Liposomes. Methods and Protocols. Biological Membrane Models

Methods in Molecular Biology™ Series Editor John M. Walker School of Life Sciences University of Hertfordshire Hatfi...

Author: Volkmar Weissig

85 downloads 1249 Views 23MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Methods

in

Molecular Biology™

Series Editor John M. Walker School of Life Sciences University of Hertfordshire Hatfield, Hertfordshire, AL10 9AB, UK

For other titles published in this series, go to www.springer.com/series/7651

Liposomes Methods and Protocols Volume 2: Biological Membrane Models

Edited by

Volkmar Weissig Department of Pharmaceutical Sciences, Midwestern University College of Pharmacy Glendale, Glendale, AZ, USA

Editor Volkmar Weissig Department of Pharmaceutical Sciences Midwestern University College of Pharmacy Glendale Glendale, AZ USA [email protected]

ISSN 1064-3745 e-ISSN 1940-6029 ISBN 978-1-60761-446-3 e-ISBN 978-1-60761-447-0 DOI 10.1007/978-1-60761-447-0 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2009933261 © Humana Press, a part of Springer Science+Business Media, LLC 2010 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Humana Press, c/o Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. While the advice and information in this book are believed to be true and accurate at the date of going to press, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Cover illustration: Background art is derived from Figure 6c in Chapter 21 Printed on acid-free paper Humana Press is a part of Springer Science+Business Media (www.springer.com)

Preface Efforts to describe and model the molecular structure of biological membranes go back to the beginning of the last century. In 1917, Langmuir described membranes as a layer of lipids one molecule thick [1]. Eight years later, Gorter and Grendel concluded from their studies that “the phospholipid molecules that formed the cell membrane were arranged in two layers to form a lipid bilayer” [2]. Danielli and Robertson proposed, in 1935, a model in which the bilayer of lipids is sequestered between two monolayers of unfolded proteins [3], and the currently still accepted fluid mosaic model was proposed by Singer and Nicolson in 1972 [4]. Among those landmarks of biomembrane history, a serendipitous observation made by Alex Bangham during the early 1960s deserves undoubtedly a special place. His finding that exposure of dry phospholipids to an excess of water gives rise to lamellar structures [5] has opened versatile experimental access to studying the biophysics and biochemistry of biological phospholipid membranes. Although during the following 4 decades biological membrane models have grown in complexity and functionality [6], liposomes are, besides supported bilayers, membrane nanodiscs, and hybrid membranes, still an indisputably important tool for membrane biophysicists and biochemists. In vol. II of this book, the reader will find detailed methods for the use of liposomes in studying a variety of biochemical and biophysical membrane phenomena concomitant with chapters describing a great palette of state-of-the-art analytical technologies. Moreover, besides providing membrane biophysicists and biochemists with an immeasurably valuable experimental tool, Alex Bangham’s discovery has triggered the launch of an entirely new subdiscipline in pharmaceutical science and technology. His observation that the lamellar structures formed by phospholipids exposed to aqueous buffers are able to sequester small molecules has lead to the development of the colloidal drug delivery concept. Following initial studies of enzyme encapsulation in liposomes as an approach towards the treatment of storage diseases [7, 8], a few years later in two New England Journal of Medicine landmark papers, Gregory Gregoriadis outlined the huge carrier potential of liposomes in biology and medicine [9, 10]. The following 2 decades saw immense efforts in academia and in soon-to-be-founded start-up companies to turn Gregoriadis’ vision into clinical reality. These 20 years of intense work in liposome laboratories around the world finally culminated with the FDA (USA) approval of the first injectable liposomal drug, Doxil, in February of 1995. Today, liposomes present the prototype of all nanoscale drug delivery vectors currently under development. Lessons learned in the history of over 40 years of Liposome Technology should be heeded by new investigators in the emerging field of pharmaceutical and biomedical nanotechnology. Volume I of this book is dedicated to state-of-the-art aspects of developing liposome-based pharmaceutical nanocarriers. All chapters were written by leading experts in their particular fields, and I am extremely grateful to them for having spent parts of their valuable time to contribute to this book. It is my hope that together we have succeeded in providing an essential source of practical

v

vi

Preface

know-how for every investigator, young and seasoned ones alike, whose research area involves in one way or another phospholipids, glycolipids, and cholesterol. Last but not least, I would like to thank John Walker, the series editor of “Methods in Molecular Biology,” for having invited me to assemble this book and above all for his unlimited guidance and help throughout the whole process. Glendale, AZ

Volkmar Weissig

References 1. Bangham AD, Standish MM, Watkins JC (1965) Diffusion of univalent ions across the lamellae of swollen phospholipids. J Mol Biol 13(1):238–252 2. Chan YH, Boxer SG (2007) Model membrane systems and their applications. Curr Opin Chem Biol 11(6):581–587 3. Danielli JF, Davson H (1935) A contribution to the theory of permeability of thin films. J Cell Comp Physiol 5:495–508 4. Gorter E, Grendel F (1925) On bimolecular layers of lipoids on the chromocytes of the blood. J Exp Med 41:439–443 5. Gregoriadis G (1976) The carrier potential of liposomes in biology and medicine (second of two parts). N Engl J Med 295(14):765–770

6. Gregoriadis G (1976) The carrier potential of liposomes in biology and medicine (first of two parts). N Engl J Med 295(13):704–710 7. Gregoriadis G, Ryman BE (1971) Liposomes as carriers of enzymes or drugs: a new approach to the treatment of storage diseases. Biochem J 124(5):58P 8. Gregoriadis G, Leathwood PD, Ryman BE (1971) Enzyme entrapment in liposomes, FEBS Lett 14(2):95–99 9. Langmuir I (1917) The constitution and structural properties of solids and liquids. II. Liquids. J Am Chem Soc 39:1848–1906 10. Singer SJ, Nicolson GL (1972) The fluid mosaic model of the structure of cell membranes. Science 175(23):720–731

Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

v xi

1 Utilization of Liposomes for Studying Drug Transfer and Uptake . . . . . . . . . . . . Alfred Fahr and Xiangli Liu 2 The Use of Liposomes in the Study of Drug Metabolism: A Method to Incorporate the Enzymes of the Cytochrome P450 Monooxygenase System into Phospholipid, Bilayer Vesicles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . James R. Reed 3 Use of Liposomes to Study Cellular Osmosensors . . . . . . . . . . . . . . . . . . . . . . . . Reinhard Krämer, Sascha Nicklisch, and Vera Ott 4 Studying Mechanosensitive Ion Channels Using Liposomes . . . . . . . . . . . . . . . . . Boris Martinac, Paul R. Rohde, Andrew R. Battle, Evgeny Petrov, Prithwish Pal, Alexander Fook Weng Foo, Valeria Vásquez, Thuan Huynh, and Anna Kloda 5 Studying Amino Acid Transport Using Liposomes . . . . . . . . . . . . . . . . . . . . . . . . Cesare Indiveri 6 Use of Liposomes for Studying Interactions of Soluble Proteins with Cellular Membranes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chris T. Höfer, Andreas Herrmann, and Peter Müller 7 Liposomal Reconstitution of Monotopic Integral Membrane Proteins . . . . . . . . . Zahra MirAfzali and David L. DeWitt 8 The Reconstitution of Actin Polymerization on Liposomes . . . . . . . . . . . . . . . . . Mark Stamnes and Weidong Xu 9 Electroformation of Giant Unilamellar Vesicles from Native Membranes and Organic Lipid Mixtures for the Study of Lipid Domains under Physiological Ionic-Strength Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . L.-Ruth Montes, Hasna Ahyayauch, Maitane Ibarguren, Jesus Sot, Alicia Alonso, Luis A. Bagatolli, and Felix M. Goñi 10 Visualization of Lipid Domain-Specific Protein Sorting in Giant Unilamellar Vesicles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Martin Stöckl, Jörg Nikolaus, and Andreas Herrmann 11 Biosynthesis of Proteins Inside Liposomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pasquale Stano, Yutetsu Kuruma, Tereza Pereira de Souza, and Pier Luigi Luisi 12 Study of Respiratory Cytochromes in Liposomes . . . . . . . . . . . . . . . . . . . . . . . . . Iseli L. Nantes, Cintia Kawai, Felipe S. Pessoto, and Katia C.U. Mugnol 13 Use of Liposomes to Evaluate the Role of Membrane Interactions on Antioxidant Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Salette Reis, Marlene Lúcio, Marcela Segundo, and José L.F.C. Lima

1

vii

11 21 31

55

69 83 95

105

115 127

147

167

viii

Contents

14 Studying Colloidal Aggregation Using Liposomes . . . . . . . . . . . . . . . . . . . . . . . . Juan Sabín, Gerardo Prieto, and Félix Sarmiento 15 Assessment of Liposome–Cell Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jan A.A.M. Kamps 16 Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nejat Düzgünes¸, Henrique Faneca, and Maria C. Pedroso de Lima 17 The Use of Isothermal Titration Calorimetry to Study Multidrug Transport Proteins in Liposomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . David Miller and Paula J. Booth 18 Studying Lipid Organization in Biological Membranes Using Liposomes and EPR Spin Labeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Witold K. Subczynski, Marija Raguz, and Justyna Widomska 19 Membrane Translocation Assayed by Fluorescence Spectroscopy . . . . . . . . . . . . . J. Broecker and S. Keller 20 Interaction of Lipids and Ligands with Nicotinic Acetylcholine Receptor Vesicles Assessed by Electron Paramagnetic Resonance Spectroscopy . . . . . . . . . . Hugo Rubén Arias 21 Environmental Scanning Electron Microscope Imaging of Vesicle Systems . . . . . . Yvonne Perrie, Habib Ali, Daniel J. Kirby, Afzal U.R. Mohammed, Sarah E. McNeil, and Anil Vangala 22 Freeze-Fracture Electron Microscopy on Domains in Lipid Mono- and Bilayer on Nano-Resolution Scale . . . . . . . . . . . . . . . . . . . . . . . . . . . Brigitte Papahadjopoulos-Sternberg 23 Atomic Force Microscopy for the Characterization of Proteoliposomes . . . . . . . . Johannes Sitterberg, Maria Manuela Gaspar, Carsten Ehrhardt, and Udo Bakowsky 24 Method of Simultaneous Analysis of Liposome Components Using HPTLC/FID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sophia Hatziantoniou and Costas Demetzos 25 Viscometric Analysis of DNA-Lipid Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . Sadao Hirota and Nejat Düzgünes¸ 26 Fluorometric Analysis of Individual Cationic Lipid-DNA Complexes . . . . . . . . . . Edwin Pozharski 27 Fluorescence Resonance Energy Transfer-Based Analysis of Lipoplexes . . . . . . . . . Edwin Pozharski 28 Analysis of Lipoplex Structure and Lipid Phase Changes . . . . . . . . . . . . . . . . . . . Rumiana Koynova 29 Fluorescence Methods for Evaluating Lipoplex-Mediated Gene Delivery . . . . . . . Henrique Faneca, Nejat Düzgünes‚ , and Maria C. Pedroso de Lima 30 FRET Imaging of Cells Transfected with siRNA/Liposome Complexes . . . . . . . . Il-Han Kim, Anne Järve, Markus Hirsch, Roger Fischer, Michael F. Trendelenburg, Ulrich Massing, Karl Rohr, and Mark Helm 31 Spectral Bio-Imaging and Confocal Imaging of the Intracellular Distribution of Lipoplexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sebastian Schneider and Regine Süss

189 199

209

233

247 271

291 319

333 351

363 369 385 393 399 425 439

457

Contents

32 Techniques for Loading Technetium-99m and Rhenium-186/188 Radionuclides into Pre-formed Liposomes for Diagnostic Imaging and Radionuclide Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Beth Goins, Ande Bao, and William T. Phillips 33 Fluorescence Correlation Spectroscopy for the Study of Membrane Dynamics and Organization in Giant Unilamellar Vesicles . . . . . . . . . . . . . . . . . . Ana J. García-Sáez, Dolores C. Carrer, and Petra Schwille 34 Liposome Biodistribution via Europium Complexes . . . . . . . . . . . . . . . . . . . . . . . Nathalie Mignet and Daniel Scherman 35 Biosensor-Based Evaluation of Liposomal Binding Behavior . . . . . . . . . . . . . . . . . Gerd Bendas 36 Use of Liposomes to Study Vesicular Transport . . . . . . . . . . . . . . . . . . . . . . . . . . Kohji Takei, Hiroshi Yamada, and Tadashi Abe Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

ix

469

493 509 519 531 543

Contributors Tadashi Abe • Department of Neuroscience, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama, Japan Hasna Ahyayauch • Unidad de Biofisica (CSIC-UPV/EHU), Leioa, Spain Habib Ali • School of Life and Health Sciences, Aston University, Birmingham, UK Alicia Alonso • Unidad de Biofisica (CSIC-UPV/EHU), Leioa, Spain Hugo Rubén Arias • Department of Pharmaceutical Sciences, College of Pharmacy, Midwestern University, Glendale, AZ, USA Luis A. Bagatolli • Unidad de Biofisica (CSIC-UPV/EHU), Leioa, Spain Udo Bakowsky • Department of Pharmaceutical Technology and Biopharmacy, Philipps-Universität Marburg, Marburg, Germany Ande Bao • Department of Radiology, University of Texas Health Science Center, San Antonio, TX, USA Andrew R. Battle • Molecular Biophysics Laboratory, School of Biomedical Sciences and Institute for Molecular Bioscience, The University of Queensland, Brisbane, QLD, Australia Gerd Bendas • Department of Pharmacy, Rheinische Friedrich Wilhelms University Bonn, Bonn, Germany Paula J. Booth • Department of Biochemistry, University of Bristol, Bristol, UK Jana Broecker • Leibniz Institute of Molecular Pharmacology FMP, Berlin, Germany Dolores C. Carrer • BIOTEC, Technische Universität Dresden, Dresden, Germany Costas Demetzos • Department of Pharmaceutical Technology, School of Pharmacy, University of Athens, Athens, Greece David L. DeWitt • Department of Biochemistry and Molecular Biology, Michigan State University, East Lansing, MI, USA Nejat Düzgünes¸ • Department of Microbiology, Arthur A. Dugoni School of Dentistry, University of the Pacific, San Francisco, CA, USA Carsten Ehrhardt • School of Pharmacy and Pharmaceutical Sciences, University of Dublin, Trinity College Dublin, Dublin, Ireland Alfred Fahr • Department of Pharmaceutics, Friedrich-Schiller-University, Jena, Germany Henrique Faneca • Faculty of Science and Technology, Center for Neuroscience and Cell Biology, University of Coimbra, Coimbra, Portugal Roger Fischer • German Cancer Research Center (DKFZ), Heidelberg, Germany Alexander Fook Weng Foo • Molecular Biophysics Laboratory, School of Biomedical Sciences and Institute for Molecular Bioscience, The University of Queensland, Brisbane, QLD, Australia Ana J. García-Sáez • BIOTEC, Technische Universität Dresden, Dresden, Germany Maria Manuela Gaspar • Unidade Novas Formas de Agentes Bioactivos, iMed, Faculdade de Farmácia, Universidade de Lisboa, Lisboa, Portugal

xi

xii

Contributors

Beth Goins • Department of Radiology, University of Texas Health Science Center, San Antonio, TX, USA Felix M. Goñi • Unidad de Biofisica (CSIC-UPV/EHU), Leioa, Spain Sophia Hatziantoniou • Department of Pharmaceutical Technology, School of Pharmacy, University of Athens, Athens, Greece Mark Helm • Department of Chemistry, Institute of Pharmacy and Molecular Biotechnology, University of Heidelberg, Heidelberg, Germany Andreas Herrmann • Mathematisch-Naturwissenschaftliche Fakultät I, Institut für Biologie/Biophysik, Humboldt Universität zu Berlin, Berlin, Germany Sadao Hirota • Department of Material Science, School of Engineering, Tokyo Denki University, Tokyo, Japan Markus Hirsch • Department of Chemistry, Institute of Pharmacy and Molecular Biotechnology, University of Heidelberg, Heidelberg, Germany Chris Höfer • Institut für Biologie/Biophysik, Humboldt Universität zu Berlin, Berlin, Germany Thuan Huynh • Molecular Biophysics Laboratory, School of Biomedical Sciences, The University of Queensland, Brisbane, QLD, Australia Maitane Ibarguren • Unidad de Biofisica (CSIC-UPV/EHU), Leioa, Spain Cesare Indiveri • Dipartimento di Biologia Cellulare, Università della Calabria, Arcavacata di Rende, CS, Italy Anne Järve • Department of Chemistry, Institute of Pharmacy and Molecular Biotechnology, University of Heidelberg, Heidelberg, Germany Jan A.A.M. Kamps • Laboratory for Endothelial Biomedicine & Vascular Drug Targeting Research, Medical Biology Section, Department Pathology & Medical Biology, University Medical Center Groningen, Groningen, The Netherlands Cintia Kawai • Centro Interdisciplinar de Investigação Bioquímica CIIB, Universidade de Mogi das Cruzes, Mogi das Cruzes, S.P., Brazil Sandro Keller • Leibniz Institute of Molecular Pharmacology FMP, Berlin, Germany Il-Han Kim • Department of Bioinformatics and Functional Genomics, German Cancer Research Center (DKFZ), Institute of Pharmacy and Molecular Biotechnology, University of Heidelberg, Heidelberg, Germany Daniel J. Kirby • School of Life and Health Sciences, Aston University, Birmingham, UK Anna Kloda • Molecular Biophysics Laboratory, School of Biomedical Sciences, The University of Queensland, Brisbane, QLD, Australia Rumiana Koynova • Northwestern University, Evanston, IL, USA Reinhard Krämer • Institute of Biochemistry, University of Cologne, Cologne, Germany Yutetsu Kuruma • “Enrico Fermi” Study and Research Center, Rome, Italy José L.F.C. Lima • REQUIMTE, Faculdade de Farmácia, Universidade do Porto, Porto, Portugal Xiangli Liu • Department of Pharmaceutics, Friedrich-Schiller-University, Jena, Germany Marlene Lúcio • REQUIMTE, Faculdade de Farmácia, Universidade do Porto, Porto, Portugal Pier Luigi Luisi • Biology Department, University of RomaTre, Rome, Italy

Contributors

xiii

Boris Martinac • Molecular Biophysics Laboratory, School of Biomedical Sciences, The University of Queensland, Brisbane, QLD, Australia Ulrich Massing • Department of Clinical Research, Tumor Biology Center, Freiburg, Germany Sarah E. McNeil • School of Life and Health Sciences, Aston University, Birmingham, UK Nathalie Mignet • Unité de Pharmacologie Chimique et Génétique; CNRS, UMR 8151, Paris, France; Inserm, U 640, Paris, France; Faculté des Sciences Pharmaceutiques et Biologiques, Université Paris Descartes, Paris, France; ENSCP, Paris, France David Miller • Department of Biochemistry, University of Bristol, Bristol, UK Zahra MirAfzali • Encapsula NanoSciences LLC, 441 Donelson, Pike, Suite 345, Nashville, TN 37214, USA Afzal U. R. Mohammed • School of Life and Health Sciences, Aston University, Birmingham, UK L.-Ruth Montes • Unidad de Biofisica (CSIC-UPV/EHU), Leioa, Spain Katia C.U. Mugnol • Centro Interdisciplinar de Investigação Bioquímica CIIB, Universidade de Mogi das Cruzes, Mogi das Cruzes, S.P., Brazil Peter Müller • Institut für Biologie/Biophysik, Humboldt Universität zu Berlin, Berlin, Germany Iseli L. Nantes • Centro Interdisciplinar de Investigação Bioquímica CIIB, Universidade de Mogi das Cruzes, Mogi das Cruzes, S.P., Brazil Sascha Nicklisch • Institute of Biochemistry, University of Cologne, Cologne, Germany Jörg Nikolaus • Mathematisch-Naturwissenschaftliche Fakultät I, Institut für Biologie/Biophysik, Humboldt-Universität zu Berlin, Berlin, Germany Vera Ott • Institute of Biochemistry, University of Cologne, Cologne, Germany Prithwish Pal • Molecular Biophysics Laboratory, School of Biomedical Sciences, The University of Queensland, Brisbane, QLD, Australia Brigitte Papahadjopoulos-Sternberg • NanoAnalytical Laboratory, San Francisco, CA, USA Maria C. Pedroso de Lima • Department of Biochemistry, Faculty of Science and Technology, Center for Neuroscience and Cell Biology, University of Coimbra, Coimbra, Portugal Tereza Pereira de Souza • Biology Department, University of RomaTre, Rome, Italy Yvonne Perrie • School of Life and Health Sciences, Aston University, Birmingham, UK Regine Süss • Department of Pharmaceutical Technology and Biopharmacy, AlbertLudwigs University, Freiburg, Germany Felipe S. Pessoto • Centro Interdisciplinar de Investigação Bioquímica CIIB, Universidade de Mogi das Cruzes, Mogi das Cruzes, S.P., Brazil Evgeny Petrov • Molecular Biophysics Laboratory, School of Biomedical Sciences, The University of Queensland, Brisbane, QLD, Australia William T. Phillips • Department of Radiology, University of Texas Health Science Center, San Antonio, TX, USA Edwin Pozharski • Department of Pharmaceutical Sciences, University of Maryland, Baltimore, MD, USA

xiv

Contributors

Gerardo Prieto • Biophysics and Interfaces Group, Department of Applied Physics, Faculty of Physics, University of Santiago de Compostela, Santiago de Compostela, Spain Marija Raguz • Department of Biophysics, Medical College of Wisconsin, Milwaukee, WI, USA James R. Reed • Department of Pharmacology, Louisiana State University Health Science Center, New Orleans, LA, USA Salette Reis • REQUIMTE, Faculdade de Farmácia, Universidade do Porto, Porto, Portugal Paul R. Rohde • Molecular Biophysics Laboratory, School of Biomedical Sciences, The University of Queensland, Brisbane, QLD, Australia Karl Rohr • Department of Bioinformatics and Functional Genomics, German Cancer Research Center (DKFZ), Institute of Pharmacy and Molecular Biotechnology, University of Heidelberg, Heidelberg, Germany Juan Sabín • Biophysics and Interfaces Group, Department of Applied Physics, Faculty of Physics, University of Santiago de Compostela, Santiago de Compostela, Spain Félix Sarmiento • Biophysics and Interfaces Group, Department of Applied Physics, Faculty of Physics, University of Santiago de Compostela, Santiago de Compostela, Spain Daniel Scherman • Unité de Pharmacologie Chimique et Génétique; CNRS, UMR 8151, Paris, France; Inserm, U 640, Paris, France; Faculté des Sciences Pharmaceutiques et Biologiques, Université Paris Descartes, Paris, France; ENSCP, Paris, France Sebastian Schneider • Department of Pharmaceutical Technology and Biopharmacy, Albert-Ludwigs University, Freiburg, Germany Petra Schwille • BIOTEC, Technische Universität Dresden, Dresden, Germany Marcela Segundo • REQUIMTE, Faculdade de Farmácia, Universidade do Porto, Porto, Portugal Johannes Sitterberg • Department of Pharmaceutical Technology and Biopharmacy, Philipps-Universität Marburg, Marburg, Germany Jesus Sot • Unidad de Biofisica (CSIC-UPV/EHU), Leioa, Spain Mark Stamnes • Department of Molecular Physiology & Biophysics, Roy J. and Lucille A. Carver College of Medicine, University of Iowa, Iowa City, IA, USA Pasquale Stano • Biology Department, University of RomaTre, Rome, Italy Martin Stöckl • Mathematisch-Naturwissenschaftliche Fakultät I, Institut für Biologie/Biophysik, Humboldt-Universität zu Berlin, Berlin, Germany Witold K. Subczynski • Department of Biophysics, Medical College of Wisconsin, Milwaukee, WI, USA Kohji Takei • Department of Neuroscience, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama, Japan Michael F. Trendelenburg • German Cancer Research Center (DKFZ), Heidelberg, Germany

Contributors

xv

Anil Vangala • School of Pharmacy and Chemistry, Kingston University, London, UK Valeria Vásquez • Biochemistry Department, Gordon Center for Integrative Science, The University of Chicago, Chicago, IL, USA Justyna Widomska • Department of Plant Physiology and Biochemistry, Faculty of Biochemistry, Biophysics and Biotechnology, Jagiellonian University, Krakow, Poland Weidong Xu • Department of Molecular Physiology & Biophysics, Roy J. and Lucille A. Carver College of Medicine, University of Iowa, Iowa City, IA, USA Hiroshi Yamada • Department of Neuroscience, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama, Japan

Chapter 1 Utilization of Liposomes for Studying Drug Transfer and Uptake Alfred Fahr and Xiangli Liu Abstract On entry into the body of the patient, drugs have to overcome many barriers in order to reach the target. The knowledge of the ability of drugs to cross these barriers, which mostly consist of lipid membranes, is of utmost interest in pharmacy. High values of lipophilicity of a drug might be a good pre-requisite for crossing these barriers. It also led liposomologists to think that highly lipophilic drugs may “stick” in the lipophilic interior of liposomal phospholipid membranes and therefore these liposomes may act as a retard formulation of the lipophilic drug. The presented method here estimates the transfer time of lipophilic drugs between liposomal lipid bilayers. This may help to judge the presumed retardation function of a specific liposomal delivery system for a chosen lipophilic drug. Key words: Membrane transfer, Liposome drug delivery system, Lipophilic drug, Retardation, Mini column method

1. Introduction In the pharmaceutical field, liposomes are commonly known as drug carrier systems, but appeared at first in the scientific community as models of biological membranes (1). Similar to the early use in membrane biophysics, liposomes can also be very useful in pharmaceutical sciences, as the passive diffusion of drugs through physiological barrier membranes (e.g. epithelial cells in the gut) can be estimated with their help by different kinds of experimental setups (2). Another drug-delivery related problem can also be assessed by applying membrane biophysics methods to pharmaceutical V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_1, © Humana Press, a part of Springer Science+Business Media, LLC 2010

1

2

Fahr and Liu

liposomology: the often not verified, but very often cited assumption that a lipophilic drug should bind to a high degree to the lipophilic interior of the liposomal bilayer simply because of its similarities in lipophilicity. Rather, as we had found for several lipophilic compounds, these compounds are transferred quite rapidly to other lipophilic binding places (3), which can also be reasoned by theoretical considerations (4). Methods like dialysis (5) have their value for estimating the thermodynamics of the transfer processes, but for analysing the kinetics in the minute time-range with absolute mass transfer values, other methods should be used. Not for all drugs the elegant method of spin-label measurements (6) can be used to get an estimate of drug retardation. Here we provide a protocol for measuring the transfer of lipophilic drugs between liposomal membranes. The drug is transferred during a mixing process from donor liposomes to acceptor liposomes; the donor liposomes are removed from the sample using a micro ion exchange column method. We describe this method under the assumption, that the drugs are available with a radioactive marker. This very comfortable situation regarding the transfer analysis is typically present in big pharmaceutical companies with their own radiochemistry labs ready for labelling the drugs of interest. A chromophoric group in the drug moiety helps as well in making the analysis simple, especially if this leads to a fluorescence drug. In the author’s experience, a sensitive HPLC-method can also be a sufficient method for analysis, but it may be time-consuming for reliable results. The obtained data can in most cases be easily analysed. By incorporating an un-exchangeable marker, one can not only determine the retention of donor liposomes on the column, but can also validate the whole method.

2. Materials 2.1. Substances

1. Egg phosphatidylcholine (EPC) (Lipoid KG, Ludwigshafen, Germany). Synthetic phosphatidylcholines (e.g. from Avanti Polar Lipids, Alabaster, AL, USA.) are highly recommended. 2. Dicetylphosphate (DCP) (Sigma-Aldrich, Taufkirchen, Germany). 3. Cholesterol and 1-palmitoyl-1,2-oleoyl phosphatidylcholine (POPC) (Genzyme, Liestal, Switzerland or Avanti Polar Lipids). 4. 14C-labelled cholesteryl-oleoyl-ether and 3H-labelled cholesteryl-oleate (GE Healthcare UK Ltd, Buckinghamshire, UK). The lipids were stored in –20°C before use. Typically small aliquots of a delivered batch are stored in the freezer to avoid degradation caused by repeated thaw and freeze.

Utilization of Liposomes for Studying Drug Transfer and Uptake

3

5. The 14C labelled cyclosporin A (CyA) was synthesised and kindly provided by the radiochemistry group of Sandoz Pharm Ltd., now Novartis Inc. (Basel, Switzerland). 1–5 are either bought as chloroformic solution or the obtained powder is dissolved in chloroform in order to make stock solutions. These stock solutions should be kept at −20 to −80°C and should not be used after three months of storage. Always allow the frozen stock solution to reach ambient temperature in order to avoid water contamination of the solution which leads to lipid degradation. 6. DEAE Sepharose CL-6B (GE Healthcare Bio-Sciences AB, Uppsala, Sweden) is supplied pre-swollen in 20% ethanol. 7. Tris, saccharose, sodium chloride, and sodium azide (all in the highest purity available) (e.g. Sigma-Aldrich, Taufkirchen, Germany). 8. The two buffers used throughout the experiments are named Buffer A: 145 mM NaCl, 10 mM Tris pH 7.4, and Buffer B: 290 mM Saccharose, 10 mM Trizma pH 7.4, 0.02 % sodium azide. 2.2. Devices

1. Minicolumns were made in the workshop of the University of Perspex® and should preferably have the dimensions as indicated in Fig. 1. 2. All other devices should be available in a standard laboratory for liposomal research and are mentioned in the following text for reference.

3. Methods 3.1. Preparation of Donor Liposomes

1. Egg phosphatidylcholine (EPC), Dicetylphosphate (DCP), and cholesterol are weighed into individual test tubes and dissolved in chloroform to get the concentrations 20 mg/ml for EPC, 10 mg/ml for DCP, and 10 mg/ml for cholesterol. Appropriate volumes of each lipid with a mole ratio of each lipid component 7:1:2 (EPC:DCP:Cholesterol = 7:1:2) and 14 C-CyA (phospholipid:CyA = 300:1) are transferred to a single tube (see Note 1). 2. The mixture is evaporated using a Rotavapor device rotating at slow speed and heated (30°C) and is then subjected to vacuum to form a thin layer of lipid on the wall of the flask (see Note 2). This cyclosporin-lipid layer is hydrated with the appropriate amount of Buffer A (to get a phospholipid concentration of 10 mg/ml) for 2 h using the same Rotavapor

4

Fahr and Liu

Fig. 1. Schematic drawing of the device for measuring the transfer of substances between liposomes by the microcolumn technique

device after degassing the buffer and purging it with nitrogen gas. The final milky suspension is vortexed for 30 s at moderate speed, until all lipids have been removed from the glass vessel. This can be easily checked by optical inspection of the glass vessel from the outside. 3. This suspension is extruded 21 times (or more, but always use an uneven number!) through polycarbonate membranes with pore size 100 nm using a commercially available extruder

Utilization of Liposomes for Studying Drug Transfer and Uptake

5

(LiposoFast, Avestin Inc., Canada) in order to get homogenously sized liposomes. 4. The resulting liposomes are assessed for size distribution and zeta-potential and stored at +4°C before use. Liposomes formed by this method typically display a diameter range of 110–130 nm and a zeta-potential of −40 to −50 mV. 3.2. Preparation of Donor Blank Liposomes

1. Donor blank liposomes (without any transferrable drug included) are prepared using the same procedures as in Subheading 3.1. 2. The donor blank liposomes are labelled by 14C-cholesteryloleoyl-ether at 1 mCi/ml (see Note 3) as a non-exchangeable lipid marker.

3.3. Preparation of Acceptor Liposomes

1. 1-palmitoyl-1,2-oleoyl phosphatidylcholine (POPC) and cholesterol are weighed into individual test tubes and dissolved in chloroform to get the concentrations 20 mg/ml for POPC and 10 mg/ml for cholesterol. Appropriate volumes of each lipid with a mole ratio of the two lipids (POPC:cholesterol = 8:2) are transferred to a single tube. 3H-labelled cholesteryl-oleate is incorporated into the formulation at 1 mCi/ml. 2. The mixture is evaporated, hydrated with an appropriate amount of buffer A (to get a phospholipid concentration of 10 mg/ml), and extruded following the procedures as described in Subheading 3.1 to get the homogenously sized liposomal suspension. 3. Liposomes display a diameter range of 130–150 nm and zetapotential of ~0 mV.

3.4. Preparation of Liposomes for Saturation

1. 1-palmitoyl-1,2-oleoyl phosphatidylcholine (POPC), and cholesterol are weighed and mixed at a mole ratio of 8:2 following the procedures as described in Subheading 3.3. 2. The mixture is evaporated and hydrated (to get a phospholipid concentration of 10 mg/ml) following the procedures as described in Subheading 3.1 to get the liposomal suspension. 3. 2 ml of the liposomal suspension is transferred to a glass vial. This vial is mounted in a MSE-Soniprep 150 (Zivy AG, Oberwil, Switzerland) ultrasonic disintegrator with a titanium probe. The tip of the titanium probe is positioned 4 mm below the surface of the liposomal suspension. The vial is kept in an ice/water bath during the sonication, and nitrogen gas is used for purging the suspension to avoid oxidation of lipids during sonication. 4. The suspension is sonicated for 60 min (2 times 30 cycles, 1 cycle = 30 s sonicating and 30 s non-sonicating) with a 50% duty cycle using a MSE process timer (Zivy AG, Oberwil, Switzerland). The amplitude of the titanium probe is adjusted to 12 mm. The temperature of the suspension is checked to ensure that it does not rise significantly during sonication.

6

Fahr and Liu

5. This method allows the preparation of very small liposomes by introducing high amounts of energy into the phospholipid suspension. The average particle size of liposomes by sonication method is around 30–60 nm. 3.5. Size Determination of Liposomes

1. The size of the liposomes is determined by Dynamic laser light scattering using a Zetasizer Nano ZES3600 (Malvern Instruments Inc., UK). 2. Data are analysed by the Dispersion Technology software version 5.02 (Malvern Instruments Inc., UK).

3.6. Preparation of the Column Filling Gel DEAE-Sepharose CL-6B

1. The necessary amount of the DEAE-Sepharose CL-6B (depending on the number of mini-columns to be filled) is poured in an Erlenmeyer flask of appropriate volume (usual 200 mL size). 2. The gel settles down and the ethanol is carefully and slowly decanted by means of a Pasteur pipette connected via a tube to a water jet vacuum. 3. The gel is washed twice with buffer A in a ratio of 75% settled gel to 25% buffer A and kept in a last washing step with buffer B (1:1, v/v) at 4°C (see Note 4).

