler JP and Lanyi JK (1999) Structural changes in bacteriorhodopsin during ion transport at 2 Ångström resolution. Science 286: 255–261 Mancino L, Dean D and Blankenship R (1984) Kinetics and thermodynamics of the P870+QA– → P870+QB– reaction in isolated reaction centers from the photosynthetic bacterium Rhodopseudomonas sphaeroides. Biochim Biophys Acta 764: 46–54 McElroy JD, Mauzerall DC and Feher G (1974) Characterization of primary reactants in bacterial photosynthesis. II. Kinetic studies of the light-induced EPR signal (g = 2.0026) and the optical absorbance changes at cryogenic temperatures. Biochim Biophys Acta 333: 261–277 McPherson PH, Nagarajan V, Parson WW, Okamura MY and Feher G (1990) pH-dependence of the free-energy gap between DQA and D+Q A– determined from delayed fluorescence in reaction centers from Rhodobacter spheroides R-26. Biochim Biophys Acta 1019: 91–94 Mims WB (1965) Pulsed electron nuclear double resonance (ENDOR) experiments. Proc Roy Soc (London) 283: 452–457 Möbius K, Savitsky A, Schnegg A, Plato M and Fuchs M (2005) High-field EPR spectroscopy applied to biological systems: Characterization of molecular switches for electron and ion transfer. Phys Chem Chem Phys 7: 19–42 Mulkidjanian AY (1999) Conformationally controlled pK-switching in membrane proteins: One more mechanism specific to the enzyme catalysis? FEBS Lett 463: 199–204 Nabedryk E, Breton J, Sebban P and Baciou L (2003) Quinone (QB) binding site and protein structural changes in photosynthetic reaction center mutants at Pro-L209 revealed by vibrational spectroscopy. Biochemistry 42: 5819–5827 Norris JR, Druyan ME and Katz JJ (1973) Electron nuclear double resonance of bacteriochlorophyll free radical in vitro and in vivo. J Am Chem Soc 95: 1680–1681 Norris JR, Scheer H, Druyan ME and Katz JJ (1974) An electronnuclear double resonance (ENDOR) study of the special pair model for photo-reactive chlorophyll in photosynthesis. Proc Natl Acad Sci USA 71: 4897–4900 O’Halloran TV (1993) Transition metals in control of gene expression. Science 261: 715–725 Okamura MY, Paddock ML, Graige MS and Feher G (2000) Proton and electron transfer in bacterial reaction centers. Biochim Biophys Acta 1458: 148–163 Paddock ML, Flores M, Isaacson R, Chang C, Abresch EC, Selvaduray P and Okamura MY (2006) Trapped conformational states of semiquinone (D+•QB–) formed by B-branch electron transfer at low temperature in Rhodobacter sphaeroides reaction centers. Biochemistry 45: 14032–14042 Parson WW (1987). The bacterial reaction center. In: Amesz J (ed) Photosynthesis, pp 43–61. Elsevier, New York Petsko GA and Ringe D (2000) Observation of unstable species in enzyme-catalyzed transformations using protein crystallography. Curr Opin Chem Biol 4: 89–94 Pokkuluri PR, Laible PD, Deng Y-L, Wong TN, Hanson DK and Schiffer M (2002) The structure of a mutant photosynthetic reaction center shows unexpected changes in main chain orientations and quinone position. Biochemistry 41: 5998–6007 Pokkuluri PR, Laible PD, Crawford AE, Mayfield JF, Yousef MA, Ginell SL, Hanson DK and Schiffer M (2004) Temperature and cryoprotectant influence secondary quinone binding position in
972 bacterial reaction centers. FEBS Lett 570: 171–174 Poluektov OG, Utschig LM, Dubinskij AA and Thurnauer MC (2004) ENDOR of spin-correlated radical pairs in photosynthesis at high magnetic field: A tool for mapping electron transfer pathways. J Am Chem Soc 126: 1644–1645 Poluektov OG, Utschig LM, Dubinski AA and Thurnauer MC (2005) Electron transfer pathways and protein response to charge separation in photosynthetic reaction centers: Timeresolved high-field ENDOR of the spin-correlated radical pair P+865QA–. J Am Chem Soc 127: 4049–4059 Prisner TF, Van der Est A, Bittl R, Lubitz W, Stehlik D and Möbius K (1995) Time-resolved W-band (95 GHz) EPR spectroscopy of Zn-substituted reaction centers of Rhodobacter sphaeroides R-26. Chem Phys 194: 361–370 Remy A and Gerwert K (2003) Coupling of light-induced electron transfer to proton uptake in photosynthesis. Nat Struct Biol 10: 637–644 Rohrer M, MacMillan F, Prisner TF, Gardiner AT, Möbius K and Lubitz W (1998) Pulsed ENDOR at 95 GHz on the primary acceptor ubisemiquinone QA– • in photosynthetic bacterial reaction centers and related model systems. J Phys Chem B 102: 4648–4657 Schlichting I and Chu K (2000) Trapping intermediates in the crystal: Ligand binding to myoglobin. Curr Opin Struct Biol 10: 744–752 Schmid R and Labahn A (2000) Temperature and free energy dependence of the direct charge recombination rate from the secondary quinone in bacterial reaction centers from Rhodobacter sphaeroides. J Phys Chem B 104: 2928–2936 Schweiger A and Jeschke G (2001) Principles of Pulse Electron Paramagnetic Resonance. Oxford University Press, Oxford Sinnecker S, Reijerse E, Neese F and Lubitz W (2004) Hydrogen bond geometries from electron paramagnetic resonance and electron-nuclear double resonance parameters: Density functional study of quinone radical anion-solvent interactions. J Am Chem Soc 126: 3280–3290 Snyder SW and Thurnauer MC (1993). Electron spin polarization in photosynthetic reaction centers. In: Deisenhofer M and Norris J (eds) The Photosynthetic Reaction Center, pp 285–329. Academic Press, New York Srajer V, Teng T, Ursby T, Pradervand C, Ren Z, Adachi S, Schildkamp W, Bourgeois D, Wulff M and Moffat K (1996) Structure of a protein photocycle intermediate by millisecond time-resolved crystallography. Science 274: 1726–1729 Steffen MA, Lao K and Boxer SG (1994) Dielectric asymmetry in the photosynthetic reaction center. Science 264: 810–816 Stehlik D, Bock CH and Petersen J (1989) Anistoropic electron spin polarization in photosynthetic reaction centers. J Phys Chem 93: 1612–1619 Stowell MHB, McPhillips TM, Rees DC, Soltis SM, Abresch E and Feher G (1997) Light induced structural changes in photosynthetic reaction center: Implications for mechanism of electron-proton transfer. Science 276: 812–816 Takahasi E, Maroti P and Wraight CA (1992). Coupled proton and electron transfer pathways in the acceptor quinone complex of reaction centers from Rhodobacter sphaeroides. In: Muller A (ed) Electron and Proton Transfer in Chemistry and Biology, pp 219–236. Elsevier, New York Tang J, Utschig LM, Poluektov O and Thurnauer MC (1999) Transient W-band EPR study of sequential electron transfer
Oleg G. Poluektov and Lisa M. Utschig in photosynthetic bacterial reaction centers. J Phys Chem 103: 5145–5150 Thurnauer MC and Gast P (1985) Q-band (35 GHz) EPR results on the nature of A1 and the electron spin polarization in photosystem I particles. Photochem Photobiol 9: 29–38 Thurnauer MC and Norris J (1980) An electron spin echo phase shift observed in photosynthetic algae. Possible evidence for dynamic radical pair interactions. Chem Phys Lett 76: 557–561 Thurnauer MC, Poluektov O and Kothe G (2004). Time-resolved high-frequency and multifrequency EPR studies of spin-correlated radical pairs in photosynthetic reaction center proteins. In: Grinberg OY and Berliner LJ (eds) Very High Frequency (VHF) ESR/EPR (Biological Magnetic Resonance, Vol 22), pp 165–206. Kluwer Academic Publishers, Dordrecht Tiede DM, Utschig L, Hanson DK and Gallo DM (1998) Resolution of electron and proton transfer events in the electrochromism associated with quinone reduction in bacterial reaction centers. Photosynth Res 55: 267–273 Tiede DM, Vazquez J, Cordova J and Marone PA (1996) Timeresolved electrochromism associated with the formation of quinone anions in the Rhodobacter sphaeroides R26 reaction center. Biochemistry 35: 10763–10775 Treutlein H, Schulten K, Brunger AT, Karplus M, Deisenhofer J and Michel H (1992) Chromophore-protein interactions and the function of the photosynthetic reaction center: A molecular dynamics study. Proc Natl Acad Sci USA 89: 75–79 Trissl H-W, Bernhardt K and Lapin M (2001) Evidence for protein dielectric relaxations in reaction centers associated with the primary charge separation detected from Rhodospirillum rubrum chromatophores by combined photovoltage and absorption measurements in the 1–15 ns time range. Biochemistry 40: 5290–5298 Utschig LM, Bryson JW and O’Halloran TV (1995) 199Hg NMR of the metal receptor site in MerR and its protein-DNA complex. Science 268: 350–358 Utschig LM, Thurnauer MC, Tiede DM and Poluektov OG (2005) Low-temperature interquinone electron transfer in photosynthetic reaction centers from Rhodobacter sphaeroides and Blastochloris viridis: Characterization of QB– states by high-frequency electron paramagnetic resonance (EPR) and electron-nuclear double resonance (ENDOR). Biochemistry 44: 14131–14142 van den Brink JS, Hulsebosch RJ, Gast P, Hore PJ and Hoff AJ (1994) QA binding in reaction centers of the photosynthetic purple bacterium Rhodobacter sphaeroides R26 investigated with electron spin polarization spectroscopy. Biochemistry 33: 13668–13677 Van der Est A, Bittl R, Abresch EC, Lubitz W and Stehlik D (1993) Transient EPR spectroscopy of perdeuterated Zn-substituted reaction centers of Rhodobacter sphaeroides R26. Chem Phys Lett 212: 561–568 Van der Est A, Siekmann I, Lubitz W and Stehlik D (1995) Differences in the binding of the primary quinone acceptor in photosystem I and reaction centres of Rhodobacter sphaeroides R26 studied with transient EPR spectroscopy. Chem Phys 194: 349–360 Verméglio A and Clayton RK (1977) Kinetics of electron transfer between the primary and secondary electron acceptor in reaction centers from Rhodopseudomonas sphaeroides. Biochim Biophys Acta 461: 159–165
Chapter 48
Protein Environments and Electron Transfer Processes
Weil JA, Bolton JR and Wertz JE (1994) Electron Paramagnetic Resonance. Elementary Theory and Practical Applications. Wiley, New York Wraight CA (2004) Proton and electron transfer in the acceptor quinone complex of photosynthetic reaction centers from Rhodobacter sphaeroides. Frontiers in Biosciences 9: 309–337 Wuttke DS, Bjerrum MJ, Winkler JR and Gray HB (1992) Electron-tunneling pathways in cytochrome c. Science 256: 1007–1009 Xu Q, Baciou L, Sebban P and Gunner MR (2002) Exploring
973
the energy landscape for QA– to QB electron transfer in bacterial photosynthetic reaction centers: Effect of substrate position and tail length on the conformational gating step. Biochemistry 41: 10021–10025 Xu Q and Gunner MR (2001) Trapping conformational intermediate states in the reaction center protein from photosynthetic bacteria. Biochemistry 40: 3232–3241 Zouni A, Witt HT, Kern J, Fromme P, Kraub N, Saenger W and Orth P (2001) Crystal structure of Photosystem II from Synechococcus elongatus at 3.8 Å resolution. Nature 409: 739–743
Index A α-helical segment 190 α-ketobutyrate 613 α-Proteobacteria 780, 791 α/β heterodimer 137–139 α helix 511 α polypeptide 175, 188, 204 αβ LH1 subunit 513 A-branch 301 A-side electron transfer 346 aa3-type cytochrome c oxidase 407, 408, 541, 636, 788 AAA+ 69, 748 AAA+ ATPases 749 AAA proteins 69 AAA+ proteins 70 AAnP. See aerobic anoxygenic phototrophs AAP. See aerobic anoxygenic phototrophs ABC. See ATP-binding cassette (ABC) absorbance spectra in vivo 39 absorption maxima 431 carotenoid 926 absorption transitions 200 accA 122 accB 122 accC 122 accD 122 acceptor pool 530 acceptor quinone 379, 380–399 reactions 382 acceptor quinone complex 383 accessory phototrophy 50 Acetobacteraceae 603, 614 3-acetoxychlorophyllide a 68 acetyl carbonyl 204 Acidiphilium 34, 112, 113, 599, 608 Acidiphilium acidophilum 599 Acidiphilium cryptum 603, 614 Acidiphilium cryptum JF-5 790, 791, 792 Acidiphilium rubrum 112, 113 Acidisphaera 34 Acidithiobacillus 459, 608 Acidithiobacillus ferrooxidans 459, 666 acidophilic purple bacteria 11 acid pocket 715 acpP 122 acpS 122 ACP synthase 122 acrylic acid 546 AcsF 42, 72 acsF 44, 47, 51, 72, 794 actin filament 486 action spectrum 871 active P+QB– state 957, 961, 963 acyl carrier protein (ACP) 122
acyltransferase 123 adduct formation 815, 822 adenosine-5´-phosphosulfate (APS) 610, 613 adenosine-5´-phosphosulfate reductase 610, 611 adenosyl-GDP-cobinamide 91 adenosyl cobalamin 72, 73, 82, 91 adenosylcobinamide phosphate 91 adenosylcobyric acid 90 adenylylsulfate 610, 613 adenylysulfate:phosphate adenylyltransferase (APAT) 610 Adiantum capillus-veneris 814 AdoMet:diacylglycerol 3-amino-3-carboxypropyl transferase 128 ADP-inhibition 477 ADP sulfurylase 610 Aequoria victoria 845 aerobic anoxygenic phototrophs 19, 31, 32–52, 599 carbon metabolism 40–41 deep ocean vertical distribution 50 ecological roles 47–51 environment 32 evolution 38–40 marine 48–49 morphology 35 nutritional status 43 phylogeny 33, 37 taxonomy 37 aerobic bacteria 112 aerobic cobalamin biosynthetic pathway 84 aerobic conditions 58, 72, 112, 417 aerobic cyclization system 72 aerobic pathway 84 aerobic photosynthetic bacteria 98, 99, 112 aerobic phototoxicity 38 aerobic phototrophic bacteria 19 aerobic purple bacteria 19–20 aerobic repression CrtJ 716–721 aerobic repressor 784, 785 aerobic respiration 149 aerotaxis 651 AerR 785, 791 aerR 785, 786, 789 affinity chromatography 849 AFFM. See atomic force fluorescence microscopy (AFFM) AFM. See atomic force microscopy (AFM) AFM topographs 164 Agrobacteria 694 Agrobacterium 702 Agrobacterium tumefaciens 805 ALA. See δ-aminolevulinate; See 5-aminolevulinic acid (ALA) ALAD. See δ-aminolevulinic acid dehydratase ALA dehydratase 60 ALA synthase 778, 780, 782, 783, 786, 788, 791 Alcaligenes eutrophus 744
Index
976 Alcaligenes faecalis 635 alcohols 40 Alexandrium 48 algae 103 algal blooms 48 Alkalilimnicola 600 Alkalilimnicola ehrlichei 604, 615 alkaliphilic purple bacteria 10–11 Alkalispirillum 600 Allochromation vinosum 612, 745 Allochromatium 8 Allochromatium minutissimum 607 Allochromatium vinosum 3, 4, 6, 103, 384, 515, 596, 601, 604, 606, 609, 610, 613, 615, 853 Allochromatium warmingii 601 allosteric 791 alphaproteobacteria 5, 18–19, 59, 98, 99, 111, 112, 581, 597 phototrophic 597 alternative nitrogenase 769 amidase 90 amino acid distribution 918, 920 model sequence 920 motifs 918, 920 sequence contexts 922 amino acid sequence 921 model 916 5-aminolevulinic acid (ALA) 778, 780, 782, 783, 786, 788, 791 aminopropanol 82 aminopropyltriethoxysilane 862 ammonium transporter 769–770 AmtB 769–770 amphiphile 146 amphiphilic protein maquettes 905–907 AmtB 769–770 ammonium transporter 769–770 anaerobic ammonia oxidation 624 anaerobic aromatic compound degradation 589–590 anaerobic benzoate degradation 580–589 anaerobic benzoate photometabolism 580 anaerobic conditions 72, 523 anaerobic respiration 149 anaerobiosis 848 analogs 184 anaplerotic CO2 fixation 40 anaplerotic CO2 incorporation 48 AnfA 769 anfA 765 anfHDGK 769 anfHDK 761 anhydrorhodovibrin 929, 931 anisotropy decay 243 annihilation experiments 148 Antarctica 4 antenna clustering 945 complex assembly 946 domain formation 944 heterogeneity 944
hexagonally packed 945 LH2 size heterogeneity 948 packing density 946 ring size 151 antenna absorption transition 203–205 antenna complex 146–151; See also light-harvesting 1 complex; See also light-harvesting 2 complex artificial 861 synthesis 147 antenna connectivity 170 antenna domain formation 944 antenna domains 257, 945, 946, 949 antenna heterogeneity 944 antenna proteins 200–201 anthraquinone 304, 382, 391 antibiotic resistance markers 842 antimycin 519 antimycin A 435, 455, 526 antioxidative 41 anti sigma factor FlgM 645 APAT 611 APB. See aerobic phototrophic bacteria apoCyt c 409 apoCyt c heme binding 413 apocytochrome c 407 apoptosis 527 AppA 652, 718, 719, 732, 785, 818, 831 light-responding antirepressor 718 regulatory role 718 appA 785 apparent equilibrium constant 520, 524 apr 603 APS kinase 613, 615 APS reductase 615 APS reductase pathway 610 Aquifex aeolicus 749 Arabidopsis thaliana 72, 612, 847 ArcB/ArcA 753 ArcB sensor kinase 754 Archaea 609, 746 Arhodomonas aquaeolei 600 aromatic compounds 577–591 degradation 577–591 aromatic residues 916, 918, 920, 922, 924 arrA 549 arrB 549 ars determinants 662 arsABC 673, 675 arsC gene family 676 arsenate 549, 663 arsenate reductase 676 arsenate respiration 549 arsenic 548–550 arsenite 550 arsenite oxidase 459, 549, 552 arsenite oxidation 549, 664 ars operon 663 Arsukibacterium ikkense 601 artificial antenna complex 861 artificial LH1-type complex 873 Aspergillus nidulans 847