3.7. Preparation of the Ion-Exchange Microcolumn

1. The ion-exchange microcolumns were manufactured from Perspex® in the mechanical workshop of the university. The total bed volume of the microcolumn is about 0.5 ml (not very critical). 2. Some glass wool was placed at the bottom of the inner side of the column in order to ease the retaining of the gel in the column. 1.0 ml DEAE-sepharose CL-6B suspension pretreated as described in Subheading 3.6 is filled in the column (see Note 5). 3. The column is eluted with 2 ml buffer B at a speed of 30 drops in 45 s by using a pump (for example LKB Pump P-1, Pharmacia (Uppsala, Sweden)) and packed at the same time. 4. The microcolumns are saturated by applying 20 ml of the saturation liposomes prepared as described in Subheading 3.4 and eluted with 1.5 ml buffer B (see Note 6).

3.8. Validation of the Separation Efficiency of the Microcolumns 3.8.1. Donor Blank Liposomes

1. 10 ml donor blank liposomes prepared as described in Subheading 3.2 are placed on the top surfaces of the saturated microcolumns prepared as described in Subheading 3.7 and eluted with 1.5 ml buffer B at a speed of 30 drops in 45 s by using a pump. 2. The quantity of donor liposomes in the eluate is measured by liquid scintillation counting and the capturing capacity of the microcolumns for donor liposomes is estimated.

Utilization of Liposomes for Studying Drug Transfer and Uptake

7

3. The experiment is repeated nine times by using the same columns to check the reproducibility of the experiments. 3.8.2. Acceptor Liposomes

1. 10 ml acceptor liposomes prepared as described in Subheading 3.3 are placed on the top surfaces of the saturated microcolumns prepared as described in Subheading 3.7. 2. The same procedures are performed as described in Subheading 3.8.1 to measure the recovery of the acceptor liposomes in the eluate by liquid scintillation counting and the reproducibility of the experiments.

3.8.3. Mixture of Donor and Acceptor Liposomes

1. Donor blank liposomes prepared as described in Subheading 3.2 and acceptor liposomes prepared as described in 3.3 are mixed in the ratio of 1:10 (v/v). 10 ml of the mixture is placed on the surface of the saturated microcolumns and is eluted with 1.5 ml buffer B (see Note 7). 2. The same procedures are performed as described in Subheading 3.8.1 to measure the capturing capacity of the microcolumns for donor liposomes and the recovery of the acceptor liposomes in the eluate and the reproducibility of the experiments.

3.9. Example: Cyclosporin A (CyA) Transfer Between Liposomes

1. Transfer of CyA between liposomal membranes was studied using the microcolumn prepared as described in Subheading 3.7. Donor liposomes prepared as described in Subheading 3.1 and acceptor liposomes prepared as described in Subheading 3.3 are mixed in a ratio of 1:10 (v/v) at a temperature of 37°C. 2. At certain time points, 10 ml of the mixture is placed on the top surface of the saturated microcolumns and eluted with 1.5 ml buffer B as shown in Fig. 1. 3. The CyA quantity in the eluates was measured by liquid scintillation counting. The transfer kinetics of CyA, an additional model drug (cholesterol) and 3H-cholesteryl-oleoyl-ether as non-transferrable liposome-marker is shown in Fig. 2.

3.10. Liquid Scintillation Counting

The quantity of liposomes and CyA is determined by liquid scintillation counting on a Liquid Scintillation Analyzer Tri-Carb 2800 TR (Perkin Elmer, Herrenberg, Germany).

3.11. Data Analysis

Data analysis is mostly done by applying a simple mono-exponential model to the obtained data: y = y0 × (1 – e–k × t). This describes the data in all simple transfer cases quite well and delivers an understandable value k, which can easily be converted into half times. Of course, this is only a phenomenological description of the transfer processes, but elaborate models studying the mechanisms involved in the transfer from one liposome to the other are out of the scope of this chapter.

8

Fahr and Liu

Fig. 2. Transfer of CyA between liposomes. Donor liposomes (EPC:DCP:Chol = 7:1:2) were loaded with 14C-CyA (phospholipid:CyA = 300:1) Acceptor liposomes, composed of POPC:Chol = 8:2, were mixed with donor liposomes in a ratio of 10:1. For comparison, 14C-cholesterol and 3H-cholesteryl-oleoyl-ether transfer was also studied. Curve fits were done using the equation described in Subheading 3.11

4. Notes 1. The lipids and drug used here for liposome preparation are only an example; different lipids and drugs can be used for liposome preparation for different purposes, as long as the donor liposomes are strongly negatively charged and the acceptor liposomes neutral. 2. The experimental conditions (e.g. temperature, organic solvent) depend on lipids and drugs used in the study. The vacuum should be applied gradually in order to form a homogeneous lipid layer. After formation of the lipid layer, it should be kept under vacuum for more time to remove the organic solvent as much as possible. A very good control is the experimenter’s nose. If you are suffering from a cold at the time of the experiment, ask a colleague. 3. Linear calibration between scintillation counting and concentration of the radiolabelled lipid marker should be done to choose the suitable concentration of the lipid marker incorporated into liposome. After successfully establishing the whole procedure, the amount of radioactive tracer used can

Utilization of Liposomes for Studying Drug Transfer and Uptake

9

be reduced to the absolute minimum necessary for counting. But do these measurements regularly, especially, if you have not done it for a while. 4. In the washing process, any strong or even violent stirring or shaking of the gel must be avoided. We are not exactly sure, why this influences the quality of the results, but it does! Just treat the gel as a sensitive flower at Valentine’s Day. 5. Great care should be taken to avoid air bubbles to be induced or included in the microcolumn, which will influence the experimental results by delivering erratic results. This can be easily optically checked through the Perspex® material used for making the microcolumns. 6. The optimal amount of saturation liposomes depends on the type of liposomes, so this amount should be checked in different cases. This can be done by applying small amounts of saturation liposomes to the column and observing the eluate by turbidity measurements. A steep increase in turbidity means, that sufficient saturation liposomes have been applied. The saturation of the minicolumns is necessary, as the resin (gel) contains many lipophilic binding places, which could adsorb complete or part of liposomes during elution. There will be a small loss of drug from the acceptor liposomes to the pre-saturated minicolumn gel, but as the donor liposomes are also adsorbed and stay in equilibrium with the acceptor liposomes, the total net transfer is negligible for this type of measurement for the chosen set-up. 7. The buffer used can be pre-cooled to 4°C in order to minimise transfer processes that might occur (see Note 6) during the elution.

Acknowledgment We thank Rene Schaufelberger (Novartis Pharma Inc., Basel) for excellent technical assistance in setting up these procedures.

References 1. Papahadjopoulos D, Bangham AD (1996) Biophysical properties of phospholipids. II. Permeability of phosphatidyl liquid crystal to univalent ions. Biochim Biophys Acta 126:185–188 2. Flaten GE et al (2006) Drug permeability across a phospholipid vesicle-based barrier 2. Characterization of barrier structure, storage stability and stability towards pH changes. Eur J Pharm Sci 28(4):336–43

3. Shabbits JA, Chiu GN, Mayer LD (2002) Development of an in vitro drug release assay that accurately predicts in vivo drug retention for liposome-based delivery systems. J Control Release 84(3):161–70 4. Fahr A et al (2005) Transfer of lipophilic drugs between liposomal membranes and biological interfaces: consequences for drug delivery. Eur J Pharm Sci 26:251–265

10

Fahr and Liu

5. Joguparthi V, Xiang TX, Anderson BD (2008) Liposome transport of hydrophobic drugs: gel phase lipid bilayer permeability and partitioning of the lactone form of a hydrophobic camptothecin, DB-67. J Pharm Sci 97(1):400–20

6. Bienveue A et al (1985) Kinetics of phospholipid transfer between liposomes (neutral or negatively charged) and high-density lipoproteins: a spin-label study of early events. Biochim Biophys Acta 835:557–566

Chapter 2 The Use of Liposomes in the Study of Drug Metabolism: A Method to Incorporate the Enzymes of the Cytochrome P450 Monooxygenase System into Phospholipid, Bilayer Vesicles James R. Reed Abstract Although lipids are essential for the optimal activity of the cytochromes P450 monooxygenase system, relatively little is known about the membrane environment in which these enzymes function. One approach used to mimic the structural arrangement of lipids and enzymes within the endoplasmic reticulum is to physically incorporate the cytochromes P450 and their redox partners in a vesicle bilayer of phospholipids. Several methods have been devised for this purpose. This chapter describes a method in which the P450 monooxygenase system is incorporated by first, solubilizing the enzymes and lipid with sodium glycocholate. After the protein and lipid aggregates are dispersed, the detergent is removed by adsorption using BioBeads SM-2 resin which leads to the formation of bilayer vesicles of phospholipid containing incorporated cytochrome P450 and NADPH cytochrome P450 reductase. This procedure requires relatively a short preparation time, provides concentrated reconstituted systems that can be used in a wide range of applications, allows for several enzyme samples to be prepared simultaneously so that different conditions can be compared, and results in minimal loss of active enzyme. Key words: Phospholipid vesicles, Cytochromes P450, Reconstituted systems, Drug metabolism

1. Introduction The cytochromes P450 (P450) represent a ubiquitous gene superfamily comprising a diversity of isoforms that are expressed virtually in every organism in a species-specific pattern and are responsible for most xenobiotic metabolism in vivo (1). Thus the cytochromes P450 play a key role in the oxidation and clearance of most drugs and, in some instances, bioactivate toxins and promutagens to reactive intermediates that bind to cellular macromolecules V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_2, © Humana Press, a part of Springer Science+Business Media, LLC 2010

11

12

Reed

which, in turn, may lead to toxicity and/or the initiation of carcinogenesis (2–5). Because of these attributes, metabolism by the cytochromes P450 is relevant to a range of interests, including those of pharmacologists, toxicologists, and cancer researchers. The monooxygenase reactions catalyzed by these enzymes require electrons that can be delivered by various redox partners (6). The NADPH-cytochrome P450 reductase (reductase) is the primary redox partner in vivo and is capable of delivering both electrons needed in the catalytic cycle of the P450. Thus, the minimum enzyme assemblage needed to study monooxygenase reactions by the P450 system includes the reductase and the P450 isoform of interest. Much of the early information regarding the substrate specificity of the individual isoforms has come from studies in which the metabolites are identified after incubating the compound of interest with the active P450 monooxygenase enzyme assemblage and a source of NADPH (7–9). The P450 enzymes involved in xenobiotic metabolism and their redox partners are embedded in the membrane of the endoplasmic reticulum. As discussed in more depth later on, the membrane environment is essential for the functioning of the P450 monooxygenase system. Early attempts, over 30 years ago, to purify and characterize the various isoforms were limited by the dependence of enzyme activity on an unknown, heat-stable factor that was lost during the purification of the proteins. It was later found that the heat-stable factor was microsomal lipid (8, 10). Thus, it was determined that a lipid milieu was essential for the reconstitution of the catalytic activity of the purified enzymes. Subsequent studies have proposed that the lipid serves both as a “scaffold” to properly orient the P450 enzyme and the reductase for functional interaction (11, 12) and an effector that influences catalysis by the P450 enzyme (13). The effector role of phospholipid is evidenced by the modulation of P450 enzymatic activity at lipid:P450 enzyme ratios that are too low to facilitate the formation of liposomes. The “scaffold function” of lipid is ascribed to an additional level of modulation of enzyme activity observed at lipid concentrations at which liposomes form. Most enzymatic studies with the purified P450 enzymes use a short-chain (C-12), non-physiologic lipid, dilaurylphosphatidylcholine (DLPC). The reasons for this choice are the following: (1) the ease of preparation of the reconstituted systems with this lipid and (2) the lipid stimulates metabolism with most of the commonly studied P450 enzymes (14). However, studies have shown that the enzyme-lipid assemblages in the reconstituted systems with DLPC bear little structural similarity to the monolamellar, bilayer arrangement of lipid in the endoplasmic reticulum (15–17). Thus, it is clear that in order to truly appreciate the significance of the “scaffold” effect of the lipid on P450 metabolism, the enzymes must be physically incorporated in a vesicular bilayer of lipid.

Use of Liposomes in the Study of Drug Metabolism

13

Several methods used to generate these P450 vesicular reconstituted systems (VRS) have been published (18–20). Unfortunately, the methods require detergent, and this makes the procedures considerably more difficult and time-consuming than those using DLPC. However, the methods have been applied extensively and have generated interesting findings regarding the scaffold effect of the lipid bilayer on P450 metabolism (21–25). The approach common to all the methods for the preparation of VRS involves the following sequence of steps: (a) drying of the lipid in order to remove organic solvents that would inactivate or denature the enzymes, (b) detergent treatment to solubilize and disperse the enzymes and the lipids, and (c) a detergent removal procedure which causes the lipids to coalesce and form lipid bilayer vesicles, and in the process, the enzymes are incorporated in the vesicles. The thermodynamics associated with the detergent solubilization of a membrane complicate the ability of the VRS preparation methods to study the relationship between P450 activity and lipid concentrations. As mentioned above, this is an important aspect when evaluating the scaffold effect of lipids on enzymatic activity. The complexity of membrane solubilization has been reviewed in detail previously (26). At low concentrations, detergent binds to lipid and partitions between the lipid and aqueous phases (27, 28). Thus, in the simplest terms at low concentrations, two populations of detergent can be assigned – that bound to lipid and that partitioned into solution as monomers. The ratio of detergent concentrations in the two forms is dependent on the partition coefficient of the detergent for the aqueous and lipid phases. The initial stage of lipid solubilization occurs when the detergent concentration is increased to a point at which the bilayers containing a mixture of detergent and lipid are lysed into mixed micelles (26). Furthermore, it has been shown that this process occurs when the monomeric detergent concentration, in equilibrium with that bound to lipid, approaches or reaches the critical micelle concentration (CMC) of the pure detergent in aqueous solution (26, 27). As the detergent concentration is increased above the threshold required for membrane lysis, the excess detergent binds primarily to the mixed micelles of lipid and detergent, and the monomeric detergent concentration remains relatively constant at the CMC of the detergent. In the process of increasing the detergent concentration above the lysis threshold, the sizes of these micelles are reduced. Upon complete solubilization of the lipid/enzyme assemblage, each component lipid and enzyme is contained in an individual micelle of detergent. Thus, the total concentration of detergent needed to lyse the membrane will depend on both the concentration of the lipid and the extent to which the detergent partitions to the lipid phase. In applying the VRS methods at different concentrations of lipid, it is necessary to first identify the “effective” detergent to

14

Reed

lipid ratio (which is the ratio of the lipid-bound detergent concentration to the total lipid concentration) under a given set of optimized conditions that result in a desired degree of lipid/enzyme solubilization. The effective detergent to lipid ratio can then be used to calculate the detergent concentration needed to achieve comparable degrees of solubilization at different lipid concentrations. If the total detergent concentration (and not the lipid-bound detergent concentration) is adjusted in proportion to the lipid concentration in the methods for VRS preparation, the final VRS will have an excess detergent concentration (because both the monomeric and the lipid bound detergent concentrations will be scaled up). This in turn, likely will result in a high proportion of inactive enzyme because of the destructive effects of the excess detergent. Alternatively, if the detergent concentration is not adjusted when the lipid concentration is increased dramatically, the lipid will not be sufficiently solubilized to allow for the vesicular incorporation of the enzymes. We have tested two of the published methods, cholate gel filtration and cholate dialysis (17) and have found the active P450 enzyme tends to be extremely labile to the detergent, whereas the reductase tends not to incorporate into the vesicular fraction of the enzyme/lipid assemblage. Furthermore, preparations made using gel filtration are diluted and can be prepared only one at a time, limiting the opportunity to compare different experimental conditions, whereas, the dialysis procedure used to remove detergent is labor intensive and extremely time-consuming (3 × 12 h incubations against 3 L of dialysis buffer). Because of the limitations associated with these common methods for VRS preparation, we developed an improved technique for the preparation of P450 VRS (29). In the course of this work, we found a detergent (sodium glycocholate) that was less destructive to the P450 enzyme and utilized a more rapid way to remove the detergent by adsorption to BioBeads SM2 resin. We found this method to be superior to the two most commonly applied methods because of the following characteristics: (1) samples prepared using this method were also found to contain a higher proportion of the vesicular-incorporated reductase, (2) the detergent removal step could be carried out much more rapidly than the dialysis (2 h vs. 3 days), (3) the generated VRS samples could be easily prepared at relatively high enzyme concentrations (³5 mM), and (4) several samples could be prepared simultaneously thus allowing for the comparison of different conditions used in the reconstitution of the enzymes. We found that the VRS prepared under these conditions had very high catalytic activity relative to the reconstituted systems made using sonicated DLPC (17). The later sections of this chapter describe in detail the VRS preparation method. In addition, this chapter shows an approach that can be used to estimate the

Use of Liposomes in the Study of Drug Metabolism

15

effective detergent to lipid ratio in order to obtain VRS preparations with comparable levels of enzyme incorporation and activity over a range of lipid concentrations. This VRS method can be used to address general questions about the scaffold effect of the lipid. However, this type of reconstituted system also provides a more favorable structural framework with which to study the interaction of P450 enzymes with the various redox partners. Furthermore, the potential for future interest in the methods to prepare VRS with reductase and P450 is tremendous given the fact that lipids are known to separate into functional domains identified as “rafts” in studies with the plasma membrane (30). Evidence is just starting to accumulate that functional, lipid microdomains may also be present in the endoplasmic reticulum (31) and thus, may be significant in regulating P450 metabolism.

2. Materials 1. 0.5 M Hepes (pH 7.5) 2. 1 M MgCl2 3. 10% (w/v) aqueous sodium glycocholate (Calbiochem La Jolla, CA); membrane phospholipid of choice in chloroform at a concentration of 10 mg/mL. We have routinely used phosphatidylcholine from bovine liver (Avanti Polar Lipids Alabaster, AL). The lipid is both light and air sensitive with a tendency to oxidation of unsaturated acyl chains. The chloroform solutions are stored at –20°C. 4. Concentrated enzyme stock solutions (>10°mM reductase or P450 enzyme, respectively) in 100°mM potassium phosphate (pH 7.4) with 20 % glycerol (enzyme stock solutions are frozen at −80°C). 5. BioBeads SM-2 (Bio-Rad Hercules, CA). 6. 5 mm syringe filter (GE Osmonics, Minnetonka, MN).

3. Methods The general method characterized and described previously (29) results in a 0.5 mL solution of VRS containing 5 mM each of reductase and P450 enzyme and a 500:1 ratio of lipid:P450 (see Note 1). An adaptation is described below the general method which allows for adjustments in the lipid:P450 ratio of the VRS.

16

Reed

3.1. Preparation of the BioBeads SM-2 for Detergent Removal

This needs to be done at least two days before preparing the VRS. The use of BioBeads SM-2 to remove detergent from biological preparations has been described in depth previously (32). 1. Add 200 mL methanol to 30 g of BioBeads SM-2. 2. Stir for 15 min and collect beads by filtration with a Buchner funnel fitted with standard filter paper. 3. Immediately wash beads with another 500 mL of methanol; then repeat filtration. 4. Immediately wash beads with 1000 mL of water and repeat filtration. 5. Transfer beads to a chromatographic column and slowly wash beads with 2000 mL of water. 6. The moist beads are stored in ultrapure water at 4°C until required. When used, the beads are filtered as described above and added to the preparation containing detergent, lipid, and enzyme (supernatant derived from Step 5 of Subheading 3.3). Beads have been stored up to 3 months, periodically changing the water without noticeable problems in preparing the VRS.

3.2. General Method for Preparation of VRS: Drying the Lipid

1. Dry 1 mg of phospholipid (from a 10 mg/mL solution in chloroform) in a 1.5 mL microfuge tube overnight in a lyophilizer (see Note 2). 2. Release the vacuum on the lyophilizer by filling the chamber with N2 (see Note 3). 3. Add 50 mL of 0.5 M of Hepes (pH 7.5) and 50 mL of 10% sodium glycocholate to the tube containing dried lipid (see Notes 4 and 5).

3.3. Detergent Solubilization of Lipid and enzyme

1. Blow nitrogen over the tube opening of the solution from step 3.2.3 before capping, then bath-sonicate, and periodically vortex the tube until the solution is clear (usually 5 min). 2. While the lipid is being solubilized (step 3.3.1.), add P450 enzyme and reductase (2.5 nmol of each) to a second 1.5 mL microfuge tube and dilute to 482.5 mL with ultrapure water. 3. Add 7.5 mL of 1 M MgCl2 to the tube containing the P450 enzyme and the reductase. 4. Add the solution derived from step B.1 in approximate aliquots of 25 mL (approximately 1/4th the total volume of the solution) to the mixture of the P450 enzyme and the reductase (tube from step 3.3.3.). (See Note 6). Nitrogen is blown over the tube after the addition of each aliquot, and then the tube is capped and gently inverted to mix the enzymes with the detergent/lipid. 5. After the final addition of the solution from step B.1, blow nitrogen over the tube opening, cap the tube, invert several times, and incubate at 4°C for 1 h.

Use of Liposomes in the Study of Drug Metabolism

3.4. Removal of Detergent

17

1. Add 0.25 g of BioBeads SM-2 to the solution from step 5 of Subheading 3.3, blow nitrogen over the tube opening, cap the tube, and rock at 4°C for 2 h. 2. Draw off the solution from the Bio-Beads with a 26.5 gauge needle and filter through a 5 µm syringe filter. 3. The beads are rinsed twice with 0.2 mL aliquots of 50 mM Hepes (pH 7.4) and 15 mM MgCl2. These bead washes are filtered and added to the original bead filtrate. The final filtrate (containing the original filtrate in addition to the two bead washes) is typically around 0.65 mL in volume. 4. The sample is ready for use. An aliquot (0.1 mL) of the solution is routinely taken to determine the recovery of active P450 by determining the amount of enzyme capable of forming a ferrous CO-complex (33). A second 0.1–0.2 mL aliquot is run through a Superose 6 size exclusion column (MWCO 5,000 kDa) to determine the efficiency of enzyme incorporation. In this chromatographic step, it is assumed that the protein and lipid eluting in the void volume of the column represent the components that are incorporated into the bilayer, lipid vesicles. Reductase incorporation and activity is determined in the final preparation and in the fractions from the column by measuring the rate of the reductase-mediated reduction of cytochrome c (34). In addition, the concentration of phospholipid can also be determined in each fraction (35) in order to more accurately determine the lipid:protein ratio of the vesicular fraction.

3.5. Calculation Used to Adjust the Detergent Concentration Needed for Preparation of VRS with Different Lipid Concentrations

In adjusting our VRS method to changing concentrations of lipid, we assume that at the optimized conditions, the lipid bilayer is sufficiently solubilized to allow for the physical incorporation of the enzymes into the PC bilayer vesicles that form as the detergent is removed. If the monomeric concentration of sodium glycocholate in our optimized VRS preparation is equal to the CMC (as predicted by the studies that have examined the solubilization of membranes by detergents (discussed in the Introduction)) we can approximate the concentration of lipid-bound detergent from the total concentration used in the optimized conditions (those described in the protocol). More specifically, with a molecular weight of 488 g/mol, a concentration of 1% sodium glycocholate solution corresponds to 20.5 mM. The CMC of this detergent is 7.1 mM (36). Thus, we assume that the concentration of lipidbound detergent in the mixed detergent-lipid micelles is 13.4 mM (20.5 mM–7.1 mM). This value can be used to calculate the effective detergent to lipid ratio, and at any concentration of lipid, the concentration of detergent needed to achieve the same level of solubilization is then determined by first using the effective detergent to lipid ratio to calculate the amount of lipid-bound detergent and adding this amount to the CMC.

18

Reed

An example is shown to better explain the calculation involved: Amount of lipid in the defined, optimized condition (those described in the protocol) = 1.25 mmol Amount of lipid-bound detergent = 6.7 mmol (13,400 mM * 0.0005 L) Thus, the effective detergent:lipid ratio = 5.36 mole of detergent/ mole of PC Amount of monomeric detergent in the reconstitution = 3.6 mmol (7,100 mM * 0.0005 L) If one wants to make a reconstituted system with five times the lipid concentration as that in the defined, optimized condition, the calculation is as follows: Amount of PC = 6.25 mmol Amount of bound detergent = 5.36 detergent/PC × 6.25 mmol = 33.5 mmol detergent Amount of monomeric detergent = CMC*volume = 3.6 mmol Sum of monomeric and bound detergent in the 0.5 mL reconstitution = 37.1 mmol Detergent concentration = 74.2 mM = 3.62%. Thus, one should add 3.62% (w/v) sodium glycocholate to solubilize the lipid in a VRS containing 6.25 mmol of phospholipid (see Note 7).

4. Notes 1. We have found that when the lipid concentration is £1250 mM, the bilayer vesicles of lipid do not readily form. Thus, in order to make vesicles with P450: lipid ratios £250:1, the P450 concentration should be lowered (and not the concentration of lipid) from those stated in the general method. 2. We have also prepared the VRS by drying the lipid solution under a stream of N2. The lipid must be dried slowly and for a minimum of 2 h. If the flow of the nitrogen stream is too high, residual chloroform may be “trapped” under a film of lipid. This is apparent if the detergent-solubilized lipid (Step B.1) is cloudy. If this is observed, the enzymes will not incorporate properly in the lipid vesicles of the final preparation. In general, it has been found that the reductase does not incorporate into the vesicle bilayer as readily if the lipid is dried under N2 as compared to the results obtained with lyophilized lipid. 3. Our lab flushes the lyophilizer with anaerobic nitrogen that has run through a heated column of BASF palladium catalyst (BASF RO-20). 4. Solutions are bubbled with N2 (or preferably Argon) for 2–5 min before adding to the lipid.

Use of Liposomes in the Study of Drug Metabolism

19

5. We have also prepared the VRS using 0.5 M potassium phosphate (pH 7.25) without the MgCl2 and have not observed any significant differences in the quality of the final preparations. 6. Initially, we would add the enzymes from the concentrated stock solutions to the sonicated lipid and detergent solution after the latter was diluted with water. However, it was found that this method resulted in a much higher loss of the active P450. It seems the incremental addition of the detergent is less harmful to the P450. 7. The amount of BioBeads should not be adjusted to remove the excess detergent. The adsorptive capacity of BioBeads has been studied extensively (37), and it has been shown that the beads will also adsorb lipids. Thus, when the beads are scaled up with detergent, the excess lipid is more readily adsorbed and the desired increase in lipid concentration in the VRS is not attained. References 1. Porter TD, Coon MJ (1991) Cytochrome P-450. Multiplicity of isoforms, substrates, and catalytic and regulatory mechanisms. J Biol Chem 266:13469–13472 2. Guengerich FP (2001) Common and uncommon cytochrome P450 reactions related to metabolism and chemical toxicity. Chem Res Toxicol 14:611–650 3. Weng Y, Fang C, Turesky RJ, Behr M, Kaminsky LS, Ding X (2007) Determination of the role of target tissue metabolism in lung carcinogenesis using conditional cytochrome P450 reductasenull mice. Cancer Res 67:7825–7832 4. Iyanagi T (2007) Molecular mechanism of phase I and phase II drug-metabolizing enzymes: implications for detoxification. Int Rev Cytol 260:35–112 5. Rooney PH, Telfer C, McFadyen MC, Melvin WT, Murray GI (2004) The role of cytochrome P450 in cytotoxic bioactivation: future therapeutic directions. Curr Cancer Drug Targets 4:257–265 6. Hannemann F, Bichet A, Ewen KM, Bernhardt R (2007) Cytochrome P450 systems–biological variations of electron transport chains. Biochim Biophys Acta 1770:330–344 7. Guengerich FP (1989) Characterization of human microsomal cytochrome P-450 enzymes. Annu Rev Pharmacol Toxicol 29:241–264 8. West SB, Lu AYH (1972) Reconstituted liver microsomal enzyme system that hydroxylates drugs, other foreign compounds and endogenous substrates. V. Competition between cytochromes P-450 and P-448 for reductase in 3,

9.

10.

11.

12.

13.

14.

15.

4-benzpyrene hydroxylation. Arch Biochem Biophys 153:298–303 Saine SE, Strobel HW (1976) Drug metabolism in liver tumors. Resolution of components and reconstitution of activity. Mol Pharmacol 12:649–657 Strobel HW, Lu AYH, Heidema J, Coon MJ (1970) Phosphatidylcholine requirement in the enzymatic reduction of hemoprotein P-450 and in fatty acid, hydrocarbon, and drug hydroxylation. J Biol Chem 245:4851–4854 Ingelman-Sundberg M (1977) Phospholipids and detergents as effectors in the liver microsomal hydroxylase system. Biochim Biophys Acta 488:225–234 Taniguchi H, Pyerin W (1988) Phospholipid bilayer membranes play decisive roles in the cytochrome P-450-dependent monooxygenase system. J Cancer Res Clin Oncol 114:335–340 Causey KM, Eyer CS, Backes WL (1990) Dual role of phospholipid in the reconstitution of cytochrome P- 450 LM2-dependent activities. Mol Pharmacol 38:134–142 Balvers WG, Boersma MG, Veeger C, Rietjens IM (1993) Kinetics of cytochromes P-450 IA1 and IIB1 in reconstituted systems with dilauroyl- and distearoyl-glycerophosphocholine. Eur J Biochem 215:373–381 Autor AP, Kaschnitz RM, Heidema JK, Coon MJ (1973) Sedimentation and other properties of the reconstituted liver microsomal mixedfunction oxidase system containing cytochrome P-450, reduced triphosphopyridine nucleotidecytochrome P-450 reductase, and phosphatidylcholine. Mol Pharmacol 9:93–104

20

Reed

16. French JS, Guengerich FP, Coon MJ (1980) Interactions of cytochrome P-450, NADPHcytochrome P-450 reductase, phospholipid, and substrate in the reconstituted liver microsomal enzyme system. J Biol Chem 255:4112–4119 17. Reed JR, Kelley RW, Backes WL (2006) An evaluation of methods for the reconstitution of cytochromes P450 and NADPH P450 reductase into lipid vesicles. Drug Metab Dispos 34:660–666 18. Taniguchi H, Imai Y, Iyanagi T, Sato R (1979) Interaction between NADPH-cytochrome P-450 reductase and cytochrome P-450 in the membrane of phosphatidylcholine vesicles. Biochim Biophys Acta 550:341–356 19. Ingelman-Sundberg M, Glaumann H (1980) Incorporation of purified components of the rabbit liver microsomal hydroxylase system into phospholipid vesicles. Biochim Biophys Acta 599:417–435 20. Schwarz D, Gast K, Meyer HW, Lachmann U, Coon MJ, Ruckpaul K (1984) Incorporation of the cytochrome P-450 monooxygenase system into large unilamellar liposomes using octylglucoside, especially for measurements of protein diffusion in membranes. Biochem Biophys Res Commun 121:118–125 21. Ingelman-Sundberg M, Blanck J, Smettan G, Ruckpaul K (1983) Reduction of cytochrome P-450 LM2 by NADPH in reconstituted phospholipid vesicles is dependent on membrane charge. Eur J Biochem 134: 157–162 22. Bosterling B, Trudell JR, Galla HJ (1981) Phospholipid interactions with cytochrome P-450 in reconstituted vesicles. Preference for negatively-charged phosphatidic acid. Biochim Biophys Acta 643:547–556 23. Kawato S, Gut J, Cherry RJ, Winterhalter KH, Richter C (1982) Rotation of cytochrome P-450. I. Investigations of protein-protein interactions of cytochrome P-450 in phospholipid vesicles and liver microsomes. J Biol Chem 257:7023–7029 24. Schwarz D, Pirrwitz J, Ruckpaul K (1982) Rotational diffusion of cytochrome P-450 in the microsomal membrane-evidence for a clusterlike organization from saturation transfer electron paramagnetic resonance spectroscopy. Arch Biochem Biophys 216: 322–328 25. Taniguchi H, Imai Y, Sato R (1987) Proteinprotein and lipid-protein interactions in a reconstituted cytochrome P-450 dependent

26. 27.

28.

29.

30. 31.

32. 33.

34.

35. 36.

37.

microsomal monooxygenase. Biochem 26: 7084–7090 Hjelmeland LM (1990) Solubilization of native membrane proteins. Meth Enzymol 182: 253–264 Jackson ML, Schmidt CF, Lichtenberg D, Litman BJ, Albert AD (1982) Solubilization of phosphatidylcholine bilayers by octyl glucoside. Biochem 21:4576–4582 Bayerl TM, Werner G-D, Sackmann E (1989) Solubilization of DMPC and DPPC vesicles by detergents below their critical midellization concentration: high-sensitivity differential scanning calorimetry, Fourier transform infared spectroscopy and freeze-fracture electron microscopy reveal two interaction sites of detergents in vesicles. Biochim Biophys Acta 984:214–224 Reed JR, Brignac-Huber LM, Backes WL (2008) Physical incorporation of NADPHcytochrome P450 reductase and cytochrome P450 into phospholipid vesicles using glycocholate and Bio-Beads. Drug Metab Dispos 36:582–588 Pike LJ (2004) Lipid rafts: heterogeneity on the high seas. Biochem J 378:281–292 Browman DT, Resek ME, Zajchowski LD, Robbins SM (2006) Erlin-1 and erlin-2 are novel members of the prohibitin family of proteins that define lipid-raft-like domains of the ER. J Cell Sci 119:3149–3160 Holloway PW (1973) A simple procedure for removal of Triton X-100 from protein samples. Anal Biochem 53:304–308 Omura T, Sato R (1964) The carbon monoxidebinding pigment of liver microsomes. I. Evidence for its hemoprotein nature. J Biol Chem 239: 2370–2378 Phillips AH, Langdon RG (1962) Hepatic triphosphopyridine nucleotide-cytochrome c reductase: isolation, characterization, and kinetic studies. J Biol Chem 237:2652–2660 Stewart JC (1980) Colorimetric determination of phospholipids with ammonium ferrothiocyanate. Anal Biochem 104:10–14 Antonian L, Deb S, Spivak W (1990) Critical self-association of bile lipids studied by infrared spectroscopy and viscometry. J Lipid Res 31:947–951 Levy D, Bluzat A, Seigneuret M, Rigaud JL (1990) A systematic study of liposome and proteoliposome reconstitution involving BioBead-mediated Triton×-100 removal. Biochim Biophys Acta 1025:179–190

Chapter 3 Use of Liposomes to Study Cellular Osmosensors Reinhard Krämer, Sascha Nicklisch, and Vera Ott Abstract When cells are exposed to changes in the osmotic pressure of the external medium, they respond with mechanisms of osmoregulation. An increase of the extracellular osmolality leads to the accumulation of internal solutes by biosynthesis or uptake. Particular bacterial transporters act as osmosensors and respond to increased osmotic pressure by catalyzing uptake of compatible solutes. The functions of osmosensing, osmoregulation , and solute transport of these transporters can be analyzed in molecular detail after solubilization, isolation, and reconstitution into phospholipid vesicles. Using this approach, intrinsic functions of osmosensing transporters are studied in a defined hydrophilic (access to both sides of the membrane) and hydrophobic surrounding (phospholipid membrane), and free of putative interacting cofactors and regulatory proteins. Key words: Osmotic stress, Osmosensing, Transport, Reconstitution, Liposome, Proteoliposome, Membrane protein, Phospholipid, Signal transduction

1. Introduction Under steady state conditions, cells maintain a certain ratio of internal versus external osmotic pressure which results in a particular cell turgor. A change in the external osmolality leads to a change in the turgor, and cells have developed sophisticated mechanisms of osmoregulation to respond to this challenge. Changing osmolality is one of the most frequent types of environmental stress for many cells, in particular microbial cells. In order to properly respond to this stress, cells harbor sensory mechanisms which perceive stimuli related to osmotic stress, transduce these stimuli into appropriate intracellular signals, or directly respond to these stimuli by appropriate actions.

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_3, © Humana Press, a part of Springer Science+Business Media, LLC 2010

21

22

Krämer, Nicklisch, and Ott

In the case of hypo-osmotic stress, the major elements of stimulus response are mechanosensitive channels, which are dealt with in another chapter of this book. For the response to hyperosmotic stress, a major reaction is the activation of transport systems leading to the accumulation of compatible solutes. On the other hand, membrane bound sensory systems for hyperosmotic stress must be available, leading to appropriate responses in the cell on the level of transcription. In vitro systems for the study of membrane bound osmosensing transporters for compatible solutes require the presence of topologically closed vesicles and thus the use of liposomes. Functional reconstitution of transporter proteins in proteoliposomes allows the study of these systems in the absence of any other interfering mechanisms, cofactors or proteins, facilitating experimental access to both sides of the membrane, as well as to the composition of the phospholipid membrane in which the proteins are embedded. On the other hand, proteoliposomes have a number of drawbacks for the study of osmosensing in comparison to intact cells, mainly due to the inherent lack of a cell wall. As a consequence, proteoliposomes are in general more fragile than cells, and they lack turgor pressure. Consequently, osmosensory events related to turgor cannot be studied using proteoliposomes. As a further consequence, the fact that the morphological response of liposomes to hyperosmotic stress is different from cells has to be taken into consideration. Since liposomes behave as osmometers, i.e., they change their volume by water efflux according to the changing external osmotic pressure, and membranes are rigid in terms of their surface dimension; on the other, liposomes do not shrink in size but just invaginate under conditions of hyperosmotic stress to adapt the internal volume. The physical nature of stimuli, perceived by cellular osmosensors, is difficult to define, and in most cases not known (1). There are, however, a number of transport systems (1–3) in which these stimuli have been defined to a significant extent, and one of these model systems, the betaine transporter BetP from the grampositive soil bacterium Corynebacterium glutamicum, is used as an example here. In order to study osmosensing by a transport system, solute transport has to be measured. This necessarily requires some knowledge of transport kinetics and energetics, which, at least at a basic level, can be treated exactly as enzyme kinetics and energetics, and will not be further discussed here. Another basic requirement for the issue dealt with here is the proper handling of membrane proteins, their solubilization, isolation, purification, and, in particular, reconstitution. In this article, the relevant procedures of functional reconstitution of BetP from C. glutamicum will be described; however, no detailed reference to appropriate

Use of Liposomes to Study Cellular Osmosensors

23

procedures for obtaining solubilized membrane proteins in a functional active state will be made.

2. Materials 2.1. Preparation of Preformed Liposomes for Reconstitution

1. E. coli polar lipid extract (Avanti, Alabaster, USA).

2.2. Reconstitution of BetP Into Preformed Liposomes

1. Reconstitution buffer: 100 mM phosphate buffer, pH 7.5.

2. Liposome buffer: 100 mM phosphate buffer, pH 7.5.

2. Solubilized and purified BetP protein (minimal protein concentration should be around 0.3 mg/ml, detergent concentration about 0.1% dodecyl maltoside (DDM solgrade, Anatrace, Maumee, USA). The protein concentration has to be measured with an appropriate method, e.g., using amido black (4). 3. 20% v/v Triton X-100 solution in water. 4. Spectrophotometer for monitoring liposome titration with detergent. 5. Biobeads as absorbent for detergents (Biobeads SM2, Biorad, Munich, Germany; washing of the beads in methanol/water should be carried out as described by the manufacturer). The beads are stored in water. Directly before the use for absorbing detergent, the “wet beads” are prepared by placing biobeads suspended in water on filter paper, which removes the surplus water. These “wet beads” are used for weighing. 6. Minivial ultracentrifuge for harvesting liposomes. 7. Avanti Mini Extruder System (Avanti, Alabaster, USA). 8. Polycarbonate filters, pore size 400 nm (Nucleopore, Schleicher & Schuell, Dassel, Germany).

2.3. Betaine Transport as a Response of BetP to an Osmotic Shift

1. Rapid filtration unit (FH225V, Hoefer, Holliston, USA). 2. Membrane filters (Nitrocellulose, 0.45 mm pore size; Millipore, Schwalbach, Germany). 3. Radioactively labeled 14C-betaine (synthesized from 14C-choline by the use of choline oxidase)(5). 4. Transport buffer: 20 mM phosphate buffer, pH 7.5, 25 mM NaCl (if necessary, an increased osmolality is adjusted by addition of ionic solutes (e.g., NaCl), or neutral solutes, e.g., proline, if the ionic strength is not be changed significantly), 15 mM 14 C-betaine, and 0.5 mM valinomycin. 5. Washing buffer: 0.1 M LiCl. 6. Szintillation fluid and szintillation counter.

24

Krämer, Nicklisch, and Ott

3. Methods In the following, functional reconstitution of BetP, an osmosensing membrane protein, and the quantification of its response to osmotic stress is described. BetP is a secondary active glycine betaine transporter from the grampositive soil bacterium C. glutamicum (3). This carrier is energetically coupled to the co-transport of two sodium ions and thus driven by the electrochemical sodium potential. The transport activity of BetP is regulated by the actual osmotic stress, and it has been shown that the carrier protein is able to sense the extent of osmotic stress without any additional factors or components (3, 6). Isolation and purification of membrane proteins in a functionally active state requiring optimization of the kind and concentration of detergent used, of the solubilization conditions (ionic strength, type of buffer, pH, temperature, etc.), on the one hand, as well as optimization of the purification procedure (type of tag used, variation of isolation conditions, etc.), on the other. The protocol described here starts with membrane protein(s) in solubilized and functionally active form. The stability of solubilized membrane proteins in terms of functionality is in general a serious problem and needs attention (see Note 1). In order to obtain an appropriate in vitro test system, the solubilized osmosensing membrane protein is reconstituted into liposomes, often prepared from E. coli lipids. There are a large number of different methods for membrane protein reconstitution; the method of choice used in most of the successful procedures recently published is the integration into preformed liposomes developed by Rigaud and colleagues (7, 8). A number of different aspects have to be tested and optimized when trying to obtain an appropriate in vitro system for testing functional properties of a reconstituted membrane protein. The preformed liposomes may differ in the type of lipids used depending on the requirement of the particular membrane protein to be reconstituted (see Note 2). Furthermore, lipid quality is an important issue, which means purity and lipid stability. The latter critically depends on how they are handled and stored (see Note 3). 3.1. Preparation of Preformed Liposomes for Reconstitution

1. Purified lipids dissolved in organic solvent (chloroform/methanol) are mixed. The mixture is evaporated to dryness in a rotary evaporator (30°C; temperature should be above phase transition temperature of the lipids). 2. Traces of solvent are removed overnight by freeze-drying (lyophilization). 3. Liposome buffer (plus 2 mM b-mercaptoethanol) is added at a concentration of 20 mg lipid/ml and lipids are suspended by stirring at room temperature (RT) for about 2 h.

Use of Liposomes to Study Cellular Osmosensors

25

4. The lipid suspension is flushed with N2 gas, frozen in liquid N2, and stored at −80°C until use. 5. For reconstitution an aliquot of lipids is thawed gently at RT. Thereafter, liposomes are prepared by extrusion through polycarbonate filters (pore size 400 nm, multiple extrusions 17 times). 3.2. Reconstitution of BetP Into Preformed Liposomes

1. The preformed liposomes are diluted fourfold into liposome buffer. 20% v/v Triton X-100 is added stepwise (in 2–4 ml portions, thorough mixing by the use of a pipette) until detergent saturation of liposomes is achieved. Reaching the correct physical state of the liposomes for reconstitution is monitored by measuring optical density at 540 nm (7, 8). Optimal conditions for incorporation of the solubilized protein are obtained right after reaching the maximum value of light scatter at 540 nm. 2. The solubilized protein is then added drop by drop with thorough mixing. The amount of detergent added to the mixture together with the solubilized protein should be small in comparison to the amount of Triton X-100 added before. This mixture is shaken gently for 30 min at RT. 3. For removal of detergent according to the batch-procedure (7, 8), polystyrol beads (biobeads are added according to their absorption capacity for particular detergents (9), e.g., 5 mg and 10 mg of wet beads per mg of triton or DDM, respectively) are added. After shaking for 1 h at RT the same amount of beads is added and the mixture is shaken again for 1 h. The sample is then shaken overnight at 4°C with an additional twofold amount of beads. Subsequently, wet beads (same amount as for the first addition) are again added and shaken for 45 min at 4°C (see Note 4). 4. Finally, the proteoliposomes are separated from the beads with a pipette and washed twice with ice cold liposome buffer by centrifugation (Beckman Optima TLX tabletop ultracentrifuge, 350.000 g, 20 min, 4°C) before being resuspended in liposome buffer at a concentration of about 60 mg/ml lipid. The proteoliposomes are frozen in liquid N2 and stored at −80°C. 5. In order to assess the efficiency of reconstitution, the protein content of each batch of reconstituted protein has to be thoroughly measured by protein determination, e.g., using the amido black assay (4) (see Note 5).

3.3. Betaine Transport as a Response of BetP to an Osmotic Shift

1. For transport measurements, an aliquot of the proteoliposomes is thawed gently at RT, and extruded (polycarbonate filter, pore-size 400 nm) 17 times in a total volume of up to 1 ml, before being concentrated to the original volume by centrifugation (ultracentrifuge for mini vials, 350.000 g, 20 min, 20°C) and resuspended in liposome buffer. 2. In the transport assay the proteoliposomes, supplemented with internal K+ during the preparation in liposome buffer, are rapidly

26

Krämer, Nicklisch, and Ott

diluted into transport buffer containing valinomycin and the labeled substrate, betaine, and thoroughly mixed. In a standard experiment, the reaction is started by pipetting 5 ml of concentrated proteoliposomes into 1000 ml of transport buffer. According to the dilution ratio, this will lead to a K+-diffusion potential of about 140 mV. For measuring the activity of a particular transport protein, an appropriate driving force has to be provided. The driving force for BetP is the electrochemical Na+ potential, consequently, a sufficient amount of Na+ has to be present in the medium, in addition to the membrane potential, supplied here as K+-diffusion potential. 3. For each time-point in the following kinetic assay, 200 ml samples are used. These samples are withdrawn from the transport assay and instantly filtered through membrane filters in a filtration unit. The time intervals chosen depend on the velocity of the reaction under study; experienced experimenters are able to handle time points down to 5 s intervals. The filters are washed immediately twice with about 2 ml of washing buffer each (see Note 6). 4. After finishing the kinetic experiment, the filters are removed with tweezers and transferred into a scintillation vial each, which is filled with scintillation fluid, and counted. 5. The amount of betaine taken up into the proteoliposomes by the reconstituted carrier protein at the chosen time-points is calculated on the basis of the specific radioactivity applied in the transport buffer and the amount of protein present in the 200 ml sample of proteoliposomes. Protein determination should be carried out for each experiment, since variable losses during the extrusion procedure may give rise to variation. 6. Substrate uptake activity is interpreted in terms of transport kinetics, and specific transport rates are calculated (see Note 7). Beside the activity of the reconstituted protein, the result of uptake measurements may depend on a variety of other factors (see Note 8). 7. Testing a transport protein for its osmosensing function will need variation of the medium on both sides of the membrane, with respect to concentration and quality of solutes, e.g., ionic and nonionic solutes, or type of cations and anions, respectively. The solute composition in the interior of the proteoliposomes can be varied during liposome preparation (Step 1 of Subheading 3.3.). The composition of the internal compartment can also be changed after proteoliposome preparation by applying additional freezing and thawing cycles in a medium of changed composition followed by liposome generation using the extrusion procedure. The lipid composition can also be varied during the preparation of the preformed liposomes.

Use of Liposomes to Study Cellular Osmosensors

27

Additional lipids can be integrated by repeated cycles of freezing and thawing mixtures of (proteo) liposomes with different lipid compositions. 8. It should be noted, that there are two different types of experiments testing the dependence of a membrane inserted protein on variation of osmotic properties. On the one hand, the external osmolality (or solute composition) can be changed at the start of the transport experiment, leading to an osmotic gradient and a subsequent volume change of the proteoliposomes. A variation of solute concentration or composition on both sides of the membrane at the same time, i.e., during proteoliposome preparation, on the other hand, changes the osmotic pressure without introducing an osmotic gradient and without concomitant volume changes (6).

4. Notes 1. The stability of solubilized membrane (carrier) proteins is a serious issue for the success of this type of experiment. The conditions of solubilization, isolation, and storage, if necessary, have to be optimized carefully. Most transport proteins are notoriously unstable in the solubilized state, but, in general, rather stable once properly inserted into a phospholipid membrane. Consequently, they should be purified quickly and reconstituted into proteoliposomes right after purification. 2. Often, selected lipids have to be used for the reconstitution of a particular membrane protein, in order to reach optimum reconstitution efficiency or functionality.. Different membrane proteins have different preference for lipids, with respect to the phospholipid headgroup, the headgroup charge, and the fatty acid composition. Typical lipids to start with are lipids from E. coli membranes, asolectin (soy bean lipids), or egg yolk lipids. Once, a basic activity of the reconstituted protein is observed, variation of lipids will normally improve the results. An obvious strategy to overcome some of these problems is the use of synthetic phospholipids; however, for this, a balanced composition of unsaturated (essential for most membrane proteins) and saturated fatty acids is required. For unknown reasons, effective reconstitution into synthetic lipid mixtures seems to be more difficult compared with natural lipids, e.g., from E. coli. In any case, the fact that proteins in general prefer mixtures of phospholipids, both with respect to the headgroup and the fatty acid composition over the membranes prepared from single compounds has to be taken into account.

28

Krämer, Nicklisch, and Ott

3. The same as stated for handling of the protein is also true for handling of lipids. Always the best lipids available should be used. In general, Avanti is a good choice as a supplier for lipids, however, be careful for variation of lipid quality from batch to batch. After having established an optimized protocol for reconstitution, try to examine different batches of the lipids of choice and then buy a stock of lipids of the optimal batch. In general, if the proteoliposome test system, after having been successfully established, fails to work for unknown reasons, the quality of lipids is a good candidate to think about. For this reason, it is recommended always that part of a lipid batch which was proven to function is stored away, to serve as a control for later experiments. Other reasons for failure, of course, are the functionality of the solubilized protein, and, rather often, the efficiency of detergent removal by adsorbance to biobeads, which directly affects the efficiency of reconstitution and/or the permeability properties of the proteoliposomes. 4. The stability of the lipids used is a serious issue, too. The manufacturer supplies lipids in general sealed under nitrogen gas; it is recommended that the same condition is kept when storing lipids in the lab, preferably at −80°C. Useful recommendations for lipid handling and storage are found in (10–12). 5. Detergent removal is an important step during reconstitution and is prone to a number of difficulties. Complete removal of detergent from the proteoliposomal sample can simply be tested by the absence of foam after shaking. It has to be taken into account that the beads not only absorb detergents, but all amphiphilic (and even hydrophilic) compounds, in general in the order detergent > lipid > protein. It is thus important to limit the amount of biobeads used for absorption of the detergent in order not to lose too much protein. An alternative strategy partially avoiding this problem is to presaturate the biobeads with lipids (7, 8). 6. The efficiency of reconstitution is a complex problem. First, it refers to the amount of protein successfully integrated. This value may vary with the kind of detergent and lipid used, as well as with particular aspects of the reconstitution procedure. It has to be quantified by protein determination in the proteoliposomes, e.g., by the amido black method (4). In general, the efficiency of protein insertion decreases with increasing protein/lipid ratio during reconstitution. Second, reconstitution efficiency also means the fraction of functionally active protein in comparison to the total amount of integrated protein. Only the latter is measured by protein determination. This ratio is rather difficult to determine and in general needs additional methods, e.g., quantification of substrate binding. Third, protein orientation may be an issue. In most cases, the

Use of Liposomes to Study Cellular Osmosensors

29

orientation of the reconstituted membrane protein is not known. This, however, may be of crucial significance for interpreting its function in the in proteoliposomal in vitro system, in particular for transport proteins being dependent on oriented driving forces. There are a number of methods available for defining the orientation of a membrane-inserted protein, e.g., the use of antibodies, proteases, and impermeable labeling reagents for specific amino acids. 7. Although seeming apparently simple, fast, and efficient liposome filtration is a crucial step in the whole procedure. It should be taken into account that filtration, as used here, is to a large extent adsorption to the filter material, in view of the pore size of the filters used. Before application to the filtration unit, the membrane filters have to be presoaked in washing buffer, in order to avoid labeled substrate to be soaked into the external ring of the filter which is fixed in the filtration unit. Before running the kinetic experiment, the filters should be briefly soaked by application of vacuum. When applied to the filter during transport kinetics, the proteoliposome sample should quickly be distributed over the entire filter surface. The same holds true for the washing solution, which preferentially should be applied along the walls of the filtration chamber. 8. Transport kinetics has been elaborated in relevant text books. In general, when establishing transport kinetics of a particular carrier protein, a time course including several experimental points (five at least) has to be used in order to find out in which time window the uptake kinetics is reasonably linear with time. Once established, experiments are frequently carried out using two time points only within the linear kinetic phase, because of the requirement of rapid assays and/or frequently because of a limited range of linearity (mainly due to the small size of the proteoliposomes used). In order to provide sufficient experimental reliability, these two point-kinetics have to be carried out in multiple sets, e.g., fivefold. When changing experimental conditions, multiple time point kinetics has to be applied again to verify linearity. It should furthermore be mentioned that the linearity of a transport assay also depends on the amount of substrate taken up during the time course monitored in the kinetic experiment. For this reason, the amount of labeled substrate taken up at the end of the time course should not exceed one third of the added label. This amount, however, may sometimes be much lower, if higher substrate concentrations have to be used because of a relatively low substrate affinity of the transport protein under study. 9. It should be pointed out that the result of the kinetic analysis of a reconstituted transport protein not only depends on the functionality of the protein, and the experimental conditions

30

Krämer, Nicklisch, and Ott

applied (e.g., kind and extent of driving force, pH, ionic strength, kind of lipid provided, etc.), but to a large extent also on the quality of liposomes used. This refers, on the one hand, to the integrity and stability of liposomes, which in the presence of residual detergent, or high amounts of protein integrated, can be an issue. The integrity of the proteoliposomes used can be tested by the incorporation of calceine (13). Another reason for unexpected failure of transport experiments is partial leakiness of the proteoliposomes to small ions, e.g., protons, leading to a quick collapse of the driving force. This will in general not lead to leakiness of the larger solutes, like substrates (or calceine), and is frequently caused by residual amounts of detergent present. Furthermore, the size of the liposomes used is of critical significance. An early deviation of transport kinetics from linearity may indicate the presence of liposomes too small in size for the particular transport experiment. The impact of the liposome size on the shape of the transport kinetics observed is also a major reason for using liposomes sized by the extrusion procedure through polycarbonate filters. All other methods of proteoliposome formation lead to mixtures of vesicles with a broad size distribution. This fact will have a complex influence on the resulting transport kinetics. References 1. Wood JM (1999) Osmosensing by bacteria: signals and membrane-based sensors. Microbiol Mol Biol Rev 63:230–262 2. Poolman B, Spitzer JJ, Wood JM (2004) Bacterial osmosensing: roles of membrane structure and electrostatics in lipid-protein and protein-protein interactions. Biochim Biophys Acta 1666:88–104 3. Morbach S, Krämer R (2003) Impact of transport processes in the osmotic response of Corynebacterium glutamicum. J Biotechnol 104:69–75 4. Schaffner W, Weissmann C (1973) A rapid, sensitive and specific method for the determination of protein in dilute solution. Anal Biochem 56:502–514 5. Landfald B, Strøm AR (1986) Choline-glycine betaine pathway confers a high level of osmotic tolerance in Escherichia coli. J Bacteriol 165:849–55 6. Rübenhagen R, Morbach S, Krämer R (2001) The osmoreactive betaine carrier BetP from Corynebacterium glutamicum is a sensor for cytoplasmic K+. EMBO J 20:5412–5420

7. Rigaud J-L, Levy D (2003) Reconstitution of membrane proteins into liposomes. Methods Enzymol 372:65–86 8. Paternostre MT, Roux M, Rigaud J-L (1988) Mechanism of membrane protein insertion into liposomes during reconstitution procedures involving the use of detergents. Biochemistry 27:2668–2677 9. Rigaud J-L, Mosser G, Lacapere J-J, Olofsson A, Levy D, Ranck J-L (1997) Bio-beads: an efficient strategy for two-dimensional crystallization of membrane proteins. J Struct Biol 118:226–235 10. Zuidam NJ, Crommelin DJ (1995) Chemical hydrolysis of phospholipids. J Pharm Sci 84:1113–9 11. Hernández-Caselles T, Villalaín J, GómezFernández JC (1990) Stability of liposomes on long term storage. J Pharm Pharmacol 42:397–400 12. Torchilin VP, Weissig V (2003) Liposomes, 2nd edn. Oxford University Press, New York 13. Allen TM, Cleland LG (1980) Serum-induced leakage of liposome contents. Biochim Biophys Acta 597:418–426

Chapter 4 Studying Mechanosensitive Ion Channels Using Liposomes Boris Martinac, Paul R. Rohde, Andrew R. Battle, Evgeny Petrov, Prithwish Pal, Alexander Fook Weng Foo, Valeria Vásquez, Thuan Huynh, and Anna Kloda Abstract Mechanosensitive (MS) ion channels are the primary molecular transducers of mechanical force into electrical and/or chemical intracellular signals in living cells. They have been implicated in innumerable mechanosensory physiological processes including touch and pain sensation, hearing, blood pressure control, micturition, cell volume regulation, tissue growth, or cellular turgor control. Much of what we know about the basic physical principles underlying the conversion of mechanical force acting upon membranes of living cells into conformational changes of MS channels comes from studies of MS channels reconstituted into artificial liposomes. Using bacterial MS channels as a model, we have shown by reconstituting these channels into liposomes that there is a close relationship between the physico-chemical properties of the lipid bilayer and structural dynamics bringing about the function of these channels. Key words: MscL, MscS, NMDA, Liposome reconstitution, Patch clamp, EPR spectroscopy, FRET spectroscopy, Confocal microscopy

1. Introduction MS channels present a classical example of ion channels for which the composition and properties of the surrounding lipid matrix are crucial for their function. In a nutshell, the MS channels’ leitmotif is “force from lipids” (1, 2). Studies on prokaryotic (MS) channels over many years have demonstrated that the lipid bilayer transmits the mechanical force directly to this type of MS channels enabling them to detect changes in osmotic forces acting upon these microorganisms (3, 4). Among prokaryotic MS channels, the best characterized channels are MscL and MscS of

V. Weissig, (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_4, © Humana Press, a part of Springer Science+Business Media, LLC 2010

31

32

Martinac et al.

Escherichia coli. Bacterial MS channels were the first channels shown to sense directly the membrane tension in the lipid bilayer caused by the external mechanical force applied to bacterial cell membrane (5,6). Recent evidence indicates that, similar to prokaryotic MS channels, some MS ion channels found in membranes of a variety of eukaryotic cells, such as TREK-1, TRAAK and TRPC1, can also be gated by the bilayer mechanism (7–10). The glutamate-gated NMDA receptor channel has also been shown to exhibit mechanosensitivity upon reconstitution into artificial liposomes (11). The question of how MS channel proteins detect mechanical stresses in the lipid bilayer of cellular membranes and what makes these channels undergo conformational changes in response to membrane tension has been addressed in several studies examining the physical principles underlying gating of MscL channels by bilayer deformation forces (12–14). In one study, a combination of functional patch-clamp experiments with structural electronparamagnetic resonance (EPR) spectroscopic experiments allowed the drawing of the following conclusions about forces from lipids that affect both MS channel structure and function. First, the importance of the hydrophobic surface match for MscL mechanosensitivity was shown to result from a membrane-tension-induced bilayer thinning, which stabilizes intermediate conformations of MscL leading to channel opening due to a better hydrophobic match of the thinner bilayer with the open channel compared with that of the closed conformation. Second, geometric shape inequalities between lipid molecules asymmetrically distributed between the two leaflets of the lipid bilayer were shown to be critical in triggering the opening of the MscL pore of >25 Å in diameter (12, 13). These results have been confirmed by a FRET spectroscopic study in which asymmetric distribution of lipid molecules between the two leaflets of the lipid bilayer was shown to open MscL (14). Both studies thus indicate that the mechanism of mechanotransduction in MS channels is defined by both local and global asymmetries in the transbilayer pressure profile at the lipid protein interface. The role of bilayer thickness and the concept of hydrophobic mismatch in MscL functionality have further been investigated in studies exploring the effect of static magnetic fields (SMF) of moderate intensity (~400 mT) on the open probability of MscL reconstituted in phospholipid bilayers (15, 16). Results of these studies suggest that due to co-operative superdiamagnetism of phospholipid molecules, SMF most likely cause a rearrangement of the lipids resulting in a change of bilayer thickness. Given that not only ion channels but also other signaling membrane proteins may experience local change in bilayer thickness and/or curvature as a consequence of association with lipid microdomains or chemical modification of the lipid bilayer, the findings from studies of bacterial MS channels may also have implications

Studying Mechanosensitive Ion Channels Using Liposomes

33

for the functioning of all kinds of signaling membrane proteins including ion channels. The production of the MscL or MscS channel entails the protein expression of its subunit which then readily self-assimilates into the homo-multimeric functional channel within the bacterial membrane, without any post-translational requirements. Subunit expression with a fused purification tag allows simplified isolation of the channel (subunits) from bacterial membrane preparations using commonly available commercial purification supplies. Suitable tag expression vectors are described further below. Two affinity expression tag systems commonly used include charged affinity and specific enzymatic affinity. Charged affinity relies on a gene construct to produce a fusion with six histidine amino acids to either the amino or carboxyl terminal of the protein. This allows the fusion protein to bind to nickel or cobalt chelated to sepharose beads. Capture to beads allows further washing (purification) and final elution with a charged competitor (imidazole) solution. Alternatively for higher purity at lower yields and for a product with no artificial charges, enzymatic affinity may be used. This involves a fusion product with (typically) glutathione-S-transferase (GST) which allows specific tight binding to sepharose beads coupled to the GST target, glutathione. The protein of interest is cleaved off GST with (typically) thrombin via a target site that has been designed between the two fusion products from the onset. It is also possible to design affinity histidine tags that are also cleavable (not described). It should be noted that thrombin cleavage will leave part of its recognition sequence on the final product, and other amino acids from cloning vectors are typically also attached to the final protein, including any (uncleaved) 6 × His sequence.

2. Materials 2.1. MscL Production with GST or 6 × His Affinity Fusion Tags

1. Luria Broth (LB) (10 g/l bacto-tryptone, 5 g/l yeast extract, 10 g/l NaCl) autoclave sterilized.

2.1.1. Materials Common for Both MscL Production Methods

3. Ampicillin dihydrate antibiotic, 100 mg/ml in water, 0.22 mm filter sterilized.

2. Autoclave sterilized 100 ml and ~2,800 ml culture flasks.

4. 1 M Isopropyl b-D-1-thiogalactopyranoside (IPTG) in water, 0.22 mm filter sterilized. 5. Phosphate buffered saline (PBS) (10×): 80 g NaCl, 2 g KCl, 14.4 g Na2HPO4 2H2O and 2.4 g KH2PO4 into 1 L of water. On diluting adjust pH with HCl or NaOH.

34

Martinac et al.

6. Flat bristle brushes (in the order of approximately 5 mm and 8 mm width). 7. Phenylmethanesulphonylfluoride (PMSF), 20 mg/ml in isopropanol. 8. Deoxyribonuclease I (DNase) (e.g., Sigma DN25). 9. n-Dodecyl-b-D-Maltopyranoside (DDM), 200 mM in water (Anatrace D310). 10. Bradford Reagent (Coomassie dye binding protein assay, Protein dye reagent for quantitation). 2.1.2. MscL Production with GST Affinity Fusion Tag

1. AW737KO MscL knock-out strain transformed with N-terminus gluthione-S-transferase tagged MscL (pGEX-2T plasmid, GE) expression construct. 2. GST Wash Buffer: 1× PBS pH 7.4 with 1 mM DDM 3. 10 ml chromatography columns (e.g., Biorad 731–1550) 4. Glutathione sepharose (Bioworld or GE), washed following manufacture’s protocol, but with 1 mM DDM detergent. 5. Thrombin. (GE or Sigma) Dissolved and stored as per supplier’s instructions.

2.1.3. MscL Production with 6×His Affinity Fusion Tag

1. M15 E.coli [with pREP4 plasmid (Qiagen)] and with N-terminus 6×His-tagged MscL (pQE30 plasmid, Qiagen) or C-terminus 6×His-tagged MscL (pQE70 plasmid, Qiagen) expression construct. 2. Kanamycin sulphate antibiotic, 25 mg/ml ml in water, 0.22 mm filter sterilized. 3. Talon metal affinity resin (Clonetech 635502), washed following manufacture’s protocol, but with 1 mM DDM detergent. 4. 20 ml chromatography columns (e.g., Biorad 732–1010) 5. His Wash Buffer, 1× PBS pH 6.0, 1 mM DDM, 5 mM imidazole. 6. His Elution Buffer, 1× PBS pH 6.0, 1 mM DDM, 500 mM imidazole. 7. Millipore Amicon Ultra-15 Ultracel 10 k centrifugal filter device. Washed and used to manufacturer’s instructions.

2.2. MscS Production 6×His Affinity Fusion Tags

1. M15 E. coli [with pREP4 plasmid (Qiagen)] and transformed with pRARE plasmid (Merck), made competent ready for transformation. 2. N-terminus 6 × His-tagged MscS (pQE30 plasmid, Qiagen), or C-terminus 6 × His-tagged MscS (pQE70 plasmid, Qiagen) expression construct. The C-terminus construct gives greater yield.

35

Studying Mechanosensitive Ion Channels Using Liposomes

3. Luria Broth (LB): 10 g/l bacto-tryptone, 5 g/l yeast extract, 10 g/l NaCl, autoclave sterilized. 4. Autoclave sterilized 100 ml and ~2800 ml culture flasks. 5. Ampicillin dihydrate antibiotic, 100 mg/ml in water, 0.22 µm filter sterilized. 6. Kanamycin sulphate antibiotic, 25 mg/ml ml in water, 0.22 µm filter sterilized. 7. Chloramphenicol antibiotic 25 mg/ml in ethanol. 8. Glycerol. 9. Isopropyl b-D-1-thiogalactopyranoside (IPTG) 1 M in water, 0.22 mm filter sterilized. 10. Flat bristle brushes (in the order of approximately 5 mm and 8 mm width). 11. Phenylmethanesulphonylfluoride (~115 mM) in isopropanol.

(PMSF),

20

mg/ml

12. Deoxyribonuclease I (DNase) (e.g., Sigma DN25). 13. n-Dodecyl-b-D-Maltopyranoside (DDM), 200 mM in water (Anatrace D310). 14. Talon metal affinity resin (Clonetech 635502), washed following manufacture’s protocol, but with 1 mM DDM. 15. 20 ml chromatography columns (e.g., Biorad 732–1010). 16. Phosphate buffered saline (PBS), (10×): 80 g NaCl, 2 g KCl, 14.4 g Na2HPO4 2H2O and 2.4 g KH2PO4 into 1 L of water. On diluting to make 1× adjust pH with HCl or NaOH. 17. Solubilization Buffer: 1× PBS pH 7.5, 1 mM PMSF, 10% v/v glycerol. 18. MscS Wash Buffer 1: 1× PBS pH 7.5, 1 mM DDM, 10% v/v glycerol. 19. MscS Wash Buffer 2: 1× PBS pH 6.0, 1 mM DDM, 10% v/v glycerol, 5 mM imidazole. 20. MscS Elution Buffer: 1× PBS pH 6.0, 1 mM DDM, 10% v/v glycerol, 300 mM imidazole. 21. Millipore Amicon Ultra-15 Ultracel 10 k centrifugal filter device. Washed and used to manufacturer’s instructions. 2.3. NMDA Protein Expression and Purification 2.3.1. Cell Culture and NMDA Receptor Channel Constructs

1. Cell culture used for expression of NMDA receptor NR1 and NR2 subunits. (a) SF9 Spodoptera frugiperda (insect cell culture) (1 or 2 L). (b) XL99, a suspension CHO-K1 (Chinese Hamster Ovary cells) cell line adapted to growth in suspension in serum-free media (1 L).

36

Martinac et al.

2. Growth media (a) Sf-900II (Invitrogen) serum-free media for insect cell culture. (b) CHO-S-SFMII (Invitrogen) media for CHO mammalian cell culture. 3. NMDA receptor channel constructs. (a) 6 × His-tagged NR1a subunit (C-terminus, pENTR/ D-TOPO or pcDNA 3.3-TOPO plasmids). (b) 6 × His-tagged NR2A subunit (C-terminus, pENTR/ D-TOPO or pcDNA 3.3-TOPO plasmids). 2.3.2. Buffers for Purification of NMDA Receptor Channel Proteins

1. Cell Lysis and Solubilization Buffer (100 ml, pH 7.3): 20 mM sodium phosphate 500 mM NaCl, 10 mM imidazole, 3% octyl glucoside (OG), 20 mg/ml DNase protease inhibitor tablet (Roche), 1.25 ml of TALON resin (Clontech) added after solubilization. 2. Wash Buffer (50 ml, pH 7.3): 20 mM sodium phosphate, 500 mM NaCl, 20 mM imidazole, 1.5% OG, protease inhibitor tablet (Roche). 3. Elution Buffer (4 ml, pH 7.3): 20 mM sodium phosphate, 500 mM NaCl, 150 mM imidazole, 1.5% OG, protease inhibitor tablet (Roche). 4. Gel Filtration Buffer (50 ml, pH 7.5): 20 mM sodium phosphate, 300 mM NaCl.