Index assembly 195 core complexes 172 assembly factor LhaA 174 assembly factor for LH2 174 assembly pathways 407 assembly protein Surf1 545 assimilatory nitrate reductase (Nas) 548, 637 assimilatory sulfate reduction 610, 612–615, 613, 616 asymmetrical ζ-carotene 104 AT-rich codons 850 ATB binding Walker A and B motifs 411 ATCC17023 wild-type strain 849 ATCC17025 780, 782, 791 ATCC17029 780, 782 atomic force fluorescence microscopy (AFFM) 270 atomic force microscopy (AFM) 137, 146, 150, 163, 165, 201, 254, 255, 257, 258, 262, 270, 286, 478, 513, 520, 528, 861, 862, 864–865, 941–951 fast scanning 950 scan range 950 tip as nano-dissector 950 atomic level structural model photosynthetic unit 287 atomic models 950 atovaquone analog 445 ATP 90, 613 ATP-binding cassette (ABC) 410 ATP-binding cassette transporter complex 410 ATP-synthase 475–493, 942 ATP:sulfate adenylyltransferase 610 ATP analog 806 ATPase 419, 748 CPx-type 661 P-type 661 ATPase domain 69 ATP binding site 69 ATP dependent oligomers 70 ATP hydrolysis 66, 70 ATP production 425 ATP regenerating system 67 ATP regeneration system 74 ATP sulfurylase 610, 611, 613, 615, 616 ATP synthase 254, 262, 264, 279, 476–488, 510, 538 proton translocation 476–488 ATP synthesis 510 ATP synthesis/hydrolysis 478–486 auto-induce 844 auto-oxidization 438 autophosphorylation 800, 802, 806 RegB 713 Avena sativa 815 average excitation lifetime 280 avoided level crossings 283 Azoarcus 581, 588 Azotobacter vinelandii 637 azoxystobin 442
977
B β-barrel 416 β-carotene 98, 113 β-carotene ketolase 113 β-dodecyl maltoside 260 β-galactosidase 780, 789 β-hydroxydecanoyl-ACP dehydratase 122 β-hydroxyl-ACP dehydratase 122 β-ketoacyl-ACP reductase 122 β-ketoacyl-ACP synthase I 122 β-ketoacyl-ACP synthase II 122 β-ketoacyl-ACP synthase III 122 β-octyl-glucoside 185 critical micelle concentration 185 micelle 185 β-octylglucoside (β-OG) 140, 146, 864 β-oxidation 580, 581, 584, 585, 586, 587 β helix 511 β mutant 338 β mutation 347 β polypeptide 175, 188, 205 ‘b’ position 454, 456 B-branch 301, 306 electron transfer 348, 958 B-branch electron transfer 348, 958 B-side electron transfer 346–348 b-type heme 452, 752 B1020 150 B780 934 B798-832 35, 46 B800 142, 150, 203, 214, 225–226, 878, 882–885 B800-814 46 B800-820 46, 150 B800-830 148, 150 B800-850 46, 150 B800-B850 920 fluorescence up-conversion 225 B800 molecule 203 B806 46 B820 182, 188, 932, 934 dimeric structure 186 hydrophobic surface area 183 NMR experiments 186 oligomerization 187 reversible dissociation 183 B820-type complex 188 minimal requirements 190 B820 complex 157, 175, 183 chemically synthesized polypeptides 189 heterodimeric B820 189 homodimeric B820 189 membrane-spanning middle segment 189 N-terminus 189 protein interactions 189 proteolysis 189 shorter synthetic polypeptides 189 B850 142, 214, 878, 885–889 B850 ring elliptical deformation 207 B870 934
978 B875 150, 927 B880 926, 932, 935 B880 formation 931 B890 150 Bacillus 457, 669 Bacillus PS3 479 Bacillus selenitireducens 549, 664 Bacillus subtilis 413, 813, 852 back-reaction 148 bacterial artificial chromosome (BAC) 45 bacterial plasma membrane 264 bacteriochlorin 58, 897 bacteriochlorin exclusion 301 bacteriochlorin replacement 302 bacteriochlorin ring 141 (bacterio)chlorophyll 916 (bacterio)chlorophyll binding pocket 916 statistical analysis 916 binding 916 binding-motifs 916 binding pockets 919 binding site re-design 920 bioinformatics 916 Chl binding 918–919 hydrogen bonding 916, 918 interactions 917 ligation 916, 918 motifs 922 pigment-protein 916 pockets 918 substituent 918 interactions 918 α-ligated states 918 α-ligation 918 β-ligated 923 β-ligated states 918 β-ligation 918 bacteriochlorophyll (BChl) 21, 42, 58–75, 138, 182, 184, 186, 202, 232, 337, 356, 699, 703, 778, 780, 783, 787, 794, 895–907, 914, 916, 920; See also chlorophyll (Chl) aggregation 897 B800 142 B850 142 bacteriochlorophyll-protein hydrogen bonds 186 (B)Chl ligation 915 binding 202, 919, 922 model 919 binding pocket distribution 918 biosynthesis 58–75, 164, 787 C132 keto group 186 C3-acetyl group 139, 140, 144 C31 acetyl group 186 exchange 915 histidine ligand 185, 344 isocyclic ring 915 ligand 922 ligation α-type 915 β-type 915
Index Mg coordination 184–188 binding energy 185 modified pigments 915 molecule 204, 205 distorted conformation 205 monomer 338, 341 π-π interactions 190 Qy-band 897 replacement 305 spectra 897 stereochemical aspects 915 surface area 897 tetrapyrrole ring 42 transition dipoles 168 water ligand 344 bacteriochlorophyll-B850/protein interface 922 bacteriochlorophyll-binding proteins synthetic 895, 896–907 bacteriochlorophyll/protein interface 915, 916, 922 packing interactions 923 bacteriochlorophyll/transmembrane helix interface 923 bacteriochlorophyll a 42, 58, 65, 68, 112, 214, 338, 873 zinc-substituted 873 bacteriochlorophyll aGG 68 bacteriochlorophyll b 58, 65, 112 bacteriochlorophyll b-containing B800-1020 46 bacteriochlorophyll biosynthesis 83 cyclase enzyme 83 Mg chelatase 90 bacteriochlorophyll dimer 337, 344 bacteriochlorophyll exchange light-harvesting 1 complex 928 bacteriochlorophyllide a 68 bacteriochlorophyll molecules coordination to Mg 933 bacteriochlorophyll pathway 147 bacteriochlorophyll protein assembly 918 bacteriochlorophyll protein platforms de novo design 895, 896–907 bacteriochlorophyll synthase 67, 74, 147 bacteriochlorophyll synthesis 83, 92 bacteriopheophorbide 902 Fe 902 Ni 902 Zn 902 bacteriopheophytin (BPhe) 44, 193, 341, 356, 392 bacteriopheophytin a 338 bacteriophytochrome 43, 267, 719, 728–730, 799–807, 827–831 Agrobacterium tumefaciens 805 Bradyrhizobium ORS278 800 CBD 805 chromophore binding domain 803 Cph1 804 Deinococcus radiodurans 800, 805 dimerization domain 803 flash photolysis 805 Fremyella displosiphon 800 GAF domain 803, 804, 805 Kineococcus radiotolerans 804
Index PAS domain 803, 805 Pfr 804 photocycle 829 PHY domain 803 Pr 804 Rhodobacter sphaeroides 804 Rhodospirillum centenum 800 Synechocystis PCC 6803 800, 804 bacterioplankton 37 bacteriorubixanthinal 113 Badger-Bauer relationship 390 band shift 554 Banff thermal springs 51 barrel 416 basal body 646 base-catalyzed mechanism 815 bc1:RC stoichiometric ratio 512 BchB 74 bchB 74 BchC 73 bchC 73 BchD 67, 717 bchD 787 BchE 42, 72 bchE 44, 47, 51, 72, 783, 794, 795 BchF 73 bchF 73 BchG 268 bchG 74 BchH 67, 69, 71 protoporphyrin IX binding 70 bchH 787 BchI 67, 70, 717 bchI 787 bchJ 72 BChl. See bacteriochlorophyll (BChl) BchL 74 bchL 74 BChl-B850 915, 923 BChl/BChl coupling 207 Bchla GG reductase 307 BChl synthase 268 BChl triplet 300 BchM 71 bchM 71 BchN 74 bchN 74 bchP 74, 164 BchX 74 bchX 74 bchXYZ 73 BchY 74 bchY 74 BchZ 74 bchZ 74 bciA 73 bd-type oxidase 540, 542, 543 behavior 646–653 behavioral responses 646–653 benzamidine 146 benzene rings 577 benzoate degradation 580–589, 591, 702
979 benzoate photometabolism 580 benzoquinone 382 benzoyl-CoA reductase 583 betaine 119 betaine lipid 120, 128–129 betaproteobacteria 5, 22, 23, 98, 100, 111, 112, 603 bH–bL heme distances 430 bH cytochrome 517 bilin 729, 804–805, 828 ATP analog 806 biliverdin IXα 804, 805 E-Z isomerization 804 Lumi-F 804, 805 Lumi-R 804, 805 meta-Ra 805 meta-Rc 805 photo-isomerization 804 phycocyanobilin 804 phytochromobilin 804 Z-E isomerization 805 ZZZssa conformation 805 bilin chromophore 828 biliverdin 829 biliverdin IXα 804, 805 binding 16-fold LH1 ring 163 binding change mechanism 477 binding energy 185 binding helix 902, 903, 917 binding motifs 916 binding pocket statistical analysis 924 binding site CrtJ 717 RegA 715 binuclear iron-sulfur clusters 439 biodiversity 51 bioenergetic enzymes 703 biofilm 51, 696, 701 biofilm formation 820, 824, 828 biogenesis 407, 408–421, 846 core complexes 174–175 cytochrome complex 415–421 intermediates 408 photosynthetic complexes 175 biogenesis pathway 408 biogeochemical cycling 51 bioinformatics 916 biomass 50 biomass production 126 bionanotechnology 862 biosynthesis antenna complex 147 bacteriochlorophyll 787 carotenoid 97–114 gene cluster 103 genes 102 carotenoid gene cluster 101 heme 415, 780 membrane lipid 119–132 spirilloxanthin 104–108 biotin carboxylase subunit of ACC 122
Index
980 biotin carboxyl carrier protein subunit of ACC 122 biphasic growth 600 biphasic photooxidation 517 5-bisphosphate carboxylase/oxygenase (Rubisco) 41, 564 bisphosphatidylglycerol 125 bL–c1 heme distances 430 Blastochloris 440 Blastochloris sulfoviridis 7, 579, 597, 598 Blastochloris viridis 9, 45, 104, 161, 239, 256, 258, 259, 296, 299, 343, 356, 360, 380, 383, 384, 387, 388, 394, 397, 512, 514, 518, 942–944, 960 reaction center-light-harvesting 1 complexes 162, 258, 943 reaction center X-ray structure 383, 386 Blastomonas 34 Blastomonas natatoria 34 Blastopirellula marina 458 bL cytochrome 517 blooms stratified lakes 7–8 BlrB 818, 821 BluB 91 BluE 92 blue-shift 42, 45, 148, 188, 203, 204, 245–247, 306, 823, 827, 829, 922, 930, 932 blue light photoreceptor 719 blue light sensor 652 blue native polyacrylamide gel electrophoresis (BN-PAGE) 418526 BluF 92 BLUF domain 731, 732–734, 811, 816–823 crystal structure 733 BLUF photocycle 818 BN-PAGE. See Blue-Native-polyacrylamide gel electrophoresis (BN-PAGE) Bordetella bronchiseptica 415 Bordetella parapertussis 415 Bordetella pertussis 414 Born-Oppenheimer approximation 885 Bos taurus 479 Box I 439 Box II 439 BPhe. See bacteriopheophytin (BPhe) BPhe replacement 306 Bradyrhizobiaceae 23, 603, 614 Bradyrhizobium 19, 22–27, 100, 113, 439, 627, 667, 693, 720, 721, 728, 729 PpsR2 719 Bradyrhizobium denitrificans 32 Bradyrhizobium japonicum 418, 540, 581, 708, 709, 744, 766, 781 Bradyrhizobium sp. BTAi1 23, 627, 629, 630, 719, 789, 790, 792 Bradyrizobium ORS278 102, 113, 699, 719, 729, 790, 800, branched four-helix bundle proteins 901–904 broad phonon side band (PSB) 884 Brownian ratchet 485 bta genes 120 btaA 121 btaB 121 1 – Bu state 224, 227 energy transfer 224
bucket brigade hydrogen transfer 518 buoyancy regulation 34 Burkholderiales 599 bypass reactions 460
C c position 456 C-5 pathway 59 c-di-GMP phosphodiesterase 820 c-ring 478 C-terminal light-harvesting 1 complex 871 C-terminal processing 150 C-terminal regions 187 C-terminus of LH1 α progressive deletion 165 c-type apocytochromes 269 c-type cytochrome 409–415 maturation 436 pentaheme 547 C131 keto group 144 C3-acetyl group 139, 140, 144 bacteriochlorophyll 144 C30 molecule 41 C4 pathway 778, 791 C5 pathway 778, 791 caa3-type enzyme 542 CadA/CadC resistance system 661 caldariellaquinone 469 Calvin-Benson-Bassham cycle 564–566, 568–570, 572 Calvin Benson Bassham response regulators/sensor kinase system (CbbRRS) 572–574 Calvin Cycle 19, 25, 27, 38 Calyptogena magnifica 609 Candidatus kuenenia 458 canthaxanthin synthesis 113 carbohydrates 40 carbon-fixation 24–27 carbon availability 789 carbon cycling 48 carbon dioxide metabolism 564–575 carbon fixation 707 carbon metabolism aerobic anoxygenic phototrophs 40–41 carbon monoxide 752 carbon sources 40 carbonyl 387, 389, 390, 392, 395, 396 carboxyltransferase subunit (CT-α) of Acetyl-CoA carboxylase 122 carboxyltransferase subunit of ACC 122 cardiolipin 120, 125, 526, 527 carotenal pathway 108, 109, 112 carotene isomerase 104 carotenogenesis 92, 101–111 pathways table 99–101 carotenoid 41, 97–114, 138, 183, 193, 300, 340, 348, 699, 700, 703, 914, 915, 916, 919, 921, 924 absorption 915 absorption maxima 926
Index acids 98 acyclic 98 assembly 915 binding 916, 920, 924 aromatic residues 924 mutations 924 statistical analysis 919 binding-motifs 916 binding pockets 919, 924 bioinformatics 916 biosynthesis 97, 98–114, 800, 854 gene cluster 103, 113 genes 102 carotenoid pathways 99 table 99–101 composition 98 conformations 924 conjugation length 930 cyclic 98 effect on LH1 formation 930 energy transfer to BChl 915 Erythrobacter-type 113, 114 glucoside 98 fatty acid ester 98 glucoside fatty acid ester 98 hydrophobic interactions 915, 924 methoxy 98 mutant 107 non-photosynthetic 98, 113 π-π 924 photoactive 38 pigment-protein sites 916 plant 113 polyene chain 919, 924 resonance Raman studies 924 role in assembly of RC-LH1 complexes 193 S* state 222–224 energy transfer 222–224 S1 state 219–222 energy transfer 219–222 S2 state 216–219 energy transfer 216–219 stabilizing effect on LH1 931 structural stabilization 915 sulfates 98 carotenoid-B820 interactions 933 carotenoid-bacteriochlorophyll energy transfer 213–230, 927, 931, 933 carotenoid-band shift 554 carotenoid-less mutant 165 carotenoid-LH1 subunit interactions 931 carotenoid acids 98 carotenoid binding in LH1 931 hydrophobic interactions 933 π-π stacking 933 carotenoid biosynthesis genes 102 gene cluster 101 carotenoid excited state 215–216 carotenoid glucoside 98, 109–110 carotenoid glucoside ester 109 carotenoidless mutant 163
981 carotenoidless mutants 193, 926 carotenoid pathways table 99–101 carotenoid sulfate 98, 113, 114 carotenoid sulfates 38 carotenoid synthesis 175, 800 carotenoid triplet 298 catalysis Rieske/Cytochrome b complex (RB) 455–456 cation-π interaction 329, 444 cation diffusion facilitator (CDF) 661 cation diffusion facilitator system 661 CBA efflux pumps 658 cbb 564–575 cbb3 cytochrome oxidase 700, 735, 754, 782, 788 cbb3-Cox 415–420 cbb3-type cytochrome c oxidase 407, 540–541, 543, 545, 553, 629, 636, 781 cbbI operon 564–570, 572, 573 cbbII operon 564–570, 572, 573 cbbLS gene 564, 565, 573, 574 cbbM gene 564, 566, 573 CbbR 563–572 cbbR 564 cbbR 564–567, 569, 572 CbbRRS. See Calvin Benson Bassham response regulators/ sensor kinase system (CbbRRS) CbiE 89 CbiT 89 CC1HL 410 CcdA 410 ccdA 409 Ccm-system I 407, 409–415 CcmABCDEFGHI 409, 410 CcmHIF-containing complex 414 ccoGHIS operon 540 CcoN 540 CcoNOQP 419 ccoNOQP 417, 781 ccoNOQP operon 540 CcoO 540 CcoP 540 Ccs-system II 409–415 CcsA 410 CcsB 410 CcsX 410 cd1 surface helix 436 cd1 type nitrite reductase 635 cd2 surface helix 436 CDP-diacylglycerol 123 cdsA 121 cell-free expression system 852 cellular BChl content 42 cellular redox 708 cellulase 854 Cellulomonas fimi 854 census techniques 47 central metal binding selectivity 902 coordination 898, 902 coordination number 904
982 cfp 649 chaperone 411 copper metallo 417 chaperones 701 chaperonins 269 charge-transfer character 208 charged residues 329, 331 charge recombination 343, 345–347, 383 charge separation 136, 355, 356–370, 362 reaction center 356–370 charge transfer 446 chelatase 58 chemiolithoautotrophic 564 chemoheterotrophic 564 chemoheterotrophic culture 842, 844, 848 chemoheterotrophic growth 267 chemosensory machinery 645 chemosensory pathway 643 chemosensory proteins 649 chemotaxis 646, 698 Rhodobacter sphaeroides 647–651 chimeric RCs 298 Chlamydomonas 457, 814–815 Chlamydomonas reinhardtii 72, 468, 813 chloramphenicol 70 chlorate reductase 550 Chlorella 65 chloride 312 chlorin 897 chlorin macrocycles 58 chlorin reductase 67, 73, 74 chlorobactene glucosyltransferase 110 chlorobactene lauroyltransferase 110 Chlorobaculum tepidum 103, 110 chlorobenzoate 577 3-chlorobenzoate 578 Chlorobiaceae 380, 596 Chlorobium 746 Chlorobium chlorochromatii 676 Chlorobium ferrooxidans 676 Chlorobium limicola 103, 549 Chlorobium limicola DSM245 676 Chlorobium phaeobacteroides 549 Chlorobium tepidum 73, 103, 110, 410, 676 Chlorobium vibrioforme 607 Chloroflexaceae 380, 596 Chloroflexi 746 Chloroflexus 8, 679 Chloroflexus aggregans DSM 9485 676 Chloroflexus aurantiacus 103, 385, 549, 550, 666, 813 Chloroflexus aurantiacus J-10-fl 676 chlorophyll (Chl) 67, 914, 918. See also bacteriochlorophyll (BChl) binding sites 898 covalent binding 901 induced protein structure 905 spectra 897 stereochemistry 898 chlorophyll-protein interactions 897–899, 898 chlorophyll a 343 chlorophyll a´ 343