2.4. SDS-Polyacrylamide Gel Electrophoresis (SDS-PAGE)

1. An electrophoresis vertical mini electrophoresis rig system and if desired compatible pre-cast 12% polyacrylamide denaturing gels of 1 mm thickness, ten well for protein analysis, as well as associated running buffers as per manufacture’s specifications. For manually cast SDS-PAGE gels, the following materials are required. 2. Distilled water (not from a source utilizing ultraviolet light sterilization.) 3. Tris(hydroxymethyl)aminomethane (TRIS), electrophoresis grade. 4. 1.5 M Tris-HCl pH 8.8. 5. 0.5 M Tris-HCl pH 6.8. 6. 10% Sodium Dodecyl Sulfate (Sodium lauryl sulfate) (SDS) 7. 30% Acrylamide/Bis Solution, 37.5:1 (e.g., Biorad 161–0158). The non-polymerized acrylamide monomer solution is a potent neurotoxin hazard and must be handled appropriately. Nonvolatile when supplied as a solution. 8. 10% (in water) Ammonium persulphate (APS), stored in ~500 ml aliquots stored at −20°C. 9. Tetramethylethylenediamine (TEMED).

Studying Mechanosensitive Ion Channels Using Liposomes

37

10. 5× Elpho Buffer: 15.1 g Tris, 94 g Glycine, double distilled water to 900 ml, mix, add 50 ml 10% SDS, top up to 1 L. 11. 4× Sample Buffer: 1.52 g Tris-HCl, 20 ml glycerol, 2 g SDS, 2 ml 2-mercaptoethanol, 1 mg bromophenol blue. 12. Protein molecular weight marker (suitable is Fermentas #SM0431). 13. Coomassie Stain: 1.25 g (Coomassie) Brilliant Blue R-250, 225 ml methanol, 50 ml glacial acetic acid, double distilled water to 500 ml. 14. Destain (300 ml methanol, 10% glacial acetic acid, water to 500 ml). 2.5. Liposome Reconstitution 2.5.1. Dehydration/ Rehydration (D/R) Method

1. Chloroform, AR grade. 2. Lipid. Azolectin (L-a-Phosphatidylcholine from soybean, Type II-S, Sigma P5638, or Type IV-S, Sigma P3644) or pure lipids (e.g., Phosphatidylcholine (PC) with acyl lengths of 16 to 20, or PC mixture with phosphatidylglycerol (PG) and/or Phosphatidylethanolamine (PE) pure lipids from Avanti Polar Lipids, Inc, Alabaster) 10 mg/ml in chloroform, stored at −20°C. 3. Nitrogen (N2) Industrial Grade Dry Gas supply, fed through a glass pipette for a nitrogen stream. 4. Purified mechanosensitive channel protein, as described within this chapter. 5. Dehydration/Rehydration Buffer Solution (D/R Buffer): 200 mM KCl, 5 mM HEPES (pH 7.2 adjusted with KOH). 6. Bio-Beads SM-2 Adsorbent (Biorad #152–8920) prepared by washing thrice (on shaker/rocker) in methanol for 30 min each. Remove methanol. Rinse briefly in distilled water by shaking three times, followed by two 30 min washes in distilled water. Decant and cover beads with fresh water. Degas by vacuum for 3 h to overnight. Store at 4°C. Before use, blot the required amount of beads with tissue paper to remove excess moisture. 7. Ethanol, AR grade. 8. Glass microscope slides, cleaned with ethanol, and dried. 9. Filter paper or similar absorbent paper/tissue type.

2.5.2. Sucrose Method

In addition to materials 1 to 6 listed above in Subheading 2.5.1, sucrose is also required.

2.6. Patch-Clamp Recording From MS Channels Reconstituted into Liposomes

1. Borosilicate glass calibrated microcapillary tube (e.g Drummond Scientific Co., Broomall, PA, 100 ml Calibrated Pipets [sic] cat: 2-000-100).

38

Martinac et al.

2. Recording Solution: 200 mM KCl, 40 mM MgCl2, 5 mM HEPES (pH 7.2 adjusted with KOH). 3. 0.22 mm syringe filter. 2.7. Confocal Microscopy of Liposome Reconstituted MS Channels

1. Single cysteine mutant expression construct and the resulting protein purified using deoxygenated Wash Buffer and Elution Buffer (see text). 2. Labeling Buffer: Deoxygenated PBS pH 7.5, with 1 mM DDM (n-Dodecyl-b-D-Maltopyranoside (DDM) (Anatrace D310)) 3. 40 mM tris(2-carboxyethyl)phosphine (TCEP) in distilled water. 4. Alexa Fluor® 488 C5-maleimide (AF488) (Invitrogen Molecular Probes A10254). Resuspended in water, separated in 100 mg aliquots which are then lyophilized/desiccated and stored at −20°C in a light sealed container. 5. Optional b-mercaptoethanol. 6. Dialysis cassette, 10 K molecular weight cut off, 3 ml capacity (e.g., Pierce, Thermo Fisher Scientific Inc, # 66380). 7. Dialysis Buffer: PBS pH 7.5 with 0.5% Triton X-100, chilled to 4°C for use. A 10× stock may be preferentially made, adjusted to pH 7.5 on dilution. 8. Confocal microscope with 60× objective lens, and appropriate excitation laser. 9. Cover slips and glass slides.

3. Methods Different constructs of MscL and MscS (GST fusion proteins or 6×His-tagged proteins) have been used for liposome reconstitution. The former is more suitable for testing channel function by the patch-clamp technique while the latter is appropriate for structural studies using EPR or FRET spectroscopy. In the case of NMDA receptor channels a 6×His-tagged protein has been used. The amount of protein required for reconstitution depends on the protein-to-lipid ratio (w/w for patch clamp experiments or mol/mol for structural EPR and FRET experiments), reconstitution method, and the type of experiment. For MscL reconstitution we have also found that the lipid-to-protein ratio depends on which construct is being used. Particularly, for MscS we have found that the lipids commonly used for electrophysiological measurements, such as DOPC or E. coli polar lipids cannot be used for MscS structural determination through spectroscopy, since they promote two-dimensional aggregation, regardless of the expression and purification conditions (20). This has not

Studying Mechanosensitive Ion Channels Using Liposomes

39

been tested in detail for experiments with MscL or NMDA. Assuming that dehydration/rehydration (D/R) method is used (see further) the number of active channels per liposome patch varies with the protein-to-lipid ratio (e.g., 1:1000 w/w) and is higher when the GST-MscL construct is reconstituted compared to 6 × His-tagged MscL constructs. If a sucrose reconstitution method is used (see further) less protein is required and higher percentage of liposome patches with active channels become available for experiments. Using the Vesicle PrepPro (Nanion Technologies GmbH) (see further) MscL (1/900 w/w protein-to-lipid ratio) can be reconstituted into liposomes in less than three hours total preparation time. Usually 5 micrograms of any of the MS channels described here is sufficient for several liposome preparations using the D/R method. If protein is scarce, to minimize the amount of protein required for reconstitution smaller amounts of lipids (ca. 2 mg) can be used per single liposome prep, which usually gives enough material for two days of patch clamping. 3.1. MscL- GST or His Tag Fusion Protein Purification

This procedure takes several days, thus it is advisable to start at the beginning of a week. Days 2 and 3 may be combined into a single long day.

3.1.1. Day 1: Overnight Starter Culture

1. Add 10 ml glycerol stock of bacterial strain (with expression construct) into a 100 ml culture flask with 12 ml LB and 12 ml 100 mg/ml Ampicillin. For pQE His constructs in M15 strain, also add 12 ml 25 mg/ml kanamycin sulphate. 2. Incubate at 37°C at 240 rpm (for ¾ inch orbit) in an orbital shaker incubator.

3.1.2. Day 2: Growth, Expression, and Cell Collection

1. Add 10 ml of overnight culture into a ~2,800 ml culture flask with 1000 ml LB and 1 ml 100 mg/ml Ampicillin. For pQE 6 × His constructs in M15 strain, also add 1000 ml 25 mg/ml kanamycin sulphate. 2. For aeration, seal the flasks with tissue paper (such as low lint fine task wipers) and incubate at 37°C at 150 rpm (for 2 inch orbit) in an orbital shaker incubator. 3. When optical density of the culture (OD) A600 = 0.8 (occurs in ~2–2.5 h) induce expression with the addition of 1000 ml 1 M IPTG. Also add 500 ml 100 mg/ml ampicillin. (Hint: For multiple samples, hold a considerably faster culture(s) at 4°C until others catch up before adding IPTG etc.). 4. Incubate with shaking for a further 4 h. 5. Centrifuge culture at around 8,000 ´ g for 15 min at 4°C. Discard supernatant. 6. For every 10 g of cell pellet resuspend in 10 ml PBS (with a ~8 mm wide brush). 7. Store the resuspension at −20°C or continue with day two.

40

Martinac et al.

3.1.3. Day 3: Cell Disruption and Membrane Processing

1. Thaw sample and keep on ice. 2. Add roughly a “match head” amount of DNase, and 1/100 volume of 20 mg/ml PMSF. 3. French press (Thermo Electron Corporation) twice at 16,000 PSI. 4. Remove debris, unbroken cells, or inclusion bodies by high speed centrifugation at 8,000 ´ g (k-Factor ~3,600) for 30 min at 4°C. Collect supernatant. 5. Pellet membranes by ultracentrifugation at 235,000 × g (k-Factor ~133) for 3 h at 4°C. Discard supernatant. 6. Suspend pellet in 5 ml PBS with a 5 mm width brush. 7. Add a further 17 ml PBS and 880 ml 200 mM DDM. 8. Solubilize on a roller wheel overnight at 4°C.

3.1.4. Day 4: Clarification and Column Binding and Purification

1. Clarify the solubilization by ultracentrifugation at 10,000 for 20 min at 4°C. Recover supernatant to use. 2. For 6×His fusion constructs add 2 ml prepared cobalt-resin beads. For GST fusion constructs add 1 ml prepared glutathione sepharose beads. 3. Incubate on a rocker or roller for 2 h at room temperature. 4. Pour the cobalt resin suspension into 20 ml chromatography columns. Pour the GST resin suspension into 10 ml chromatography columns. Release bottom caps when resin has settled somewhat after several minutes, and allow all liquid to flow out as discard. 5. Wash the cobalt resin twice with 25 ml His Wash Buffer. Wash GST resin twice with 10 ml GST Wash Buffer. For GST fusion protein, recap column and mix beads by pipetting with 1 ml GST Wash Buffer and transfer to a 2 ml centrifugation tube (see Note 1). Add 20 units of thrombin, and incubate overnight at room temperature. (B) For His tag protein elute with 15 ml His Elution Buffer into/or then transfer to a centrifugal filter device for concentration down to or just under 1 ml. Centrifuge to concentrate. If desired, imidazole can be removed, or reduced, by careful layering (to avoid mixing) with ~10 ml Wash Buffer and concentrated again. Repeat the “reverse decanting” process with another 10 ml Wash Buffer. Protein is ready for analysis and / or for use. Ensure protein is at less concentration than 500 µg / ml to avoid precipitation. Store at 4 °C. 6. Protein concentration can be determined by the Bradford assay at an absorbance of 595 nm, (typically measure 33.3 ml in 1 ml reagent) following the supplier’s instructions, referenced to a standard. Imidazole does not affect the assay.

Studying Mechanosensitive Ion Channels Using Liposomes 3.1.5. Day 5: Protein Collection and Determination of Protein Concentration

41

1. For GST fusion protein, centrifuge the beads at 1,000 × g (with slow down ramp) for 5 min. Collect supernatant as purified protein. Store at 4°C. 2. Protein concentration can be determined by the Bradford assay at an absorbance of 595 nm, (typically measure 33.3 ml in 1 ml reagent) following the supplier’s instructions, referenced to a standard. 3. For GST fusion protein, (optional), if poor thrombin cutting is suspected, analyze uncleaved protein present by loading ~15 ml of leftover beads by polyacrylamide gel electrophoresis (PAGE), or by first eluting protein off 15 ml leftover beads with 15 ml of 20 mM glutathione. Collect supernatant for PAGE analysis.

3.2. MscS- His Tag Fusion Protein Purification

This procedure takes several days, thus it is advisable to start at the beginning of a week. Membrane processing is required to be the same day as cell disruption, thus day 2 is a longer than usual work day.

3.2.1. Day 1: Overnight Starter Culture

1. Using sterile technique, on ice, to 50 ml competent M15 cells (with pRARE and pREP4) add 1 ml expression construct, mix, and let stand for 20 min. Heat shock at 42°C for 45 s, and place on ice for 2 min. 2. Add the transformation mixture into a 100 ml culture flask with 12 ml of LB and 12 ml Ampicillin (100 mg/ml- to maintain pQE plasmid), 12 ml kanamycin sulphate (25 mg/ml- to maintain pREP4 plasmid), and 12 ml chloramphenicol (25 mg/ ml-to maintain pRARE plasmid). 3. Incubate at 37°C at 240 rpm (for ¾ inch orbit) in an orbital shaker incubator.

3.2.2. Day 2: Growth, Expression, Cell Disruption, and Membrane Processing

1. Add 10 ml of overnight culture into a ~2,800 ml culture flask containing 1000 ml LB, 1 ml 100 mg/ml ampicillin, 1 ml 25 mg/ ml kanamycin sulphate, and 1 ml 25 mg/ml chloramphenicol. 2. For aeration, seal the flasks with tissue paper (such as low lint fine task wipers) and incubate at 37°C at 150 rpm (for 2 inch orbit) in an orbital shaker incubator. 3. When optical density of the culture (OD) A600 = 1.0 to 1.2 (occurs in ~2–2.5 h), let the culture stand at room temperature without shaking for an hour. 4. Induce expression with the addition of 800 ml 1 M IPTG and 4 ml glycerol (0.4% final). Incubate within shaker as previously, but at 25 °C for 4 h. 5. Centrifuge culture at around 8,000 × g for 15 min at 4°C. Discard supernatant.

42

Martinac et al.

6. For every 10 g of cell pellet resuspend (with a ~8 mm wide brush) in 10 ml PBS with 145 ml 20 mg/ml PMSF. Continue with cell disruption on the same day. 7. Add roughly a “match head” amount of DNase, mix. 8. French press (Thermo Electron Corporation) twice at 16,000 PSI. 9. Remove debris, unbroken cells, or inclusion bodies by high speed centrifugation at 8,000 × g (k-Factor ~3,600) for 22 min at 4°C. Collect supernatant for next step. 10. Pellet membranes by ultracentrifugation at 235,00 × g (k-Factor ~133) for 3 h at 4°C. Discard supernatant. 11. Suspend pellet in 5 ml Solubilization Buffer with a ~5 mm width brush. 12. Transfer suspended pellet to a 50 ml tube. Add 2.24 ml 200 mM DDM. Make up to final volume 50 ml using Solubilization Buffer (8 mM final DDM). 13. Solubilize on a roller wheel overnight at 4°C. 3.2.3. Day 3: Clarification and Column Binding, Protein Purification and Determination of Protein Concentration

1. Clarify the solubilization by ultracentrifugation at 100,000 × g for 20 min at 4°C. Recover supernatant to use. 2. Add 1.5 ml prepared cobalt-resin beads to the supernatant. Incubate on a rocker or roller for 3 h at 4°C. 3. Pour the cobalt resin suspension into a 20 ml chromatography column. Release bottom caps when resin has settled somewhat after several minutes, and allow all liquid to flow out and discard. 4. Wash the cobalt resin twice with 7.5 ml MscS Wash Buffer 1. 5. Wash the cobalt resin with 20 ml MscS Wash Buffer 2. 6. Elute with 4.5 ml MscS Elution Buffer into/or then transfer to a centrifugal filter device for concentration down to or just under 1 ml. Centrifuge to concentrate. Avoid over concentration that will result in precipitation. 7. If desired, imidazole can be removed, or reduced, by careful layering (to avoid mixing) with ~10 ml MscS Wash Buffer 1 and concentrate again as per step 6. Repeat the “reverse decanting” process with another 10 ml of MscS Wash Buffer 1. Protein is ready for analysis and/or for use. 8. Protein concentration can be determined by the Bradford assay measuring absorbance at 595 nm, (typically measure 33.3 ml in 1 ml reagent) following the supplier’s instructions, referenced to a standard. 9. When expressing from E. coli MS-channel knockout cells, ensure the protein concentration is lower than 500 µg/ml, to avoid precipitation. Expression in M15 (pQE32 or pQE70) or BL21-

Studying Mechanosensitive Ion Channels Using Liposomes

43

Rosetta cells (pET28) appears to allow further concentration of wild type MscS up to 10–15 mg/ml without precipitation. However, for some mutants, it is not recommeded to go above 4–5 mg/ml. Store at 4°C, do not freeze. 3.3. NMDA-6 × His-Tagged Protein Purification 3.3.1. Protein Expression in Sf9 cells

1. Grow Sf9 cells in Sf-900II media with 120 rpm shaking at 27°C. 2. For co-expression of NR1A and NR2A subunits, the cells are co-infected at the same MOI (multiplicity of infection) with both NR1A and NR2A recombinant baculoviruses at 3 × 106 cells/ml, using a MOI of 2–5 plaques forming unit per cell. The optimal time of harvest for the co-expression of NR1A and NR2A subunits is 72 h post infection. 3. Centrifuge the culture at 8,000 × g for 15 min at 4°C. Store the cell pellet at −80°C until required.

3.3.2. Protein Expression in CHO Cells

1. CHO cells are planted in 500 ml flasks at a density 1 × 106 cells/ml in CHO-S-SFMII (serum free, low protein media). 2. Cell culture is incubated at 37°C in a humidified atmosphere of 37°C, 7.5% CO2 in air. Samples are taken daily for determination of viable cell density using tryptan blue exclusion. 3. For transfection, cells are cultured to 80–90% confluence. 4. DNA constructs (1.6 mg/ml of transfection volume) and lipofectamine (1.6 ml) are diluted in 40 ml CHO-S-SFMII media and mixed gently. 5. After 5 min of incubation at room temperature the diluted DNA constructs and lipofectamine are combined and incubated for 20 min at room temperature to allow complex formation to occur. 6. The DNA-Lipofectomine complexes are slowly added to the flask containing CHO-S-SFMII media and gently mixed by swirling the flask. 7. Cells are incubated at 37°C, 7.5% CO2 on shaker at 150 rpm for 24 or 48 h.

3.3.3. Protein Purification

1. For cell lysis, the cell pellet from a 1 L expression is resuspended in 100 ml cell lysis and solubilization buffer without the detergent. 2. Cells are disrupted by sonication using a probe at 75 W for 15 s with 10–20 s cooling periods on ice (3×). 3. The OG concentration is adjusted to 3% after sonication and a sample is placed on rotating wheel for 20 min. 4. The lysed sample is clarified by centrifugation at 18,000 × g for 15 min at 4°C.

44

Martinac et al.

5. The clarified lysate is applied to Talon resin at a ratio 10:1 (e.g., 12.5 ml of lysate is applied to 1.25 ml talon) and placed on a rotating wheel for 20–30 min. 6. Unbound proteins are removed by centrifugation at 700 × g for 2 min at 4°C. 7. After binding, each 1.25 ml talon is washed for 10 min using 50 ml wash buffer. 8. Elution of the protein is performed by transferring 2.5 ml talon to 10 ml disposable column (Biorad, Hercules, CA) and protein is eluted with 4 × 1 ml of Elution Buffer. 9. For gel filtration, 500 ml of Talon purified protein sample is loaded onto a Superdex 75 10/300 GL column (GE Healthcare, Rydalmere NSW) at 0.25 ml/min using gel filtration buffer, and fractions (25 × 0.5 ml) are collected for analysis by Western blot and reconstitution into liposomes. 3.4. SDS-PAGE

Typically, a 12 % SDS PAGE gel is used for gel analysis of both MscL and MscS. 1. Assemble the electrophoresis vertical mini electrophoresis rig system to manufacturer’s instructions and use precast gels as prescribed. If manual gels are to be cast, the following gel recipe is generally suitable for a 1 mm thick mini protein gel: 2. In a disposable tube, make 10 ml 12% Running Gel by mixing together (without undue aeration): 3.35 ml distilled water, 2.5 ml 1.5 M Tris-HCl pH 8.8, 100 ml 10% SDS, 4 ml 30% acrylamide/Bis solution, 50 ml 10% APS, mix, 5 ml TEMED, mix. 3. With a pipette fill the bottom ~3/4 of constructed glass plate structure with the freshly prepared Running Gel before it sets. Usually two gels are prepare with the 10 ml solution, as usually two gels are required for electrophoresis within the rig. 4. Without mixing, carefully drop about 500 ml of water to the top to act as an air barrier. Allow the gel to set, which occurs within about half an hour. 5. In a disposable tube, make 10 ml 12% Stacking Gel by mixing together (without undue aeration): 3 ml distilled water, 1.25 ml 0.5 M Tris-HCl pH 6.8, 50 ml 10% SDS, 650 ml 30% acrylamide/Bis solution, 25 ml 10% APS, mix, 5 ml TEMED mix. 6. Turn gel setting rig upside down to remove water, use lint free tissues to remove remaining water if required. Return rig to normal position. 7. With a pipette, add the Stacking Gel on top of the Running Gel, to the top of the constructed glass plate structure (see Note 2). Place the well-combs into the top to allow sample

Studying Mechanosensitive Ion Channels Using Liposomes

45

well moulds to set. Place combs without introducing air bubbles. 8. Once the Stacking Gel layer has set, place the gel(s) within the running rig (as per manufacturer’s instructions) and add 1 × Elpho Buffer (at the same time to stop the gel from drying) to the top/full level, so that buffer is above the well layer. The apparatus will have inner and outer buffer reservoirs – both should be full to the same level to halt buffer from possible leak below the gel top and to allow uniform dissipation of heat. Remove well-combs. 3.4.1. Preparation of Protein Samples for PAGE and Electrophoresis

1. To ~20 ml protein sample mix in ¼ volume 4 × Sample Buffer in a microcentrifuge tube. Cap. 2. Heat (denature) sample using a hot block at 95°C for 5 min. Denature protein molecular weight marker only if instructed by manufacturer. (Fermentas #SM0431 requires denaturing.). 3. Spin any condensation down, and load sample(s) and molecular weight marker into separate gel wells (see Note 3). 4. To manufacturer’s instructions, cover rig and run electrophoresis at 100 V (or as appropriate) until the bromophenol blue marker dye is near the end of the gel. 5. Remove gel from glass plate and incubate, covered, with rocking in Coomassie Stain for several hrs to overnight. 6. Remove Stain solution (can be reused), rinse briefly with Destain and incubate covered with rocking in Destain for several hrs to overnight or until sample bands are distinct, and background is down to a desired level. Refresh Destain if saturated with stain. Adding lint free tissues to the Destain solution will sequester excess brilliant blue diffusing from the gel.

3.4.2. Gel Analysis: MscL

Wild-type MscL will appear at a molecular weight position of around 15 kDa, or theoretically higher at 16 kDa depending on how many further extra amino acids are added due to the cloning construct used. For the Fermentas #SM0431 marker, MscL will migrate in between the two smallest marker bands at 18.4 and 14.4 kDa. For reasons not yet defined, MscL often appears as a doublet at its size position. His-tag reliant purifications may yield faint contamination of multiple bands, whilst GST-tag reliant purifications may yield specific contaminants potentially identified as: 37 kDa bovine thrombin 29 kDa of the GST (Glutathione S-transferase) cleaved protein 58.5 kDa of the GST dimer 44 kDa MscL-GST fusion protein

46

Martinac et al.

3.4.3. Gel Analysis: MscS

The MscS subunit is about 31 kDa and will migrate at this size. Despite the denaturing conditions of SDS PAGE, an intense dimer band may exist. If denaturing is suboptimal, trimer and tetramer bands may also be detected. Of notable importance, degraded MscS will yield a band (a sub-band) just slightly smaller than the 31 kDa band, which may be reflected in slightly smaller multimers. Although the sub-band would normally not be present in fresh preparations, its presence, which may become the predominant product, will be reflective of compromised MscS function, and is a detrimental indicator. MscS, which can be stored for several months or to a year or so, should be analyzed on a gel periodically (or by other characterization methods) to detect the presence of the sub-band.

3.4.4. Gel Analysis: NMDA Receptor Channel Proteins

The metal affinity chromatography resulted in several low molecular weight mass bands visible on SDS/PAGE gel indicating nonspecific bindings and/or breakdown products of NR1a and NR2A subunits. Gel filtration chromatography purification is required to separate two main protein bands corresponding to 170 kDa for NR2A subunit and 100 kDa for the NR1a subunit from the contaminants.

3.5. Liposome Reconstitution

We have used several methods for reconstitution of MS channels into artificial liposomes (Fig. 1). Details are described below.

3.5.1. Dehydration/ Rehydration (D/R) Method

This method was the first developed for studies of MS channels by reconstitution into liposomes (17) and has since been the most frequently used method for characterization of biophysical and pharmacological properties of MS channels (7, 11, 13, 18) (Fig. 1a).

3.5.1.1. Preparation of Liposome Vesicles Using Bio-Beads SM2 (BioRad)

1. Rinse glass test-tube with chloroform, discard, and dry with nitrogen stream. 2. Dissolve 20 mg lipid into 1 ml chloroform in the glass tube. 3. Obtain an even lipid film on the surface of the tube by swirling the tube rapidly under a nitrogen stream until all the chloroform has evaporated. 4. Completely remove all chloroform by placing the tube under a high nitrogen stream for at least 15 min. 5. Resuspend lipid(s) in 2 ml of D/R Buffer and vortex for 60 s until cloudy. 6. Bath sonicate within a standard sonicator cleaner for 10–20 min. 7. Within a sealable tube, mix 200 ml (2 mg lipid) of liposome vesicles (from the sonication) with an amount of purified mechanosensitive protein to obtain the desired protein-tolipid wt/wt ratio (see Note 4). Fill to 2 ml with D/R Buffer.

47

Studying Mechanosensitive Ion Channels Using Liposomes

a D/R buffer

Purified MscL protein Rehydrated lipids in D/R buffer

Sonicate

3 hr rocking/Remove detergent (Biobeads), centrifuge

Dried lipid film

b

Dehydration/ Rehydration

Patch clamp pipette Blister

Pellet

MscL or MscS

0.4 M Sucrose

Oven incubation, 3 hrs at 45 ºC

Reconstituted proteoliposome

3 hr rocking (or overnight rocking)

Patch clamp pipette Lipid cloud

Dried lipid film

c

Reconstituted proteoliposome

O-ring

Purified MscL protein*

NMDAR protein*

Blister

Reconstituted proteoliposome

AC, 3 V, 5 Hz, 2 hours Patch clamp pipette

Electroformation Dried lipid film

GUVs

Blister

Fig. 1. Diagram showing different liposome reconstitution methods used for studies of MS channels (Subheading 3.5.3). (a) Dehydration/rehydration method. (b) Sucrose method. (c) Electric field method (Vesicle PrepPro, Nanion Technologies). (* Please note MscL is added before formation of GUVs, whereas NMDAR protein is added after GUV formation.)

8. Roller mix at room temperature for 1 h. 9. Add 320 mg of tissue blotted Bio-Beads per ml of protein solubilized in 1 mM DDM. (The amount of Bio-Beads used is determined by total detergent content.) The mixture is gently stirred for 3 h. 10. Allow beads to settle to the bottom of the tube, and use the supernatant which contains detergent-free proteoliposomes. 11. Micro ultracentrifuge the supernatant at 250,000 × g for 45 min at 4°C. 12. Remove supernatant completely. Resuspend pellet in 80 ml D/R Buffer by pipetting.

48

Martinac et al.

13. Place 2–4 spots of 20 ml of proteoliposome suspension on a cleaned microscope slide. 14. Place slide(s) in vacuum desiccator for at least 6 h at 4°C. (Dehydration step.). 15. Add 20 ml D/R Buffer to each spot and leave overnight at 4°C within a Petri dish containing filter paper moistened with distilled water. (Rehydration step.) Sample is now ready to transfer into a patch recording chamber. 3.5.2. Sucrose Method

Our laboratory recently developed this reconstitution method (19), which has several advantages over the classical D/R method (Fig 1b). Requiring preparation times of 6 h or less, this method significantly saves time compared with the D/R procedure, which on average requires two days preparation time. It also represents the first highly reproducible method for incorporation of the MscS channel protein, which does not incorporate readily into azolectin liposomes or liposomes made of mixtures of pure lipids. Therefore, this new method has the potential to be used for studies of ion channels that may be difficult to study by liposome reconstitution using D/R method.

3.5.2.1. Day 1

1. Dissolve 10 mg of PC lipid (azolectin) in 1 ml of chloroform. 2. Take 200 mL aliquots and place into small test tubes (12 × 75 mm is ideal). 3. Remove solvent under a stream of N2 whilst swirling tube. 4. After continued drying of the lipid film under a stronger stream of N2 add 2 mL of H2O as a prehydration step (5 min)place drop of H2O directly onto the top of the lipid film; after prehydration the clear film develops opaqueness. 5. To the prehydrated lipids add 1 ml of 0.4 M sucrose and place in an oven or water bath at 45°C for 3 h. After this time the lipid will peel off the surface of the glass. 6. Add required protein amount to achieve a desired protein/lipid ratio. Important: add to bottom of tube, and do not shake. 7. Place on an orbital mixer at approximately 100 rpm and shake for 3 h. 8. Protein has now incorporated and can be used for experiments. However to ensure complete incorporation it is recommended to mix overnight.

3.5.2.2. Day 2

1. The lipids should have formed a cloud floating in the middle of the solution. Add approximately a cm³ of prepared BioBeads, and return to the mixer for 3 h. 2. Add 2 ml of the lipid cloud to recording bath, and they will form very large blisters.

Studying Mechanosensitive Ion Channels Using Liposomes

49

3.5.3. Electric Field Method (Vesicle PrepPro, Nanion Technologies)

20 ml of 5 mM lipid dissolved in chloroform was dried on the conductive side of an ITO-slide. A rubber O-ring acts as a spacer in between the two slides (Fig 1c).

3.5.3.1. MscL Protein Reconstitution

1. Once dried, add 250 ml of 210 mM D-sorbitol into the confines of the O-ring. 2. Add required protein/lipid ratio (typically 1:900 w/w) 3. The settings used are: Temperature = 23°C Frequency = 5 Hz Amplitude = 3 V 4. Allow to run for 2 h 5. After process is completed, extract GUVs with a 1 ml pipette.

3.6. Patch-Clamp Recording From MS Channels Reconstituted Into Liposomes

The results from studies of MS channels reconstituted into artificial liposomes have been described in numerous publications. Several reviews and original research papers (2, 3, 7, 11, 13, 17, 18, 20–22) highlight major findings of the structural and functional properties of MS channels using liposome reconstitution techniques.

3.6.1. Patch-Clamp Pipettes

1. Recording micropipettes are formed from borosilicate glass microcapillaries by using a pipette puller (e.g., P-87 Flaming/ Brown, Sutter Instrument Co., Novato, CA). Pulled recording pipettes should be ~1 mm in diameter corresponding to a pipette resistance in the range of 3.0–6.0 MW in Recording Solution. 2. To reduce electrical noise, pipette tips can be coated using Sylgard 184 (23) or transparent nail enamel (24) but is not required if single channel current is more than around 15 pA.

3.6.2. Recording

1. A 2–5 ml aliquot of liposomes made using any of the above methods is introduced into the recording chamber filled with the Recording Solution at 22°C. Recording pipettes are back filled with the same recording solution cleaned through a 0.22 mm filter. (Symmetrical solutions are preferentially used for determining single channel conductance of MS channels.). 2. Unilamellar blisters will be visible from several minutes up to 30 min (17) by which time the liposomes will have settled to the bottom of the chamber. The most common causes of “blistering” failure are incorrect pH or temperature of the Recording Solution. 3. Lower recording pipette to a blister and form a gigaohm seal (>1 GW) by applying suction to the patch pipette with a syringe. Suction is halted when a sudden decrease in pipette current occurs. MscL and MscS activities are usually recorded in excised liposomes patches (18, 19) which are obtained by next exposing the pipette tip to air for an instant only. (Fig. 2b.).

50

Martinac et al.

Fig. 2. (a) Confocal imaging of liposomes containing AlexaFluor488-labeled MscS proteins (Subheading 3.7) reconstituted by the Sucrose method at 1:100 (w/w) ratio (Subheading 3.5.2). Images were recorded about 10 min after addition of liposomes to the imaging buffer (same as buffer used for patch-clamp recording) and were obtained by an Olympus FluoView1400 Laser Scanning Microscope (Olympus) using a Nikon 60× water-immersion objective, NA 1.00 (Nikon, Japan), with excitation at 488 nm by an Argon laser (Melles–Griot). (b) Patch-clamp recording of the AF488-MscS liposomes prepared by the same method showing that the channel activity (recorded at pipette voltage VP = +30 mV) is not compromised by labeling. Patch-clamping was performed as per protocols described in Subheading 3.6

NMDA receptor channel activity is recorded after inclusion of agonists in the pipette solution or adding agonists directly to the bath solution (11). 4. Ion currents arising from activation of MS channels using negative pipette pressure are recorded with a patch clamp amplifier (e.g., Axon 1D or Axopatch 200B, Axon Instruments). The suction is monitored with a piezoelectric pressure transducer (Omega Engineering, Stamford, USA). Currents are usually filtered at 2 kHz and digitized at 5 kHz for offline analysis. 5. Single channel recordings can be analyzed using software such as pCLAMP (Axon Instruments) or in-house applications. 6. For NMDA receptor channels the protein is mixed with GUVs at required ratio depending on protein concentration (usually 1:10 or 1:20). 3.7. Confocal Microscopy of Liposome Reconstituted MS Channels

Confocal imaging of fluorescent-tagged MS proteins reconstituted into liposomes is often useful to observe protein expression, clustering phenomena, or for analytical studies such as Föster Resonance Energy Transfer (FRET) studies. MS proteins can be fluorescently tagged either genetically or by generating a single cysteine mutant for site-specific attachment of thiol-reactive probes (Fig. 2a). While an MscL protein tagged genetically with Green Fluorescent Protein has been expressed and characterized (25), this approach is more problematic for MscS due to its large cytoplasmic domain. Here we describe a protocol for confocal imaging of liposome reconstituted MS proteins tagged with the fluorescent dye, Alexa Fluor 488 C5-maleimide (AF488).

Studying Mechanosensitive Ion Channels Using Liposomes 3.7.1. Protein Labeling

51

1. Generate expression constructs of single cysteine mutants using site directed mutagenesis. Suitable kit systems include QuickChange (Stratagene) or Phusion (Finnzymes). 2. Express and purify the six histidine-tagged single cysteine mutants using protocols similar to wild-type proteins described in Subheading 3.1, except that the Wash and Elution buffers are previously deoxygenated thoroughly (before addition of DDM to avoid frothing) and supplemented with 10 mM tris(2-carboxyethyl)phosphine (TCEP) to reduce any protein disulfide bonds that may have formed. A suitable method of deoxygenation employs percolating nitrogen gas for 15 min through a fish-tank aerator placed at the bottom a buffer. 3. Set up a labeling reaction consisting of 10 nmole of protein (usually ~300 mg), to a tenfold excess of AF488 (i.e., use a 100 mg prepared aliquot) made up to a final volume of 2 ml with Labeling Buffer (see Notes 5–7). Rock for 2 h at room temperature or overnight at 4°C. 4. Optional: stop reaction using 5 mM b-mercaptoethanol final concentration. 5. Transfer reaction to a dialysis cassette and place in 2 L chilled Dialysis Buffer with stirring at 4°C. Replace the buffer twice, after 2 and 4 h, and then dialyse overnight. 6. Collect protein from dialysis cassette and concentrate if necessary. 7. Reconstitute labeled proteins into azolectin lipids by either the D/R or Sucrose Method (Subheading 3.5) at desired concentrations. For imaging of reconstituted liposomes, protein to lipid ratios of 1:100 to 1:250 (w/w) usually suffices. However, for applications such as FRET, it is advisable to use ratios of 1:20 to 1:50 (w/w). 8. Add liposomes to the buffer used for patch-clamp recordings (Subheading 3.6) and wait about 10 min for blisters to form. 9. Image liposomes using a 60× lens and appropriate laser excitation (488 nm for AF488) and power settings (See Fig. 2a).

4. Notes 1. End of pipette may be cut off for a bigger hole to avoid clogging. Also do not place tip to very bottom of column. Allow room to enable beads to enter the tip. 2. Avoid introducing air bubbles because bubbles would cause mould disruption. 3. Long narrow gel load micropipette tips are available.

52

Martinac et al.

4. Usually, a 1:4000 to 1:6000 protein-to-lipid ratio is used for the purpose of single channel recordings. To obtain a membrane fraction containing many mechanosensitive proteins a 1:100 to 1:200 protein-to-lipid ratio may be used. 5. Handling of Alexa Fluor dye should be done in subdued light. 6. In case the protein is not labeled immediately after purification, it should be pre-treated with a tenfold excess of TCEP for 1 h at room temperature. 7. The pH of the labeling reaction solution should be between 7.2 and 7.5; therefore, it is advisable to ensure that the added protein (previously eluted at pH 6.0) be at an adequate concentration.

Acknowledgments We wish to thank Dr Stephen Hughes for his contribution to studies of magnetic field effects on the MscL channels using liposome reconstitution technique. This research has been supported by grants of the Australian Research Council and the National Health and Medical Research Council of Australia to B. Martinac and A. Kloda. References 1. Kung CA (2005) A possible unifying principle for mechanosensation. Nature 436:647–54 2. Martinac B (2005) Force from lipids: physical principles of gating mechanosensitive channels by mechanical force revealed by chemical manipulation of cellular membranes. The Chemical Educator 10(2):107–14 3. Hamill OP, Martinac B (2001) Molecular basis of mechanotransduction in living cells. Physiol Rev 81:685–740 4. Martinac B (2001) Mechanosensitive channels in prokaryotes. Cell Physiol Biochem 11:61–76 5. Markin VS, Martinac B (1991) Mechanosensitive ion channels as reporters of bilayer expansion. A theoretical model. Biophys J 60:1120–7 6. Martinac B, Adler J, Kung C (1990) Mechanosensitive ion channels of E. coli activated by amphipaths. Nature 348:261–3 7. Maroto R, Raso A, Wood TG, Kurosky A, Martinac B, Hamill OP (2005) TRPC1 forms the stretch-activated cation channel in vertebrate cells. Nat Cell Biol 7:179–85 8. Maingret F, Patel AJ, Lesage F, Lazdunski M, Honoré E (2000) Lysophospholipids open the two-pore domain mechano-gated K+

9.

10.

11.

12.

13.

channels TREK-1 and TRAAK. J Biol Chem 275:10128–33 Zhang Y, Gao F, Popov VL, Wen JW, Hamill OP (2000) Mechanically gated channel activity in cytoskeleton-deficient plasma membrane blebs and vesicles from Xenopus oocytes. J Physiol 523(Pt 1):117–30 Zhou XL, Batiza AF, Loukin SH, Palmer CP, Kung C, Saimi Y (2003) The transient receptor potential channel on the yeast vacuole is mechanosensitive. Proc Natl Acad Sci USA 100:7105–10 Kloda A, Lua L, Hall R, Adams DJ, Martinac B (2007) Liposome reconstitution and modulation of recombinant N-methyl-D-aspartate receptor channels by membrane stretch. Proc Natl Acad Sci USA 104:1540–5 Perozo E, Cortes DM, Sompornpisut P, Kloda A, Martinac B (2002) Open channel structure of MscL and the gating mechanism of mechanosensitive channels. Nature 418:942–8 Perozo E, Kloda A, Cortes DM, Martinac B (2002) Physical principles underlying the transduction of bilayer deformation forces during mechanosensitive channel gating. Nat Struct Biol 9:696–703

Studying Mechanosensitive Ion Channels Using Liposomes 14. Corry B, Rigby P, Liu ZW, Martinac B (2005) Conformational changes involved in MscL channel gating measured using FRET spectroscopy. Biophys J 89:L49–51 15. Hughes S, El Haj AJ, Dobson J, Martinac B (2005) The influence of static magnetic fields on mechanosensitive ion channel activity in artificial liposomes. Eur Biophys J 34:461–8 16. Petrov E, Martinac B (2007) Modulation of channel activity and gadolinium block of MscL by static magnetic fields. Eur Biophys J 36:95–105 17. Delcour AH, Martinac B, Adler J, Kung C (1989) Modified reconstitution method used in patch-clamp studies of Escherichia coli ion channels. Biophys J 56:631–6 18. Häse CC, Le Dain AC, Martinac B (1995) Purification and functional reconstitution of the recombinant large mechanosensitive ion channel (MscL) of Escherichia coli. J Biol Chem 270:18329–34 19. Battle AR, Petrov E, Pal P, Martinac B (2009) Rapid and improved reconstitution of bac terial mechanosensitive ion channel pro teins MscS and MscL into liposomes using

53

a modified sucrose method. FEBS Letters 583: 407–412 20. Vasquez V, Cortes DM, Furukawa H, Perozo E (2007) An optimized purification and reconstitution method for the MscS channel: strategies for spectroscopical analysis. Biochemistry 46:6766–73 21. Sukharev S (2002) Purification of the small mechanosensitive channel of Escherichia coli (MscS): the subunit structure, conduction, and gating characteristics in liposomes. Biophys J 83:290–8 22. Martinac B, Kloda A (2003) Evolutionary origins of mechanosensitive ion channels. Prog Biophys Mol Biol 82:11–24 23. Sakmann B, Neher E (1995) Single-channel recording. Plenum Press, New York and London 24. Martinac B, Buechner M, Delcour AH, Adler J, Kung C (1987) Pressure-sensitive ion channel in Escherichia coli. Proc Natl Acad Sci USA 84:2297–301 25. Norman C, Liu ZW, Rigby P, Raso A, Petrov Y, Martinac B (2005) Visualisation of the mechanosensitive channel of large conductance in bacteria using confocal microscopy. Eur Biophys J 34:396–402

Chapter 5 Studying Amino Acid Transport Using Liposomes Cesare Indiveri Abstract The transport of amino acid across the membranes has great importance in cell metabolism. Specific experimental methodologies are required for measuring the vectorial reactions catalyzed by the membrane transporters. So far, the most widely used technique to study amino acid transport was the measure of amino acid flux in intact cell systems expressing a specific transporter. Some limitations in this procedure are caused by the presence of endogenous transporters and intracellular enzymes and by the inaccessibility of the intracellular compartment. Alternative experimental strategies which allow to reducing the interferences and improving the handling of the internal compartment would be useful to the amino acid transport knowledge. An experimental protocol, which makes use of liposomes to study the transport of amino acid mediated by the glutamine/amino acid (ASCT2) transporter, solubilized from rat kidney brush borders, is described. The procedure is based on the reconstitution of the transporter in liposomes by removal of detergent from mixed micelles of detergent, solubilized protein, and phospholipid. The transport is assayed in the formed proteoliposomes measuring the Na+ dependent uptake of l-[3H]glutamine in antiport with internal l-glutamine. This method allows measuring the transport activity under well controlled experimental conditions and permits performing experiments which cannot be realized in intact cell systems. Key words: Liposomes, Reconstitution, Transport, Membrane, Amino acids, Glutamine

1. Introduction The interest in the study of amino acids transport across the cell membrane increased exponentially in the last decade. This was also due to the relevance of this research field in human physiology and pathology. In mammalian cells the transport of amino acids is carried out by a large number of proteins which are different in primary structure and function. Most of these transport systems have been functionally characterized using intact eukaryotic cell systems in which specific transporters were expressed. Other V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_5, © Humana Press, a part of Springer Science+Business Media, LLC 2010

55

56

Indiveri

transporters, which have been predicted on the basis of gene sequences, are still orphan with respect to the function. The knowledge of the state of the art in this field is clearly and systematically described in a number of reviews (1–8). Even though intact cell experimental models allowed classification of a large number of amino acid transporters, technical limitations of the models made several aspects of the structure, function, and regulation of the transporters still unclear or unknown. Thus, the development of alternative experimental strategies would give benefits to the knowledge of amino acid transport. A suitable model for transport studies is the liposome system, which has often been helpful in defining and clarifying functional properties of membrane transporters of different types. Thanks to this model it is possible to perform experiments which are forbidden in intact cells. An important advantage of reconstituting membrane transporters in liposomes with respect to the study in cell systems is the control of experimental conditions in the internal compartment, which help to determine internal parameters like the Km for substrates or the influence of internal effectors on the transport. In this respect, a reconstitution method which allows the insertion of the protein in the liposomal membrane in the same orientation of the cell membrane is particularly reliable since it mimics the physiological situation, i.e., the correspondence of the intraliposomal and extraliposomal with the intracellular and extracellular sides. Other advantages are the reduction of interferences due to the absence of enzymes which could modify the substrates and the absence of different transporters in the same vesicles that is guaranteed by the higher phospholipids/protein ratio (one transport protein per liposome) with respect to the cell membrane (several transport proteins per cell); this feature has also the effect of prolonging the time course of the substrate uptake, leading to better resolution of the initial transport rate and the kinetics. The liposome model provides the possibility of modifying the lipid composition of the membrane to study the influence of specific lipids on the transport. To get further insights into the transport of amino acids, we have pointed out a procedure of reconstitution of the glutamine/amino acid transporter ASCT2 extracted from apical plasma membranes (brush-borders) of rat kidney where the transport of glutamine is very active being involved in essential physiological functions. The previous studies in intact cells expressing the ASCT2 protein (9–14) showed that it cata lyzes antiport of glutamine with neutral amino acids which is dependent on extra cellular Na+ in kidney, intestine, lung, muscle,

Studying Amino Acid Transport Using Liposomes

57

testis, as well as in nervous tissue. The external Km’s for glutamine, other amino acids, and Na+ were determined and it was proposed that, differently from other neutral amino acid transporters, ASCT2 could transport also the anionic amino acid glutamate. The reconstitution of the glutamine/amino acid transporter in liposomes, besides confirming the previous data, permitted a deeper understanding of the ASCT2 transporter (15, 16). The reconstituted transporter, which is inserted in the liposomal membrane with the same orientation of the cell membrane, is inhibited by externally added Cys and Lys specific reagents. The internal Km’s for the amino acids were determined; their values are 20 times higher than the external ones. The mechanism of the complex transport reaction was found to be random simultaneous. The transporter is regulated by internal (intracellular) ATP. It also catalyzes a glutamate/glutamine antiport mode in which, glutamate, not the zwitterion glutamic acid, is transported with lower Km at acidic pH. The protocol for performing the reconstitution and the transport assay of the glutamine/amino acid transporter ASCT2 from rat kidney is described here. This protocol requires, as starting materials, brush border membranes from rat kidney which are prepared according to a previously described method (17). The transporter is solubilized from the brush borders with the non ionic detergent C12E8 and reconstituted in liposomes. The reconstitution procedure derives from a method which was originally pointed out for mitochondrial transport proteins (18, 19). It is based on the slow removal of detergent from mixed micelles of detergent phospholipid and protein by repeated passages through chromatography columns filled with hydrophobic resin. This leads to the formation of unilamellar phospholipid vesicles with the protein inserted in the membrane. The method has been adapted for the reconstitution of plasma membrane transporters. This goal has been achieved by changing the type of hydrophobic resin used and the critical parameters of the reconstitution like the protein concentration, the detergent/phospholipid ratio,and the number of repeated passages through the chromatography columns (15, 20). The transport is finally assayed by measuring the uptake of radioactive l-glutamine externally added to the proteoliposomes. The transport is active only in the presence of external Na+ and internal unlabeled l-glutamine or other neutral amino acids, since ASCT2 catalyzes an obligatory antiport of neutral amino acids which is dependent on extraliposomal Na+. Using this protocol, several types of experiments can be performed, by changing or adding some components or modifying some steps of the procedure.

58

Indiveri

2. Materials All the solutions and suspensions must be prepared in bi-distilled water. 2.1. Preparation of Brush Border Membranes

1. Buffer 1: 0.3 M sucrose, 5 mM EGTA 12 mM Tris-HCl pH 7.4. 2. Buffer 2: 0.15 M sucrose, 2.5 mM EGTA 6 mM Tris-HCl pH 7.4. 3. 50 mM MgCl2.

2.2. Solubilization of the Transporter

1. Brush borders. Aliquots of 50 µl brush borders prepared as described in Subheading 3.1 (containing from 150 to 300 µg protein depending on the preparations) stored at −20°C. 2. Solubilization buffer: 1.95 % C12E8 (octaethylene glycol monododecyl ether from Fluka). The solution can be stored at 0–4°C for one week.

2.3. Reconstitution of the Brush Border Extract

1. Stock buffer pH 7.0: 500 mM Hepes-Tris pH 7.0. Prepare the buffer dissolving Hepes in a volume of water about 70% of that final; then add Tris powder up to pH 7.0 and water up to the final volume. Use this solution to prepare all the diluted solutions. The stock buffer pH 7.0 can be stored at 0–4°C for 4–6 weeks. 2. 100 mM l-glutamine: Prepare fresh and use this solution to prepare all the diluted solutions. 3. Amberlite resin. Swell Amberlite XAD-4 resin (20-50 mesh from Fluka) adding 2 volumes of methanol to about 50 g resin. Mix gently using a glass stick; then leave for about 30 min at room temperature. Remove the supernatant methanol and repeat the procedure from 5 to 8 times until the supernatant methanol remains limpid. Then wash the resin with excess water 5 times (leave the resin in water 15 min among the washings). Store the resin at 0–4°C in 2 volumes of water. The resin can be stored for 4 weeks without adding conservatives. 4. Amberlite columns. Prepare 4 columns of about 0.5 cm internal diameter. Pasteur pipettes whose lower outlets are closed by cotton can be used (see Note 1). Fill the columns with the Amberlite resin up to 2.5 cm height. The Amberlite columns cannot be reused. 5. Amberlite equilibrating buffer. Freshly prepare 15 ml of a solution of 30 mM glutamine, 20 mM Hepes/Tris at pH 7.0 using the 100 mM l-glutamine, and the stock buffer pH 7.0. The concentration of glutamine may vary depending on the experiments (see Subheading 3.4, item 12 and related Notes).

Studying Amino Acid Transport Using Liposomes

59

6. Solution of 10% C12E8 (see also Subheading 2.2, item 2). The solution can bnume stored at 0–4°C for 4 weeks. 7. Preformed liposomes. Prepare the liposomes dissolving 1 g phospholipid (egg yolk phospholipids from Fluka) in a final volume of 10 ml water. Vortex to homogeneity, i.e., until no solid particles are present. Then sonicate 2 ml of the suspension in a glass tube for 2 min in pulse, i.e., 1 s sonication, 1 s intermission using a sonicator (at least 100 W of maximum power equipped with a 3 mm diameter micro tip) set at 40 W output power. The tube must be maintained at low temperature in an ice/water bath during sonication (see Note 2). The preformed liposomes can be stored at 0–4°C for 3–4 days. 2.4. Transport Assay

1. Sephadex resin. Swell about 2 g of Sephadex G-75 (40-120 µm dry bead diameter from Sigma-Aldrich) in excess water overnight. Then, eliminate excess air by vacuum for 15 min. 2. Columns A: Fill 4 columns of 0.7 cm internal diameter up to 15 cm height (Econo-Columns from Bio-Rad are suitable for this purpose) with Sephadex resin. The columns can be reused (see Notes 3 and 4). Store the columns at 0–4°C. 3. Buffer A: 30 mM sucrose and 20 mM Hepes-Tris pH 7.0 (use the stock buffer pH 7.0 solution). Prepare fresh. 4. Columns B: Fill 16 columns of 0.6 cm internal diameter up to 8 cm height (Glass Columns-Reusable from Pierce are suitable for this purpose) with Sephadex resin. The columns can be reused (see Notes 3 and 4). Store the columns at 0–4°C. 5. Buffer B: 50 mM NaCl. The solution can be stored for 1–2 weeks. 6. 100 mM l-glutamine (see Subheading 2.3, item 2). 7. 2 M NaCl can be stored at room temperature for several weeks. 8. Stock buffer pH 7.0 (see Subheading 2.3, item 1) 9. Labeled glutamine. Prepare freshly the solution with 2 µl of 100 mM l-glutamine, 50 µl of 2 M NaCl, 4 µl of stock buffer pH 7.0, 2 µl radioactive l-[3H]glutamine (l-[G-3H]glutamine from GE Healthcare, 1 mCi/ml), and 42 µl water (see Note 5). The concentrations of labeled glutamine may vary depending on the experiment (see Subheading 3.4, item 12 and related Notes). 10. Inhibitor. Freshly prepared solution of 0.8 mM mersalyl (mersalyl acid from Sigma-Aldrich) in 20 mM Hepes/Tris pH 7.0 (use the stock buffer pH 7.0).

60

Indiveri

3. Methods The method for the study of the transport mediated by the glutamine/amino acid transporter ASCT2 in proteoliposomes consists of three main steps: the solubilization of brush borders, the reconstitution, and the transport assay. The brush borders are prepared using a procedure previously described by other authors (17) with few modifications: details to carry out this procedure are described in Subheading 3.1. The brush border extract is inherently labile. To obtain reliable results it is important to prepare all the materials needed for the whole procedure before thawing the brush borders for the solubilization step. Initial reconstitution mixtures are then prepared mixing the brush border extract with phospholipids and further detergent to obtain mixed micelles. The mixtures are repeatedly chromatographed on Amberlite XAD-4 columns. After repeated passages through the resin, the detergent is removed; thus, phosholipids and proteins form the proteoliposomes. These vesicles contain, in the internal and external compartments, the solutes added in the initial reconstitution mixture and in the Amberlite equilibrating buffer, i.e., the substrate glutamine and the buffer. The reconstituted proteoliposomes are stable up to 4 h at 0–4°C. The external substrate is removed by size exclusion chromatography on Sephadex G-75, before starting the transport assay. This procedure permits the starting of the transport with a concentration of external labeled glutamine (0.1 mM) lower than the concentration of the internal unlabeled glutamine (30 mM). In the case of antiport systems like the glutamine/amino acid transporter ASCT2, the low external/internal concentration ratio leads to the accumulation of the radioactive substrate (added outside) in the internal compartment of the proteoliposomes. 50 mM Na+ (as NaCl) is added together with 0.1 mM l-[3H]glutamine. The transport reaction is stopped by the inhibitor mersalyl (20 µM) added at different times. In control samples the inhibitor is added together with the l-[3H]glutamine and Na+. The l-[3H]glutamine taken up by these controls represents the unspecific permeability of the proteoliposomes that will be subtracted from the l-[3H] glutamine taken up by the samples. After stopping the transport at various times, the external radioactivity is removed by Sephadex G-75 chromatography and the radioactivity entrapped inside the vesicles is counted. The transport activity is calculated from the entrapped radioactivity data. The method described allows obtaining reliable time courses of the transport process which represent the basis for every type of experiment (see Subheading 3.4, step 12) - to study the kinetics or the dependence of the transport on different types of inhibitors and effectors (specific reagents, nucleotides, ions, phospholipids).

Studying Amino Acid Transport Using Liposomes

3.1. Preparation of the Brush-Border Membranes

61

1. Cut thin slices (1–2 mm thickness) from kidney cortex (16–20 kidneys from male wistar rats of about 200 g) upto collect about 10 g of material. Put the slices in cold buffer 1 (previously kept at 0°C for 1 h) and wash 3 times to eliminate the blood. 2. Add 60 ml of cold buffer 1 and homogenize in a blender (good quality kitchen blender of about 500 ml) at maximal speed for 3 min. 3. Add cold MgCl2 solution up to reach a final concentration of about 12 mM (see Note 6), mix and keep 15 min at 0°C. 4. Centrifuge the homogenate in 50 ml tubes at 2,600 g in a preparative centrifuge for 15 min at 0°C. Collect together the supernatants. 5. Centrifuge the supernatant in 50 ml tubes at 18,500 g in a preparative centrifuge for 30 min at 0°C. Collect the sediments and resuspend together with 30 ml of cold buffer 2. 6. Homogenize with a 30 ml potter and then add cold MgCl2 solution up to reach a final concentration of about 12 mM (see Note 6), mix and keep 15 min at 0°C. 7. Centrifuge the suspension in 50 ml tubes at 2,600 g in a preparative centrifuge for 15 min at 0°C. Collect together the supernatants. 8. Centrifuge the supernatant in 50 ml tubes at 18,500 g in a preparative centrifuge for 30 min at 0°C. Collect the sediments and resuspend together in 30 ml of cold buffer 2. 9. Homogenize with a 30 ml potter and centrifuge in 50 ml tubes at 18,500 g in a preparative centrifuge for 30 min at 0°C. Collect the sediments and resuspend together in 2 ml of cold buffer 2. 10. Store the preparation at −20°C, in aliquots of 50 µl (150– 300 µg protein) in disposable tubes of 1 or 1.5 ml. The preparation can be stored without loss of activity up to 2 months.

3.2. Solubilization of the Brush-Border Membranes

1. Thaw 50 µl of brush border in ice/water bath. When the membranes are liquid, vortex the tube and add 100 µl of the solubilization buffer. Vortex for 30 s and keep at 0°C for 2 min. 2. Centrifuge the solubilized brush border at about 13,000 g for 4 min at 0°C. Collect the supernatant (brush border extract) in a disposable tube and keep for not more than 10 min, at 0°C until the preparation of the initial reconstitution mixture (Subheading 3.3, item 2). 3. After use (see Subheading 3.3, item 2) store the remaining extract at 0°C for the determination of the protein concentration. 4. The protein concentration in the extract is determined by a standard procedure like the BioRad or the Lowry method (stock solutions from Bio-Rad). This can be done after the transport assay.

62

Indiveri

Fig. 1. Columns holders. The photo shows examples of Plexiglas home- made columns holders for Amberlite (left), A (middle) and B (right) columns. Holes of appropriate sizes permit the insertion of the columns and the tubes. In the case of the holder for columns B, a movable support for vials is inserted below the columns. This support can be removed and replaced by a container that will be marked “radioactive” (see Subheading 3.4, step 10). Movable containers for equilibrating or washing the columns are present in the holders for Amberlite and A columns. Note the black sheet beyond the column tips in the left and middle holders to facilitate the view of the turbid drops

3.3. Reconstitution of the Brush Border Extract

1. Equilibrate 4 Amberlite columns applying 3 ml of the Amberlite equilibrating buffer. Columns holders are useful in this and the following sections (Fig. 1 and see Note 4). 2. Prepare 4 initial reconstitution mixtures (see Note 7) in 1.5 ml disposable tube adding the different solutions in the following order: 25 µl of brush border extract, 75 µl of the 10% C12E8, 100 µl of preformed liposomes, 210 µl of 100 mM l-glutamine (final concentration 30 mM), 28 µl of the stock buffer pH 7.0 (final concentration 20 mM), 262 µl water (to reach a final volume of 700 µl); vortex for 30 s. These operations must be performed at 0–4°C, for example in ice/water bath. 3. Apply each reconstitution mixture onto an equilibrated Amberlite column. Discard the first 5 drops and then collect the eluate. 4. Apply the eluate onto the same Amberlite column. Collect the eluate without discarding drops and repeat this passage further 14 times on the same column (16 total passages). These operations must be performed at room temperature. After these passages, active proteoliposomes are formed which are stable at 0°C up to 4 h.

Studying Amino Acid Transport Using Liposomes

3.4. Transport Assay (8 + 8 Samples Time Course Experiment)

63

1. Prepare a sample tube with 44 µl of labeled glutamine and a control tube with 44 µl of labeled glutamine and 22 µl of inhibitor. 2. Prepare 8 inhibition tubes with 2.5 µl of inhibitor. 3. Load 550 µl of each of the 4 proteoliposome samples on each of the 4 columns A and wait for the absorption of the samples into the resin. Gently, but quickly, apply small volumes (200– 400 µl) of the Buffer A on the columns, till the proteoliposomes are in the middle of the column. Then add 10 ml of buffer A. Collect 600 µl of proteoliposomes into disposable 1.5 ml tubes starting from the second turbid droplet (see Note 8 and Fig. 1). 4. Collect 500 µl of proteoliposomes eluted from each of the 4 columns A and mix together in a 2 ml disposable tube. 5. Transfer 814 µl proteoliposomes into both the sample and control tubes and mix quickly with the pipette tip. This represents the start of the transport reaction for both the samples (S) and the controls (C). A delay of 15–30 s can be practised between the two transfers (see also Note 9). 6. After 5, 10, 20, 30, 40, 60, 80, and 100 min, rapidly transfer (in such a way as to mix the proteoliposomes with the inhibitor) 100 µl from the (S) to an inhibition tube obtaining the inhibited sample (IS). At each time quickly apply 100 µl of the (IS) and 100 µl of the (C) on two columns B (see Note 9). Wait for adsorption of the suspension into the resin and elute adding 100 µl, 200 µl, 200 µl, 400 µl, 500 µl, 500 µl of 50 mM NaCl, waiting for the complete adsorption of each aliquot of the solution before adding the following. Discard the first 900 µl eluate and collect the following 1,000 µl into vials of the appropriate size to be analyzed in a b-counter. Also in this step a column holder will be very useful (Fig. 1). 7. In a separate vial (total radioactivity) add 5 µl of the labeled glutamine for the determination of the total radioactivity added to each sample. 8. After addition of at least 4 ml of scintillation cocktail (PicoFluor 40 from PerkinElmer) vortex each vial and count the radioactivity. 9. Regenerate the columns A with 10 ml buffer A and then 50 ml water. Equilibrate the columns applying 20 ml 0.1 % NaN3 (see Note 3). 10. Regenerate the columns B, with 10 ml buffer B and then 50 ml water. Equilibrate the columns applying 10 ml 0.1 % NaN3 (see Note 3). Very important: the solution eluted from the columns B contains radioactivity (3H); it must be collected

64

Indiveri

in special containers and treated according to the local regulation concerning the use of radioisotopes (see also Note 5). 11. Calculate the specific transport using the following equation: transportt (nmol/mg protein) = (cpmsample– cpmcontrol)/(SR × mg protein) (see Note 10) where transportt is the specific transport at time t, cpmsample and cpmcontrol are the cpm measured in the (IS) and in the respective (C), SR is the specific radioactivity (cpm/nmol) calculated as: total radioactivity/nmoles of glutamine per sample (see Note 11), and mg protein is the amount of proteins per sample derived from the concentration of the protein added to the initial reconstitution mixture (see Note 12). 12. Experimental data expressed as nmol/mg protein obtained above are interpolated by non linear regression analysis using the first order rate equation: y = A(1 – e– kt) where y are the specific transport values at the different times, A are the nmoles of glutamine taken up at infinite time (extrapolated from the fitting), k is the first order rate constant of the uptake process. The initial rate of the transport process is calculated as k · A (nmol/min/mg protein).

Fig. 2. Time course of l-[3H]glutamine uptake in the proteoliposomes. 0.1 mM l-[3H] glutamine is added at time zero to proteoliposomes containing 30 mM internal glutamine in the presence of 50 mM external NaCl. The transport reaction is stopped at the indicated times using 20 µM mersalyl. In the control samples the inhibitor is added at time zero (see Subheading 3.4). The curve is derived from the interpolation of the data points in a first order rate equation (see Subheadings 3.4, item 11 and 3.4, item 12)

Studying Amino Acid Transport Using Liposomes

65

A non linear regression analysis software should be used (Grafit from Sigma-Aldrich is suitable). From the interpolation of the experimental data in the described equation a time course is obtained like the example reported in Fig. 2. The time course experiment described can be used, with some modifications, to study the effect of inhibitors or activators (see Note 13) and the transport kinetics (see Note 14).

4. Notes 1. Cut the Pasteur pipettes leaving a short tip of about 2 cm. Put the cotton in the upper part of the tip. The elution rate of the column depends mainly on the compression of the cotton. Use medium compression to obtain elution rate of 30–50 µl/s (about 1 drop/s). 2. The glass tube should have a rounded bottom. It is very important to use a Pb free glass tube. If you are not sure about the feature of the tube, then use a plastic 2 ml disposable tube; this may lead to some variability in the liposome preparations. Ensure that the temperature of the suspension in the tube does not exceed 10°C during sonication. 3. The filled columns A and B can be reused, normally for 10 to 20 times. In any case, the columns must be refilled with new swollen resin when the elution rate has evidently decreased. Before refilling with the new resin, the glass columns must be cleaned with common detergents used for dishes and then washed with bidistilled water to completely remove any trace of detergent. 4. The number of Amberlite columns, columns A and columns B depends on the type of the experiment. In the protocol for the time course described here the requirement is 4 Amberlite columns and columns A; 16 columns B. The operator has to consider that: (1) a column A is needed for each Amberlite column; (2) the proteoliposomes eluted from column A are normally sufficient to perform 4–5 transport assay samples; (3) a column B is necessary for each transport assay sample. Column holders are of great help in the experimental procedures requiring the simultaneous elution of several columns (see Fig. 1). 5. The 3H isotope is not particularly hazardous; follow the local regulation concerning the use of radioisotopes. For safety measures, it is sufficient to use disposable gloves and take care not to directly touch the solutions containing radioactive glutamine. To this aim, centrifuge by short spin the tubes which contain radioactive solutions in order to avoid some radioactive drops that may stick to the tube edge or closure.

66

Indiveri

6. The volume of the MgCl2 solution (50 mM) required to be added must be calculated on the basis of the volume of the suspended sediment, which may vary in different preparations. The final concentration of MgCl2 (12 mM) is not very critical. 7. In this protocol four identical initial reconstitution mixtures are prepared. However, the number or the composition of initial reconstitution mixtures may change depending on the type of experiment. Two general types can be distinguished: (1) experiments in which compounds or concentrations in the external liposomal space are changing. In this case identical initial reconstitution mixtures are prepared; (2) experiments in which compounds or concentrations in the intraliposomal space are changing. In this case different reconstitution mixtures are required. The Amberlite columns must be equilibrated with different solutions containing the same compounds and buffer of the corresponding reconstitution mixture. 8. To clearly distinguish the turbid droplets place a sheet of black plastic or other material beyond the outlet of the columns (see Fig. 1). 9. To be faster use two different pipettes, one for the samples (Subheading 3.4, item 5) or inhibited samples (Subheading 3.4, item 6),and the other for the controls. A constant delay of 15–30 s can be maintained, after the Subheading 3.4, item 5 procedure (start of the transport reaction), between the two different (S) and (C) samples. The delay among different samples will be very useful in the case of more complex experiments requiring large number of samples (see Notes 13 and 14). 10. Depending on the type and settings, ß-counters will give the radioactivity measurements as cpm or dpm units. Both units can be used equally if the radioactivity of the samples and the specific radioactivity are expressed in the same units. In a well performed experiment, cpmcontrol must be lower than 10% of the respective cpmsample. 11. Calculate the nmoles from the concentration of externally added l-[3H]glutamine in the100 µl sample (each sample in the present experiment contains 10 nmoles of glutamine). 12. The protein content in each sample of 100 µl should correspond to 1/7 of the amount of protein in an initial reconstitution mixture (measured in the extract). However, a dilution of the samples occurs during the chromatography on column A. We have empirically calculated that the amount of protein in the 100 µl sample actually corresponds to 1/12 of that in the initial reconstitution mixture. It has to be stressed that the protein cannot be directly assayed in the proteoliposome samples since lipids interfere with the assay methods. 13. To study the influence of effectors on the transporter on the external side, these molecules must be added to the

Studying Amino Acid Transport Using Liposomes

67

proteoliposomes before the transport assay together with the labeled glutamine in the sample and in the control tubes. To study the influence of effectors on the internal side, the molecules must be included in the initial reconstitution mixture and in the Amerlite equilibrating buffer at the same concentration. In this case, several different initial reconstitution mixtures are needed, each with a different compound or different internal concentration of the effector (see also Note 7). The concentration of each effector should be chosen according to previous data (15, 16). In the case of investigation with novel molecules, a preliminary experiment should be performed in advance with few concentrations chosen in a wide range (for example 0.1, 1, 10, 100 µM). 14. To obtain kinetic constants like Km or Vmax several time courses must be performed at different substrate concentrations and the initial transport rate calculated. In experiments of this type which require several data points, the time course can be reduced to 4 points (5, 10, 30 and 60 min). For reliable constant determinations at least five different glutamine concentrations within an appropriate range must be used and each experiment should be repeated at least three times. In particular, for the determination of external Km, the l-glutamine in the labeled glutamine solution must be varied from 0.05 to 5 mM (final concentration in the proteoliposome samples; the other components do not vary) at fixed internal glutamine concentration of 30 mM. Since the transporter catalyzes an obligatory antiport of glutamine and neutral amino acids, the internal Km can be investigated by following the time course of the uptake of fixed 1 mM l-[3H]glutamine (see above for the labeled glutamine solution) in the presence of various intraliposomal glutamine or other amino acids which are accepted by the ASCT2 transporter (15, 16). The intraliposomal concentrations of the amino acid must be varied from 1 to 30 mM; the concentration of internal glutamine or other amino acids are given by the concentration in the initial reconstitution mixture and in the Amberlite equilibrating buffer (see also Note 7).