Index chlorophyll binding proteins synthetic 896–907 chlorophyll delivery 269 chlorophyll geometries robustness and optimality 285 chlorophyllide 73, 74 chlorophyllide a 68 chlorophyll synthase ChlG 74 chlorophyll synthesis 800 chloroplast 409, 476, 746 chloroxanthin 107 chromate 663 Chromatiaceae 5, 380, 596, 600 Chromatiales 600 chromatic adaptation 260, 267, 800 Chromatium 8, 543, 666 Chromatium okenii 6, 109 Chromatium purpuratum 109, 183, 201, 220 Chromatium salexigens 823 Chromatium tepidum 10, 201 Chromatium vinosum 103, 148–149, 356, 435, 441, 566, 581, 745 Chromatium weissei 6 chromatophore 71, 147, 257, 426, 484, 520, 521, 703 mini 484 chromatophore heterogeneity 520 chromatophore model 264 photosynthetic unit 287 chromatophore vesicle 262 chromophore 207–208, 823 bilin 828 biliverdin 829 cyanophycobilin 831 flavin adenine dinucleotide (FAD) 813, 816 linear tetrapyrrole 827 modified 824 phycocyanobilin 829 phytochromobilin 827 riboflavin 816 chromosome 780, 786, 791 Chromosome 1 150 ChrR 737 circadian clock 702–703 circular chromosomes 692 circular conformations 163 circular dichroism (CD) 183, 206, 237, 849, 921 spectrum of iB873 933 citric acid cycle 41 Citrobacter 669 Citromicrobium 40 Citromicrobium bathyomarinum 34 clamp region 441 climate change 40, 52 closed ring 163 closed ring structure 160 cls 121 cluster-to-cluster transfer rates 281, 288 CO 752 co-crystallization 528 CO2 fixation 563, 564 anaplerotic 40
Index photoautotrophic 40 primary producers 2 coal refuse heaps 51 cob(II)yrinic acid a,c-diamide 90 cobA 784, 787 cobA 787 cobalamin 81, 81–92 cobalamin biosynthetic genes Rhodobacter capsulatus 84 Rhodobacter sphaeroides 84 cobalamin biosynthetic operons 84 cobalt 82, 90 cobalt insertion 84 cobaltochelatase 89, 90 CobC 90 CobE 92 CobG 86 CobH 89 CobJ 87 CobL 88 CobN 89 CobP 91 CobQ 90 CobS 89 CobT 89 CobW 92 cobyrinic acid a,c diamide synthase 609 CobZ 87 codon usage 850 cofactor functional assembly 924 interactions 916 reconstitution 924 cofactor-binding motif 935 light-harvesting complex 935 cofactors 914 coherent energy transfer 882 cold habitat purple bacteria 11 colloid theory 172 Comamonadaceae 603, 614 combinatorial design 899–901 combinatorial synthesis 900 complementation rescue technique 66 complex-specific assembly factors 268 LhaA 268 PucC 268 PuhB 268 PuhC 268 PuhE 268 Complex I 510, 526, 745 reverse electron transfer 512 Complex III 526 Complex IV 526 complex lipid composition 120 complex organics 40 compliance 485 computational sequence analysis 918 computational studies 139 conductive atomic force microscopy (CAFM) 862, 863, 865–869 conformation 383
983 conformation-dependent cleavage 446 conformational motion 235 conformational relaxation 968 conformational substates 207, 957 conformation locking 967 Congregibacter litoralis 34 Congregibacter litoralis KT71 24 conjugal mating 850 conjugated double-bond system 140 conjugated macrocycle 204 conjugation 851 conjugation length 931 consensus sequence 716 RegR 716 continuous assays 69 CooA 743, 752 CooLH hydrogenase 746 Rhodospirillum rubrum 746 Coomassie Brilliant Blue 846 cooperative formation 934 cooperative formation of LH1 931 coordinate regulation 717 coordination axial 408 Mg 139, 140 Mg2+ 144 out of plane 408 coo regulon 752 cop operon 662 copper-binding motif 420 copper-binding protein Cox11p 545 copper-containing Nor 634 copper binding site 418 copper metallochaperone 417 copper transporter 420 coproporphyrinogen III 60, 64, 785, 787, 788, 794 coproporphyrinogen III oxidase 60, 64 core antenna 915 core complex 943, 946 assembly 172 biogenesis 174–175 core dimers phospholipid content 173 quinone content 173 core light-harvesting antenna complex. See light-harvesting 1 complex corrinoid synthesis 784 corrin ring 82, 84 cosmid 782, 785, 789 cosynthase 64 coupling 345, 348 coupling efficiency 486 coupling strength 205 coxII promoter 853 Cph1 804 CPO. See coproporphyrinogen III oxidase CPx-type ATPases 661 Craurococcus 34, 113 crocacin analog 445 cross-sectional area 136 cross-transfer probabilities 281
984 cross-trapping probability 289 CRP 752 crt genes 102, 104, 105 CrtA 102, 104, 105, 107, 108 crtA 102 CrtB 102, 105 crtB 717 CrtC 102, 104, 105, 107, 108 CrtD 102, 104, 105, 107, 108 CrtE 102, 105 CrtF 102, 104, 105, 107, 108 CrtH 104 CrtI 102, 103, 105, 108, 113 crtI 717 CrtISO 104 CrtJ 707, 708, 716–720, 785–787, 791, 795 aerobic repression 716–721 binding site 717 oxidized 717 recognition sequence 717 crtJ 785, 786, 787, 789 CrtJ/PpsR 716, 719 phylogenetic analysis 719, 720 CrtR 104 CrtW 102, 104, 105, 113 CrtY 102, 104, 105, 113 crude oil 587 cryo-electron microscopy 254, 256, 286, 942, 947 cryo-EM projection map 169 cryptochrome 734 crystallization 140 lipidic cubic phases 140 crystals Type II 146 crystal structure protein maquette 904 uroporphyrinogen III synthase 64 CuA 542 CuA center 416, 418, 419 Cupriavidus metallidurans 612 Cupriavidus necator 744 current-voltage curves reaction center 867 current-voltage measurements 863, 868 current rectification 868 curvature mismatch 173 CXXCH heme binding motif 408, 409 cyanide 82 cyanobacteria 103, 104, 267, 380, 384, 409, 596, 694 phycobilisomes 384 cyanobacterial phytochrome 800 cyano cobalamin 72, 82 cyanophycobilin 831 cyclic electron flow 263 cyclic electron transfer 266, 510, 512 cyclo-octasulfur 608 cyclohexanecarboxylate degradation 587–588 CydA 540 CysDN 613 cysDN 614
Index cysH 613, 614 CysI 613 cysI 614 CysJ 613 cysJ 614 cysPTWA 613, 614 cystathionine 613 cystathionine γ-lyase 613 cysteine 613 cysteine persulfide 605 cysteine sulfane 605 Cyt. See cytochrome (Cyt) cytidylyltransferase 124 cytochrome (Cyt) 45, 258, 408, 514, 700, 703 a-type 408 b-type 408 c-type 408 d-type 408 o-type 408 soluble 45 tetraheme 942, 943 cytochrome aa3-type terminal oxidase 541 cytochrome b 426, 427, 431–436 di-heme 745 surface 435 topology 432 cytochrome b558/566 459 cytochrome b561 270 cytochrome b6 432 cytochrome b6 f complex 427, 432, 453 evolution 456–458 cytochrome bc1 complex 46, 173, 254, 257, 259, 262–265, 279, 425–447, 453–455, 509–530, 539, 625, 903, 907, 942, 950 dimeric structure 525, 528 distal niche 454 evolution 456–458 extended π-orbital system 445 homodimeric organization 512 large-scale domain movement 431 proximal niche 455 Qi site 467–468 Qo site 455 Rhodobacter mutants 428–430 Rhodobacter sphaeroides 455 structure 427–446, 453–455 subunit IV 262, 432, 453 topology 432 cytochrome bd-type quinol oxidase 539–540 cytochrome bH 453 cytochrome b hemes 452 cytochrome bL 453, 455 cytochrome c 408, 452, 625 large 437 maturation 408 small 437 tetraheme 45 tethering 527 cytochrome c heme lyase 409 cytochrome c oxidase (COX) 131, 426, 527, 672
Index cytochrome c synthetase 409 cytochrome c´ 408 cytochrome c1 263, 408, 426, 427, 436–438 cytochrome c2 46, 173, 263, 279, 323–333, 345, 348, 408, 426, 452, 456, 514–516, 527–528, 542–543, 703, 844 co-crystallization 528 confinement 521 diffusion and confinement 527–528 Rhodobacter sphaeroides 323, 324–333 soluble periplasmic 45 cytochrome c2:reaction center complex 323–333 distal conformation 516 mutation effects 329–332 Rhodobacter sphaeroides 323–333 structure 325–327 cytochrome c2m 515 cytochrome c3 458 cytochrome c4 458 cytochrome c8 514 cytochrome cbb3-type terminal oxidase 540–541 cytochrome cbb3 oxidase 515, 713, 717 cytochrome co 408 cytochrome complex 415–421 biogenesis 415–421 cytochrome cp 408 cytochrome cy 408, 426, 453, 456, 515, 517, 521, 527, 542–543, 845, 846 membrane-bound 542–543 cytochrome f 458 cytochrome maturation c-type 436 cytochrome oxidase 510, 521 cytochrome subunit 298 cytoplasm 407, 409, 412, 413, 419, 420 cytoplasmic membrane 147, 262, 264, 267, 409, 410, 412, 415, 418, 841, 846 invagination sites 267 CzcCBA system 658 Czr system 675
D δ-aminolevulinate 59 δ-aminolevulinate synthase 57 δ-aminolevulinic acid dehydratase 62. See also porphobilinogen synthase δ-aminolevulinic acid synthase 60 ∆µ~H+ See proton electrochemical potential difference 2-D photon echo 242 damage oxidative 545 dark growth 6, 7 Davydov splitting 886 DCCD 480 DCCD-binding protein 480 ddhA 547 ddhB 547 ddhC 547 ddhD 547 deactivation channel 928 Dead Sea 10
985 Debye-Waller factor 885 Dechloromonas 415 deep ocean 50, 52 aerobic anoxygenic phototrophs vertical distribution 50 Dehalococcoides ethenogenes 746 4-dehydrolycopene 112 Deinococcus radiodurans 800, 829, 831 delay after laser flash (DAF) 964 deletion strain PUC705-BA 844, 846 PUF∆LMX21 844, 846 ∆∆11 844, 846 delocalization 234, 237–238, 238, 241 denitrification 624–636 denitrification gene clusters 628 de novo design bacteriochlorophyll protein platforms 895–907 de novo pigment proteins 895 de novo polypeptides 915 density-matrix 355, 368–370 density functional theory 221, 393, 394 Deriphat 160 855 Deriphat 160-C 308 Desulfovibrio vulgaris 611 detergent 145–146, 182, 855 detergent micelle 157 detoxification enzymatic 662–665 detrapping probability 281 deuteroporphyrin IX 69 Dexter energy transfer 218 Dhp domain 806 di(acyl-glucosyl)-diapocarotene-dioate 110, 113 di-heme cytochrome b 745 diacylglycerol-N,N,N,-trimethylhomoserine 120, 128 diacylglycerol glycosyltransferase 121 diacylglycerolhomoserine 128 diacylglycerolhomoserine N-methyltransferase 121 diacylglycerolhomoserine synthase 121 diacylglycerol moiety 122 diagonal disorder 881, 883 diapocarotenoic acid 113 4,4´-diapocarotene-4,4´-dioate 41 diazotroph 761 diazotrophic growth 700 dicarboxylic acids 586 dielectric effects 361 dielectric properties 203 dielectric response P+QA– formation 967 dielectric screening 358 differential centrifugation 846, 855 diffusion limit 328 diheme cytochrome c1 counterparts 436, 437 diheptanoyl-sn-glycero-3-phosphocholine (DHPC) 170 2-dihydro-3 112 4-dihydroanhydrorhodovibrin 108 2-dihydroneurosporene 112 17,18-dihydroporphyrin 897. See also chlorin 3,4-dihydrospheroidene 107
Index
986 4-dihydrospirilloxanthin 108 dihydroxylycopene diglucoside 110, 112 diketospirilloxanthin 111 2,2´-diketospirilloxanthin 108 dimeric core complex comparison with monomer core complex 169 excitation sharing in array 173 reaction center-light-harvesting 1-PufX complex 168–174 dimerization 419, 947 dimer of supercomplexes 529 dimers 946 dimethyl diselenide 674 dimethyl ditelluride 674 dimethyl selenenyl sulfide 674 dimethyl selenide 674 dimethyl sulfide 40, 546, 600, 601 dimethylsulfoniopropionate (DMSP) 40, 546, 600 dimethyl sulfoxide (DMSO) 546, 601, 780, 781, 788, 853 dimethyl sulfoxide reductase 547, 548, 552 dimethyl telluride 674 2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) 870 dinitrogenase reductase activating glycohydrolase (DraG) 767 dinitrogenase reductase ADP-ribosyl transferase (DraT) 767 dinoflagellates 48 Dinoroseobacter 34 Dinoroseobacter shibae 41, 43, 603, 614, 627 Dinoroseobacter shibae DFL 12 676 dioleoyl-9 160 dioxygenase 459 diphenylamine 103 diphosphatidyl-glycerol 527 dipole-dipole approximation 280 dipyrromethane cofactor 63 directed modification reaction center 338–349 disorder 235, 237, 244, 246 light-harvesting complexes 205–207 disordered exciton model 238, 240 dispersed polaron theory 367 dispersive forces 203 disproportionation 607 dissipation probability 280 dissociation constant 327 dissolved organic matter 50 dissymmetry 528 disulfide 606 disulfide bond 413, 714 disulfide bridge 438 divinyl-protochlorophyllide 71, 72 DMS dehydrogenase 548, 550 DMSO. See dimethylsulfoxide (DMSO) DMSO/DMS couple 547 DMSO/TMAO reductase 552 DMSO reductase 547, 548, 552 DMSP-producing algae 50 DNA-binding activity 715 DNA binding motif 715 DnaK 269 DNA recognition helix 716 DNA recognition sequences 715
DNase I footprinting 715 DNR/NNR 633 domain mobility 469 dorA gene 547 dorC 548 dorC gene 547 dor promoter 853 DOXP pathway 103 D pathway 544, 546 DraG 767–768 DraT 767–768 DsbA-DsbB pathway 412 DsbB link 674 dsbD 409 DsbD 410 dsr genes 602, 603, 607, 608, 609 dsrAB 602, 609 DsrAB 607, 609, 616 DsrC 609 dsrEFH 609 DsrEFH 609, 610 DsrJ 609 DsrK 609 DsrL 609 DsrN 610 DsrO 609 DsrP 609 Dsr proteins 608, 610 DsrR 609 duroquinone 390 dynamic disorder 206, 207 dynamic Stokes shift 239
E ecological roles aerobic anoxygenic phototrophs 47–51 Ectothiorhodosinus 600, 601 Ectothiorhodosinus mongolicum 600, 601 Ectothiorhodospira 10, 46, 543, 549, 600, 601 Ectothiorhodospiracea 604, 615 Ectothiorhodospiraceae 5, 112, 596, 600 Ectothiorhodospira haloalkaliphila 600, 601 Ectothiorhodospira halochloris 512 Ectothiorhodospira halophila 652, 823 Ectothiorhodospira marina 601 Ectothiorhodospira marismortui 600, 601 Ectothiorhodospira mobilis 601 Ectothiorhodospira shaposhnikovii 515, 600, 601 Ectothiorhodospira vacuolata 601, 853 effective Hamiltonian 279, 280 light-harvesting assembly 279 ef loop 436, 443 elastic coupling 485–486 elastic power transmission 485 electrical charge separation 426 electric capacitance 484 electric functionality 862 electroabsorption 344 electrochemical proton gradient 269
Index electrochromism 483 electrode 862, 863, 870 APS-ITO 870 PSU-assembled 870 counter 870 potential 871 RC-assembled 871 reference 870 working 870 electromotor 478–481 structure 478–481 electron-nuclear double resonance (ENDOR) 342, 380, 389, 391, 393, 396, 954, 954–969 1 H 959, 960 2 H 959 high-field 396 high-frequency 966 matrix high-frequency 960, 969 Mims-type 959, 960 Q-band 396 spin-correlated radical pair (SCRP) 965, 966, 968 time-resolved 965, 967 time-resolved high-frequency 963, 968 electron-phonon coupling 884 electron-phonon coupling strength 885 electron acceptors 41 electron crystallography 137 electron density map 140 electron donor 4, 41, 149, 597 reaction center 514–515 sulfur compounds 149 electron flow 46 electronic absorption 207 electronic coupling 279, 343, 881 strong 881 weak 881 electronic coupling factor 361 electronic matrix element 345 electronic structure 342 electronic transitions 200, 204, 208 electron injection barrier 868 electron microscopy 844, 941 analysis 163 negative stain analysis 168 electron paramagnetic resonance (EPR) 183, 380, 389, 393, 395, 439, 455, 955 analysis 87 high-frequency 955, 969 spectra 444, 445 pulsed 955 time-resolved 955 electron spin densities 342 electron spin echo (ESE) 957 electron spin echo envelope modulation (ESEEM) 393, 439 electron transfer 39, 266, 287, 327–330, 337, 339–349, 355, 366, 368, 370, 382, 396, 398, 413, 510, 674, 822, 868, 921, 954–973 –20 °C 521 A-side 346 B-branch 348, 958 B-side 346–348
987 electron spin echo (ESE) 957 exponential dependence on distance 345 first order rate 327, 332 low-temperature QA– QB → QAQB– 957–959 QA– QB → QAQB– 956 reactions 327–329, 363–370 second order rate 332–333 sequential 963 surface-hopping model 363–367 uphill steps 514 electron transfer pathways Rhodobacter sphaeroides 382 electron transport 426, 735 electron transport chain 674 electrostatic confinement 527 electrostatic coupling 904 electrostatic environment 208 electrostatic forces 340 electrostatic interaction 327, 341, 514, 528, 657 elemental sulfur 596–597, 601, 602, 605–607, 616 S7 rings 607 S8 rings 607 ellipsis 944 elliptical deformation B850 ring 207 elliptical LH116 assembly 945 Em 438, 439, 445, 554 encounter complex 527 endoplasmic reticulum (ER) 269 endoprotease GluC 189 endosymbiosis 433 energetic disorder 888 correlated 888 random 888 energy-conserving H2 evolving membrane-associated hydrogenase 746 energy barrier 944 energy conduits 147 energy dissipation 898 energy transduction 41 energy transfer 149, 216–226, 233, 239, 258, 266, 862, 881, 915, 943 1 – Bu state 224 carotenoid-bacteriochlorophyll 927 coherent 882 domains 943 electronic coupling 881 hopping 235, 243, 246 incoherent 881 light-harvesting 1 complex 226–227 light-harvesting 2 complex 216–226 S* state 222–224 S1 state 219–222 S2 state 216–219 single photosynthetic unit 889–891 energy transition tuning 205 energy trapping 243–244, 260 robustness graceful degradation 285 parameter insensitivity 285
Index