Acknowledgments This work was supported by the PRIN (Progetti di Ricerca Scientifica di Rilevante Interesse Nazionale) 2006 grant n. 2006054479 from MiUR (Ministero dell’Università e della Ricerca). The author is indebted to Dr. Francesca Oppedisano, Dr. Lorena Pochini, and Dr. Michele Galluccio for help in preparing the manuscript.

68

Indiveri

References 1. Palacin M, Estevez R, Bertran J, Zorzano A (1998) Molecular biology of mammalian plasma membrane amino acid transporters. Physiol Rev 78:969–1054 2. Bode BP (2001) Recent molecular advances in mammalian glutamine transport. J Nutr 131:2475S–2485S 3. Broer S (2002) Adaptation of plasma membrane amino acid transport mechanisms to physiological demands. Pflugers Arch 444: 457–466 4. Mackenzie B, Erickson JD (2004) Sodiumcoupled neutral amino acid (System N/A) transporters of the SLC38 gene family. Pflugers Arch 447:784–795 5. Kanai Y, Hediger MA (2004) The glutamate/ neutral amino acid transporter family SLC1: molecular, physiological and pharmacological aspects. Pflugers Arch 447:469–479 6. Verrey F, Ristic Z, Romeo E, Ramadan T, Makrides V, Dave MH, Wagner CA, Camargo SM (2005) Novel renal amino acid transporters. Annu Rev Physiol 67:557–572 7. McGivan JD, Bungard CI (2007) The transport of glutamine into mammalian cells. Front Biosci 12:874–882 8. Broer S (2008) Amino acid transport across mammalian intestinal and renal epithelia. Physiol Rev 88:249–286 9. Utsunomiya-Tate N, Endou H, Kanai Y (1996) Cloning and functional characterization of a system ASC-like Na+-dependent neutral amino acid transporter. J Biol Chem 271:14883–14890 10. Torres-Zamorano V, Leibach FH, Ganapathy V (1998) Sodium-dependent homo- and hetero-exchange of neutral amino acids mediated by the amino acid transporter ATB. Biochem Biophys Res Commun 245:824–829 11. Broer A, Brookes N, Ganapathy V, Dimmer KS, Wagner CA, Lang F, Broer S (1999) The astroglial ASCT2 amino acid transporter as a mediator of glutamine efflux. J Neurochem 73:2184–2194

12. Dolinska M, Dybel A, Zablocka B, Albrecht J (2003) Glutamine transport in C6 glioma cells shows ASCT2 system characteristics. Neurochem Int 43:501–507 13. Dolinska M, Zablocka B, Sonnewald U, Albrecht J (2004) Glutamine uptake and expression of mRNA’s of glutamine transporting proteins in mouse cerebellar and cerebral cortical astrocytes and neurons. Neurochem Int 44:75–81 14. Lim J, Lorentzen KA, Kistler J, Donaldson PJ (2006) Molecular identification and characterisation of the glycine transporter (GLYT1) and the glutamine/glutamate transporter (ASCT2) in the rat lens. Exp Eye Res 83:447–455 15. Oppedisano F, Pochini L, Galluccio M, Cavarelli M, Indiveri C (2004) Reconstitution into liposomes of the glutamine/amino acid transporter from renal cell plasma membrane: functional characterization, kinetics and activation by nucleotides. Biochim Biophys Acta 1667:122–131 16. Oppedisano F, Pochini L, Galluccio M, Indiveri C (2007) The glutamine/amino acid transporter (ASCT2) reconstituted in liposomes: transport mechanism, regulation by ATP and characterization of the glutamine/glutamate antiport. Biochim Biophys Acta 1768:291–298 17. Biber J, Stieger B, Haase W, Murer H (1981) A high yield preparation for rat kidney brush border membranes. Different behaviour of lysosomal markers. Biochim Biophys Acta 647:169–176 18. Krämer R, Heberger C (1986) Functional reconstitution of carrier proteins by removal of detergent with a hydrophobic ion exchange column. Biochim Biophys Acta 863:289–296 19. Palmieri F, Indiveri C, Bisaccia F, Iacobazzi V (1995) Mitochondrial metabolite carrier proteins: purification, reconstitution, and transport studies. Methods Enzymol 260:349–369 20. Pochini L, Oppedisano F, Indiveri C (2004) Reconstitution into liposomes and functional characterization of the carnitine transporter from renal cell plasma membrane. Biochim Biophys Acta 1661:78–86

Chapter 6 Use of Liposomes for Studying Interactions of Soluble Proteins with Cellular Membranes Chris T. Höfer, Andreas Herrmann, and Peter Müller Abstract Methods are described that have been used for characterizing the interaction of the soluble bovine seminal plasma protein PDC-109 with liposomes. PDC-109 binds to bull sperm cells upon ejaculation and is an important modulating factor of sperm cell maturation. The binding of the protein to sperm cells is mediated via lipids of the sperm plasma membrane. Most of our current knowledge about the molecular mechanisms of PDC-109–membrane interaction has been obtained by studies employing lipid vesicles. The experimental strategy described here can be applied to investigate the interaction of soluble proteins with membranes in order to understand their physiological role. Key words: Protein–membrane interaction, Membrane integrity, Phospholipid, Seminal plasma protein, PDC-109, Fluorescence spectroscopy, ESR spectroscopy

1. Introduction The interaction of soluble proteins with membranes is crucial for the regulation of biological activities. Proteins become recruited to the leaflets of plasma or intracellular membranes by binding to protein receptors or to lipids which results in adhesion and/or insertion of the protein to the membrane. Thereby, the protein and/or membrane structure and dynamics are influenced, finally resulting in a modified biological activity. Some examples for soluble proteins that modulate physiological functions upon binding to membranes are (1) phospholipases influencing phospholipid metabolism (1), (2) annexins which are supposed to play a role in membrane trafficking events such as exocytosis, endocytosis, and cell–cell adhesion (2), (3) bacterial toxins having deleterious impact on cells (3, 4), (4) mammalian seminal plasma proteins V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_6, © Humana Press, a part of Springer Science+Business Media, LLC 2010

69

70

Höfer, Herrmann, and Müller

which interact with the plasma membrane of sperm cells modulating the genesis of the cells, and (5) amyloid proteins like a-synuclein which is probably involved in the genesis of Parkinson’s disease (5, 6). With regard to proteins of seminal plasma, a protein family which is characterized by an Fn type II domain has gained special importance. This domain was originally described for the extracellular matrix protein fibronectin, in which it constitutes part of the collagen binding region and is implicated in a variety of extracellular binding events (7). Bull seminal plasma contains various Fn type II proteins, the most prominent representative being PDC-109 (also named BSP-A1/A2) (8). PDC-109 binds to sperm cells upon ejaculation and influences the capacitation process of the cells (9). To understand the physiological role of PDC109 on a molecular level, numerous studies, in particular using liposomes, have been perfomed (for reviews see (10, 11)). It has been found that the interaction with spermatozoa is realized via binding to membrane phospholipids carrying a choline head group, which are phosphatidylcholine (PC) and sphingomyelin (SM). The lipid-binding sites have been characterized from the protein crystal structure (12). Upon binding to vesicles, the protein influences membrane structure and dynamics, in that e.g., (1) the mobility of lipids is reduced, (2) membrane integrity is impaired, and (3) lipids are extracted from the membrane. Here, the application of various methods is described to identify and characterize important aspects of the interaction of PDC-109 with membranes by using liposomes.

2. Materials 2.1. Isolation of PDC-109

1. PDC-109 is purified from the seminal plasma of reproductively active Holstein bulls by a combination of affinity chromatography on heparin-Sepharose and DEAE-Sephadex chromatography as described by Calvete et al. (13). 2. The purity of the protein is verified by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) as well as reverse-phase HPLC, N-terminal sequence amino acid, and mass spectrometric analyses. 3. For the experiments, a 1 mM stock solution of the protein is prepared in Hepes-buffered salt solution (HBS) containing 145 mM NaCl and 5 mM HEPES, pH 7.4.

2.2. Lipids

1. Lipids are obtained from Avanti Polar Lipids (Alabaster, AL), if not stated otherwise and used without further purification. Stock solutions of lipids are prepared in chloroform or chloroform/ methanol (1:1) and stored at −80°C (see Notes 1 and 2).

Protein-Membrane Interaction

71

2.3. Fluorescence Measurements

1. All fluorescence measurements are done at 37°C using an Aminco Bowman Series 2 spectrofluorometer (SLM-AMINCO, Rochester, NY) with 4 nm slit width for both excitation and emission.

2.4. Electron Spin Resonance (ESR) Measurements

1. The spin-labeled analog of PC, 1-palmitoyl-2-(4-doxylpentanoyl)sn-glycero-3-phosphocholine (SL-PC), is synthesized according to (14). 2. ESR spectra are recorded at 4°C using a Bruker EMX spectrometer (Bruker, Karlsruhe, Germany) with the following parameters: modulation amplitude 1 G, power 20 mW, scan widths 100 G, accumulation nine times.

3. Methods 3.1. Preparation of Large Unilamellar Vesicles (LUVs)

1. Appropriate quantities of the lipid stock solutions (including when necessary fluorescent or spin-labeled lipids) are transferred into a glass tube, and the solvent is removed under a stream of nitrogen and subsequently under vacuum for 1 h. For labeled vesicles, one thereby obtains symmetrically labeled membranes, i.e., the analog is localized on both leaflets. 2. Lipids are resuspended in a small volume of ethanol (giving a final ethanol concentration below 1% (v/v)). HBS is added and the mixture is vortexed to induce the formation of multilamellar vesicles (see Note 3). 3. For the preparation of LUVs, the lipid suspension is subjected to five freeze–thaw cycles and extruded 10 times at 40°C through a 100-nm-diameter polycarbonate filter (Nucleopore GmbH, Tübingen, Germany) using either an extruder (Extruder, Lipex Biomembranes Inc., Vancouver, Canada) or a mini-extruder (Avanti Polar Lipids, Alabaster, AL) (15) (see Note 4).

3.2. Binding of PDC-109 to Liposomes Measured by Flotation Assay

1. As a first approach to characterize membrane binding of a protein, ultracentrifugation techniques can be used to separate the liposome-bound fraction of the protein from the free unbound protein. A variety of methods based upon this principle exists, and the appropriate one has to be chosen according to individual requirements. Either liposomes and bound protein can be pelleted using multilamellar vesicles (16) or sucrose-loaded vesicles (17), or liposomes and bound protein are floated on a sucrose density gradient (18). The flotation technique allows a clear separation of membrane-bound floating proteins especially for studying proteins that tend to aggregate or form large complexes that might be pelleted independently of the binding to liposomes. The assay can

72

Höfer, Herrmann, and Müller

serve for qualitative as well as quantitative analysis and can be utilized to investigate binding conditions such as pH, ionic strength, and lipid specificity. 2. LUVs are prepared consisting of 2 mM 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) and 0.5 mol% 1,2-dipalmitoylsn-glycero-3-phosphoethanolamine-N-(7-nitro-2-1,3-benzoxadiazol-4-yl) (N-NBD-DPPE) (see Notes 3 and 5). 3. Sucrose solutions (75% and 25% w/v) are prepared in HBS. 4. PDC-109 (12 µM) is mixed with different amounts of LUVs in a total volume of 150 µl in 1.5-ml Polyallomer Microfuge tubes for ultracentrifugation (Beckman Instruments Inc., Palo Alto, CA). 5. The solution is mixed thoroughly with 100 µl of 75% sucrose giving a final sucrose concentration of 30%. This bottom fraction is carefully overlayered with 200 µl of 25% sucrose and 50 µl of HBS. 6. The tubes are centrifuged for 1 h at 240,000 g and 4°C using an ultracentrifuge TL-100 and rotor TLA-100.3 (Beckman Instruments Inc., Palo Alto, CA). 7. Immediately after centrifugation, the fractions of the gradient are collected manually (see Note 6). 8. Aliquots of all fractions are analyzed by SDS-PAGE followed by silver staining. 9. The fluorescence of N-NBD-DPPE can be detected in the front part of the gel by scanning the gel before or after silverstaining on a fluorescent image analyzer FLA-3000 (FUJIFILM, Düsseldorf, Germany) with excitation at 473 nm. 10. Flotation of PDC-109 with DOPC-LUVs shows an increase in unbound protein with decreasing lipid concentration (Fig. 1). Above a protein-to-lipid ratio of 1:40, all protein is membrane-bound, whereas at 1:20 part of the protein is unbound. Since only the lipids of the outer leaflet of LUVs are accessible to protein binding, these data indicate a binding stoichiometry of about 10–20 lipid molecules per one protein molecule, which is in agreement with other studies (19). 3.3. Binding of PDC-109 to Liposomes Measured by Intrinsic Protein Fluorescence

1. Upon binding to membranes, protein structure is often modified. Those changes might be followed by measuring the intrinsic protein fluorescence if the protein contains fluorescent amino acids. Among these, tryptophan is most useful because of its comparatively high quantum yield. Since fluorescence is dependent on the local environment of the fluorophore, conformational changes can be monitored at least for those parts of the protein containing fluorescent amino acids. 2. PDC-109 is diluted from the stock solution to a concentration of 2.5 µM into a fluorescence cuvette containing lipid vesicles and

Protein-Membrane Interaction sucrose gradient

0% 25 %

= LUVs

NBD-labeled LUVs + protein

73

NBD fluorescence 1h

Collecting

240 000 g

the fractions

30 % 0% 25% 30%

Fig. 1. Top: Schematic diagram of the flotation assay. Bottom: Flotation assay analysis of PDC-109 binding to lipid vesicles. PDC-109 was floated with DOPC-LUVs in a sucrose step gradient at different protein-to-lipid ratios. After centrifugation, the fractions of the gradient with 30% sucrose (bottom fraction), 25% sucrose (middle fraction), and 0% sucrose (floating liposome fraction) were analyzed by SDS-PAGE, and subsequent silver staining showed a decrease of floating liposomebound protein with decreasing lipid concentration. The fluorescence of the membrane-anchored liposome label N-NBD-DPPE in the gel front was detected on a fluorescent image analyzer, and indicated the presence of LUVs in 0% and 25% fractions if the overall lipid concentration was sufficient. The bottom fraction was devoid of lipids 0.6

fluorescence intensity [a.u.]

3 0.5

2 0.4

1 0.3 0.2 0.1 0.0 300

320

340

360

380

400

420

wave length [nm] Fig. 2. Influence of lipid vesicles on the intrinsic fluorescence of PDC-109. The fluorescence spectra (excitation at 280 nm) of 2.5 µM PDC-109 were recorded at 37°C in the absence of liposomes (curve 1) and in the presence of 25 µM eggPC/ eggPE(2:1)-LUVs (curve 2 ) or eggPC-LUVs (curve 3). Note the shift of the wavelength of the maximum fluorescence intensity upon addition of LUVs

fluorescence spectra are recorded in the range of 300–400 nm (excitation wave length 280 nm) (Note 7). 3. Upon interaction with eggPC-LUVs, the intrinsic fluorescence of PDC-109 is increased and shifted to shorter wave lengths (Fig. 2, curves 1 and 3) (see Note 8). These data indicate a localization of tryptophan residues in a more hydrophobic

74

Höfer, Herrmann, and Müller

environment in the presence of vesicles which is probably caused by an intercalation of (part of) the protein into the bilayer (see Notes 9 and 10). 4. Mixing of PDC-109 with eggPC/eggPE(2:1)-LUVs results in lower changes of protein fluorescence compared to those of pure PC liposomes (Fig. 2, curve 2). This indicates that PDC-109 has a specificity for phosphorylcholine-containing phospholipids (19, 20). 3.4. Influence of PDC-109 on Membrane Integrity Measured by Calcein Leakage

1. LUVs are loaded with a high concentration of the nonpermeable fluorophore calcein, resulting in low fluorescence intensity due to self-quenching. A perturbation of membrane integrity results in a release of the fluorophore from the vesicles, which can be measured by the increase of fluorescence. 2. LUVs are prepared in HBS containing additionally 70 mM of calcein (Fluka Feinchemikalien, Neu-Ulm, Germany) (see Note 11). 3. Calcein-filled vesicles are separated from bulk calcein using NAP-5 columns (GE Healthcare, Freiburg, Germany) at room temperature and HBS as elution buffer (see Note 12). 4. Calcein-filled LUVs are diluted with HBS in a fluorescence cuvette (giving a final lipid concentration of 10–20 µM, see Note 13) while continuously stirring the solution. The timedependent fluorescence is monitored at 515 nm (excitation wave length 490 nm). After about 30 s, 5 µM PDC-109 is added and maximal leakage is determined by addition of 0.5% (w/v) Triton X-100 after about 230 s. 5. The addition of PDC-109 to eggPC-LUVs induces a release of calcein, indicating that the protein is able to disturb the membrane integrity of these vesicles (Fig. 3) (see Note 14). In contrast, the perturbation of eggPC/eggPE membranes in the presence of PDC-109 is significantly lower as seen from the low calcein leakage, a result again supporting the specificity of PDC-109 for phosphorylcholine-containing lipids.

3.5. Verification of PDC-109 Intercalation into Membrane Bilayer Measured by Förster Resonance Energy Transfer (FRET)

1. In order to see whether PDC-109 intercalates into the hydrophobic lipid phase upon binding to LUVs, FRET was recorded from tryptophan residues of the protein (donor) to the fluorescent lipid 1-hexadecanoyl-2-(1-pyrenedecanoyl)-sn-glycero-3phosphocholine (pyrPC, Invitrogen Ltd, Karlsruhe, Germany). This analog bears the pyrene group (acceptor) in the hydrophobic part of the membrane. A FRET signal strongly indicates a penetration of tryptophan residues into the lipid bilayer since for such signal both fluorophores have to approach below 10 nm. 2. EggPC-LUVs are labeled with 1 mol% pyrPC (see Subheading 3.1). Fluorescence spectra of labeled LUVs (final lipid concentration 20 µM) are recorded between 300 and

Protein-Membrane Interaction

75

normalized fluorescence

1.0

Triton X-100 0.8

0.6

PDC-109

1

0.4

0.2

2 0.0 0

50

100

150

200

250

time [s] Fig. 3. Influence of PDC-109 on the leakage of calcein from lipid vesicles. The time-dependent fluorescence of calcein was measured at 37°C for LUVs composed of eggPC (curve 1) or eggPC/eggPE (2:1) (curve 2). Leakage was induced by adding 5 µM PDC-109 to the vesicles at time zero, giving a final lipid-to-protein ratio of about 4. After 230 s, Triton X-100 was added (final concentration 0.5 % (w/v)) to obtain complete leakage of calcein, which was set to 1

fluorescence intensity [a.u.]

2.5

8 6

2.0

4 2

1.5

1.0

2

0 300

350

400

450

3 4

1

3

0.5

0.0 300

350

400

450

500

550

wave length [nm] Fig. 4. Förster resonance energy transfer (FRET) from tryptophan residues of PDC-109 to pyrene-labeled PC. eggPC-LUVs were labeled with 1 mol% pyrPC, and the fluorescence spectra of PDC-109 and pyrPC were recorded at 37°C in the absence (curve 1) and in the presence of unlabeled (L/P = 10, curve 2) and labeled PC-LUVs (L/P = 10, curve 3) with excitation at 280 nm as well as in the presence of labeled LUVs (L/P = 10, curve 4) with an excitation at 345 nm, i.e., direct excitation of pyrene. The inset shows the spectra with the entire range of the y-axes

500 nm in the absence and in the presence of 2 µM PDC-109 with excitation at 280 nm (Fig. 4). For comparison, fluorescence spectra of PDC-109 without vesicles and with unlabeled PC-LUVs were also recorded (see Notes 15 and 16).

76

Höfer, Herrmann, and Müller

3. The spectra reveal that part of the PDC-109 excitation energy is transferred to the pyrene moiety as seen from a decrease in tryptophan fluorescence at around 340 nm and an increase of pyrene (monomer) fluorescence between 380 and 450 nm (Fig. 4, compare unlabeled and labeled LUVs in the presence of protein, curves 2 and 3, respectively). 4. From these data, it can be concluded that upon binding of PDC-109 to lipid membranes (part of) the protein bearing tryptophan residues intercalates into the membrane phase (see Notes 9 and 17). 3.6. Interaction of PDC-109 with Spin-Labeled Lipids

1. ESR spectra of spin-labeled lipids localized in membranes reflect the lipid mobility and are sensitive to lipid–protein interactions (21). 2. EggPC-LUVs are labeled with 2.5 mol% of SL-PC (see Subheading 3.1). 3. Labeled LUVs (2 mM) are mixed with 0.2 mM PDC-109 and, after a 5 min incubation on ice, ESR spectra are recorded at 4°C. 4. Figure 5 (spectrum 1) shows a typical membrane spectrum of SL-PC in eggPC-LUVs, reflecting a partially restricted motion of the analog within the membrane. 5. In the presence of PDC-109, one observes an additionally superimposed spectrum which is especially visible in the region of the low-field peak (Fig. 5, spectrum 2, see arrow).

10 G 1

2

3

Fig. 5. Influence of PDC-109 on the ESR spectrum of spin-labeled membranes. 2 mM eggPC-LUVs were labeled with 0.05 mM SL-PC, and the ESR spectra were recorded at 4°C in the absence (spectrum 1) and in the presence (spectrum 2) of 0.2 mM PDC-109. The immobilized component caused by PDC-109 (see arrow) was extracted by spectra subtraction, giving spectrum 3

Protein-Membrane Interaction

77

6. The spectral component that originated in the presence of PDC-109 can be obtained by subtracting the ESR spectrum in the absence of protein from that in the presence of protein (Fig. 5, spectrum 3) (see Note 18). This spectrum reflects a decreased mobility of the analog as seen from the peak broadening, indicating that PDC-109 causes an effective restriction of the lipid mobility in PC membranes (22) (see Note 19). 3.7. S ummary

The experiments described here allow characterization of the interaction of a soluble protein with membranes. Other approaches partly using liposomes are described in the literature. To study PDC-109, the following methods have been also applied: Fouriertransform infrared (FTIR) spectroscopy, CD spectroscopy, and calorimetric methods (differential scanning calorimetry, isothermal titration calorimetry) (see (10, 11)). The results of those experiments, in combination with in vitro and in vivo studies on the biologically relevant cellular system, allow the elucidation of the physiological role(s) of the respective protein (11, 12).

4. Notes 1. In case of storage for longer times, we prefer to store the lipids in the dry form, i.e., after evaporation of the solvent, at −80°C in order to prevent the lipids from any decomposition. 2. During handling with the organic solutions of lipids, the usage of any plastics should be prevented and solely glassware (tubes, pipette tips etc.) should be used since the solvents may dissolve substances (e.g., softeners) from the plastics. 3. When preparing liposomes, the detaching of lipids from the glass wall and their dissolution in the buffer is, in our opinion, an important point to be considered. Certain lipid species may be incompletely dissolved, resulting in a deviation of the desired lipid composition of liposomes. In general, unsaturated PC species or the often-used eggPC mixture can be easily dissolved by the addition of buffer and subsequent vortexing. However, problems may arise when using other lipid such as PE, SM, or cholesterol (also when used in a mixture with PC species) due to their distinct physicochemical parameters. For example, it is extremely difficult to dissolve pure cholesterol in an aqueous buffer. The dissolution of lipids could be facilitated by either pre-resuspending the dried lipids in a small volume of ethanol and/or by performing all steps of liposome production at higher temperatures. For example, when using SM species (which are characterized by

78

Höfer, Herrmann, and Müller

long fatty acid chains) we work at 65°C. After preparation of liposomes, lipid concentration of vesicles can be checked by measuring the phospholipid and, if necessary, the cholesterol content (see [23] and references therein). For measuring the lipid composition, the use of thin-layer chromatography or mass spectrometric methods (ESI, MALDI-TOF) is recommended. 4. Basically, different kinds of lipid vesicles can be used: small unilamellar vesicles (SUVs), large unilamellar vesicles (LUVs), or multilamellar vesicles (MLVs). SUVs are prepared by sonication of aqueous lipid dispersion. These vesicles are rapidly prepared. However, owing to their small size (diameter of about 30 nm), they have a large surface curvature questioning their use as a model for a biological membrane. Moreover, SUVs are comparatively unstable and fuse. Therefore, those vesicles should be used shortly after preparation. MLVs can also be prepared easily (see Subheading 3.1). However, due to the presence of more than one bilayer, these vesicles have an undefined accessible surface area. Therefore, use of MLVs could be unfavorable for certain problems. LUVs, although their preparation takes comparatively more effort, are appropriate for investigating numerous aspects of protein–membrane interaction, e.g., their membranes mimick biological membranes with regard to membrane curvature. After preparation, LUVs are supposed to be stable for several days; however, this might be checked. 5. The addition of the fluorescent lipid N-NBD-PE allows (1) visualization of the liposome fractions on a fluorescent image analyzer and (2) estimation of the loss of lipids during the preparation of LUVs by comparing the fluorescence of the lipid solution before and after vesicle preparation. For an accurate quantitative analysis, one has to determine the phospholipid and, if necessary, cholesterol content of LUVs (see Note 3). 6. There are two strategies for collecting the fractions. The three fractions of 250, 150, and 50 µl are collected from the bottom to the top with a Hamilton syringe, permitting a very clean separation of the fractions and restricting the liposomes mainly to the 0% sucrose floating fraction which is suitable for a quantitative analysis. But, if the protein adsorbs to the surface of the tubes as membrane-binding proteins often do, it is preferable to collect the fractions from the top to the bottom, as otherwise the floating fraction can become contaminated with protein sticking to the bottom of the tube. This can be checked by rinsing and boiling the emptied tube with 50 µl sample buffer and subsequent analysis by SDS-PAGE. After collecting the 0% and 25% fractions from the top, the 30% bottom fraction is taken from the bottom of the tube leaving

Protein-Membrane Interaction

79

a rest of 20 µl in the tube to prevent an incompletely removed lipid film from mixing with the bottom fraction. Therefore, the latter technique is not very suitable for quantitative analysis, but, nevertheless, can qualitatively show whether a protein binds to liposomes. If the utilized lipid concentration is high enough, one can visualize the localization of liposomes in the tube under UV light due to the fluorescence of N-NBDDPPE and estimate whether the floating of liposomes was succesful (Fig. 1). 7. In order to correct fluorescence spectra for the influence of light scattering caused by liposomes, the spectra of lipid vesicles in the absence of protein are also recorded and subtracted from those in the presence of protein. 8. For quantification of spectra, the fluorescence intensity at 333 nm is determined and normalized to the protein fluorescence in the absence of vesicles. By measuring the fluorescence changes at different lipid-to-protein ratios, the stoichiometry of protein lipid interaction can be determined which is about 10 for PDC-109 (19). Moreover, one can also estimate from the spectra the wavelength of the fluorescence maximum which is shifted to shorter wavelength (blue shift) upon change of tryptophan residues to a more hydrophobic environment. 9. If the protein to be investigated contains more than one tryptophan residue (for PDC-109 five residues), the exact determination which residue(s) intercalate(s) into the membrane bilayer is hampered since the measured fluorescence spectrum is the superposition of the spectra of each residue. The stepwise replacement of tryptophan residues by nonfluorescent amino acids using molecular biological approaches and measuring fluorescence spectra of those mutants in the presence of liposomes could allow the determination of the region(s) of the protein that are involved in membrane intercalation. Moreover, measurement of the life-time or quenching of fluorescence may allow distinguishing several tryptophan residues (usually 2–3). 10. In principle, an increase of fluorescence intensity and a blue shift of fluorescence reflect a change of the fluorophore to a more hydrophobic environment. However, changes of fluorescence with opposite tendencies in the presence of liposomes do not argue inevitably against a protein–membrane interaction since tryptophan fluorescence could also be influenced by other interactions, e.g., leading to fluorescence quenching. Moreover, if the protein solely adheres to the membrane surface, it is possible that the intrinsic fluorescence remains unchanged.

80

Höfer, Herrmann, and Müller

11. For preparing the calcein solution, first add only half of the necessary volume of buffer to calcein, then adjust the pH of the solution with NaOH to pH = 7.5, and finally fill up with buffer to the desired volume. The high calcein concentration (70,000 times the concentration of general fluorescence measurements) causes an extreme contamination of all labware that comes into contact with this solution. Therefore, all labware has to be cleaned thoroughly after contact with calcein solutions protecting other colleagues from undesirable calcein measurements. This especially holds for the extruder. To prevent its expensive cleaning, we used to reserve one mini-extruder in the lab solely for calcein measurements. 12. NAP-5 columns calibrated for the addition of 500 µl solution are used according to the manufacturer’s instructions. If the volume of liposome solution is smaller and has to remain undiluted, one can load the original solution onto the column and allow it to enter the gel bed completely. Subsequently, the eluting buffer is added. The running front containing the liposomes can be easily followed by eye because of the intense color of calcein and the eluant is collected by dropwise sampling. 13. In case LUVs are diluted after running on NAP-5 columns, the final lipid concentration varies resulting in different molar lipid-to-protein ratios in different experiments. Therefore, in order to estimate the exact ratio, the lipid concentration of calcein-loaded LUVs has to be measured after column chromatography (see [23]). 14. The degree of leakage mediated by PDC-109 increases with rising protein concentrations (23). The leakage kinetics is quantified by determining the amount of leakage (L) according to

L=

Ft − F0 Fmax − F0

where F0 and Fmax refer to the initial fluorescence intensity of calcein-filled LUVs before the addition of PDC-109 and the fluorescence intensity after addition of Triton X-100, respectively (see Fig. 3). Ft denotes to the fluorescence intensity reached in the plateau phase after addition of PDC-109 (in the experiment of Fig. 3 after 200 s). Intensities are corrected for dilution due to the addition of PDC-109 and Triton X-100. 15. Since pyrene is sensitive to quenching by oxygen, it is recommended to degas buffers with nitrogen before using. 16. As in our experiments, one may observe some pyrene fluorescence by excitation of a control sample (i.e., labeled LUVs in the absence of protein) at 280 nm. Therefore, this spectrum is subtracted from that measured in the presence of protein to extract the component solely caused by FRET (see Fig. 4).

Protein-Membrane Interaction

81

17. As a negative control, in order to exclude that FRET occurs from protein dissolved in the bulk buffer onto pyrPC in the membrane, a search was made for FRET between the soluble, tryptophan-containing protein bovine serum albumin and pyrPC labeled LUVs. Upon mixing bovine serum albumin with labeled vesicles, no FRET-dependent pyrene fluorescence was detected at all (24). 18. The extent of immobilization, i.e., the amount of spin-labeled lipids affected by the protein, can be estimated by quantifying the subtraction (22). However, this quantification presumes that (1) solely two fractions of lipids exist in the presence of the protein (i.e., a part of molecules influenced by the protein and the other not influenced) and (2) the exchange rate between these states is negligible in the timescale of ESR. 19. The spin-labeled PC used here is characterized by a short fatty acyl chain bearing the spin moiety at the sn2 position of the glycerol backbone. This feature allows incorporation of the analog also into preformed membranes labeling solely the outer leaflet of LUVs membranes (22). However, those experiments can also be performed by using other spin-labeled phospholipid analogs having different head groups and/or two long fatty acyl chains (long-chain PC analogs available from Avanti Polar Lipids, Alabaster, AL) or spin-labeled steroids (partly available from Sigma-Aldrich Chemie GmbH, Taufkirchen, Germany) in order to characterize the lipid specificity of the protein (22, 25). Adding these analogs during liposome preparation results in symmetrically labeled vesicles (see also Subheading 3.1); however, the labeling of preformed membranes might be difficult.

Acknowledgments The work was supported by the Deutsche Forschungsgemeinschaft (Mu 1017/2). The fruitful cooperation during this project with Edda Töpfer-Petersen is kindly acknowledged. We thank Anja Arbuzova for critical reading of the manuscript. References 1. Winget JM, Pan YH, Bahnson BJ (2006) The interfacial binding surface of phospholipase A2s: PLA2. Biochim Biophys Acta 1761:1260–1269 2. Waisman DM (1995) Annexin II tetramer: structure and function. Mol Cell Biochem 150:301–322

3. Cabiaux V, Wolff C, Ruysschaert JM (1997) Interaction with a lipid membrane: a key step in bacterial toxins virulence. Int J Biol Macromol 21:285–298 4. Palmer M (2004) Cholesterol and the activity of bacterial toxins. FEMS Microbiol Lett 238:281–289

82

Höfer, Herrmann, and Müller

5. Davidson WS, Jonas A, Clayton DF, George JM (1998) Stabilization of alpha-synuclein secondary structure upon binding to synthetic membranes. J Biol Chem 273:9443–9449 6. Beyer K (2007) Mechanistic aspects of Parkinson’s disease: alpha-synuclein and the biomembrane. Cell Biochem Biophys 47:285–299 7. Skorstengaard K, Thogersen HC, Petersen TE (1984) Complete primary structure of the collagen-binding domain of bovine fibronectin. Eur J Biochem 140:235–243 8. Manjunath P, Sairam MR (1987) Purification and biochemical characterization of three major acidic proteins (BSP-A1, BSP-A2 and BSP-A3) from bovine seminal plasma. Biochem J 241:685–692 9. Therien I, Bleau G, Manjunath P (1995) Phosphatidylcholine-binding proteins of bovine seminal plasma modulate capacitation of spermatozoa by heparin. Biol Reprod 52: 1372–1379 10. Wah DA, Fernández-Tornero C, Sanz L, Romero A, Calvete JJ (2002) Sperm coating mechanism from the 1.8 A crystal structure of PDC-109-phosphorylcholine complex. Structure 10:505–514 11. Manjunath P, Therien I (2002) Role of seminal plasma phospholipid-binding proteins in sperm membrane lipid modification that occurs during capacitation. J Reprod Immunol 53:109–119 12. Ekhlasi-Hundrieser M, Müller P, TöpferPetersen E (2008) Male secretory proteins sperm tools for fertilisation. In: Glander HJ, Paasch U (eds) Biology of male germ cells. Shaker Publisher GmbH, Aachen, Germany 13. Calvete JJ, Varela PF, Sanz L, Romero A, Mann K, Töpfer-Petersen E (1996) A procedure for the large-scale isolation of bovine seminal plasma proteins. Protein Expr Purif 8:48–56 14. Fellmann P, Zachowski A, Devaux PF (1994) Synthesis and use of spin-labeled lipids for studies of the transmembrane movement of phospholipids. Methods Mol Biol 27:161–175 15. Mayer LD, Hope MJ, Cullis RP, Janoff AS (1985) Solute distributions and trapping efficiencies observed in freeze-thawed multilamellar vesicles. Biochim Biophys Acta 817:193–196

16. Heuer K, Arbuzova A, Strauss H, Kofler M, Freund C (2005) The helically extended SH3 domain of the T cell adaptor protein ADAP is a novel lipid interaction domain. J Mol Biol 348:1025–1035 17. Buser CA, McLaughlin S (1998) Ultracentrifugation technique for measuring the binding of peptides and proteins to sucrose-loaded phospholipid vesicles. Methods Mol Biol 84:267–281 18. Bigay J, Casella JF, Drin G, Mesmin B, Antonny B (2005) ArfGAP1 responds to membrane curvature through the folding of a lipid packing sensor motif. EMBO J 24: 2244–2253 19. Desnoyers L, Manjunath P (1992) Major proteins of bovine seminal plasma exhibit novel interactions with phospholipid. J Biol Chem 267:10149–10155 20. Müller P, Erlemann KR, Müller K, Calvete JJ, Töpfer-Petersen E, Marienfeld K, Herrmann A (1998) Biophysical characterization of the interaction of bovine seminal plasma protein PDC-109 with phospholipid vesicles. Eur Biophys J 27:33–41 21. Marsh D, Horvath LI (1998) Structure, dynamics and composition of the lipidprotein interface. Perspectives from spinlabelling. Biochim Biophys Acta 1376: 267–296 22. Greube A, Müller K, Töpfer-Petersen E, Herrmann A, Müller P (2001) Influence of the bovine seminal plasma protein PDC-109 on the physical state of membranes. Biochemistry 40:8326–8334 23. Tannert A, Töpfer-Petersen E, Herrmann A, Müller K, Müller P (2007) The lipid composition modulates the influence of the bovine seminal plasma protein PDC-109 on membrane stability. Biochemistry 46:11621–11629 24. Greube A, Müller K, Töpfer-Petersen E, Herrmann A, Müller P (2004) Interaction of fibronectin type II proteins with membranes: the stallion seminal plasma protein SP-1/2. Biochemistry 43:464–472 25. Müller P, Greube A, Töpfer-Petersen E, Herrmann A (2002) Influence of the bovine seminal plasma protein PDC-109 on cholesterol in the presence of phospholipids. Eur Biophys J 31:438–447

Chapter 7 Liposomal Reconstitution of Monotopic Integral Membrane Proteins Zahra MirAfzali and David L. DeWitt Abstract In spite of considerable progress in the methodology for reconstitution of membrane proteins into the liposomes, a successful reconstitution still appears to be more an art than a science. Reconstitution of membrane proteins into bilayers is required for establishing several aspects of the functions of membrane proteins and lipids and for elaborating models of naturally occurring membranes. Cyclooxygenase (COX)-1 and -2 (also prostaglandin endoperoxide H2 synthase, PGHS-1 and -2) belong to the class of monotopic membrane proteins. Membrane-binding domains of both COX-1 and -2 contain four short, consecutive, amphipathic a-helices (A, B, C, and D). Crystal structures of the COXs indicate that basic, hydrophobic, and aromatic residues in the membrane-binding domain are oriented away from the protein core and form a surface on the enzyme, which has been proposed to interact with the lipid bilayer (1). In this chapter, we describe a fast and efficient method for direct incorporation of COX-1 and -2 isozymes – as models for monotopic integral membrane proteins – into preformed liposomes containing fatty acids without loss of activity. Key words: Monotopic membrane protein, Proteoliposomes, Liposome reconstitution, Direct incorporation, Membrane defect, Incorporating impurities into membranes, Cyclooxygenase enzyme

1. Introduction Monotopic membrane proteins such as cyclooxygenase (COX)-1 and -2 interact only with one leaflet of the lipid bilayer. Detergent removal is a classical method that is used for incorporation of integral membrane proteins into liposomes. However, this technique does not work with monotopic membrane proteins such as COX isozymes. Thus, it seems that a more generally applicable method is needed for this class of proteins. Medium-length-chain V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_7, © Humana Press, a part of Springer Science+Business Media, LLC 2010

83

84

MirAfzali and DeWitt

(C12–28) fatty acids, cholates, and cholesterol have been shown to promote the reconstitution of integral membrane proteins into liposomes. Therefore, one of the key features for successful incorporation of solubilized membrane proteins into preformed liposomes appears to be the state of organization of the lipid bilayer. Bilayers conducive to direct incorporation of large membrane proteins are achieved by incorporating “impurities”, such as fatty acids, lysophospholipids, mixtures of structurally different phospholipids, and cholesterol. The organizational defects in a bilayer decreases the activation energy for insertion of proteins and increases the free energy change associated with incorporation of proteins [2]. In this chapter, we explain the step by step process of incorporating a monotopic membrane protein into preformed liposomes using direct incorporation method.

2. Materials 2.1. Expression of Recombinant COX into SF21 Insect Cells

1. Spodoptera frugiperda (SF21) insect cells (Invitrogen Corpo ration, Carlsbad, CA). 2. Recombinant baculovirus expressing His-tagged COX-1 or COX-2 isozymes. 3. Media: HyQ-SFX-Insect serum-free insect cell media (Hyclone, Logan, UT) is supplemented with 0.1% pluronic F-68, 1× lipid concentrate, and 0.2% glucose. 4. Fernbach flasks (Bellso Biotechnology Inc., Vineland, NJ). 5. Innova model 4000 benchtop gyrotory incubator shaker (New Brunswick Scientific Co. Inc, Edison, NJ). 6. Hausser Nageotte Bright-Line Hemacytometer (Thermo Fisher Scientific, Waltham, MA). 7. Zeiss light microscope (Thermo Fisher Scientific, Waltham, MA). 8. Beckman refrigerated floor centrifuge (Beckman-Coulter, Fullerton, CA). 9. DuPont-Sorvall HS-4 centrifuge rotor (Thermo Fisher Scientific, Waltham, MA). 10. Nalgene centrifuge bottles for Sovall HS-4 centrifuge rotor (Thermo Fisher Scientific, Waltham, MA).

2.2. Purification of His-tagged COX from Baculovirus-Infected SF21 Insect Cells

1. Bioneb cell disruption system (Glas-Col, Terre Haute, Indiana). 2. Nitrogen cylinder. 3. Nickel-homogenization buffer: 25 mM NaPO4, 100 mM NaCl, 20 mM imidazole, pH = 7.4.

Liposomal Reconstitution of Monotopic Integral Membrane Proteins

85

4. Nickel-wash buffer: 25 mM NaPO4, 300 mM NaCl, 20 mM imidazole, pH = 7.4. 5. Nickel-elution buffer: 25 mM NaPO4, 100 mM NaCl, 200 mM imidazole, pH = 7.4. 6. Dialysis buffer: 25 mM Tris-HCl, 50 mM KCl, pH 7.4. 7. Fast-flow Ni-NTA Sepharose resin (Qiagen Inc, Valencia, CA). 8. Polypropylene centrifuge tubes 50 ml with caps (Thermo Fisher Scientific, Waltham, MA). 9. Beckman refrigerated floor ultracentrifuge (BeckmanCoulter, Fullerton, CA). 10. Sorvall T865 ultracentrifuge rotor (Thermo Fisher Scientific, Waltham, MA). 11. Ultracentrifuge tubes for Sorvall T865 ultracentrifuge rotor (Beckman-Coulter, Fullerton, CA). 12. Slide-A-Lyzer dialysis cassette with 10,000 MWCO (Thermo Fisher Scientific, Waltham, MA). 13. Amcon ultracentrifuge filter device with 3,000 MWCO (Millipore, Billerica, MA). 2.3. Preparation of Unilamellar Liposomes

1. Buchi rotavapor with temperature-controlled water bath connected to a vacuum pump (Buchi Corporation, New Castle, DE). 2. Round-bottom flask (250 ml). 3. 1,2-Dioleoyl-sn-glycero-3-phosphocholine (DOPC) (Avanti Polar Lipids, Alabaster, AL). 4. 1,2-Dioleoyl-sn-Glycero-3-[Phospho-L-Serine] (Sodium Salt) (DOPS) (Avanti Polar Lipids, Alabaster, AL). 5. Avanti mini liposome extruder with heating block (Avanti Polar Lipids, Alabaster, AL). 6. Two gas-tight Hamilton syringes (1 ml) (Hamilton Company, Reno, NV). 7. Nuclepore polycarbonate membrane filter (100 nm pore size, 19 mm) (Thermo Fisher Scientific, Waltham, MA). 8. Liposome buffer: 25 mM Tris–HCl, 50 mM KCl, pH 7.4.

2.4. Incorporation of COX Enzyme into Preformed Liposomes

1. Microfuge tubes (2 ml) (Thermo Fisher Scientific, Waltham, MA).

2.5. Separation of Proteoliposomes from Free Proteins

1. Beckman refrigerated floor ultracentrifuge (BeckmanCoulter, Fullerton, CA).

2. Dry block temperature controlled incubator (Thermo Fisher Scientific, Waltham, MA).

2. Sorvall SW 55 Ti ultracentrifuge rotor (Thermo Fisher Scientific, Waltham, MA).

86

MirAfzali and DeWitt

3. Ultracentrifuge tubes for Sorvall SW 55 Ti ultracentrifuge rotor (Thermo Fisher Scientific, Waltham, MA). 4. Syringe and needle (Thermo Fisher Scientific, Waltham, MA). 2.6. Protein Quantification Using Bicinchoninic Acid(BCA)

1. Bicinchoninic acid (BCA) protein assay reagent (Thermo Fisher Scientific, Waltham, MA). 2. Albumin standard (2 µg/µl) (Thermo Fisher Scientific, Waltham, MA). 3. Thermo Fisher UV/vis micro-plate reader (Thermo Fisher Scientific, Waltham, MA). 4. Disposable plates (96-wells) (Thermo Fisher Scientific, Waltham, MA). 5. Incubator oven (37°C) (Thermo Fisher Scientific, Waltham, MA).

2.7. Phosphorous Quantification

1. Required solutions a. 8.9 N H2SO4: 123.5 ml of concentrated H2SO4 is added to 376.5 ml of deionized water. b. Ascorbic acid (10%): 5 g of ascorbic acid is put into a 50-ml volumetric flask. Fifty milliliters of deionized water is added. c. Ammonium molybdate(VI) tetrahydrate (2.5%): 1.25 g of ammonium molybdate(VI) tetrahydrate is put into a 50-ml volumetric flask. Fifty milliliters of deionized water is added. d. Phosphorus standard solution (0.64 mM): 0.1 ml of 32 mM phosphorus standard solution is added to 4.9 ml of deionized water. 2. Metal rack for test tubes (Thermo Fisher Scientific, Waltham, MA). 3. Glass test tubes (Thermo Fisher Scientific, Waltham, MA). 4. Thermo Fisher laboratory oven (Thermo Fisher Scientific, Waltham, MA). 5. Thermo Fisher UV/vis spectrophotometer (Thermo Fisher Scientific, Waltham, MA). 6. Spectrophotometer cuvettes (Thermo Fisher Scientific, Waltham, MA).

2.8. Cyclooxygenase Assay

1. Assay buffer: 0.1 M Tris–HCl, pH 8.0. 2. Required solutions a. Heme in DMSO: 1 ml solution of 8 mM of heme in DMSO is made and kept in the freezer. b. Arachidonic acid buffered solution: 1 ml solution of 100 µM arachidonic acid in assay buffer is made and kept in the refrigerator.

Liposomal Reconstitution of Monotopic Integral Membrane Proteins

87

3. YSI 5300A biological oxygen monitor (YSI, Yellow Springs, OH) 4. Hamilton syringe (10 µl) connected to a long needle (Hamilton Company, Reno, NV). 2.9. Gold Labeling Proteoliposomes for Electron Microscopy

1. Beckman refrigerated floor ultracentrifuge (Beckman-Coulter, Fullerton, CA). 2. Sorvall SW 55 Ti ultracentrifuge rotor (Thermo Fisher Scientific, Waltham, MA). 3. Ultracentrifuge tubes for Sorvall SW 55 Ti ultracentrifuge rotor (Thermo Fisher Scientific, Waltham, MA). 4. Syringe and needle (Thermo Fisher Scientific, Waltham, MA).

2.10. Transmission Electron Microscopy on Gold Labeled Proteoliposomes

1. Affinity-purified rabbit anti-COX antibody (custom made). 2. Gold immunoprobe NANO-GOLD-goat anti-rabbit IGGNRF (1.4-nm particles) (Immunoprobe, Yaphank, NY). 3. Beckman refrigerated floor ultracentrifuge (Beckman-Coulter, Fullerton, CA). 4. Sorvall SW 55 Ti ultracentrifuge rotor (Thermo Fisher Scientific, Waltham, MA). 5. Ultracentrifuge tubes for Sorvall SW 55 Ti ultracentrifuge rotor (Thermo Fisher Scientific, Waltham, MA). 6. Syringe and needle (Thermo Fisher Scientific, Waltham, MA). 7. Formvar (polyvinylformaldehyde)- carbon coated copper grids (G-300 mesh) (Electron Microscopy Sciences Inc, Fort Washington, PA). 8. Whatman filter paper #1 (Thermo Fisher Scientific, Waltham, MA). 9. JOEL transmission electron microscope (JOEL, Tokyo, Japan).

3. Methods Large quantities of COX enzyme are expressed in insect cells using baculovirus system containing the DNA of His-tagged COX enzyme. The purity of COX enzyme that is used for incorporation studies is more than 95%. The enzyme is purified using nickel-affinity chromatography method. In order to use this purification method, the protein is tagged with six histidines. In vitro mutagenesis is used to introduce a six-residue histidine sequence (His-tag) near the amino terminal end of the human COX-1 and -2. These isozymes are expressed using the baculovirus system. The His-tags are located one and two amino acids beyond the signal peptide cleavage sites of COX-1 and COX-2, respectively,

88

MirAfzali and DeWitt

positions that do not affect their activities or sensitivities to nonsteroidal anti-inflammatory drugs. When expressed in sf-21 cells, the His-tagged enzymes have Km values for arachidonate, and IC50 values for inhibition by nonsteroidal anti-inflammatory drugs that are similar to values reported for the nontagged enzymes. The His-tags allowed purification of the COX isozymes by a simplified protocol involving nickel-affinity chromatography [3]. Unilamellar liposomes containing impurities such as oleic acid are made in buffer and sized to 100 nm by extrusion using polycarbonate membrane filters. Various amounts of enzyme are added to fixed amount of liposomes in order to find the optimum protein-to-lipid ratio. Proteoliposomes (protein-incorporated liposomes) are separated from free liposomes (unincorporated proteins) using ficoll density gradient and ultracentrifugation methods. The incorporated proteins can be observed on the surface of the liposomes using immunogold labeling in combination with high-resolution transmission electron microscopy [4]. 3.1. Expression of Recombinant COX into SF21 Insect Cells

1. Spodoptera frugiperda (SF21) insect cells are added to sterile media in 1 L cultures at 27°C in 2.8-L Frenchback flasks shaken at 120 rpm using Innova model 4000 benchtop gyrotory incubator shaker. 2. The insect cells are counted every 8 h using Hausser Nageotte Bright-Line Hemacytometer and light microscope. 3. When cells reach a density of 1.5–2.0 × 106 cells/ml, baculovirus containing the DNA of COX enzyme is added at a multiplicity of infection of 1.0, and the infection is allowed to proceed for 72–96 h. 4. During the 72–96 h wait period, the cells are counted and lysed every 8 h and assayed for COX activity in order to monitor the progress of protein expression (see Note 1). 5. To harvest the cells, the cells and media are poured into Nalgene centrifuge bottles for Sorvall HS-4 centrifuge rotor. 6. The cells are pelleted at the bottom of the centrifuges tube using the Beckman refrigerated floor centrifuge operating at 10,000× g for 30 min. 7. The cells are washed using phosphate buffered saline (PBS) buffer and pelleted again using centrifugation. 8. The buffer is separated from the pelleted cells. 9. The tubes containing the cells are flash-frozen using liquid nitrogen and stored in a −80°C freezer.

3.2. Purification of His-tagged COX from Baculovirus-Infected SF21 Insect Cells

1. The tubes containing frozen cells are thawed at room temperature. 2. SF21 cell pellets are resuspended in 3 ml/g wet weight of nickel-homogenization buffer containing (1% v/v) of Tween 20 detergent.

Liposomal Reconstitution of Monotopic Integral Membrane Proteins

89

3. The cells are disrupted by Bioneb cell disruption system. The Bioneb is connected to high pressure nitrogen at 150 psi. The cells are passed through the system at least three times in order to achieve 100% disruption of all cells. 4. The cells are checked under the light microscope to make sure that they are completely lysed. 5. After low-speed (10,000× g for 1 h) centrifugation to pellet cell debris, the supernant which contains the solubilized COX is poured into cramp-seal ultracentrifuge tubes. 6. The insoluble material is removed by ultracentrifugation at 100,000× g for 2 h. 7. The solubilized extract is next incubated with fast-flow Ni-NTA resin overnight at 4°C with rocking. 8. The resin is poured into a column and washed with five volumes of nickel-homogenization buffer containing 0.1% Tween 20 detergent. 9. Next the resin is washed with three volumes of nickel-wash buffer containing 0.1% Tween 20. 10. The protein is eluted with nickel-elution buffer containing 0.2% Tween 20. 11. Fractions with high specific COX activity are pooled and concentrated in the Amicon ultracentrifugal filter device with 30,000 MWCO. 12. The purified COX enzyme is injected into a Slide-A-Lyzer (10,000 MWCO) dialysis cassette using a syringe. 13. The dialysis cassette is dialyzed against 2 L dialysis buffer. 14. The dialysis buffer is changed every 6 h for 24 h (see Note 2). 3.3. Preparation of Unilamellar Liposomes

1. Unilamellar liposomes are prepared using a DOPC: DOPS molar ratio of 3:7 and 9.2% (w/w) oleic acid. DOPC, DOPS, and oleic acid are dissolved in 1:1 (v/v) mixture of chloroform and methanol in a round bottom flask. 2. The round bottom flask is dried using a Buchi rotary evaporator connected to a vacuum pump for 4 h. 3. The resulting film is hydrated in a liposome buffer for 2 h. The total concentration of lipid is 38 mM. 4. The liposome dispersion is extruded 20 times through a polycarbonate membrane with 100 nm pore diameter using a handheld mini-extruder (see Note 3). Extruded liposomes form a transparent milky suspension.

3.4. Incorporation of COX Enzyme into Preformed Liposomes

1. The COX protein is added to the preformed liposomes. The concentration of the lipids in the liposome solution is 38 mM. A 1:500 protein: lipid molar ratio is used. 2. The mixture is incubated for 20 min at 37°C.

90

MirAfzali and DeWitt

3.5. Separation of Proteoliposomes from Free Proteins

1. A 0.5-ml liposome suspension is mixed with 1 ml of 30% (w/v) Ficoll in liposome buffer to give a final concentration of 20% (w/v) Ficoll. 2. The liposome suspension is transferred to a 5-ml ultracentrifuge tube. 3. Three milliliters of 10% (w/v) Ficoll is gently layered on top of the liposome suspension. 4. The swing-out rotor (Beckman SW 50.1) is used for 30 min at 100,000× g at 4°C. 5. The liposomes and proteoliposomes create a band at the interface between the 0% Ficoll and 10% Ficoll layers. A 1-ml syringe is used to isolate the proteoliposomes band from the buffer solution. The activity and concentration of the enzyme in the isolated proteoliposomes band are measured in order to calculate the incorporation efficiency of COX enzyme. 6. The unincorporated protein is retained in the 30% Ficoll layer. The activity and the concentration of the free protein are measured in order to calculate the incorporation efficiency of the COX enzyme.

3.6. Protein Quantification Using Bicinchoninic Acid (BCA)

1. The oven is turned on at 37°C. 2. Seven 1.5-ml tubes 0, 0.2, 0.4, 0.6, 0.8, 1, and 1.2 are labeled for the standard curve. 3. For the standard curve, 0, 10, 20, 30, 40, 50, and 60 µl of (2 µg/µl of albumin stock solution) is added to 100, 90, 80, 70, 60, 50, and 40 µl of water. 4. Samples for standard curve are added in triplicates at a volume of 10 µl to the wells in the 96-well plate. 5. Ten microliters of diluted aliquots of COX samples is added in triplicates to the wells in the 96-well plate. 6. Two hundred microliters of the BCA working reagent is added to each of the wells (some of the wells may turn purple after addition of the working reagent). 7. The plate that is covered by a lid is incubated for 30 min in the 37°C oven. 8. The plate is read at 562 nm. 9. The concentration of COX enzyme is calculated by linear regression of the standard curve.

3.7. Phosphorous Quantification

1. Liposome samples (~ 0.1 µmoles phosphorus) are placed in glass tubes. 2. Samples for standard curve are prepared by adding 0 (0 µmoles), 50 (0.032 µmoles), 100 (0.064 µl), 175 (0.112 µmoles), 250

Liposomal Reconstitution of Monotopic Integral Membrane Proteins

91

(0.160 µmoles), and 350 µl (0.224 µmoles) of phosphorus standard solution to each glass tube. 3. The sample and standard tubes are placed in a metal test tube rack. 4. To each of the tubes, 0.45 ml 8.9 N H2SO4 is added. Each tube is capped with a glass marble. 5. The samples are kept in oven at 200–215°C for 25 min. Temperature must be above 200°C. 6. Tubes are removed from the oven and cooled for 5 min. 7. To each tube, 150 µl H2O2 is added. 8. The tubes are returned to the oven and heated for an additional 30 min. The samples are colorless at this point (see Note 4). Tubes are cooled to ambient temperature. 9. To each tube, 3.9 ml deionized water is added. 10. To each tube, 0.5 ml of ammonium molybdate(VI) tetrahydrate solution is added. 11. The tubes are vortexed for 1 min. 12. To each tube, 0.5 ml ascorbic acid solution is added . 13. The tubes are vortexed for 1 min. 14. The tubes are heated at100°C for 7 min. Each tube is capped with a glass marble. 15. The tubes are cooled to room temperature. 16. The absorbance of each standard and sample is read at 820 nm. 17. The total amount of phosphorous in sample is calculated by linear regression of the standard curve. 3.8. COX Assay

1. To each chamber of YSI 5300A biological oxygen monitor device, 3 ml of assay buffer is added. The buffer is warmed up in to 37°C using the temperature control unit on the device. 2. The buffer inside the chamber is stirred using the magnet inside the chamber. 3. To the chamber, 50 µl of COX protein in added. 4. To the chamber ,5 µl of heme solution in DMSO is added (see Note 5). 5. The oxygen probe is inserted into the chamber. 6. Using a Hamilton syringe, 10 µl of arachidonic acid solution in buffer is added to the chamber. 7. The recorder that is connected to the oxygen probe records the rate of oxygen consumption. 8. The activity of the COX enzyme is calculated on the basis of the oxygen consumption rate.

92

MirAfzali and DeWitt

3.9. Gold Labeling Proteoliposomes for Electron Microscopy

1. About 100 ml of proteoliposomes is placed into an Eppendorf tube and 10 ml of primary antibody is added. The molar ratio of COX to antibody is 1:4. 2. The mixture is incubated for 30 min in room temperature with occasional shaking. 3. The nonattached primary antibodies are separated from the complex of proteoliposomes–primary antibodies using the Ficoll gradient technique, which was described previously. 4. The liposomes and the complex of proteoliposomes–primary antibodies are separated at the interface between the 0% Ficoll and 10% Ficoll layers. The unattached antibody will remain in the 30% Ficoll layer. 5. The complex of proteoliposomes–primary antibodies is placed into an Eppendorf tube and about 100 ml of gold immunoprobe NANOGOLD-anti rabbit (NRF) (1.4 nm particle attached to affinity-purified Fab fragment, raised in goat, against rabbit IgG (whole molecule)) is added to the tube and incubated for 30 min in room temperature with occasional shaking. 6. The nonattached gold-labeled secondary antibodies are separated from the complex of proteoliposomes–primary antibodies– gold labeled secondary antibodies using the Ficoll gradient method. 7. The liposomes and the complex of proteoliposomes–primary antibodies–gold labeled secondary antibodies are separated at the interface between the 0% Ficoll and 10% Ficoll layers. The unattached gold-labeled secondary antibody will remain in the 30% Ficoll layer.

3.10. Transmission Electron Microscopy on Gold Labeled Proteoliposomes

1. A drop of liposome solution (concentration about 5 mM of lipid) is placed on Formvar (polyvinylformaldehyde)-carbon coated copper grid G-300 mesh, and after 1 min the excess is removed with a Whatman filter paper #1. 2. A thin film is left on the grid and allowed to air dry. One drop of 0.1% solution of uranyl acetate is placed the grid (see Note 6). 3. After 1 min, this drop is again removed with the filter paper, and the resulting stained film is dried in a dust-free place. 4. A JEOL transmission electron microscope operating at 100 kV under vacuum is used to observe the liposomes (Figs. 1 and 2). Fixation and dehydration of liposome for electron microscopy flattens the liposomes and allows one to observe gold labeling on both sides of the liposomes. The average number of gold spots observed in each proteoliposomes is 123 ± 33 (n = 3). Assuming that average proteoliposomes are spheres of 100 nm, and that the two leaflets essentially double the lipid surface area, the total

Liposomal Reconstitution of Monotopic Integral Membrane Proteins

93

Fig. 1. Negative stain transmission electron micrograph of liposomes reconstituted with COX-2 and immunogold labeled with anti-COX-2 antisera (bar = 100 nm, magnification = 200,000). 0.1% uranyl acetate stain was used. Reproduced with permission from (4)

Fig. 2. Negative stain transmission electron micrograph of liposomes reconstituted with COX-2 (bar = 200,000 nm, magnification = 67,000). 1% uranyl acetate stain was used. Reproduced with permission from (4)

surface area of the proteoliposomes is (4p × 500 Å 2) × 2 = 6,280,000 Å2 of phospholipid. DOPC has been calculated to have a surface area of 59 Å2 per molecule. To simplify these calculations, we assumed that DOPS has the same surface area as DOPC. Using these assumptions, it calculated that there are 106,440 molecules of phospholipid per liposome. With our estimated incorporation ratio of 1,000:1 for phospholipid to COX per liposome, this would predict 106 molecules of COX per liposome, close to the 123 average spots observed.

94

MirAfzali and DeWitt

4. Notes 1. COX-2 is expressed in significantly higher level than COX-1 using the baculovirus system. 2. The buffer is exchanged to a nonphosphate-based buffer. The concentration of lipid in proteoliposomes is measured by phosphorous assay and therefore the buffer solution should be phosphate-free. 3. Liposomes should be extruded at a temperature that is above the liquid-to-gel phase transition temperature (Tc) of the lipid mixture. Attempts to extrude below the Tc will be unsuccessful, as the membrane has a tendency to foul with rigid membranes that cannot pass through the pores. 4. H2O2 is added in order to bleach the brown solution. If any brown color persists, 50 µl H2O2 is added to all cooled tubes and the tubes are heated again for 15 min until they become colorless. 5. During the purification process, the heme is stripped off from COX enzyme. Therefore, heme should be added to the enzyme before assaying the enzyme. 6. Transmission electron microscopy on gold-labeled liposome is carried out using a diluted stain (0.1–0.2 % solution of uranyl acetate). Because the NANOGOLD particles are small, overstaining with uranyl acetate may tend to obscure direct visualization of individual NANOGOLD particles. Therefore a diluted stain is used.

Acknowledgments The authors would like to thank Dr. Alicia Pastor and Mr. Robert Pcionek from Michigan State University Center for Advanced Microscopy for their help with the electron microscopy work.

References 1. Picot D, Loll PJ, Garavito RM (1994) The X-ray crystal structure of the membrane protein prostaglandin H2 synthase-1. Nature 367: 243–249 2. Jain MK, Zakim D (1987) The spontaneous incorporation of proteins into preformed bilayers. Biochim Biophys Acta 906: 33–67

3. Smith T, Leipprandt J, DeWitt D (2000) Purification and characterization of the human recombinant histidine-tagged prostaglandin endo peroxide H synthases-1 and -2. Arch Biochem Biophys 375:195–200 4. Mirafzali Z, Leipprandt J, DeWitt D (2005) Fast, efficient reconstitution of cyclooxygenases into proteoliposomes. Arch Biochem Biophys 443:60–65

Chapter 8 The Reconstitution of Actin Polymerization on Liposomes Mark Stamnes and Weidong Xu Abstract Membrane-associated actin polymerization is of considerable interest due to its role in cell migration and the motility of intracellular organelles. Intensive research efforts are underway to investigate the physiological role of membrane-associated actin as well as the regulation and mechanics of actin assembly. Branched actin polymerization on membranes is catalyzed by the Arp2/3 complex. Signaling events leading to the activation of the guanosine triphosphate (GTP)-binding protein Cdc42 stimulate Arp2/3dependent actin polymerization. We have studied the role of Cdc42 at the Golgi apparatus in part by reconstituting actin polymerization on isolated Golgi membranes and on liposomes. In this manner, we showed that cytosolic proteins are sufficient for actin assembly on a phospholipid bilayer. Here we describe methods for the cell-free reconstitution of membrane-associated actin polymerization using liposomes and brain cytosol. Key words: Liposome, Actin, Arp2/3, Wiscott–Aldrich syndrome protein (WASP), Cdc42

1. Introduction The actin cytoskeleton plays two basic roles connected to cell migration and the motility of intracellular organelles. First, actin microfilaments serve as tracks for myosin-motor-based motility. The best characterized role for an actomyosin complex is in muscle contraction (1). Myosin-based movement along actin is also utilized for vesicle targeting, organelle positioning, and the retraction of the cell’s trailing edge during migration (2). A second role for actin is to provide a propulsive force directly through Arp2/3catalyzed polymerization at a membrane (3). The most dramatic examples of this are the comet-tail type of motility used by endosomes that is also usurped by pathogenic bacteria such as Listeria. Similarly, actin polymerization is used directly to force the extension

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_8, © Humana Press, a part of Springer Science+Business Media, LLC 2010

95

96

Stamnes and Xu

of the leading edge membrane during cell migration. The ability to utilize actin polymerization to direct motility requires precise temporal and spatial regulation. A key regulatory mechanism for Arp2/3-dependent actin polymerization involves Cdc42, a member of the Rho-family of guanosine triphosphate (GTP)-binding proteins, and its effectors N-WASP or WAVE (4). GTP-bound Cdc42 causes a conformation change in N-WASP that exposes a C-terminal Arp2/3-binding domain. The actin polymerization activity of Arp2/3 is stimulated when bound to N-WASP. Thus, Cdc42 activation can stimulate actin polymerization at a distinct time and place. In many cell types, Cdc42 is localized to the Golgi apparatus through a binding interaction with the vesicle-coat protein, coatomer (5). There, Cdc42 serves to regulate both actin and microtubule-dependent trafficking events. We have investigated the role of Cdc42 at the Golgi complex, in part through the cell-free reconstitution of Cdc42-regulated Arp2/3dependent actin polymerization on isolated membranes and liposomes (6–10). The ability to reconstitute actin polymerization in vitro has served as a powerful tool for dissecting the mechanisms and regulation of this process at the molecular level. Indeed, the primary catalyst of actin polymerization, the Arp2/3 complex, was discovered through a biochemical purification using cell-free actin polymerization as an assay (11). The reconstitution of actin polymerization on biological and liposomal membranes has revealed details about the regulation and role of actin in the secretory pathway (5). It was recently demonstrated that the vesicle-associated GTP-binding protein ARF1 causes comet-tail-like motility of liposomes (12). We have exploited ARF1 and Cdc42 dependent actin polymerization on liposomes to demonstrate that this process involves the recruitment of Arp2/3 to the membrane (8). Here we provide detailed methods for reconstituting actin polymerization on membranes using brain cytosol and liposomes.

2. Materials 2.1. Preparation of Bovine-Brain Cytosol

1. Brain buffer: 25 mM Tris-HCl, pH 7.4, 320 mM sucrose. 2. Protease inhibitor stock solutions: 10 mg/ml aprotinin in H2O, 10 mg/ml leupeptin in H2O, 1 mM pepstatin A in dimethyl sulfoxide, 50 mM 1,10-phenanthroline in H2O, pH 5.0, 100 mM phenylmethanesulfonyl fluoride (PMSF) in 2-propanol.

Actin Polymerization on Liposomes

97

3. Breaking buffer: 25 mM Tris-HCL, pH 8.0, 500 mM KCl, 250 mM sucrose, 2 mM EGTA. Add dithiothreitol (DTT) and protease inhibitors (from the stock solutions) immediately before use to the following final concentrations, 1 mM DTT, 2 mg/ml aprotinin, 0.5 mg/ml leupeptin, 2 mM pepstatin A, 500 mM 1,10-phenanthroline, 1 mM PMSF. 4. Dialysis buffer: 25 mM Tris-HCl, pH 8.0, 50 mM KCl, add 1 mM DTT just prior to use. (see Note 1). 5. Dialysis tubing with 12,00–14,000 molecular weight cut off and 2.8 cm diameter (Membra-Cel®). 2.2. Preparation of Liposomal Membranes

1. Phospholipids: 10 mg/ml phospholipid stock solutions are prepared in chloroform. A bovine liver lipid extract (Avanti Polar Lipids) is an inexpensive source to provide a complex phospholipid bilayer. (see Note 2). 2. Resuspension buffer: 20 mM HEPES, pH 7.2, 150 mM potassium acetate, 250 mM sucrose.

2.3. Actin Polymerization Reactions

1. 5× Reaction buffer: 125 mM HEPES, pH 7.2, 12.5 mM magnesium acetate, 75 mM potassium chloride. 2. 100× ATP regenerating stock solution: 100 mM creatine phosphate, 25 mM UTP, 5 mM ATP. 3. Creatine phosphokinase stock solution: 1,000 units/ml in H 2O. 4. GTPgS solution: 10 mM in H2O. 5. 45% Sucrose/Tris (ST) solution: 10 mM Tris-HCl, pH 7.4, 45% sucrose (weight/weight, i.e. 45 g sucrose plus 55 ml water). 6. 35% ST solution: 10 mM Tris-HCl, pH 7.4, 35% sucrose (weight/weight). 7. 15% ST solution: 10 mM Tris-HCl, pH 7.4, 15% sucrose (weight/weight). 8. TCA stock solution: 100% trichloroacetic acid (weight/volume) in water.

2.4. Characterizing Actin Polymerization by SDS-PAGE

1. Laemmli sample buffer: 100 mM Tris-HCl, pH 6.8, 3% (weight/volume) sodium dodecyl sulfate (SDS), 10% (volume/volume) glycerol, 715 mM-mercaptoethanol, 0.03% (weight/volume) bromophenol blue.

2.5. Characterizing Actin Polymerization by Western Blotting

1. TBS-Tween solution: 20 mM Tris-HCl, pH 7.6, 137 mM NaCl, 0.5% Tween-20 (volume/volume). 2. Blocking solution: 5% non-fat milk powder (weight/volume) in TBS-Tween solution.