988 robustness and optimality 285 optimization 173 enoyl-ACP reductase I 122 ensemble averages 283 ensemble reactions 950 Enterobacter 669 Enterococcus hirae 661 enumeration 47 environmental limits to photosynthesis 12 enzymatic detoxification 662–665 equilibrium constant 520 ER. See endoplasmic reticulum (ER) Erb-type. See Erythrobacter-type carotenoid Erwinia herbicola 110, 854 carotenoid biosynthetic pathway 854 Erwinia uredovora 103 Erythrobacter 34, 40, 113 Erythrobacter-type carotenoid 99, 113, 114 Erythrobacter litoralis 676 Erythrobacter longus 102, 104, 113 Erythrobacter sp NAP1 19, 790, 792 Erythromicrobium 40, 113 Erythromicrobium hydrolyticum 599 Erythromicrobium ramosum 34 Erythromonas 34, 113 erythroxanthin sulfate 41 Escherichia 702 Escherichia coli 21, 40, 66, 71, 74, 104, 108, 268, 411, 476, 479, 588, 646–647, 695, 701, 744–757, 778, 781, 785, 786, 795, 820, 844, 847, 850, 852, 856 Fnr protein 753 genome 848 hemC mutants 63 hydroxymethylbilane synthase 63 membrane proteome 848, 851 paradigm 646–647 YcgF 820 Escherichia coli K-12 795 esterifying alcohol 187, 899 ethylbenzene dehydrogenase 547, 550 8-ethyl group 58 8-ethylidene 58 Euglena 817 Euglena gracilis 817, 820 eutectic model 949 eutectic phase behavior 945 Eutreptia 817 evolution aerobic anoxygenic phototrophs 38–40 cytochrome b6 f complex 456–458 cytochrome bc1 456–458 physical constraints of 284–286 purple bacteria 17–28 Rieske/Cytochrome b complex 456–458 transhydrogenase 497 exchange chromatography 851 excitation delocalization 935 excitation energy 136, 166, 206 excitation energy transfer 170, 200, 205, 950 excitation relaxation 929
excitation sharing 173, 282, 289 dimer array 173 PufX 172–174 excitation spectra 921 excitation trap 928, 935 excited state 206 excited triplet state 357 exciton 232, 235–239 exciton-phonon coupling 233, 237 exciton delocalization 929 exciton dynamics wavelike motion 246 excitonically-coupled systems 206 BChl systems 204 excitonic coupling 886 excitonic interactions 205, 207 light-harvesting complexes 205–207 exciton migration 147 lake model 147 puddle model 147 exciton model 205 exciton relaxation 246 exciton spectra 235–239 exciton state 233, 235, 238, 245 exciton transfer 45 expression plasmids 842, 844, 849 external sulfur uptake 607–608 extra-cytoplasmatic function (ECF) 737 extracellular sequestration 665 extracellular sulfur globules 600, 602, 608 extreme environments 9–12, 51 extremophilic purple bacteria 9–12 extremophily 51, 52 extremotolerance 51 extrinsic domain 430
F φ mutant 338 φB mutant 346 F1 475 F1Fo ATP synthase 538 fabA 122 fabB 122 fabD 122 fabF 122 fabG 122 fabH 122 fabI 122 fabZ 122 FAD. See flavin adenine dinucleotide (FAD) farnesyl hydroxyethyl side chain 408 farnesyl pyrophosphate 103 fast exchange limit 328 fatty acid diesters 110 fatty acid ester carotenoid glucoside 98 fatty acids 122–123, 586 FccA 606 FccAB 602, 606
Index fccAB 603 FccB 606 Fe-bacteriochlorophyll. See Fe-bacteriopheophorbide Fe-bacteriopheophorbide 902 Fe-removal/Zn-replacement 962, 965, 967 Fe-S center 87, 539, 721, 794 See also iron-sulfur cluster [2Fe2S] cluster 438, 439, 452, 753 Em7 value 439 [4Fe-4S]2+ cluster 753 [FeFe]-hydrogenase 744, 746 [NiFe]-hydrogenase 743, 745–746, 754 H2-sensing 750 femtosecond resonance Raman 227 Fe removal 305 Fe replacement 304, 305 fermentation 149 fermentative bacteria 591 ferredoxin 74, 581, 584 ferrochelatase 58, 86, 787 FhlA 752 filamentous anoxygenic phototrophs 596 finite temperature quantum theory 282 thermal disorder 282 first order electron transfer rate 327, 332 fix 781 fixed nitrogen sensing 763–766 FixK1 753 FixK2 753 FixL 753 flagella 644 flagellar-specific sigma factor (σ28) 645 flagellar helix 644 flagellar proteins 644 flagellin 644 flash photolysis 805 flavin 87, 611 flavin adenine dinucleotide (FAD) 813, 816 flavin cofactor 539 flavin mononucleotide (FMN) 813, 816 flavocytochrome c (FccAB) 602, 606, 616 flavocytochrome c sulfide dehydrogenase (FccAB) 606, 645 flavoprotein 90 flexible hinge region 157 fluorescence 233, 235, 237–240, 243–246, 357 fluorescence depolarization studies 205 fluorescence emission 933 fluorescence emission properties 928 fluorescence excitation spectra 207, 226, 882 fluorescence fluctuations 244 fluorescence properties 206 fluorescence resonance energy transfer (FRET) 478 fluorescence up-conversion 217, 225, 239, 240 B800-B850 225 FMN. See flavin mononucleotide (FMN) Fnr 707, 721, 722, 752, 753, 781, 782 regulation by 721–722 Escherichia coli 753 FNR consensus-like sequences 781, 786, 787, 794, 795 FNR consensus sequence 781, 783
989 FnrL 721, 781–786, 790, 795 oxygen inactivation model 721 fnrL 781–786, 789, 794, 795 mutant 721 null mutant 722 Rubrivivax gelatinosus 722 FnrN 753 FnrT 753 FOF1-ATP synthase 452, 475–488 gearshift 479 rotary catalysis 476–478 slip 486 torque 475 torsional spring constant 485 structure 476–478 folding membrane protein 914 foreign gene expression 839–856 formate dehydrogenase 548 Förster-type mechanism 218 Förster formula 233 generalized 233 Förster radius 276 Förster theory 242, 243, 280 fosmid library 51 four-helix bundle 903, 904, 934 design 903 lipophilic 906 modeling 903 reorientation 905 four-helix bundle proteins 895, 896, 899, 900, 901, 905 Fourier transform infra-red spectroscopy (FTIR) 380, 389–391, 397, 731, 957 Franck-Condon factor 367 free-energy perturbation 358 free energy 358, 359, 363 free energy calculations 255 free radical SAM 72 free radical SAM enzymes 65 freeze-fracture electron micrograph 262 freeze-fracture electron microscopy 168, 257, 260 freeze-thaw sonication 261 freeze thawing 946 Fremyella displosiphon 800 French press 846 Frenkel-excitons 880 freshwater lakes 51 fructose-1,6-bisphosphate 571 fructose-6-phosphate 571 fructose 1,6/sedoheptulose 1,7-bisphosphatase 566 fructose 1,6/sedoheptulose 1,7-bisphosphate aldolase 566 Fth/FtsY 269 Fulvimarina 19 Fulvimarina pelagi 24 fumarate-nitrate reduction (FNR) 708 fungi 847 funiculosin 455
Index
990
G γ-carotene 113 γ-carotene synthase 109 G+C content 850 GAF domain 728 galactosyltransferase 127 gammaproteobacteria 5, 24 Gammaproteobacteria 22, 98, 100, 109, 112 gas vesicle protein 702 gel electrophoresis blue native polyacrylamide 526 gel filtration 849 Geminicoccus 34 gene duplication 696 gene expression 744–755, 799, 839–856 auto-induction 848 hydrogenase regulation 743–755 regulation 707, 707–722, 744–755 gene homologs 696–701 general membrane assembly factors 268, 269 DnaK 269 GroEL 269 SecA 269 gene regulation light-dependent 727–737 gene split 457 gene transfer 45 horizontal 45 lateral 38 genome 691–704 architecture 692–696 characteristics 692–696 sequences 45, 101, 579 sizes 692 genomics 17, 18, 27, 28, 69–704 structural 847 geranylgeraniol 58, 74, 164, 306 geranylgeranyl 187 geranylgeranyl pyrophosphate 74, 103 geranylgeranyl pyrophosphate synthase 103 geranylgeranyl reductase 67 German Collection of Microorganisms and Cell Cultures 150 GFP. See green fluorescent protein (GFP) GGDEF/EAL domain 728 glass substrate 864 GlnB 764 global fitting data analysis 217, 223 global oxygenation 38 global regulatory system 712, 716 glucosyltransferase 127 glutamate-1-semialdehyde aminotransferase 778, 793 glutamine 764 glutamyl-tRNA reductase 59, 778, 793 glutamyl-tRNA synthetase 59 glutathione 41, 608, 613 glutathione amide 609 glutathione selenopersulfide 669 glyceraldehyde-3-phosphate dehydrogenase (cbbGII) 566 glycerol 3-phosphate 125 glycerol 3-phosphate 1-O- acyltransferase 121 glycogenin 128
glycoglycerolipid 126–128 glycolipid 119 glycosyltransferase 127 GMP-adenylyl-cyclase-FhlA (GAF) 728 gold-coated silicon probe 2-mercaptopyridine modified 869 gold substrate 861, 865–869 graceful degradation 285 Gram-positive bacteria 409 grazing pressure 50 green filamentous bacteria 109 green fluorescent protein (GFP) 845, 846, 850 green nonsulfur bacteria 746 green plant 577 green sulfur bacteria 101–104, 109, 110, 596 Chlorobium tepidum 410 GroEL 269 ground state conformation 383 ground state interaction 203–205
H H+ output 430 1 H ENDOR 959 2 H ENDOR 959, 968 H2 744–757 H2-regulation 754 H2-sensing [NiFe]-hydrogenase 750 H2-sensor 750 H2-specific signal transduction system 747 H2 sensor 743, 745, 751 H2 signal 750 H2 transduction 751 habitats 7-9, 149 Halobacterium salinarum 851 Halochromatium 10 halophilic purple bacteria 10–11 Halorhodospira 10, 110, 600, 601 Halorhodospira abdelmalekii 112, 600, 601 Halorhodospira halochloris 112, 600, 601 Halorhodospira halophila 600, 601, 604, 606, 609, 611, 615, 823, 827 Halorhodospira halophila SL1 790, 791, 792 Halorhodospira halophilum 24 Halorhodospira halophilum SL1 24 Halorhodospira neutriphila 600, 601 Halorhodospira sp. SL1 19 halotolerance 51 Hamiltonian 206 Hansenula polymorpha 852 HBA. See hydrogenobyrinic acid (HBA) HbrL 789 hbrL 789 HCCS/CCHL 410 heat-shock response 701 Helicobacter 414 heliobacteria 109 Heliobacteriaceae 380 Heliobacterium chlorum 149 helix binding 902, 903 shielding 902, 903 310 helix 137
Index helix-helix interactions 137 helix-hinge-helix solution 157 helix-turn-helix DNA binding domain 717 helix-turn-helix motif 721 helix W 166 HemA 780, 782, 789 hemA 61, 780–783, 788, 789 HemB 783, 791, 794 hemB 783, 784, 788, 789, 791 hemC 63, 784, 785, 787–789, 794 hemC mutants Escherichia coli 63 hemD 64 heme 407–420, 778, 780, 787, 789 attachment 269 stereo-selectivity 413 biosynthesis 57, 58, 65, 415, 780 delivery pathway 412 high-spin heme a3-CuB binuclear center 416 high-spin heme b3-CuB binuclear center 416 ligation 409 low-spin heme a 416 lyase 411 periplasmic ligation 410 reservoir 415 transporter 412 vinyl-2 group 411 hemE 784, 785, 788, 789, 794 heme-binding protein 899 heme-copper oxidase 416, 540 heme a 408, 409 heme a3-CuB center 544 heme bH 431, 433 heme bL 431 heme ci 457, 468 heme d 408, 409 heme macrocycle 438 heme o 408 heme reductase 410 heme b 436 HemF 64, 65, 785, 786, 794 hemF 64 HemG 795 hemG 65, 794 hemH 787, 788, 794 HemN 64, 65, 785, 786, 794 hemN 51, 64, 629 HemN-type coproporphyrinogen III oxidase 794 HemN-type motif 794 hemomolybdoprotein 602 HemT 780, 782 hemT 61, 780–782, 789 hemY 65 HemZ 786 hemZ 64, 786, 788, 789, 794 heptane-1 146 heterodisulfide reductase 609 heterologous expression 847, 848, 850–852 Heterosigma akashiwo 669 hexachlorobenzene 64 hexameric rings 70
991 HF Mims-type ENDOR 961 high-fequency 1H-ENDOR 961, 962 high-frequency EPR 955 high-light 255, 945, 948 high-light light-harvesting 2 complex 255 high-potential 2Fe-2S Rieske protein 942 high-potential iron sulfur protein (HiPIP) 514, 515, 543 high-resolution AFM 949 high aeration conditions 175 higher-throughput methodologies 851 higher plants 103, 104 high oxygen 780, 785, 786, 788 high performance liquid chromatography (HPLC) 49, 193, 927 hinge protein 438 HiPIP. See high-potential iron sulfur proteins (HiPIP) histidine 190, 344, 933 histidine kinase 652, 751, 806 histidine protein kinase 708, 712 6-histidine tag 170 HMB. See hydroxymethylbilane HMBS. See hydroxymethylbilane synthase HMB synthase 60 Hoeflea phototrophica 34 hole-burning 236, 361 holocomplex 147 holocyt c production 408 holocytochrome 407 homocysteine 613 homogeneous line shape 208 homologous recombination 850 RegA 709–711 RegB 709–711 Homo sapiens 479 horizontal gene transfer 45, 694 HoxA 748, 751 hoxE 746 hoxF 746 hoxH 746 HoxJ 748 HoxJ/HoxA 750 hoxU 746 hoxY 746 HQNO. See 2-n-heptyl-4-hydroxyquinoline-N-oxide (HQNO) Hückel molecular orbital model 342 hup locus 746 HupR 743, 748, 749, 751, 752 HupR/HoxA 747 hupR gene 749 hupR mutants 750 hupSLC genes 753 hupS promoter 754 HupT 743, 748, 750, 751, 752 HupT/HoxJ 747 HupT/HupR 750 hupT gene 749 hupT mutants 750 hupTUV operon 749, 750 HupU 750 HupUV 743, 751, 752 HupUV/HoxBC 747
992 HupUV hydrogenase 750 HvrA 736 hvrA 770 hybrid reconstitution 191 hybrid sensor kinase 572 hydride transfer 495 transhydrogenase 503 hydrogenase 743–755 gene expression regulation 743–755 hydrogen bond 137, 140–143, 161, 186, 187, 203, 204, 330, 340–345, 347, 349, 384, 385, 387, 390–394, 396, 915, 916, 918, 922–924 network 175, 208 polypeptide 141 quinone binding sites 389 hydrogen evolution 744–757 hydrogenobyrinic acid (HBA) 84, 89 hydrogenobyrinic acid a,c-diamide 89 Hydrogenophilus thermoluteolus 566 hydrogenosome 746 hydrogen transfer 518 bucket brigade 518 hydropathy analysis 148 hydrophobic effect 900 hydrophobic groups 190 hydrophobic interaction 327, 331, 915 hydrophobic mismatch 267, 946 hydrophobic mutations 331 hydrophobic residues 329, 331 hydrophobic surface 183 hydrostatic pressure 208 hydrothermal vent 32, 39, 50 hydrothermal vent plumes 47 4-hydroxybenzoate 578, 579, 586–587 hydroxy chlorophyllide 73 4-hydroxy cinnamic acid 823 3-hydroxyethyl bacteriochlorophyllide a 68 3-hydroxyethyl bacteriochlorophyllide a dehydrogenase 67 3-hydroxyethyl chlorophyllide a 68 hydroxymethylbilane 60 hydroxymethylbilane synthase 63 Escherichia coli 63 hydroxyneurosporene-O-methyltransferase 107 hydroxyneurosporene synthase 107 hydroxy protochlorophyllide 73 hynS 753 hyp 753 hyperfine coupling 396 hyperfine interactions 391, 393 hypersaline springs 42, 51 hypersaline spring system 39 hyp genes 745
I iB873 175, 932–934 CD spectrum 933 iB873 complex 934 ICM. See intracytoplasmic membrane (ICM)
Index Idiomarina loihiensis 824 IHF. See integration host factor (IHF) Ilyobacter tartaricus 478, 479 IMAC. See immobilized metal affinity chromatography (IMAC) image processing 948 imidazole 815 immobilized metal affinity chromatography (IMAC) 842, 849 immunoblotting 846 inclusion bodies 69, 841, 842, 847, 852, 853 incoherent energy transfer 881 incoherent hopping 206 indium tin oxide 862 electrode 863 individual complexes 943 inducible ICMs 852 induction 267 infrared epifluorescence microscopy (IREM) 47, 49 infrared fast repetition rate fluorometry (IRFRR) 47,49 infrared spectra 394, 396 inhibitors 70 inorganic sulfur compounds 600 integration host factor (IHF) 749, 752–754, 770 interaction motif prediction 916 interactions cofactor 916 intercellular connective structures 34 intercomplex disorder 883 intercomplex heterogeneity 883 intermediate iB873 932 intermediates of LH1 assembly 932, 933 intermolecular disulfide bond 714 internal transcribed spacer analysis 35 interspecific complementation 854 intracellular sequestration 665 intracellular sulfur globules 602, 607, 608 intracomplex disorder 883 intracomplex heterogeneity 883 intracytoplasmic membrane (ICM) 136, 137, 147, 172, 200, 254, 257, 258, 264, 525, 841–860 biogenesis 846 domains 261 formation 267 inducible 852 induction of assembly 267 spherical vesicles 845 synthesis 844 tubular membranes 844 ultrastructure 844 inverse-Mollweide projection 287 invertebrates 410 in vitro transcription-translation system 175 ion exchange chromatography 61, 849, 869, 926, 929 ionic strength 516, 527, 950 IR difference spectroscopy 391 iron 193, 415 limitation 415 iron-reducers 591 iron-sulfur cluster 744–757; See also Fe-S center
Index iron-sulfur flavoprotein 609 iron-sulfur protein 426, 427, 431, 432, 438–441 extrinsic domain 432, 435 macro-movement 441 topology 432 iron-sulfur protein subunit 512 Isochromatium buderi 601 isocitrate dehydrogenase (ICDH) 500 isocyclic ring 898, 915 isocytochrome c2 542–543 isoelectric points 848 isomerase 82 isomerization 804, 825, 827, 829 isoprene 384, 387, 388 isoprenoid alcohol 897 isoprenoid chain 518 iterative design 899–901 iterative redesign 904 IUPAC-IUB nomenclature 98 IUPAC numbering scheme 58
J Jannaschia 19, 440 Jannaschia sp. CCS1 790, 792, 819 Juan de Fuca Ridge 50 Jα helix 815
K keto carbonyl 204 Ketogulonicigenium 38 ketolation 108 keystone consumer 50 KH2PO4 571 kinase activity 713 kinase domain 806 Kineococcus radiotolerans 804 kinetic analysis 60, 74 kinetic efficiency 476 kinetic fluorometry 49 kinetic mechanism 71 kinetic phase 515 kinetics 355 Klebsiella pneumoniae 748 Kleinfeld effect 956 Km values 69, 462 K pathway 544, 546
L laboratory culture ease of growth 4 Labrenzia 34 lacZ 780, 783, 785–787, 789 reporter activity 568 Lake Fryxell 4 Lamprobacter modestohalophilus 667 Lamprocystis roseopersicina 108 Langmuir-Blodgett film 863, 864 lateral gene transfer 38
993 lauryldimethylamine oxide (LDAO) 140, 146, 300, 926, 929 LD. See linear dichroism (LD) LDAO. See lauryldimethylamine oxide (LDAO) Leigh syndrome 418 leucine zipper-like motif 414 level of hydration 825 LH1. See light-harvesting 1 complex LH1-only strain 161 LH1α polypeptide 157 N-terminal domain 175 solution structure 157–158 LH1β polypeptide NMR studies 157 solution structure 157–158 LH2. See light-harvesting 2 complex LH2– mutant M21 263 LH3. See light-harvesting 3 complex LH4. See light-harvesting 4 complex LhaA assembly factor 174 lhaA 148, 174 liganding histidine 917 ligase 585, 586, 588, 589 ligation-independent cloning 842, 843, 851 light, oxygen or voltage (LOV) 813. See also LOV domain photocycle 814 light-dark difference spectrum 193 light-dependent reversal of the electron flow 554 light-driven proton gradient 175 light-harvesting absorption 922 assembly 922 BChl-B800 921 BChl-B850 921, 923 mutations 922 scanning permutagenesis 922 light-harvesting 1-PufX complex 524 S-shaped chain 524 light-harvesting complex 45–46, 511, 699, 703, 861–873, 877–891, 913, 914–936. See also specific complex artificial 862–873 assembly on solid substrates 861–873 assembly 913, 916, 920, 921, 924–935 model systems 924–926 carotenoid-binding 919, 920, 924-926 charge-transfer character 208 cofactor-binding motif 935 design 913–924 disorder 205–207 dynamic disorder 206 electronic transitions 208 electrostatic environment 208 energy transfer 921 excitation energy 206 incoherent hopping 206 excitation spectra 921 excited state 206 excitonic interactions 205–207 homogeneous line shape 208 hydrostatic pressure 208 insertion of polypeptides 268
994 membrane 268 model sequences 921 pigment tuning 935 Qy transitions 208 solvation mechanisms 207 spectral properties 199, 200–209 Stark spectroscopy 208 static disorder 206 structural variability 202 light-harvesting 1 complex 10, 45–46, 136, 156, 160–162, 182–184, 186–188, 206–208, 214, 223, 226–227, 232, 235–239, 243–244, 254, 257, 269, 511, 523, 841, 846, 871, 873, 915, 916, 942–944 β polypeptide 188 16-fold LH1 ring 163 2-D crystals of carotenoidless complex 160 2-D crystals of Rhodospirillum rubrum complex closed ring 160 cryo-EM projection map 160 structure of complex 160 absorption spectra 925 assembly 924–935 assembly intermediates 933 B780 930 B820 930 B870 930 association constants 186 atomic force microscopy variety of structures 161 B820 subunits 926 bacteriochlorophyll exchange 928 bacteriochlorophyll coordination state 930 blue-shift 45 building blocks 175 C-terminal region 187 C-terminal side 871 CD spectra 188, 925, 927 chemically-synthesized β polypeptide 186 circular dichroism 206 cofactor interactions 188 core antenna 915 effect of carotenoids on assembly 930 ellipiticity 207 elliptical conformations 163 elliptical ring 166 energy transfer 226–227 energy trapping 243–244 EPR 188 excitonic coupling 188 first purification 160 flexibility 160 fluorescence polarization 188 formation 930 fractionation into oligomers 160 fundamental B820 subunit 183 fundamental subunit 183 histindine ligand mutants 185 hydrogen-bonding 161 effect on B820 161 effect on flexibility 161
Index effect on oligomers 161 LH2 comparison 161 role in quinone exchange 161 hydrogen bond formation 187 hydrogen bond mutants 186 in vitro assembly intermediates 192 RC-LH1 complex 192 LH1-only mutant 161 modified carotenoids 929 native membrane environment 916 Ni-bacteriochlorophyll a 928 NMR studies 156, 188 Phaeospirillum molischianum 191 polypeptides 156 quinone exchange 157–159, 161, 165–167 quinone passage 523 reassociation with reaction center 192 reconstitution 181–198, 916, 925 carotenoids 929-934 red-shift 45 reversible dissociation 182 bacteriochlorophyll analogs 184 bacteriochlorophyll ligand 184 cofactor requirements 184 dissociation constant 184 metal requirements 184 ring binding 163 gaps 166 role of carotenoids 929 Rhodospirillum rubrum 2-D crystals 160 self-organized on mica 873 sequence alignment 185 singlet-singlet annihilation spectroscopy 188 steric requirements for BChl binding 187 subunits 926 supramolecular structure 873 Thermochromatium tepidum 10 unit size 928 wavelength shift 188 light-harvesting 2 complex 45–46, 110, 136, 143, 145–147, 149, 165, 185, 188, 189, 191, 202, 207, 208, 214–227, 232, 235–239, 244–247, 254, 255, 257, 258, 267–269, 279, 511, 513, 783, 801, 803, 841, 882–889, 915, 918, 920, 921, 942, 944, 946, 948 α polypeptide 188 β polypeptide 188, 202 absorption 921, 923 assembly 923 assembly factor 174 B800 band 279, 882–885 B800 molecule acetyl carbonyl 203 B850 band 279, 885–889 B850 BChl coupling strength 205 excitonic interactions 205 pure excition model 205
Index BChl-B850 915 BChl molecule distorted conformation 205 binding sites 920 circular dichroism 923 CD signal 921 conformational substates 207 conservative CD signal 206 crystal structure 202 densely packed hexagonal arrays 267 domains 260 energy transfer 216–226 fluorescence excitation 923 spectra 882 fluorescence excitation spectra 207 low-light Rhodopseudomonas palustris 203, 948 model 920–921, 922 mutants lacking 168, 171 Phaeospirillum molischianum 166, 189 polypeptides 783 Qy electronic transitions 205 resonance Raman spectroscopy 922 Rhodobacter capsulatus 205 Rhodobacter sphaeroides 203, 205 B850 BChl 204 site-directed mutagenesis 204 α polypeptide 204 β polypeptide 205 Rhodopseudomonas acidophila 203 Rhodopseudomonas molischianum 165 sequence alignment 185 size heterogeneity 255, 948 spectral heterogeneity 255 spectral properties 922 stability 922 structure of Rhodopseudomonas acidophila complex 185 light-harvesting 3 complex 150, 214 light-harvesting 4 complex 150, 801, 803 light-harvesting assembly effective Hamiltonian 279 light-induced cyclic electron flow 43, 511 light-induced oxygen uptake 543 light-induced respiration 543 light-input domain 812 light-protection 898 light-responding antirepressor AppA 718 light control 783 light intensity 98, 945, 948 light oxygen voltage (LOV) 731 light regulation 717–718 lignin 578, 581 lignin monomer 577 linear dichroism (LD) 168, 169, 172, 258, 261, 286 linear electron-phonon coupling 884 linear electron pathway 510 linear plasmids 45 linear tetrapyrrole 800, 827 lipid 935 function 130–131 membrane 914
995 lipid-deficient genetic mutants 120 lipid bilayer 854, 855 lipid biosynthesis 119, 120–132 lipid glycosyltransferase 126 lipidic cubic phases 140 crystallization 140 lipid profiles 120 lipids 193 lipid tubules 262 lipophilic four-helix bundle 906 liposomes 257 liquid state conception 519 local dielectric constant 204 Loktanella 19 Loktanella vestfoldensis SKA53 790, 792 Lon protease 850 lotic 51 LOV. See light, oxygen or voltage (LOV) LOV domain 731, 811, 813–816, 822–823 low-light conditions 143–153, 149, 150, 945, 948 low-light growth 148 low-light intensities 150 low-light light-harvesting 2 complex 255, 261, 945, 948, 949 low aeration conditions 58 lower energy transition tuning position 205 lower eukaryotes 410 low light 255 low oxygen 785–788 low potential chain 453 lycopene 98, 103–104, 108–110, 112, 113, 141, 218, 219, 225, 929, 931, 934 all trans 103 lycopene cyclase 113 lyso-ornithine beta-hydroxy acyltransferase 121 lyso-ornithine lipid 130 LysR 789 LysR-type transcriptional regulators (LTTRs) 565
M M2 influenza proton channel 906 macro-movement 431, 441, 444 macrocycle 58, 438, 899 distortion 899 magnesium chelatase 58, 66, 67, 69, 70, 90, 717, 787 magnesium ion 139 magnesium protoporphyrin IX 67, 68, 71 magnesium protoporphyrin IX monomethyl ester 68, 794 magnesium protoporphyrin IX monomethyl ester oxidative cyclase 42, 67, 71 origin of oxo-group 71 magnetic fields 357 magnetic resonance 389, 393–396 magnetic resonance spectroscopy 954 Major Facilitator Superfamily 148 MALDI-TOF 167 malonyl-CoA:ACP-trasnacylase 122 manganese 349 oxidation 312 manganous protoporphyrin IX 71
996 Marcus equation 364 Marcus theory 345, 355, 363–367, 382 Marichromatium 10 Marichromatium purpuratum 109, 515 marine environments 47 marker rescue techniques 66 MarR family 589 massive transfer 694 mass spectrometry 417 master equation 280 Mastigocladus laminosus 468 matrix ENDOR 959, 962 maturation cytochrome c 408 maturation process 407, 408 McMurdo Dry Valleys, Antarctica 4, 11 MCP chemoreceptors 647 mechanoenzymes 70 membrane 495 anchor/linker domain 843 cytoplasmic side 137, 164 eutectic phase behavior 260 incorporation of foreign protein 842 intracytoplasmic membrane (ICM) 136, 147 LH2 protomer 138 lipid-rich areas 266 N-terminal anchor 412 native 942, 949 n side 426 organization 949 periplasmic side 164 p side 426, 436 reconstituted 943 remodeling 120 ring size 137 topology 146, 158 translocation 410 vesicles 946 membrane-bound cytochrome cy 542–543 membrane-bound nitrate reductase 552 membrane assembly factors 269 membrane associated complexes 120 membrane bilayer 172 dilution 257 membrane curvature 170, 172 membrane development 262 membrane domains PufX 172–174 membrane fragments 261 membrane growth early stages 172 membrane lipid biosynthesis 119, 120–132 membrane lipids 899, 914 membrane organization 170 membrane phospholipids 267 membrane potential 266, 517 membrane protein 840, 841, 845, 846, 847–849, 854, 855, 924 folding 914 monotopic 411 membrane proteome Escherichia coli 851
Index membrane sheets 855 membrane tubules 258 menahydroquinone oxidase 418 menaquinol oxidation 548 menaquinone 394, 511, 530 mercaptomalate 601 mercaptopyridine 862 meromictic lakes 40, 51 mer operon 663 MerP protein 663 Mer system 675 Mesorhizobium loti MAFF303099 794 mesoscopic conditions 826 metabolic channeling 530 metabolic engineering 854 metabolic versatility 696–701 metal 656–682 bacterial interactions 666–675 resistance 658 genes 659–661 toxicity 657–658 metal-binding sites 63 metal clusters 444 metalloid 656–682 bacterial interactions 666–675 metallothionein 665, 668, 678 in SmtA 665 metallotolerance 52 methanogenesis 7 methionine synthase 83 methoxy-hydroxylycopene glucoside 112 methoxylycopenal 108 methoxyneurosporene 107 methoxyneurosporene dehydrogenase 107 methyl accepting chemotaxis proteins (MCPs) 647, 649, 651 methylamine dehydrogenase 853 methylation 674 methylcobalamin 82 Methylibium petroleiphilum PM1 26 methylmalonyl CoA mutase 83 Methylobacterium 19, 20, 113 Methylobacterium extorquens 32, 102, 113 Methylobacterium extorquens AM1 790, 792 Methylobacterium radiotolerans 113 Methylobacterium rhodinum 110, 113 methyltransferase 82, 84, 87, 89 mevalonate pathway 103 Mg. See also magnesium binding motif 791 coordination 139, 140, 144, 184–188, 898 Mg-chelatase 90, 717, 787 Mg-protoporphyrin IX monomethylester 794 Mg-protoporphyrin monomethyl ester cyclase 42 Mg 3,8-divinylphaeoporphyrin a5 68 mica substrate 861, 864–865 mica supports 259 mica surface 864, 947 micelle 146, 157 mixed micelles 183 microaerophilic photosynthesis 46 microarray 781, 783, 784, 786, 787
Index microarray expression 696 microbial mats 8 microfluidizer 846 microscopic examination 8 mid-point potential 520 QA 39, 46 Mims-type pulsed 1H ENDOR 959 mine drainage systems 41 minimal medium 844 mitochondria 409, 410, 416–418, 420, 476 human 409 yeast 409 mitochondrial Complex IV 408 mitochondrial inner membrane 264 mitochondrial respiratory chain 264 mitochondrial supercomplex 526, 527 Mn 419 Mn/Fe superoxide dismutase 674 MOA-stilbene 455, 462 MOA-type inhibitor 435 mobile cytochromes 515–517 mobile loop 504 models monomeric core complexes 165 model sequence 923 modified Q cycle mechanism 426 modified Redfield theory 237, 242, 244, 245 modularity 284, 285 modular organized Chl protein 901–904 modular organized proteins 896 molecular-dynamics (MD) 358 simulations 265, 358, 364, 370, 397 with quantum chemistry 282 molecular assembly 869–873 molecular chaperone 69 molecular devices 862 molecular orbital 342 molecular voltmeter 483 molybdenum 610, 761 transport 768 molybdoenzymes 611 molydopterin cofactor 611 monoglycosyldiacylglycerol glycosyltransferase 121 monomer bacteriochlorophylls 338, 341 monomeric/dimeric complex comparison 169 monomeric core complexes models 165 monomeric RC-LH1-helix W core complex 171 monooxygenase 84 monotopic membrane protein 411 Monte-Carlo simulations 172, 265 MopA 768 mopA-modABCD 765 MopB 768 motif 4F-4S cluster binding 417 apoCyt c heme binding 413 ATB binding 411 CFCF 417 conserved copper-binding 417, 420 conserved metal binding 417
997 CXXCH heme binding 409 HemN-type 794 leucine zipper-like 414 limitation 417 Mg-binding 791 tetratricopeptide repeat (TPR)-like 414 tryptophan-rich WWD 411, 414 Zn-binding 791 mucidin 455 multi-domain proteins 831 multi-subunit complex 852, 853 multidrug efflux pumps 702 multiphasic kinetics 361–362 multiple chromosome-like replicons 694 multiple domain motifs 701 mutagenesis 338, 915, 916 mutant carotenoid 107 DLL mutant 347, 339 fnrL 721 G1c 218, 222, 930 holocyt c production 408 LH2– mutant M21 168, 263 mutant lacking B800 BChl 206 protochlorophyllide-reduction 854 PrrC 418 PUC705BA 866 R26 165, 303, 383, 386 R-26 302, 304, 339 R26.1 218–220, 223, 225, 387, 667, 668, 924 R-26.1 299, 304 suppression 413 reaction center-light-harvesting 1 complexes 164 strain sym-1 308 strain QAQA 309 strain DLL 307 strain A6D1 310 mutants carotenoidless mutants 193 cytochrome bc1 Rhodobacter 428–430 deletion mutants 839-860 strain DD13 307 strain DD13/G1 307 lacking LH2 complex 168, 171 lipid-deficient genetic 120 PUF∆LMX21 844, 846 Rhodobacter cytochrome bc1 428–430 Rhodobacter sphaeroides reaction center-light-harvesting 1 complexes 164 mutations 359, 360, 363 myxothiazol 435, 442, 455, 521
N N-dimethyundecylamine-N-oxide 146 2-n-heptyl-4-hydroxyquinoline-N-oxide (HQNO) 445 n-octyl β-D-glucopyranoside (β-OG) 182, 925 N-terminal 6xHis-Tag 69 N-terminal amino acid 581
Index
998 N-terminal domain of LH1α 175 N-terminal membrane anchor 412 N-terminal signal sequence 843 N-terminal Strep-tag 74 N´-Dimethyldodecylamine N-oxide (LDAO) 869 N2O 780 NAD 746 NADH 495, 501, 571 NADH: ubiquinone oxidoreductase 745 NADH dehydrogenase (NDH-1) 538–539 NADH dehydrogenase (NDH-2) 539 Nadi reagents 408 NADP 495, 501, 746 NADPH 88, 500, 571 NADPH:protochlorophyllide oxidoreductase 73, 854 nanodevices 861 nanodissection 259 nanoscale apparatus 862 naphthoquinone 382 native membrane 941, 942, 949 native membrane environment 923 NDH-1. See NADH dehydrogenase (NDH-1) NDH-2. See NADH dehydrogenase (NDH-2) negatively stained thin sections 171 negative stain analysis 168 electron microscopy 168 network hydrogen bonds 175 Neurospora crassa 813, 828 neurosporene 98, 103–105, 107–110, 112, 216, 218, 219, 221, 222, 225, 227, 854, 931, 934 neutral semiquinone 820 neutron diffraction 145 Ni-bacteriochlorophyll. See Ni-bacteriopheophorbide Ni-bacteriochlorophyll a 303, 928 Ni-bacteriopheophorbide 902, 907 Nicotiana tabacum 64 NifA 744, 753, 766–767 post-translational control 766–767 nifA2 765 NifA regulon 766–767 NifD 74 NiFe active site 750 nif genes 760 NifH 74 nifHDK 761 NifK 74 NifL 766 nirS 635 NirT family 548 nitrate 670, 780, 781 nitrate assimilation 624, 636–638 nitrate reductase 548, 611, 625, 637, 670 nitrate reduction 623–639 nitrate respiration 630 nitric oxide reductase 625–627 nitric oxide toxicity 627 nitrite 632 nitrite oxide reductase 626 nitrite reductase 625, 635, 637, 670
cd1 type 635 nitrite reduction 632 nitrogenase 74, 582, 745, 760 alternative 769 post-translational control 767–768 nitrogen availability 789 nitrogen cycle 624 nitrogen fixation 7, 624, 637, 707, 760–772 genes 762 regulation 760–772 regulatory cascade control 763 Rhodobacter capsulatus 762–768 Rhodobacter sphaeroides 770 Rhodopseudomonas palustris 771 Rhodospirillum rubrum 771 nitrogen oxide reduction 624–639 Nitrosomonas 415 Nitrosospira 415 nitrous oxide reductase 626 NNR orthologs 634 nnrS 630 nnrU 630 Nobel prizes 276 non-phosphorous lipid 120 non-photosynthetic bacteria 109 non-photosynthetic carotenoids 113 non-sulfur bacteria 111 nonameric rings 944 nonphosphorylated RegA 715 nonsulfur purple photosynthetic bacteria 563–575 carbon dioxide metabolism regulation 564–575 normal spirilloxanthin pathway 104–106, 110 Nos 636 NosR 634 nostoxanthin 113 NQNO 468 n side 426 NtrC 748, 752, 764 NtrC regulon 765–766 Ntr systems 764 nuclear magnetic resonance (NMR) 155–157, 183, 389, 391, 478, 529, 731, 904, 905, 925, 955 2-D 584 protein maquette 904 nuclear motion 345 nuclear quadrupole 396 nuclear quadrupole coupling 393 nucleic acid probing studies 11 nucleophilic attack 815 nucleotide-binding sites 477 nuo operon 539 nutritional status 43
O O-acetyl-l-serine 613 O-acetyl-l-serine-(thiol)-lyase 613 1-O-acylglycerol 3-phosphate 2-O-acyltransferase 121
Index O-acetylhomoserine 613 O2 750, 752–753 O2 regulation 752–753 O2 sensor 753 obligately aerobic anoxygenic phototrophs 32 Oceanicola 440 okenone 98, 109, 110, 216, 218 okenone pathway 97, 99, 101, 108, 110, 112 oligomycin sensitivity conferring protein (OSCP) 477 oligonucleotide 716 oligonucleotide affinity trapping 716 oligotrophic ocean 50 olsA 121 olsB 121 one-exciton coherence 243 optical absorption maxima 431 optical mapping 694 optical spectra 339–340 optical spectroscopy 877–891, 954 optimality 285 orf1148 784 Orf162b 174 orf162b 174 orf1696 174 orf214 174 orf428 148 organic acids 40 organic carbon 50 organic solvent extraction 193 organosulfur compounds 596 organyl polysulfanes 608 organylsulfanes 607 orientational selectivity 961 ornithine 2-N-acyltransferase 121 ornithine lipid 119, 120, 129–130, 527 orthogonal synthesis 896, 900, 901 orthologs 693, 696 oscillations 362 oscillator strength 207 OSCP. See oligomycin sensitivity conferring protein (OSCP) osmotic pressure 657 osmotic stress 702 outer membrane protein 608 output domains 813, 817, 828, 831 oxidase aa3-type Cyt c oxidase 408 heme-copper 416 HemF-type coproporphyrinogen III 794 HemN-type coproporphyrinogen III 794 menahydroquinone 418 ubihydroquinone 416 oxidation/reduction midpoint potential 341 oxidation/reduction midpoint potentials 340–342 oxidative cyclase 72 oxidative damage 38, 545 oxidative protein folding 412 oxidative stress 417 oxo-group origin 71 2-oxoglutarate 764 oxygen 777–791, 794. See also O2
999 oxygen-dependent CPO 64 oxygen-independent CPO 64 oxygenase 578, 579 oxygen concentration 98 oxygen control 777, 778, 781–784, 787, 789 oxygenic light-harvesting apparatus 276 oxygenic photosynthesis 380 oxygen tension 707, 780–789, 844
P π-conjugated thiol 868 π-π interactions 190 bacteriochlorophyll 190 π-π stacking 915, 931, 934 P+700A1– 968 P+865QA– 965 P+QA– 297, 299, 304, 305, 343, 347, 382, 386, 387, 393, 396, 398, 954, 956, 967, 969 P+QA– formation dielectric response 967 P+QB– state 304, 345, 347, 386, 388, 396, 954, 956, 963 P+ reduction 515–517 P+ reduction kinetics 515–517 p-aminosalicylic acid 70 p-coumaric acid 823 P-type ATPases 661 P680 343 P700 343 Painter reaction 670, 671 pair correlation function 945 palandromic sequence 717 Pantoea ananatis 103 Pantoea stewartii 110 PAPS reductase 613 ParA chromosome partitioning protein 649 Paracoccus 38 Paracoccus denitrificans 264, 418, 479, 612, 695 Paracoccus pantotrophus 602 Paracoccus versutus 602 Paracraurococcus 34, 113 paralogs 696, 699 paramagnetic metal clusters 444 parameter insensitivity 285 partial denitrification 629, 631, 638 partial denitrifier 626, 631 PAS-fold 823 PAS domain 717, 728, 735, 750 PAS motif 572, 717 PAS superfamily 753 pathogenicity 701 pathway asymmetrical ζ-carotene 104 bacteriochlorophyll 147 carotenal 97, 108, 109, 112 carotenoid biosynthesis 101 carotenoid pathways 99 DOXP 103 DsbA-DsbB 412 mevalonate 103
1000 normal spirilloxanthin 104–106 okenone 97, 99, 101, 108, 110, 112 periplasmic apoCyt c thioreduction 412 quinol oxidase 554 R.g.-keto 97, 108 Sec-dependent 412 spheroidene 97, 107–108 spirilloxanthin 97, 99, 101, 104–106, 110, 112 tetrapyrrole 778 unusual spirilloxanthin 106 ζ-carotene 104 pbr operon 662 PCR 44, 606, 842, 851, 852 quantitative 47, 48 real-time 47 pcs 121 pea 854 Pelodictyon 746 pentaheme c-type cytochrome 547 pentaheme protein 548 Per Arnt Sim (PAS) 728 domain 728 motifs 717 perchlorate reductase 550 percolation 528 peripheral antenna 917 peripheral light-harvesting antenna complex. See light-harvesting 2 complex peripheral substituents 903 periplasm 413, 521, 606 periplasmic apoCyt c thioreduction pathway 412 periplasmic heme ligation 410 periplasmic ICM surface 259 periplasmic nitrate reductase 552 periplasmic space 413, 843 pernicious anemia 81 peroxiredoxin 418 persulfide sulfur 608 perthiol 609 PGC. See photosynthesis gene cluster pgpA 121 pgsA 121 Phaeospirillum fulvum 6, 110, 112, 579 Phaeospirillum molischianum 136, 137, 148, 166, 183, 185, 189, 191, 255, 259, 942, 945–946; See also Rhodospirillum molischianum light-harvesting 1 complex 191 light-harvesting 2 complex 137–150, 165, 166, 182–191, 202, 214, 218, 222, 225, 235, 242, 255, 259, 260, 278, 282, 283, 878–880, 884, 885, 946–948 photosynthetic apparatus 259, 260, 945, 946 reaction center-light-harvesting 1 complex 163, 259, 260, 945, 946 pH buffer 484 pH dependence 826, 348 phenol 578, 579, 581 phenylacetate 579, 581, 588, 589 pheophytin a 356 pheophytinization 42 phoB region 673 phonon 235
Index phonon-induced relaxation 206 phosphate-limited growth 120 phosphate deficiency 120 phosphate exchange 70 phosphatidate 124 phosphatidic acid 123 phosphatidylcholine 120, 125 10-phosphatidylcholine 160 phosphatidylcholine synthase 121 phosphatidyl cytidyltransferase 121 phosphatidylethanolamine 120, 124 phosphatidylethanolamine N-methyltransferase 121 phosphatidylglycerol 120, 124 phosphatidylglycerol 3-phosphate 125 phosphatidylglycerol 3-phosphate phosphatase 121 phosphatidylglycerol 3-phosphate synthase 121, 125 phosphatidylserine decarboxylase 121, 124 phosphatidylserine synthase 121, 125 phosphatidyl Tris 120 phosphoadenosine phosphosulfate 613 3´-phosphoadenosine-5´-phosphosulfate (PAPS) 610, 613 3´-phosphoadenosine-5´-phosphosulfate reductase 610 phosphoenolpyruvate 571 phosphoenolpyruvate carboxylase 41 2-phosphoglycerate 571 3-phosphoglycerate 571 phosphoglycerolipid 123–126 2-phosphoglycolate 571 phospholipid 119, 146, 173, 183 phosphorescence 357 phosphoribulokinase 566, 569 phosphorylated regulators 830 phosphorylation 715, 806 RegA 715 phosphorylation site 715, 806 phosphotransfer 802 photo-conversion 800, 804, 805 photoactive carotenoids 38 photoactive yellow protein (PYP) 652, 735, 811, 823 photoautotrophic CO2 fixation 40 photoautotrophy 40, 50 primary producer 2 photocurrent generators 862 photocurrent responses 862 photocycle 814, 823 bacteriophytochrome 829 BLUF 818 light, oxygen or voltage (LOV) 814 photoactive yellow protein 823, 826 photocycle reaction 733 photoheterotrophic growth 6, 564, 596 photoheterotrophy 6–7 photoinduced electron transfer –20 °C 521 photolithoautotrophic growth 564, 596 photomorphogenesis regulators 728 photon-echo 239 photon capture 136 photooxidation 517 photooxidative damage 71 photooxidative stress 736
Index photophosphorylation 2, 702 photoprotection 41, 97, 111, 348 photoreceptor 728–735 photoreceptor families 812 photoreceptor proteins 811–832 photoreversibility 828 photosynthesis gene cluster 20, 44, 65, 148, 699 photosynthetic apparatus 44–47, 844, 942–951 supramolecular assembly 942–951 photosynthetic CO2 fixation 48 photosynthetic complexes biogenesis 175 photosynthetic culture 842, 848 photosynthetic electron donor 4, 597 photosynthetic electron flux 50 photosynthetic electron transport 735 photosynthetic membranes 162 photosynthetic methylotrophs 32 photosynthetic pigment synthesis regulatory factor 43 photosynthetic rhizobia 32 photosynthetic unit (PSU) 146–147, 172, 257–263, 275–290, 862 atomic level structural model 287 chromatophore model 287 cordon sanitaire 285 energy transfer 889–891 organization 257–267 stoichiometry 286 supramolecular organization 286, 288 Photosystem I 267, 276, 343, 380, 915, 916, 918, 919, 968 high resolution structures 917 P700 343 structure 916 Photosystem II 265, 267, 343, 349, 380, 384, 916, 918, 919 high resolution structures 917 P680 343 structure 916 phototaxis 652, 735 phototoxicity 38, 41, 849 phototransfer 649 phototrophic alphaproteobacteria 597 phototropin 731–732, 811 phycobilisome 384 cyanobacteria 384 phycocyanobilin 804, 829 Phyllobacteriaceae 35 phylogenetic analysis 719–721 CrtJ/PpsR 719, 720 purple bacteria 4 phylogenetic surveys 47 phylogenetic tree 456 transhydrogenase 498–500 phylogeny 17, 21, 22 physical constraints of evolution 284–286 physiology purple bacteria 4–7 phytochrome 267, 799, 800–807, 811, 827, 828 phytochromobilin 804, 827 phytoene 103–104, 107, 108 15-cis form 103 all trans form 103
1001 phytoene dehydrogenase 717 phytoene desaturase 105 phytoene synthase 103, 717 phytol 74, 164, 300 phytyl-pyrophosphate 74 phytyl chains 946 pigment 139–142, 144, 146 pigment-pigment interactions 934 pigment-protein complex 44 pigment-protein sites 916 pigment exchange 359 pigment substitution 360 pigment synthesis regulation 42–44 pigment tuning light-harvesting complex 935 Pisum sativum 854 pKa 439, 445 plant carotenoids 113 plasmid broad host-range 842 pBLM2 40 pBBR1 842 pRK404 842, 843 plasticity 295–314 reaction center 295–314 platform vectors 843, 844, 845 pleomorphism 34 plsB 121 plsC 121 pmf. See proton motive force (pmf) pmtA 121 point-dipole calculation 393 Poisson-Boltzmann equation 358 Poisson distribution 520 polarized absorption 477 polar lipids 119 polonium 675 polyene 924 polyene chain 919 carotenoid 919 polyhistidine tag 842 polyhydroxyalcanoic acid 602 polymeric sulfur 607, 608 polyol ABC transporter 670 polysulfide reductase 549, 552, 612 polysulfides 596, 600, 601, 606, 607 porphobilinogen 60, 783, 784, 791 porphobilinogen deaminase 63 porphobilinogen synthase 62, 783, 791. See also δ-aminolevulinic acid dehydratase porphyrin 58, 164, 277, 288, 409, 896, 897, 899, 900 porphyrin binding site 69 porphyrinogen 58 Porphyrobacter 40, 113 Porphyrobacter meromictius 34 Porphyrobacter neustonensis 34 post-translational control nitrogenase 767–768 PpaA 785, 791 ppaA 785, 789 PpsR 708, 716–720, 729, 784–787, 791, 795, 819
Index
1002 PpsR/CrtJ binding site 784, 785, 795 PpsR/CrtJ core consensus sequence 795 PpsR2 807 Bradyrhizhobium 719 Rubriviax gelatinosus 719 precorrin-2 86 precorrin-3A 86 precorrin-4 87 precorrin-5 87 precorrin-6A 87 precorrin-8 89 primary donor 340, 382 primary electron donor 356 primary producer CO2 fixation 2 photoautotrophy 2 primary productivity 48 Prochlorococcus 47 populations 47 product inhibition 522 progressive deletion 165 C-terminus of LH1 α 165 projection map 160, 169 Propionibacterium acnes KPA171202 778 Propionibacterium freudenreichii subsp. shermanii CIP103027 778 Propionibacterium shermanii 778 propionic acid side chain 897 Propionigenium modestum 479 protease thermolysin 446 protein-Chl/BChl assembly 896 protein-chlorophyll interactions 897–899 protein-pigment interactions 905 protein-protein interactions 190, 426 protein-quinone conformation 383 protein conformational changes 954–973 protein dynamics 337, 338, 345, 346 protein maquette 904–905 crystal structure 904 design 896 NMR 904 protein packing 903 protein purification 849, 855 Proteobacteria 32, 98 α 409 α-2 34 α-3 34 α-4 34 β 409 δ 409 ε 409 γ 409 proteoliposomes 870 proteomic analysis 589 proteomics 269 proteorhodopsin 812 Proteus mirabilis 851 protochlorophyllide 68, 73 protochlorophyllide-reduction 854 protochlorophyllide reductase 67, 72, 73, 74
protomer 139, 144, 147 proton-motive force 495 protonation 815 protonation changes 826 proton electrochemical potential difference 510, 517 proton environment 962 proton motive force (pmf) 425, 452, 476 proton permeability 484 proton transfer 830 proton translocation 426, 476–488 ATP synthase 476–488 transhydrogenase 501 protoporphyrin IX 58, 60, 66, 778, 787, 794 protoporphyrin IX-Fe 436 protoporphyrin IX binding BchH 70 protoporphyrinogen 64 protoporphyrinogen IX 60, 785–787, 794 protoporphyrinogen IX oxidase 60, 65 Prr 782 PrrA 715, 754, 782, 783, 784, 785, 786, 789, 795 PrrA/B histidine kinase system 651 PrrA/PrrB 541. See also RegA/RegB PrrA/RegA binding sites 783, 795 PrrA binding sites 783 PrrB 652, 782, 785, 789, 790 PrrB/PrrA 633, 735, 754, 770 two-component system 735, 736 PrrC 418 PS I. See Photosystem I PS II. See Photosystem II psd 121 pseudoazurin 627 Pseudomonas 669, 702 Pseudomonas aeruginosa 63, 658, 701, 828 Pseudomonas denitrificans 83 Pseudomonas fluorescens 696, 699 Pseudomonas putida 816 Pseudomonas radiora 113 Pseudomonas syringae 665 pss 121 PSU. See photosynthetic unit PtdCho synthase 125 PtdEtn N-methyltransferase 125 puc 841 puc multigene cluster 147 puc operon 46, 92, 147, 149, 782, 842 puc promoter 844, 845, 848 PUC705-BA 844, 846, 866 pucA 842 pucB 842 pucBAC operon 150 pucBACDE 147 pucBA gene 149 multiple copies 149–150 PucC 147, 148 pucC 148, 174 pucC gene 801 puf 841 puf operon 92, 268, 732
Index puf promotor 844–848, 854 pufA 842 pufB 842 pufBA 45 pufC 45 pufL 842 pufL gene 40 pufLM 44, 47, 51 pufM 11, 40, 45, 842 nucleic acid probing studies 11 PufQ 268 pufQBALMX operon 256 PufX 45, 167, 191–192, 201, 207, 256–258, 263, 265, 269, 278, 525, 530, 942, 946, 947 dimerization of core complex 172 excitation sharing 172–174 location 169 membrane domains 172–174 NMR structure 167 quinone sequestration 172–174 PufX-minus strain 164 Rhodobacter sphaeroides 164 PufX– mutant quinone traffic 522–524 pufX gene 45 PufX polypeptide 157–160, 513 assembly into core complex 159 association with LH1 α 159 BChl binding 191 in vitro 159 C-terminal processing 191 chemical synthesis 159, 191 core segment 191 functional role 192 glycine-rich region 159 homodimerization 160 inhibition of LH1 formation in vitro 191 labeling with fluorescence 191 location within dimer complex 191–192 membrane topology 158 N-terminal region 169 deletion of residues 169 PufX mutants 191 purification 159, 191 quinol/quinone exchange 158 role in core complex dimerization 160, 191 solution structure 158, 159, 160 alternative solution structure 159 C-terminus 159 glycine-rich sequence 159 intersection with LH1 α 159 N-terminal region 159 transmembrane region 159 suppressor mutants 158 TOXCAT analysis 160 puh 841 puhA gene 175, 268 puhB 174 pulse-field gel electrophoresis 694 pulsed EPR 955 pump-dump-probe 227
1003 pump-probe 238, 239 pure exciton model 205 purple bacteria evolution 17–28 extremophilic 9 habitat 7–9, 149 overview 2–12 phylogenetic analysis 4 physiology 4–7 purple nonsulfur bacteria 4–6, 9, 10, 37, 42, 577–591, 596–600 purple sulfur bacteria 4, 5 laboratory culture 4 pyridoxal phosphate 778 pyridoxal phosphate-dependent 778 pyridoxal phosphate cofactor 61
Q Q-cycle 426, 455–456, 460–469, 510, 511, 517, 518 framework 455–456 Q-flip 822 Q/QH2 binding site 430, 433–435 QA 46, 517 midpoint potential 46 QA 384, 385, 392, 398 methoxy groups 384 QA– 522 QB 385, 387, 395, 396, 397, 517, 522 distal 388, 396, 956 methoxy groups 388 proximal 387, 388, 396 QB– binding sites 961 QB site 256, 258, 265 QH2:Cyt c oxidoreductase activity 672 QH2 oxidation 430 Qi 512. See also quinone reduction site Qi site 433, 434, 435, 436 cytochrome bc1 467–468 qmo genes 611 Qo 512. See also quinol oxidation site Qo site 430, 435, 442, 460–467 distal niche 462 molecular mechanism 460–467 proximal niche 455, 462 Qp 454 quantitative polymerase chain reaction 49 quantum chemical calculations 205, 815 quantum chemistry molecular dynamics 282 quantum coherence 278 quantum efficiency 276, 862 quantum mechanics 276, 502 description of photosynthesis 276 quantum theory finite temperature 282 quantum yield 193, 280, 360 quenching of LH1 emission 928, 929 quinol 254, 257 quinol/quinone exchange 157, 158, 947 quinol oxidase 717
1004 quinol oxidase (Qo) 454. See also proton motive force (pmf) quinol oxidase pathway 46, 554 quinol oxidation site 452 quinone 173, 254, 257, 264, 379–399, 510, 518–519, 523, 754 methoxy 385, 388, 389 pool 529, 554 primary 380 proximal 397 reconstitution 382 redox signal 754 secondary 380 quinone/quinol 258 quinone/quinol diffusion 949 quinone/quinol exchange 947 quinone/quinol gate 948 quinone/quinol passage 944 quinone binding 712 quinone binding pockets 46 quinone binding sites 942 hydrogen bonding 389 quinone confinement 528–529 Rhodobacter sphaeroides 522 quinone diffusion 263, 289 quinone exchange 161 quinone exclusion 301 quinone passage light-harvesting 1 complex 523 quinone pool 46, 170, 395, 426, 517 Em 554 redox poise 554 quinone reactions 518–519 quinone reduction site 452 quinone replacement 303 quinone sequestration PufX 172–174 quinone traffic PufX– mutant 522–524 quorum sensing 702 Qy absorption band 136, Qy absorption band shift 46 Qy electronic transitions 205, 232 Qy transition 203, 204, 208 Qy transition dipoles 145
R R26 165, 303, 383, 386 R-26 302, 304, 339 R26.1 218–220, 223, 225, 387, 667, 668, 924, R-26.1 299, 304 R-26 RC 299 radiationless excitation transfer 276 radical-pair 356 radical-pair recombination 815 radical pair intermediates 357–359 radical pair state P+QA– 956 radical S-adenosylmethionine superfamily 794 Ralstonia eutropha 566, 744–757, 749, 750, 752 Ralstonia metallidurans 658, 662 Raman spectroscopy 46, 166, 175, 184, 185, 203, 204, 221, 224, 227, 362, 389, 390, 394, 396, 733, 804, 819, 821, 922–924
Index random collision model 520, 526 random matrix theory 282 rare codons 850 rate-zone centrifugation 267 RB. See Rieske/Cytochrome b complex (RB) RB operons 458 RC. See reaction center (RC) RC-H subunit 6-Histidine tag 170 RC-LH1-PufX. See reaction center-light-harvesting 1-PufX complex RC-LH1-PufX core dimer 3-D reconstruction 170 RC-LH1 core structures 267 re-isomerization 827 reaction center (RC) 44–45, 193–195, 232, 254, 263, 264, 295–314, 323, 337–349, 355–370, 380, 452, 509–530, 699, 703, 841, 942 A-branch 301 B-branch 301, 306 bacteriochlorin exclusion 301 bacteriochlorin replacement 302 bacteriochlorophyll histidine ligand 344 replacement 305 water ligand 344 BPhe replacement 306 charge separation 356–370 chimeric RCs 298 current-voltage curves 867 cytochrome subunit 298 directed modification 338–349 DLL mutant 339, 347 electron donor 514–515 electron transfer B-branch 348 new reactions 348 pH dependence 348 exponential dependence on distance 345 Fe removal 305 Fe replacement 304, 305 heterodimer 137, 305, 310, 338 heterodimer mutant 340 H-polypeptide 297, 298 inactive P+QB– 957, 958, 962 in vitro reconstituion 193 LM complex 297 orientation 168 plasticity 295–314 QB site 168, 173, 256, 258 distal position 956 proximal vs. distal shift of QB– 962 quinone exclusion 301 methoxy 385, 388, 389 replacement 303 proximal quinone 397 R-26 299, 302, 304 R-26.1 299, 304 RC-H subunit 6-Histidine tag 170
Index reconsititution 193–195 Rhodobacter sphaeroides 323, 324–333 solvent extraction of subunit 193 special pair 39, 46, 163, 166, 168, 169, 238, 243, 280, 301, 356, 381, 514, 550, 866, 871, 956, specific assembly factor 174 transmembrane helices 164 type I 380 type II 380, 385 zinc substituted 305, 396 β mutation 347 reaction center-light-harvesting 1-PufX complex 147, 166–174, 201, 254, 261, 263, 269, 278 dimeric complexes 168–174 helical arrays 168 Rhodobacter veldkampii 166–168 reaction center-light-harvesting 1-W complex Rhodopseudomonas palustris 166 reaction center-light-harvesting 1 complex 136, 155–175, 162–166, 201, 255, 257–259, 268, 512, 523, 807, 862, 863, 870, 935, 942, 949 AFM and EM model 165 Blastochloris viridis 162 deletion of C-terminus of LH1α 165 hexagonal packing 162, 165, 168, 171 monomeric complexes 162–166 AFM topographs 162 cryo-EM 162 electron microscopy 162 hexagonal packing 162 Phaeospirillum molischianum 163 Rhodobacter sphaeroides mutants 164 Rhodopseudomonas palustris 261 Rhodospirillum photometricum 165 Rhodospirillum rubrum 163 role of carotenoid in assembly 193 structures 267 reaction center H 260, 268 reaction center proteins 21 reaction center QB site 157 reaction mechanism 60 reactive oxygen species (ROS) 666, 668 production 674 reassociation 864 recA 850 recognition sequence CrtJ 717 reconstituted membranes 943 reconstitution 382 cofactor 924 experiments 186 reaction center 193–195 rectification 862, 868 red-shift 45, 188, 193, 200, 203–205, 208, 235–247, 729, 733, 818, 820–823, 825, 827, 879, 921–923, 926–934 red-shifted signaling state 820 Redfield theory 281 modified Redfield theory 237, 242, 244, 245 redox cellular 708
1005 redox-active cysteine 713, 714 redox-box 714 redox-regulation 707 redox active tyrosine 349 redox balancing 630, 631 redox cofactors 853, 856 redox mid point potentials 431 redox poise 46 redox pool 631 redox potential 175 redox processes 657 redox reaction 495 redox regulation 717–718, 753–754 redox sensing 652, 713–714 redox signal 712, 714, 754 quinone 754 reduction of geranylgeraniol to phytol 74 reductive dehalogenases 82 refolding process 864 RegA 567, 707, 708, 712–717, 753, 754, 783, 786, 789, 795 binding site 715 homologs 709–711 nonphosphorylated 715 phosphorylation 715 response regulator 714–725 regA 783, 785, 787, 790 RegA/RegB 541, 651. See also PrrA/PrrB RegAB/PrrAB 563, 564, 567 RegB 652, 707, 708, 712–716 autophosphorylation 713 homologs 709–711 regB 567, 708, 712, 715 RegB/PrrB/RegS-RegA/Prra/RegR 743 RegB/RegA 633, 708, 708–716, 735, 744, 754, 770 two component system 735, 736 RegR 715 consensus sequence 716 RegS-RegR 754 regulation gene expression 707–722 nitrogen fixation 760–772 pigment synthesis 42–44 synthesis 707 regulation by Fnr 721–722 regulatory factor photosynthetic pigment synthesis 43 regulatory proteins 762 relaxation 233, 235, 239, 241, 246, 361, 363, 365 remodeling membranes 120 renewable energy 771 reorganization energy 233, 343, 347, 348, 358, 359, 364, 382 rescue-mutagenesis 922 resistance-nodulation-cell division (RND) 658 respiration 149, 510, 526, 707, 713, 745 respiratory arsenate reductase 552 respiratory chain 264, 510, 526, 527 respiratory membrane 262 respiratory uncouplers 417 response regulator 751 RegA 714–725 response regulators 572, 819
1006 restriction/modification system 850 reverse electron transfer 380 Complex I 512 Rheinheimera 601 Rhizobacteria 694 Rhizobales 598 rhizobia 35 Rhizobiales 442, 597 Rhizobium 100, 113, 549, 693, 702 Rhizobium leguminosarum 716, 744, 753 Rhizobium leguminosarum bv viciae 3841 479 Rhizobium meliloti 700 rhodanese 605 Rhodobaca 11 Rhodobaca bogoriensis 3, 108, 599 Rhodobacter 3, 11, 34, 666, 669, 841, 850–856 cell-free expression system 852 cytochrome bc1 mutants 428–430 expression system 842 Rhodobacteraceae 23, 380, 603, 614 Rhodobacterales 112, 597, 598, 614 Rhodobacterales bacterium 627 Rhodobacter azotoformans 597, 598 Rhodobacter blasticus 168–171, 256, 261, 513, 598, 668, 946–947 core dimers 168–171 dimeric core complex 170 Rhodobacter capsulatus 3, 6, 7, 9, 19, 40, 58, 61, 65, 71, 72, 102, 103, 107, 112, 175, 191, 205, 267, 338–353, 356, 360, 384, 397, 408, 417, 426, 427, 437, 440, 479, 512, 513, 515, 521, 527, 538–542, 554, 563–565, 567, 569, 570, 572, 597, 598, 606, 627, 629, 633, 634, 635, 637, 666–668, 670, 672–674, 700, 708, 735, 744–757, 760–768, 777, 780, 782–791, 794, 795, 824, 843, 845, 847, 851, 853, 854, 942 cobalamin biosynthetic genes 84 cytochrome biogenesis 407–423 cytochrome bc1 complex 425–450 cytochrome cy 845 gene regulation oxygen 707–725 light 727–741 hydrogenase 743–757 nitrogen fixation 762–768 Q sites 434 RegB/RegA 754 strain A6D1 310 strain DLL 307 strain QAQA 309 strain sym-1 308 supramolecular organization 519–522 U43 strain 307 Rhodobacter capsulatus SB1003 780, 782–792, 794, 795 Rhodobacter capsulatus strain PAS100 783 Rhodobacter sphaeroides 3, 7, 9, 19, 21, 22, 58, 61–65, 71, 102, 103, 107, 108, 112, 137, 144, 145, 157, 159, 160, 164, 168–171, 183, 184, 191, 201, 205, 215, 217–226, 239, 254, 256, 257, 260, 264, 267, 268, 277, 278, 286, 296, 323, 338–353, 356–358, 360, 365, 368, 380, 381, 383–385, 388, 397, 408, 426, 427, 437, 455, 479, 512, 513, 515–517, 521, 527, 538–543, 545, 546, 550, 554,
Index 563–569, 572, 573, 597, 598, 603, 614, 626, 635, 637, 638, 643, 647–651, 667, 668, 671, 672, 674–676, 692, 708, 728, 732, 735, 736, 745, 747–749, 754, 761, 770, 777, 780–789, 791, 794, 804, 813, 817, 819, 824, 828, 841–847, 849, 850, 851, 853, 854, 855, 878, 920, 925–927, 942, 947, 955, 956, 960, 964, 969 AppA 652, 718, 732-736, 785, 817-821, 831 bacteriochlorophyll biosynthesis genes 57-79 carbon dioxide metabolism 563-576 chemotaxis 647–651 Chromosome 1 122, 150, 691–706 cobalamin biosynthetic genes 84 core dimers 168–171 cytochrome bc1 455 cytochrome c2 323, 324–333 cytochrome c2:reaction center complex 323–333 electron transfer pathways 382 G1c mutant 218, 222, 930 gene regulation oxygen 707–725 light 727–741 tetrapyrrole biosynthesis 777–798 genome 86, 417, 538, 539, 542, 626, 627, 633, 646, 649, 652, 676, 691–706, 712, 731, 780, 786, 790, 791, 813, 819, 828, 850, 920, heterologous gene expression 839-860 hydrogenase 743-757 inclusion bodies 847 light-harvesting 2 complex 135–153, 156, 161, 165, 166, 172, 173, 186, 187, 203, 204, 206, 214, 218, 220, 222, 223, 225–257, 260, 262, 279, 513, 524, 528, 632, 667, 699, 841, 842, 854, 913-924, 947 lipid biosynthesis 119–134 mutant See Mutant/mutants N-terminal signal sequence 843 nitrate reduction 623-642 nitrogen fixation 770 PrrB/PrrA 708, 735, 754 PrrA/PrrB 541 PufX-minus strain 164 PufX polypeptide solution structure 169 quinone confinement 522 reaction center 295–321, 337–353, 379–405 R26 165, 303, 383, 386 R-26 302, 304, 339, 520 R26.1 218-220, 223, 225, 387, 667, 668, 924, R-26.1 299, 304 R-26 RC 299 strain DD13 307 strain DD13/G1 307 supramolecular organization 253–273, 275–294, 519–522, 947 Rhodobacter sphaeroides 2.4.1 539, 780–792, 795 Rhodobacter sphaeroides ATCC17025 790, 792 Rhodobacter sphaeroides ATCC17029 792 Rhodobacter sphaeroides f. sp. denitrificans 671 Rhodobacter sphaeroides RS630 780 Rhodobacter sphaeroides wild type strain NCIB8253 785 Rhodobacter sulfidophilus 3 Rhodobacter veldkampii 166–168, 270, 513, 597, 598 dimerization motifs 168
Index PufX-PufX transmembrane association 168 PufX N-terminus 167 purification 167 reaction center-light-harvesting 1-PufX complex 166–168 structure 167 Rhodobium 10 Rhodobium marinum 149, 597, 598, 930 Rhodobium orientis 597, 598 Rhodoblastus acidophilus 11, 98, 108, 109, 112, 514, 579, 792 Rhodoblastus sphagricola 112 Rhodocista centenaria 10, 728, 736 Rhodocista centenum 824 Rhodocyclales 599 Rhodocyclus 543 Rhodocyclus gelatinosus 183, 745 Rhodocyclus purpureus 34, 579, 599 Rhodocyclus tenuis 9, 98, 515, 599, 668, 674 Rhodoferax 543 Rhodoferax antarcticus 3, 11, 599 Rhodoferax fermentans 515, 543, 599 Rhodoferax ferrireducens 23, 414, 606, 614, 676 Rhodoferax ferrireducens T118 23, 26 Rhodomicrobium 666 Rhodomicrobium vannielii 104, 112, 579, 597, 598 Rhodopila globiformis 11, 104, 108, 597, 598, 612 rhodopin 110, 141, 929, 930, 931 rhodopin glucoside 142, 148, 215, 217, 220, 221 rhodopinal 108 rhodopinal glucoside 108–110, 148 rhodopinol 108, 110 Rhodoplanes elegans 597, 598 Rhodoplanes roseus 597, 598 Rhodopseudomonas acidophila 11, 46, 135–151, 185, 201, 202, 204, 214, 215, 217–223, 225, 235, 241, 254, 277, 278, 579, 878, 942 crystal structure 202 light-harvesting 2 complex 46, 110, 135–153, 165, 185, 186, 201–205, 214–225, 235, 236, 241, 245, 254, 255, 277, 278, 878, 882, 883, 886, 923, 942, 948 Rhodopseudomonas acidophila strain 10050 135–151 Rhodopseudomonas acidophila strain 7050 137, 143, 148 Rhodopseudomonas cryptolactis 10, 148 Rhodopseudomonas julia 597, 598 Rhodopseudomonas marina 160, 183 Rhodopseudomonas molischianum 202, 256, 266 crystal structure 202 Rhodopseudomonas palustris 3, 7, 9, 19, 64, 102, 103, 145, 166, 201, 254, 255, 257, 261, 265–267, 29, 513, 539, 542, 546, 563–565, 572–574, 578–581, 583–591, 597, 598, 603, 614, 626–630, 634, 637, 638, 646, 651, 676, 679, 692, 719, 728, 743, 745–749, 751, 752, 754, 761, 771, 799–807, 819, 824, 828, 829, 945, 948–949 bacteriophytochromes 799–809 degradation of aromatic compounds 577–594 genome 149, 150, 268, 578, 586, 590, 628, 629, 634, 636, 651, 692, 693, 695, 696, 699–702, 828 monomeric RC-LH1-helix W core complex 171 nitrogen fixation 771 reaction center-light-harvesting 1 core complex 261 reaction center-light-harvesting 1-W complex 166 RegS-RegR 754
1007 supramolecular organization 253-273, 948-950 Rhodopseudomonas palustris BisA53 676, 790, 792 Rhodopseudomonas palustris BisB18 676, 790, 792 Rhodopseudomonas palustris BisB5 676, 790, 792 Rhodopseudomonas palustris CGA009 479, 676, 790, 793 Rhodopseudomonas palustris HaA2 676, 790, 793 Rhodopseudomonas palustris strain 2.1.6 150 Rhodopseudomonas viridis See Blastochloris viridis 183, 201, 239, 380 rhodopsin 596, 730–731 sensory 730–731 rhodoquinone 383, 392, 511, 530 Rhodospira trueperi 106, 597, 598 Rhodospirillaceae 603, 614 Rhodospirillales 597, 598 Rhodospirillum 543, 669 Rhodospirillum centenum 10, 19, 643, 645, 728, 800, 828 Rhodospirillum fulvum 579 Rhodospirillum molischianum 165, 214, 218, 219, 221, 222, 225, 235, 242, 278, 512, 878, 930. See also Phaeospirillum molischianum light-harvesting 2 complex 137–150, 165, 166, 182–191, 202, 214, 218, 218, 222, 225, 235, 242, 255, 259, 260, 278, 282, 283, 878–880, 884, 885, 946–948 Rhodospirillum palustris 863 Rhodospirillum photometricum 9, 137, 165, 255, 256, 259, 265, 266, 942, 944, 945 AFM data 165 reaction center-light-harvesting 1 complexes 165 Rhodospirillum rubrum 3, 7, 19, 20, 58, 87, 102, 103, 106, 156, 160, 164, 182, 184, 189, 192, 201, 215, 223, 226, 239, 254, 256, 258, 259, 265, 384, 440, 441, 512, 517, 521, 539, 597, 598, 603, 614, 651, 667, 668, 674, 679, 692, 743, 746, 752, 761, 771, 789, 863, 899, 925, 926, 927, 942 carotenoidless mutants 926 containing wild-type carotenoids 160 CooLH hydrogenase 746 LH1 core complex 254 nitrogen fixation 771 puhA gene 175 reaction center-light-harvesting 1 complexes 163 reconstitution of LH1 complex 181-198, 924-935 Rhodospirillum rubrum ATCC 11170 479, 676, 789–791, 793, 794 Rhodospirillum rubrum S1 930 Rhodospirillum rubrum strain G9 304 Rhodothalassium 10 Rhodothalassium salexigens 824 Rhodovibrio 10 Rhodovibrio sodomensis 10 Rhodovulum 10, 34 Rhodovulum adriaticum 598, 612 Rhodovulum euryhalinum 598, 612 Rhodovulum iodosum 598, 612 Rhodovulum marinum 598 Rhodovulum PS88 668 Rhodovulum robiginosum 598, 612 Rhodovulum strictum 598 Rhodovulum sulfidophilum 58, 102, 137, 515, 547, 550, 598, 607, 616
Index
1008 Rhodovulum sulfidophilum W4 790 Rhodovulum sulfidophilus 202 Rhodovulum sulphidophilum 71 riboflavin 816 ribose-5-phosphate 571 ribosomal protein 418 ribosomal subunit 418 ribosome binding site 842 ribulose-1 41 ribulose 1,5-bisphosphate (RuBP) 570, 571, 572 ribulose 1,5-bisphosphate carboxylase/oxygenase (Rubisco) 19, 20, 25, 27, 564, 566, 569, 570, 572–574 Rickettsia 440, 694 Rickettsia bellii 436 Rickettsia prowazekii 694 Rieske/Cytochrome b complex (RB) 452–460, 469 catalysis 455–456 evolution 456–458 operons 458 structure 453–455 Rieske FeS protein 263 Rieske iron-sulfur center 452 Rieske protein 427, 942 Riftia pachyptila 611 rivers 51 RNA polymerase 754, 765 RNA secondary structure 843 RND. See resistance-nodulation-cell division (RND) RND protein 658 Rnf 745 robustness graceful degradation 285 parameter insensitivity 285 robustness and optimality 285 rodopinal glucoside 144 roller coaster pattern 514 Roseateles depolymerans 40 Roseibacterium 34 Roseibium 34 Roseicyclus 34, 40 Roseicyclus mahoneyensis 34 Roseiflexus castenholzii DSM13941 676 Roseiflexus sp. RS-1 676 Roseinatronobacter 34 Roseinatronobacter thiooxidans 599 Roseisalinus 34 Roseisalinus antarcticus 34 Roseivivax 34 Roseoateles 113 Roseobacter 34, 40, 113, 554, 651, 695 Roseobacter clade 34, 40, 48 Roseobacter denitrificans 19, 22, 71, 102, 104, 107, 108, 112, 113, 170, 299, 514, 515, 546, 547, 599, 603, 627, 629, 634, 635, 636, 637, 813 Roseobacter denitrificans OCh114 790, 791, 793 Roseobacter denitrificans Ohc114 676 Roseobacter gallaciensis 38 Roseobacter sp. 58 Roseobacter sp. MED193 603 Roseococcus 34 Roseococcus thiosulfatophilus 110, 113, 201, 204, 599
Roseospira 10 Roseospira mediosalina 597, 598 Roseospirillum parvum 597, 598 Roseotales depolymerans 599 Roseovarius 34, 440, 614 Roseovarius mucosus 34 Roseovarius nubinhibens 599, 603 Roseovarius sp. 19 Roseovarius sp. 217 603, 627, 790, 791, 793 rosette-forming species 34 rotary catalysis FOF1-ATP synthase 476–478 rotary ion translocation 481 rotational power spectrum 163 rotor 477, 644 Rubisco. See ribulose 1,5-bisphosphate carboxylase/oxygenase (Rubisco) rubredoxin-like metal binding fold 439 Rubrimonas 34 Rubritepida 34 Rubrivivax gelatinosa 745 Rubrivivax gelatinosus 9, 19, 24, 40, 58, 72, 102, 103, 108, 137, 147, 149, 258, 298, 414, 435, 514, 518, 599, 603, 614, 679, 719, 722, 795, 813, 819, 943 AFM 943 fnrL null mutant 722 PpsR2 719 Rubrivivax gelatinosus PM1 676 Rubrivivax gelatinosus S1 and IL144 790 Ruegeria gallaciensis 38
S σ28. See flagellar-specific sigma factor (σ28) σ54 644, 645, 646, 650, 748, 749 σ54-RNA polymerase 749, 753 σ70 748 σ70 RNA polymerase 749, 753 σ factor. See sigma factor S* state 222–224 energy transfer 222–224 S-adenosyl-L-methionine (SAM) 71, 81, 83, 86, 107, 674, 794 S-adenosyl-L-methionine:magnesium protoporphyrin IX-Omethyltransferase 71 S-adenosyl-L-methionine uroporphyrinogen III methyltransferase 86 16S rDNA sequence 33 analysis 33, 34 16S rRNA 19, 21, 22, 23 23S rRNA sequence analysis 35 S1 state 219–222 energy transfer 219–222 S2 state 216–219 energy transfer 216–219 Saccharomyces cerevisiae 479, 612, 615, 694 Salinibacter ruber 824 Salmonella typhimurium 612 salt bridge 815, 822 salt flats 51 SAM. See S-adenosyl-L-methionine (SAM) SAM O-methyltransferase 67
Index Sandaracinobacter 34, 40 Sandarakinorhabdus 34 SAR11 37 sat 614 SAXS. See small angle X-ray scattering scanning permutagenesis 922 scanning probe analysis 863 scanning tunneling microscopy 863 Schiff’s base 61, 62 Schizosaccharomyces pombe 418 Sco1 418, 419 SCRP. See spin-correlated radical pair (SCRP) SDS-PAGE. See SDS-polyacrylamide gel electrophoresis (SDS-PAGE) SDS-polyacrylamide gel electrophoresis (SDS-PAGE) 193, 846 Sec-dependent pathway 412 Sec61 protein-conducting channel 269 SecA 269 SecDF 269 secondary electrochromic markers 954 secondary quinone 380, 395 second order electron transfer 332–333 second order rate constant 325, 327–332, 516, second order reaction 325, Sec translocon 269 SecY 269 SecYEG 436 selective deuteration 961 selenate 669, 671 selenate reductase 547, 548, 550 selenate reduction 550 selenate respiration 550 selenite 669, 671 selenium 669–671 selenium oxyanions 671 selenocysteine 670 selenomethionine 670, 844, 851 self-assembled monolayer (SAM) 862, 863, 865 self-assembly 904 selfish operon theory 21 semi-aerobic 107, 785–788, 845, 848, 849, 853 semi-aerobic conditions 107 semi-automated purification 849 semiquinone (SQ) 391, 392, 394, 395, 460, 820 O2-sensitive 460 semiquinone anion 396, 467 SenC 419 sensing fixed nitrogen 763–766 sensor kinase RegB 712 sequence alignment ATP synthase 479 hydrogenase 744, 748 ALA synthase 791 light-harvesting polypeptides 145, 185, 917, 921 PucC 148 Cyt b 431, 455 sequential electron transfer 963 serine/threonine kinase domain 731 sewage 9 SgpA 608
1009 SgpB 608 SgpC 608 sgp genes 608 Sgp proteins 608 Shemin pathway 59, 778, 780 Shewanella 664 shielding helices 902, 903 short-circuit reactions 460 sigma-E 736 sigma factor 700, 701 sigma factor RpoN 766 signal generation 812 signaling 747–754 signaling state 819–822, 825, 830–831 signal transduction 43 signal transduction chain 812 signal transduction systems 43 Silicibacter pomeroyi DSS-3 479 simulations 355 single-molecule spectroscopy 166, 207, 237, 877–879, 883, 885, 891 single light-harvesting complex 244 single molecule imaging 950 single particle EM analysis 70, 170 single photosynthetic unit energy transfer 889–891 singlet-singlet annihilation 183, 188, 257, 258, 262 singlet-triplet annihilation 183 singlet-triplet quenching 257 singlet oxygen 38, 736–737 Sinorhizobium meliloti 91, 629, 781 siroheme 82, 86, 637, siroheme-[Fe4S4] 613 sirohydrochlorin ferrochelatase 86 site-directed mutagenesis 204, 439 small angle X-ray scattering 805 SodA 673 soda lakes 11 SodB 673 sodium dodecyl sulfate 846 soil 51 sojourn expansion 281 sojourn time 282 Solibacter usitatus 458 solid state conception 519 solid state NMR 391 solid support 861 solubilization 855 soluble periplasmic Cyt c2 45 solution structure 482 AppA BLUF domain 818 LH1α polypeptide 157–158 LH1β polypeptide 157–158 phot1 LOV2 813 PufX 158, 169, 270 RegA 716 solvation mechanisms 207 SorAB 611 sorAB 603, 611 sox genes 599, 602 Sox proteins 606
1010 Sox system 602–606, 612, 616 SoxB 602, 606 SoxCD 602, 605 soxCD 603 SoxE 606 soxEF 603 SoxF 606 SoxR 602 SoxS 602 SoxV 602 SoxXA 602, 605 soxXABYZ 603 SoxYZ 602, 605 Spb 736 special pair 39, 46, 163, 166, 168, 169, 238, 243, 280, 301, 356, 381, 514, 550, 866, 871, 956, specificity 360–361 spectral density 233 spectral diffusion 885 spectral hole-burning 361 spectral overlap 148, 280 spectral universality theorems 283. See also thermal disorder spherical vesicles 845 spheroidene 97, 108, 110, 215, 218, 220–222, 225, 227, 929, 931 spheroidene monooxygenase 107 spheroidene pathway 107–108, 110 spheroidenone 41, 107, 110, 216, 222, 226 spheroplast 66, 855 Sphingomonadales 113 spin-correlated radical pair (SCRP) 963, 968 electron-nuclear double resonance (ENDOR) 965, 966 – 968 P+700A1 Spinacia oleracea 479 spin boson 368 spirilloxanthin 41, 97, 104, 108, 220, 223, 226, 927, 929, 931, 933 biosynthesis 104, 104–108 spirilloxanthin pathway 99, 101, 104–106, 110, 112 Spirulina platensis 479 sqd genes 120 sqdA 121 sqdB 121 sqdC 121 sqdD 121 sqr 603 stabilization energy 190 Staleya 34 Staleya guttiformis 34 standing crop 50 Staphylococcus aureus 661 Stappia marina 34 Starkeya novella 611, 853 Stark spectroscopy 208, 344 STAS domain. See sulfate transporter and antisigma factor antagonist domain (STAS) static disorder 206, 207, 238, 246 stator 477, 644 steady-state spectra 236–237 stereo-selectivity heme attachment 413
Index steric hindrance 203 stigmatellin 389, 435, 442, 454, 462, 521 stochastic Liouville equation 369 stoichiometric excess 518 stoichiometric ratio Q:RC 523 stoichiometry 944 photosynthetic unit 286 stratified lakes 32 blooms 7–8 Streptomyces hygroscopicus 852 Streptomyces nodosus subsp. asukaensis 778 structural genomics 847, 856 structural genomics initiative 848 structural stabilization 915 structural variability 255–257 structural water 310 sucrose density gradient centrifugation 855, 856 sulfate 598, 610–612, 615 sulfate-reducers 591 sulfate assimilation 612–615 sulfate permease 612, 615 sulfate reducing bacteria 609 sulfate reductase 670 sulfate transport 612, 613 sulfate transporter and antisigma factor antagonist domain (STAS) 813, 816; See also flavin mononucleotide (FMN) sulfate transporters 616 sulfate uptake 612 sulfide 596, 598, 600, 602 sulfide:cytochrome c oxidoreductase 606 sulfide:quinone oxidoreductase (SQR) 602, 606–607 sulfide oxidation 4, 602, 606, 607 sulfide springs 32 sulfite 598, 602, 616 oxidation 610–612 sulfite:cytochrome c oxidoreductase 599, 853 sulfite dehydrogenase 610, 611 sulfite oxidase 611 sulfite oxidation 610–612 sulfite reductase 602, 605, 607, 609, 615, 616, 670 sulfolipid 613 Sulfolobus acidocaldarius 459 sulfoquinovosyldiacylglycerol 119, 120, 127 sulfoquinovosyl transferase 128 sulfur 596–616. See also elemental sulfur elemental 596–597 S7 rings 607 S8 rings 607 sulfur chain 608 sulfur compounds 596–616 sulfur cycle 40 sulfur dehydrogenase 602, 605 sulfur globule proteins 608 sulfur globules 605–609 sulfur metabolism 40, 596–616 Sulfurospirillum 669 sulfur oxidation 596, 596–602, 602–612, 616 pathways 602–612 sulfur speciation 608 sulfurtransferase 605, 606
Index supercomplex 263, 512, 520, 521, 524–526, 530 dimer 525 mitochondrial 526, 527 model 524–526 superexchange 348, 356 superoperon 44, 65, 770 superoxide production 673 superradiance 237, 244 suppressor mutants 158 supramolecular architecture 288, 943, 945 supramolecular arrangement 525-527, 891 supramolecular assembly 527, 942–951 photosynthetic apparatus 942–951 supramolecular complexes 182, 195 supramolecular organization 254, 257, 258, 267, 268, 278, 286, 517, 519–522 photosynthetic unit 286, 288 Rhodobacter capsulatus 519–522 Rhodobacter sphaeroides 519–522 supramolecular structure 873 Surf1 418 assembly protein 545 surface helix cd1 436 cd2 436 swimming 644–646 Synechocystis 66, 71, 72, 734, 800, 804, 817, 819, 820, 827, 830 synteny 629 synthetic chlorophyll protein screening 901 solid support synthesis 901 synthetic phospholipid 160
T tac promoter 854 taxonomic-phylogenetic tangles 37–38 tehAB gene pair 665 tellurate 671 tellurite 671 transport 673 tellurium 671–673 tellurium oxyanions 671 Te resistance determinants 673 tetracycline 842, 844, 849 tetraheme 258 tetraheme cytochrome 515, 942, 943 tetraheme reaction center subunit 514 7,8,17,18-tetrahydroporphyrin 897. See also chlorin tetrahydrospirilloxanthin 106, 108, 110, 112 3,4,3´,4´ tetrahydrospirilloxanthin 106 Tetrahymena thermophila 851 tetramethylbenzoquinone 390 tetrapyrrole ring bacteriochlorophyll 42 tetrapyrrole 82, 339, 700, 778–795, 804 tetrapyrrole biosynthesis pathway 777–795 regulation 777, 778–795 tetrapyrrole planes 431 tetrathionate 597, 599, 612
1011 tetrathionate hydrolase 599 tetratricopeptide repeat (TPR)-like motif 414 Thalassobacter 34 Thauera 669 Thauera aromatica 581, 583, 584, 591 thermal disorder 277, 282–284 finite temperature quantum theory 282 thermal springs 39, 51 Thermochromatium tepidum 3, 6, 8, 10, 12, 19, 356, 380, 384, 514, 728, 730, 824 light-harvesting 1 complex 10 reaction center structure 387 thermophilic purple bacteria 10 Thermosynechococcus elongatus 817, 820 T110078 820 Thermotoga maritima 805 thermotolerance 149 Thermus thermophilus 539 thio-oxidoreduction 407, 410 Thioalkalicoccus limnaeus 106 Thioalkalispira 600 Thioalkalivibrio 600 Thiobacillus acidophilus 599 Thiobacillus denitrificans 611 Thiobacillus ferrooxidans 566 Thiocapsa marina 112 Thiocapsa pfennigii 106 Thiocapsa roseopersicina 3, 6, 8, 102, 601, 613, 615, 667, 720, 745–749, 751, 753 Thiocapsa roseopersicina BBS 790 Thiocapsa roseopersicina strain BBS1 793 Thiococcus pfennigii 108 Thiocystis 8 Thiocystis gelatinosa 109 Thiocystis violacea 6, 108 Thiocystis violascens 108 thioether bond 408, 410 Thioflavicoccus mobilis 106 Thiohalocapsa 10 thiol:disulfide oxidoreductase 418, 555, 673 thiol modifying reagents 71 thiol modulation 477 Thiopedia 8 thioredoxin 613 thioredoxin reductase 669 thioredoxin TrxA 413 Thiorhodococcus drewsii 600 Thiorhodospira 600, 601 Thiorhodospira sibirica 600, 601 Thiorhodovibrio sibirica 600 Thiospirillum 8 Thiospirillum jenense 601 thiosulfate 596, 597, 600, 602, 612, 616 thiosulfate:cytochrome c oxidoreductase 599 thiosulfate dehydrogenase 612 thiosulfate oxidation 602, 605 thiouridine biosynthesis 609 thylakoid membranes 267 time-resolved fluorescence 814 time-resolved high-frequency ENDOR 963 delay after laser flash (DAF) 964
Index
1012 time–dependent infrared epifluorescence microscopy 49 time resolved EPR 955 Tll0078 818, 820 TMAO-reductase 547 TMAO/Trimethylamine couple 547 Tmp system 675 toluene 577, 578, 581 topographs 159–164, 256–265, 863–868, 943–950 TorC 548 torque 475 torsional spring constant 485 torsion angles 144 trans-cis isomerization 825 trans-sulfuration 613 transcription control 747–754 transcription-translation system 175, 269 transcriptional anti-repressor 819 transcriptional regulation 701–702, 777, 787, 795 transcriptional regulator 800 transcriptional regulator ArcA 754 transcriptional reporter 747 transcription factors 707 transcription terminators 842 transcriptomic analysis 589 transfer rates cluster-to-cluster 288 transhydrogenase 495–506 conformational changes 505 crystal structure 496 evolution 497 function 500–501 hydride transfer 501–503 mobile loop 504 phylogenetic tree 498–500 polypeptide composition 497 proton translocation 501 species distribution 496–498 X-ray structure 496, 501 transient absorption 217, 238, 239 transient absorption spectroscopy 804 transition dipoles 206 bacteriochlorophyll 168 transketolase (cbbTII) 566 translocation 407, 409, 410, 412, 414 transmembrane electro-chemical potential difference 232 transmembrane helix 137, 164, 201–202, 412, 513, 916, 918, 922 model 916, 921 prediction 918 reaction center 164 transmembrane segments 187 transmembrane voltage 484 transmembrane α-helices 914 transporter copper 420 transposon mutagenesis 65, 781 transposons 702 tree of life 19, 456, 457, 459 tributyltin chloride 478 tricarboxylic acid cycle 83 triheme reaction center subunits 515
trimethylamine-N-oxide (TMAO) 41, 546 3-triol 146 triplet bacteriochlorophyll 38 triplet formation 223 triplet state 357 Tris 120 trithionate hydrolase 599 Triton X-100 162, 182, 298, 869, 926, 928, 929 tRNAglutamate 778 trophic pyramid 50 tryptophan-rich WWD motif 411, 414 tryptophan fluorescence quenching 69 tubular membranes 171, 172, 256, 260, 263, 844 tuning position energy transition 205 tunneling matrix element 361 tunneling pathway 326, 327 twin arginine translocation system 439 specific signal sequence 439 two-component regulatory protein PrrA 782 two-component regulatory system 748–750 two-component response regulators 715 two-component signal transduction system 754 two-component system 800 two-dimensional electronic spectroscopy 227 two-dimensional nuclear magnetic resonance spectroscopy 584 two-photon absorption 221 two-photon excitation 220 type 1 histidine kinase 805 type II fatty acid synthase 122 Type I reaction center 380 Type II reaction center 380, 384, 385 tyrosyl radical 312, 349
U ubihydroquinone:cytochrome c oxidoreductase 425–447 ubihydroquinone oxidase 416 ubiquinone 44, 193, 258, 338, 383, 385, 387, 389, 392, 394, 397, 712, 713, 714, 754 ubiquinone-binding site 714 ubiquinone-ubiquinol exchange See quinone/quinol 256 ubiquinone diffusion 265 ubiquinone pool 712, 713, 714 UDP-sulfoquinovose sulfoquinovosyltransferase 121 UDP-sulfoquinovose synthase 121 UHDBT. See 5-undecyl-6-hydroxy-4,7-dioxobenzothiazole (UHDBT) ultrafast relaxation 206 5-undecyl-6-hydroxy-4,7-dioxobenzothiazole (UHDBT) 445 unitary proton conductance 482–484 unusual spirilloxanthin pathway 106, 110 upper pigmented band 147, 268 uptake hydrogenase hupSLC genes 753 ureDABCEFG 765 uroporphyrinogen III 60, 63, 81, 86, 778, 784, 794 uroporphyrinogen III decarboxylase 60, 64 bovine 64 human 64
Index uroporphryinogen III synthase 60, 63 crystal structure 64 Ustilago maydis 817
V van der Waals interactions 140 vectorial charge separation 446 vertebrates 410 vestibule for quinones 174 vibrational coherence 239, 240 vibrational modes 355, 367–368, 370 vibrational relaxation 227, 362, 370 vibrational spectroscopy 389 vibrational wavepackets 355, 362–363 Vibrio 702 vinyl-2 408 vinyl-4 408 3-vinyl bacteriochlorophyllide a 68, 74 3-vinyl bacteriochlorophyllide a hydroxylase 67 8-vinyl reductase 67, 72, 73 virulence 820 vise region 441 vitamin B12 57, 72, 81–92, 791 biosynthetic pathway 65 vnfHDK 761
W W-helix 166, 167, 169, 172, 513, 530, 948, 949 Walker A and B motifs 411, 748, 749 waste lagoon 9 water-soluble Chl protein 898, 899 water-soluble heme-binding protein maquettes 900 water-soluble protein maquettes 905 White Collar 1 813 Wolinella 669 Wolinella succinogenes 612
X X-ray absorption near-edge structure (XANES) 608
1013 X-ray crystallography 146, 202, 204, 286, 381, 383, 398, 941, 955 X-ray crystal structure 397, 954, 961 X-ray diffraction 963 X-ray diffraction spectroscopy 954 X-ray structure 383, 390, 391, 805 Blastochloris viridis 383, 386 Rhodobacter sphaeroides 385, 386 Thermochromatium tepidum 387 Xanthobacter flavus 566 xanthophyll 103 xanthopsin 811, 823–827 xenobiotic lipid 120
Y YcgF 820 yeast 409, 526 yeast extract 6 yfp 649 YidC 269
Z ζ-carotene 104 ζ-carotene pathway 104 zeaxanthin diglucoside 110 zero-phonon 235, 236, 361, 362 zero-phonon line (ZPL) 884 zinc 393, 396 Zn 419 binding motif 791 Zn-bacteriochlorophyll. See Zn-bacteriopheophorbide Zn-bacteriochlorophyll a 112 Zn-bacteriopheophorbide 902, 905 Zn-BChl 42, 45, 794 Zn-BChl a 303, 873 Zn-dependent enzymes 791 Zn-substitution 305