98

Stamnes and Xu

3. Methods The general procedure for the reconstitution of actin polymerization involves incubating the membranes with cytosol under conditions where GTP-binding proteins are active. This may be done by including the non-hydrolyzable GTP analog GTPgS in the reaction. Following the incubation, the liposomes are isolated from the reaction by centrifugation and analyzed by SDS-PAGE and Western blotting. A key to these experiments is obtaining good-quality cytosol to provide a source of actin, Arp2/3, and regulatory proteins. 3.1. Preparation of Bovine-Brain Cytosol

1. Obtain fresh bovine brains from a slaughter house. Two to three brains provide a 100–300 ml of cytosol. The brains should be placed in ice-cold brain buffer as soon as possible and stored in the buffer on ice while in transit. All of the following steps should be performed on ice or in a 4°C cold room. 2. Remove and discard the brain stem and cerebellum. Carefully peel the meninges and blood vessels from the cerebrum using forceps. Weigh the remaining cerebral tissue. 3. Chop the cerebrum into approximately 1 cm pieces using scissors. Immediately place the brain tissue into a small volume of breaking buffer (with protease inhibitors). Once the tissue is fully chopped, add additional breaking buffer such that the final volume of brain tissue plus buffer equals 1.25 L per 500 g tissue. 4. The tissue is homogenized using a blender. Blend the sample 2 times at a high setting for 30 s each. 5. Pour the homogenate into 500 ml centrifuge bottles. Centrifuge for 1 h at 14,000× g in a Sorvall SLA-3000 rotor (Thermo Scientific) or equivalent. 6. Decant the supernatant and pour into polycarbonate or quickseal tubes for Type 45 Ti rotor (Beckman-Coulter). Spin for 90 min at 140,000× g. 7. Collect the supernatant and distribute to several 1 m sections of dialysis tubing. Seal the tubing by tying knots or using clips. Be certain to leave extra space in the tubing as the retentate volume will expand considerably during the dialysis. Dialyze the sample for 4 h in 30 L of dialysis buffer. Switch the tubing to fresh dialysis buffer and dialyze for an additional 4–8 h (see Note 3). 8. Record the volume of the retentate and spin the sample for 90 min at 140,000× g using the Type 45 Ti rotor. Collect the supernatant. 9. The cytosol is concentrated by precipitation with ammonium sulfate. Slowly add 36.5 g of ammonium sulfate per 100 ml

Actin Polymerization on Liposomes

99

cytosol while stirring to obtain a final concentration of 60% saturation. Allow the sample to stir for an additional 30 min. 10. Recover the precipitate by centrifugation in 500 ml bottles in the SLA-3000 rotor 30 min. at 14,000× g. Discard the supernatant. Resuspend the pellet in dialysis buffer using one-sixth the volume recorded in step 8 (see Note 4). 11. Dialyze the sample against 30 L dialysis buffer for 4 h. Transfer the dialysis tubing to 30 L fresh dialysis buffer without DTT for an additional 4–8 h (see Note 3). 12. Collect the retentate and centrifuge for 90 min at 140,000× g using the Type 45Ti rotor. Recover the supernatant. 13. Mix the sample and determine the protein concentration of the cytosol using bicinchoninic acid (BCA) protein assay kit (Pierce). A typical preparation has a concentration of ~10 mg/ml. 14. Freeze the cytosol in aliquots using liquid nitrogen. Store the cytosol at −80°C. 3.2. Preparation of Liposomal Membranes

1. Dispense the appropriate volumes of the phospholipid stock solutions into the bottom of a glass round-bottom tube (see Note 5). 2. Evaporate the chloroform by blowing air across the tube or using an Evap-O-Rac (Cole-Parmer). 3. The phospholipids are hydrated by adding resuspension buffer and mixing using a vortex mixer (see Note 6). The final phospholipid concentration should be 2.3 mg/ml. 4. The lipid resuspension is subjected to 10 freeze-thaw cycles using a dry-ice/methanol bath. 5. Unilamellar vesicles with a uniform size can be generated by extrusion through a polycarbonate membrane. Nine passages through a LiposoFast membrane extrusion device (Avestin) containing a membrane with pore diameter of 400 nm works well. 6. The liposomes are best if used immediately, but we have successfully stored them as aliquots at −80°C for several months.

3.3. Actin Polymerization Reactions

1. The reactions are set up in 1.5 ml plastic tubes with a final volume of 200 ml (see Note 7). The final reaction conditions are as follows: 1× reaction buffer, 1× ATP regenerating solution, 8 units/ml creatine phosphokinase, 0.2 mg/ml bovine brain cytosol, 0.23 mg/ml liposomes. Actin polymerization may be stimulated by adding 20 mM GTPgS to activate GTP-binding proteins. 2. Incubate the reactions at 37°C for 20 min. 3. At the end of the incubation, isolate the liposomes by microcentrifugation for 30 min at 15,000 × g.

100

Stamnes and Xu

4. The liposomes may be separated from non-membrane-associated F-actin by flotation through a sucrose cushion. Begin by resuspending the liposome pellet in 50 ml of 45% ST solution (see Note 7). The sample is placed at the bottom of a 7 × 20 pollyallomer ultracentrifugation tube (Beckman-Coulter). The sample is overlayed with 130 ml 35% ST solution followed by 20 ml 15% ST solution. The sample is subjected to centrifugation for 45 min at 386,000× g using a TLA-100 rotor (Beckman Coulter). 5. After centrifugation, the sample tubes are frozen by careful immersion in liquid nitrogen. Slice the tube with a guillotine or heated razor blade to recover the top 100 ml of the gradient. 6. Precipitate the liposome-associated proteins by addition of trichloroacetic acid to a final concentration of 10%. Recover the precipitate by spinning in a microcentrifuge for 15 min. 7. Wash the precipitate with ice-cold acetone. 3.4. Characterizing Actin Polymerization by SDS-PAGE

1. Resuspend the precipitate using Laemmli sample buffer. 2. Load sample onto a 12% SDS-PAGE Gel (BioRad) and install the gel into an electrophoresis apparatus containing running buffer (BioRad). 3. Carry out electrophoresis at 200 V until the bromophenol blue dye front reaches the bottom of the gel. 4. Liposome-bound proteins can be visualized using Coomassie blue or silver stain as shown in Fig. 1.

Fig. 1. Shown is a Coomassie-blue-stained SDS PAGE gel of liposomes isolated by flotation from incubations carried out with brain cytosol. When the non-hydrolyzable GTP analog GTPgS is included in the reaction, multiple proteins bind to the liposomes. Actin is observed as a prominent band at 42 kD. The actin plus end-binding toxin cytochalasin D partly inhibits actin polymerization. Brefeldin A (BFA), an inhibitor of the GTP-binding protein ARF1, is a potent inhbitor of actin assembly on the liposomes. A peptide p23 that blocks Cdc42 recruitment to the membrane also inhibits actin polymerization on liposomes

Actin Polymerization on Liposomes

101

Fig. 2. Shown is a Western blot probed with antibodies against actin and tubulin. Tubulin binding to the liposomes is not affected by the activation of GTP-binding proteins. By contrast, actin levels are greatly increased in the presence of GTPgS. Actin polymerization is inhibited by cytochalasin D (CytoD), brefeldin A (BFA), and the p23 peptide as described in the legend to Fig. 1

3.5. Characterizing Actin Polymerization by Western Blotting

1. Instead of staining the SDS gel, the presence or absence of specific proteins can be assessed by Western blot analysis. 2. Transfer proteins onto a nitrocellulose membrane using an electroblot apparatus containing transfer buffer (BioRad). 3. Block the nitrocellulose by incubating in blocking solution for 1 h at room temperature. 4. Wash the membrane two times with TBS–Tween solution. 5. To detect actin, incubate the membrane in rabbit anti-actin antibody (Sigma) diluted 1/500 in TBS–Tween for 1 h at room temperature (see Note 8). 6. Wash the membrane five times with TBS–Tween and incubate with peroxidase-conjugated anti-rabbit secondary antibodies (BioRad) diluted 1/4000 in TBS–Tween for 30 min. 7. Wash the membrane five times with TBS–Tween and incubate with a chemiluminescence reagent (Pierce). 8. Expose blot to film as shown in Fig. 2.

4. Notes 1. Chill approximately 100 L of water ahead of time. 50 L cylindrical tanks with lids (Nalgene) are convenient for chilling water and carrying out the dialyses. 2. Alternatively, synthetic phospholipids may be mixed to obtain a bilayer with a defined phospholipid composition. 3. The second dialysis tank can be reused for the first dialysis in Subheading 3.1, step 11. 4. The ammonium sulfate pellets may be dislodged from the centrifuge bottles with a pipette and easily resuspended using a 100-ml dounce homogenizer.

102

Stamnes and Xu

5. A trace amount of fluorescent phospholipid may be added to the preparation so that the recovery of liposomes may be quantified using a fluorescence spectrometer. For this 10 mg/ml NBD-phosphatidylcholine (Avanti Polar Lipids) works well. 6. A bath sonicator may also be used to resuspend the phospholipids. 7. The reaction can easily be scaled up to characterize the properties of less abundant cytokeletal proteins. In this case, controls should be carried out to confirm that liposomes are successfully recovered from larger volume sucrose gradients. 8. Duplicate blots may be probed with other antibodies to characterize the properties of multiple cytoskeleton-related proteins.

Acknowledgments This work was supported by NIH grant RO1 GM068674 (M.S.). The methods for preparing cytosol and liposomes have been adapted from (13, 14). We thank Heidi Hehnly for reading the manuscript.

References 1. Cooke R (2004) The sliding filament model: 1972–2004. J Gen Physiol 123:643–656 2. Wu X, Jung G, Hammer JA III (2000) Functions of unconventional myosins. Curr Opin Cell Biol 12:42–51 3. Theriot JA (2000) The polymerization motor. Traffic 1:19–28 4. Ridley AJ (2006) Rho GTPases and actin dynamics in membrane protrusions and vesicle trafficking. Trends Cell Biol 16:522–529 5. Hehnly H, Stamnes M (2007) Regulating cytoskeleton-based vesicle motility. FEBS Lett 581:2112–2118 6. Fucini RV, Chen JL, Sharma C, Kessels MM, Stamnes M (2002) Golgi vesicle proteins are linked to the assembly of an actin complex defined by mAbp1. Mol Biol Cell 13:621–631 7. Chen JL, Fucini RV, Lacomis L, ErdjumentBromage H, Tempst P, Stamnes M (2005) Coatomer-bound Cdc42 regulates dynein recruitment to COPI vesicles. J Cell Biol 169:383–389

8. Chen JL, Lacomis L, Erdjument-Bromage H, Tempst P, Stamnes M (2004) Cytosol-derived proteins are sufficient for Arp2/3 recruitment and ARF/coatomer-dependent actin polymerization on golgi membranes. FEBS Lett 566:281–286 9. Fucini RV, Navarrete A, Vadakkan C, Lacomis L, Erdjument-Bromage H, Tempst P, Stamnes M (2000) Activated ADP-ribosylation factor assembles distinct pools of actin on golgi membranes. J Biol Chem 275:18824–18829 10. Xu W, Stamnes M (2006) The actin-depolymerizing factor homology and charged/ helical domains of drebrin and mAbp1 direct membrane binding and localization via distinct interactions with actin. J Biol Chem 281: 11826–11833 11. Welch MD, Iwamatsu A, Mitchison TJ (1997) Actin polymerization is induced by Arp2/3 protein complex at the surface of Listeria monocytogenes. Nature 385: 265–269

Actin Polymerization on Liposomes 12. Heuvingh J, Franco M, Chavrier P, Sykes C (2007) ARF1-mediated actin polymerization produces movement of artificial vesicles. Proc Natl Acad Sci U S A 104:16928–16933 13. MacDonald RC, MacDonald RI, Menco BP, Takeshita K, Subbarao NK, Hu LR (1991) Smallvolume extrusion apparatus for preparation of

103

large, unilamellar vesicles. Biochim Biophys Acta 1061:297–303 14. Wattenberg BW, Rothman JE (1986) Multiple cytosolic components promote intra-Golgi protein transport. Resolution of a protein acting at a late stage, prior to membrane fusion. J Biol Chem 261:2208–2213

Chapter 9 Electroformation of Giant Unilamellar Vesicles from Native Membranes and Organic Lipid Mixtures for the Study of Lipid Domains under Physiological Ionic-Strength Conditions L.-Ruth Montes, Hasna Ahyayauch, Maitane Ibarguren, Jesus Sot, Alicia Alonso, Luis A. Bagatolli, and Felix M. Goñi Abstract Giant unilamellar vesicles (GUVs) constitute a cell-sized model membrane system that allows direct visualization of particular membrane-related phenomena, such as domain formation, at the level of single vesicles using fluorescence microscopy-related techniques. Currently available protocols for the preparation of GUVs work only at very low salt concentrations, thus precluding experimentation under physiological conditions. In addition, the GUVs thus obtained lack membrane compositional asymmetry. Here we show how to prepare GUVs using a new protocol based on the electroformation method either from native membranes or organic lipid mixtures at physiological ionic strength. Additionally, we describe methods to test whether membrane proteins and glycosphingolipids preserve their natural orientation after electroformation of GUVs composed of native membranes. Key words: Giant unilamellar vesicles (GUVs), Electroformation, Physiological conditions, Biological membranes, Lipid domains

1. Introduction Extensive documentation has appeared in recent years describing the use of giant unilamellar vesicles (GUVs, 20 µm mean diameter) as model systems to study different physical aspects of membranes (lateral structure, mechanical properties), particularly considering the effect of not only lipid–lipid but also lipid–DNA, lipid–peptide, and lipid–protein interactions (1–6). Owing to their size, single vesicles can be directly observed using light microscopy techniques (1, 2). V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_9, © Humana Press, a part of Springer Science+Business Media, LLC 2010

105

106

Montes et al.

However, two main drawbacks regarding GUV preparation are still unsolved: (1) low salt concentration (<10 mM NaCl) must be used during the preparation process (higher ionic strengths impair GUV formation) (7) and (2) GUVs lack membrane asymmetry. Interesting methodological advances include the possibility of preparing GUVs at high salt concentrations, provided that negatively charged lipids are present in high proportions in the bilayer composition (8). Also, Baumgart et al. generated giant plasma membrane vesicles by inducing plasma membrane vesiculation or “blebbing” (9). A protocol was also published for the generation of asymmetric GUVs, but in the absence of proteins (10). Recently, on the basis of a protocol described by Pott et al. (11), a method for the electroformation of GUVs (Fig. 1) from either cell membranes (Fig. 2) or lipid extracts (Fig. 3) under physiological ionic-strength conditions was published by our laboratories (12). When the starting materials are red blood cell (RBC) membranes, the resulting giant vesicles keep the membrane asymmetry. The electroformation method and ancillary techniques are described in detail below.

2. Materials 2.1. Preparation of Human Erythrocyte Ghosts

1. Assay buffer: In all experiments, the assay buffer consists of 25 mM HEPES, 150 mM NaCl, pH 7.2. 2. Right-side out (RSO) ghosts hemolysing nonsaline buffer (A): 1.2 mM acetic acid, 4 mM MgSO4, pH 3.2. 3. Sucrose buffer (B): 43% sucrose in assay buffer. Keep at 4°C in the refrigerator. 4. Inside out (IO) ghosts hemolysing solution (C): 5 mM KH2PO4, pH 8.0. 5. IO ghosts membrane resealing buffer (D): 0.5 mM Tris– HCl, 1 mM EDTA, 2 mM DTT, pH 8.5. 6. Glucose buffer (E): 1.03 g/ml glucose in assay buffer (d = 1.03). 7. Buffer F: 10 mM Tris–HCl, 1 mM ehtylenediamide tetraacetic acid (EDTA) pH 7.8. 8. Buffer G: 20 mM HEPES, 0.5 mM dithiothreitol (DTT), 1 mM MgCl2, 150 mM KCl, pH 7.2. 9. Molibdate solution (H): 22 mg (NH4)6Mo7O24⋅4H2O, 143 ml H2SO4 in 1 L dH2O. 10. EDTA anticoagulating tubes for human blood collection (BD Vacutainer Systems, NJ, USA). 11. Three 23-gauge microlance needles (BD Biosciences, NJ, USA).

Electroformation of Giant Unilamellar Vesicles

107

12. Cholesterol concentration oxidase/peroxidase measurement kit (BioSystems, Barcelona, Spain). 2.2. Preparation of GUVs

1. DiIC18 stock solution: 1.25 mM DiIC18 in dimethyl sulfoxide (DMSO). Protect from light and store in the refrigerator. 2. Stocks of lipid mixtures in organic solvent: 0.2 mg/ml stock of the desired composition in organic solvent (generally chloroform:methanol [2:1] [vol:vol]) supplemented with 0.2 mol% of DiIC18. Keep in the freezer at −20°C. 3. Electroformation chamber that allows visualization under the microscope (homemade, see Fig. 1).

Fig. 1. Homemade GUV electroformation chamber. The Pt electrodes can be extracted from the chamber to allow sample addition and vacuum treatment. A cover glass allows direct visualization of GUVs under the microscope. The temperature is maintained constant by using a water bath

108

Montes et al.

4. 24 × 50 mm cover glasses (Menzel-Gläser, Braunschweig, Germany) to couple at the bottom of the visualization chamber. 5. TG330 function generator from Instruments (Cambridgeshire, UK).

Thurlby

Thandar

6. F12 water bath (Julabo, Seelbach, Germany) for adjusting vesicle formation temperature. 2.3. Immunostaining of RBC ghost membranes (Membrane Asymmetry Assays)

1. Primary antibodies: Mouse IgG antihuman band III (Sigma, MO, USA). Goat IgG antihuman glycophorin A and blood group A1, A2, A3 antigen mouse monoclonal IgM (Santa Cruz Biotechnologies, CA, USA). 2. Secondary antibodies: Alexa fluor 488 donkey anti-goat IgG and Alexa fluor 633 rabbit anti-mouse IgG (Invitrogen, OR, USA). 3. Alexa fluor 633 protein labeling kit (Invitrogen).

2.4. Fluorescence Microscopy

1. Zeiss-LSM 510 META NLO inverted confocal excitation fluorescence microscope (Carl Zeiss, Jena, Germany). 2. Multitrack mode software (Carl Zeiss). 3. C-Apochromat objective 40×, water immersion, NA 1.2 (Carl Zeiss).

3. Methods 3.1. Preparation of Human Erythrocyte Ghosts

Blood (group A) is collected from healthy donors, placed in EDTA anticoagulating tubes, and washed three times (or until supernatant is clear) with assay buffer by centrifugation at 1,250 × g for 10 min at 4°C. The supernatant is discarded and the erythrocyte-containing pellet is used to form ghosts. Around 5 ml of fresh blood should be enough to obtain an adequate final ghost concentration (see Note 1).

3.1.1. Preparation of RSO Ghosts

RSO ghosts are prepared following the method of Steck and Kant (13), with some modifications: 1. Osmotic shock: In order to produce hemolysis, add 10 volumes of non-saline buffer (A) to the packed RBCs. Mix gently and incubate for 15 min, 4°C, under smooth agitation. After centrifugation at 31,000 × g for 15 min at 4°C, remove the supernatant and wash the pellet two more times (or until the supernatant is clear) in the same buffer to obtain the ghost membranes. 2. Membrane resealing: Add 10 ml isoosmotic assay buffer to the precipitate and incubate for 1 h at 37°C. Centrifuge at 14,000 × g for 10 min and repeat 2–3 washes. Bring the

Electroformation of Giant Unilamellar Vesicles

109

resulting membrane suspension to a final volume of around 1.5 ml in assay buffer. 3. Sealed membrane separation by sucrose gradient centrifugation: To separate the fractions (sealed and non-sealed membranes), layer 1.5 ml of membranes over 6 ml sucrose buffer (B) in a centrifuge tube. Centrifuge using a swing-out rotor at 138,000 × g for 60 min at 4°C. Recover the band containing the resealed vesicles (RSO ghosts), which floats on top of the sucrose solution. Harvest and wash the resealed ghosts three more times in assay buffer at 14,000 × g for 10 min. Resuspend in around 1 ml assay buffer. 3.1.2. Preparation of Inside Out Ghosts

IO ghosts preparation is based on the method by Schow and Passow (14) with some modifications: 1. Osmotic shock: Dilute the washed erythrocytes in 20 volumes of hemolysing solution (C). Incubate at room temperature for 5 min to break down the cells, and pellet by centrifugation at 35,000 × g for 20 min. Repeat this wash twice with the same buffer. 2. Membrane resealing: Resuspend hemoglobin-free membranes in 10 ml of membrane resealing buffer (D). Incubate for 45 min at 4°C and for 45 min at 37°C to allow membrane resealing. Centrifuge at 28,000 × g for 30 min, and wash the pellet three times with the same buffer. After the last wash, bring the sample to a final volume of 2 ml and homogenate with a 23-gauge needle for 10–15 times. 3. Sealed membrane separation by glucose barrier centrifugation: To purify the resealed IO ghosts, layer 2 ml of membranes over 6 ml glucose buffer (E) and centrifuge at 100,000 × g for 2 h at 4°C. Collect the vesicles remaining above the barrier, wash once in 20 ml buffer F for 30 min at 40,000 × g, and resuspend in around 1 ml buffer G.

3.1.3. Determination of Lipid Concentration

In order to form GUVs from erythrocyte membranes, ghost suspensions are brought to a final lipid concentration (cholesterol + phospholipids) of ≅2.5 mM (~1 mM final concentration of cholesterol can be measured using a cholesterol concentration measurement kit). Phospholipid concentration of ghosts is determined by measuring the phosphorous content (15, 16) of lipids extracted from the ghost membranes (17).

3.1.3.1. Ghost Lipid Extraction

1. Precipitation of lipoproteic complexes: Take 250 µl of the ghost suspension and add the same volume of cold 0.6 M perchloric acid in 10-ml glass tubes. Centrifuge at 14,000 × g for 15 min, and discard the supernatant. 2. Lipid extraction: Resuspend the precipitate in 2.5 ml of cold chloroform:methanol (2:1, [vol:vol]) and incubate at room temperature for 30 min mixing every 5 min. Add 5 ml of cold

110

Montes et al.

0.1N HCl, vortex, and centrifuge at 1,700 × g for 20 min. Discard the upper phase and recover the chloroformic lipidcontaining lower phase. Repeat the extraction on the intermediate phase and mix the two chloroformic fractions together. 3.1.3.2. Phospholipid Concentration Determination

1. Prepare the calibration curve from 0 to 25 nmol of NaH2PO4 standard from a 1 mM stock. 2. Add 100 µl of 70% perchloric acid to 200 µl of sample/calibration curve, and let digestion happen at 200°C for 45 min. Mix with 500 µl distilled H2O, 100 µl molibdate solution (H), and 100 µl 10% ascorbic acid. The colorimetric reaction takes place at 100°C in 5–10 min, after which absorbance can be measured in a colorimeter at 820 nm. 3. Calculate ghost phospholipid concentration:

3.2. Preparation of GUVs from Erythrocyte Membranes (G-ghosts) and from Organic Lipid Mixtures under Physiological Conditions 3.2.1. Formation of G-ghosts

[P] × total volume 200 where [P] is the concentration calculated by this protocol and total volume is the volume of the chloroformic phase obtained after lipid extraction. [Ghost phospholipids] =

In order to be able to visualize the GUVs under the microscope, add 1 mol% DiIC18 (from the stock) with respect to total lipids to the ghost solutions before the formation. Deposit a small aliquot (~1 µl) of erythrocyte ghost suspension (RSO or IO) onto each Pt electrode in a special homemade chamber, which allows direct visualization under the microscope (Fig. 1) (18). Cover the sample to avoid light exposure and allow precipitation onto the Pt wires for 5 min at room temperature (see Notes 2 and 3).

3.2.2. Formation of GUVs Composed of Organic Lipid Mixtures

Add 2.5 µl of the appropriate lipid stock on the surface of each Pt electrode and let the solvent traces evaporate by placing the electrodes under high vacuum for at least 2 h.

3.2.3. Electroformation under Physiological Conditions

Place the electrodes on the chamber and add 500 µl of isoosmotic assay buffer. Apply an AC field using a function generator in three main steps, usually at 37°C: 1. Frequency 500 Hz, amplitude 106 mV (35 V/m) for 5 min, 2. Frequency 500 Hz, amplitude 940 mV (313 V/m) for 20 min, 3. Frequency 500 Hz, amplitude 2.61 V (870 V/m) for 90 min. The temperature used for GUV formation should ensure that the membrane under study displays a single fluid phase (above the highest phase transition temperature). Due to the design of the chamber used in vesicle formation (Fig. 1), GUVs can be directly observed under the microscope, avoiding transfer of vesicles to conventional chambered cover glasses (see Notes 4 and 5).

Electroformation of Giant Unilamellar Vesicles

3.3. Immunostaining of RBC Ghost Membranes (Membrane Asymmetry Assays)

111

The extracellular domain of glycophorin A (outer hemimembrane label) is identified with a primary goat IgG antibody. The bound antibody is detected by a secondary Alexa fluor 488-labeled donkey anti-goat IgG antibody. The cytoplasmic domain of band III (inner hemimembrane label) is recognized with a primary mouse IgG antibody and a secondary Alexa fluor 633-labeled rabbit anti-mouse IgG antibody (19). Blood group A antigen mouse monoclonal IgM was labeled with the Alexa fluor 633 protein labeling kit (Fig. 2). 1. Add each primary antibody at a final concentration of 8 µg/ml. 2. Incubate for 30 min at room temperature to allow binding of the antibodies to their specific antigen. 3. Wash three times with physiological buffer (the same buffer used during electroformation). This will reduce considerably the background fluorescent signal. 4. When necessary, add the secondary antibodies, incubate, and wash under the same conditions. Addition of primary and secondary antibodies in two steps avoids agglutination of the immunoglobulins (see Notes 6 and 7).

3.4. Fluorescence Microscopy

An inverted confocal excitation fluorescence microscope is used to visualize giant vesicles in this kind of experiments (see Note 8). Excitation wavelengths: 488, 543, and 633 nm (for Alexa 488, DiIC18, and Alexa 633 labeled antibodies). Filters: The fluorescence signals for Alexa 488, DiIC18, and Alexa 633 can be simultaneously collected using multitrack mode into three (or two, depending on the experiment) different channels using bandpass filters of 520 ± 10, 590 ± 25, and 670 ± 25 nm, respectively. Objective: Use a 40× water immersion, NA 1.2 objective.

4. Notes 1. Sometimes, RBCs break up during the centrifugation washes. This is detectable because the hemoglobin-containing supernatant becomes red, which can be avoided by decreasing the centrifugation force (g) and increasing the washing time. 2. The final volume of DMSO added to a membrane-containing sample must always be less than 2% of the final total volume in order to preserve membrane integrity. Stocks should be concentrated enough to obey this proportion. 3. Whenever working with fluorescent probes, it is convenient to cover and protect the samples from light, and to place the stocks in amber containers in order to preserve the fluorescent properties of the probes.

112

Montes et al.

4. It is important to control temperature conditions in these experiments: electroformation of GUVs must take place above the gel–fluid transition temperature of the membrane under study. Take also into account that lipid domain formation depends on the temperature at which GUVs are visualized. 5. If electroformation takes place at a high temperature, buffer might evaporate while the experiment is running. Covering the chamber with parafilm might reduce evaporation to keep the volume of the buffer constant. 6. There is a vast variety of fluorescent probes (with different partition properties than DiIC18) which allow visualization of membrane domains from different natures (see e.g., ref. (20)). There are also membrane lipid-based probes available (e.g., NBD-SM) for the study of domain formation due to specific lipid degradation (e.g., by a sphingomyelinase) (21). Take into account the dynamic properties and photobleaching susceptibility when selecting a probe for confocal fluorescence experiments. 7. In this chapter we show how to form giant vesicles from erythrocyte membranes and from some lipidic mixtures of our interest in the study of lipid domains (Figs. 2 and 3). Nevertheless, other lipid mixtures, cell types, or fractionated organelles might be susceptible to GUV formation. Their visualization and study of interaction with other molecules (such as DNA, proteins, lipids) should then be possible. The immunostaining of specific proteins present in each biological membrane can help to assess whether membrane integrity is maintained after electroformation. The experiments shown here seem to demonstrate that electroformation does not affect the presence and orientation of ghost membrane proteins, but similar experiments could be performed on membranes from different nature. 8. Here we electroform GUVs using a high-ionic strength buffer to reproduce physiological conditions. The chamber used allows direct visualization of GUVs, but if this technology is not available, the buffers could be supplemented with glucose/sucrose to facilitate vesicle sedimentation to the bottom of chambered cover glasses (18).

Fig. 3. (continued) low-fluorescence-intensity regions correspond to liquid-ordered lipid domains enriched in SM and Chol. (d) GUVs composed of PI (phosphatidylinositol):SM (1:1) (mol:mol) supplemented with DG (diglyceride); the dark areas represent gel-like SM-enriched domains. (e) GUVs composed of PI:GalCer (1:1) (mol:mol); the low-fluorescenceintensity regions correspond to gel domains enriched in GalCer. (f) GUVs composed of PI:DSPC (distearoyl phoshatidylcholine):DG (1:1:0.3) (mol:mol:mol:mol); the dark areas represent gel-like DSPC-enriched domains. All the images were obtained in assay buffer (i.e., under physiological conditions). The white bars correspond to 10 mm

Fig. 2. (a) Multicolor fluorescence image of RSO G-ghosts in the presence of band III (red, A2) and glycophorin A (magenta, A3) specific immunofluorescence markers (false color representation); the G-ghosts are labeled with DiIC18 lipophilic fluorescent probe (yellow, A1). (b) Multicolor fluorescence image of IO G-ghosts in the presence of band III (red, B2) and glycophorin A (magenta, B3) specific markers. As shown in A, the G-ghosts are labeled with DiIC18 (yellow, B1). (c) and (d) Fluorescence intensity representative images of RSO (c) and IO (d) G-ghosts in the presence of blood group A specific marker (blue); the G-ghosts are labeled with DiIC18 (green). All the images have been obtained in assay buffer (i.e., under physiological conditions). The white bars correspond to 20 mm. From ref. (12), with permission

Fig. 3. Fluorescence images (false color representation) of lipid domains appearing in DiIC18-labeled GUVs. (a) GUVs composed of red blood cell membranes (G-ghosts) after sphingomyelinase addition; the brighter regions correspond to domains enriched in ceramide produced by enzyme activity. (b) GUVs composed of POPC (1-palmitoyl-2oleoyl phosphatidylcholine)/DPPC (dipalmitoyl phosphatidylcholine) (3:2) (mol:mol); the high-fluorescence-intensity areas correspond to DPPC-rich gel phase. (c) GUVs composed of PC:PE:SM:Chol (1:1:1:1) (mol:mol:mol:mol); the

114

Montes et al.

Acknowledgements This work was supported by FNU, Denmark (272-06-0511), the Danish National Research Foundation, the Spanish Ministerio de Educación y Ciencia (BFU 2005-0695 and BFU 2004-02955), and the University of the Basque Country (GIU06/42). References 1. Bagatolli LA (2006) To see or not to see: Lateral organization of biological membranes and fluorescence microscopy. Biochim Biophys Acta 1758:1541–1556 2. Menger FM, Keiper JS (1998) Chemistry and physics of giant vesicles as biomembrane models. Curr Opin Chem Biol 2:726–732 3. Veatch SL, Keller SL (2005) Seeing spots: Complex phase behavior in simple membranes. Biochim Biophys Acta 1746:172–185 4. Ambroggio EE, Separovic F, Bowie JH, Fidelio GD, Bagatolli LA (2005) Direct visualization of membrane leakage induced by the antibiotic peptides: Maculatin, citropin, and aurein. Biophys J 89:1874–1881 5. Bernardino de la Serna J, Perez-Gil J, Simonsen AC, Bagatolli LA (2004) Cholesterol rules: Direct observation of the coexistence of two fluid phases in native pulmonary surfactant membranes at physiological temperatures. J Biol Chem 279:40715–40722 6. Kahya N, Schwille P (2006) Fluorescence correlation studies of lipid domains in model membranes. Mol Membr Biol 23:29–39 7. Bagatolli LA, Parasassi T, Gratton E (2000) Giant phospholipid vesicles: Comparison among the whole lipid sample characteristics using different preparation methods: A two photon fluorescence microscopy study. Chem Phys Lipids 105:135–147 8. Akashi K, Miyata H, Itoh H, Kinosita K Jr (1996) Preparation of giant liposomes in physiological conditions and their characterization under an optical microscope. Biophys J 71:3242–3250 9. Baumgart T, Hammond AT, Sengupta PS, Hess T, Holowka DA, Baird BA, Webb WW (2007) Large scale fluid/fluid phase separation of proteins and lipids in giant plasma membrane vesicles. Proc Natl Acad Sci USA 104:3165–3170 10. Pautot S, Frisken BJ, Weitz DA (2003) Engineering asymmetric vesicles. Proc Natl Acad Sci USA 100:10718–10721 11. Pott T, Bouvrais H, Méléard P (2008) Giant unilamellar vesicle formation under

12.

13.

14. 15. 16. 17. 18.

19.

20.

21.

physiologically relevant conditions. Chem Phys Lip 154:115–119 Montes LR, Alonso A, Goñi FM, Bagatolli LA (2007) Giant unilamellar vesicles electroformed from native membranes and organic lipid mixtures under physiological conditions. Biophys J 10:3548–3554 Steck TL, Kant JA (1974) Preparation of impermeable ghosts and insideout vesicles from human erythrocyte membranes. Meth Enzymol 31:172–180 Schow G, Passow H (1973) Preparation and properties of human erythrocyte ghosts. Mol Cell Biochem 2:197–217 Fiske HC, Subbarow Y (1925) The colorimetric determination of phosphorus. J Biol Chem 66:375–400 Böttcher C, van Gent C, Fries C (1961) A rapid and sensitive sub-micro phosphorous determination. Anal Chim Acta 1061:297–303 Bligh EG, Dyer WJ (1959) A rapid method of total lipid extraction and purification. Can J Biochem Physiol 37:911–917 Fidorra M, Duelund L, Leidy C, Simonsen AC, Bagatolli LA (2006) Absence of fluidordered/fluid-disordered phase coexistence in ceramide/POPC mixtures containing cholesterol. Biophys J 90:4437–4451 Kaufmann S, Tanaka M (2003) Cell adhesion onto highly curved surfaces: One-step immobilization of human erythrocyte membranes on silica beads. Chemphyschem 4:699–704 Sot J, Ibarguren M, Busto JV, Montes LR, Goñi FM, Alonso A (2008) Cholesterol displacement by ceramide in sphingomyelin-containing liquid-ordered domains, and generation of gel regions in giant lipidic vesicles. FEBS Lett 582:3230–3236 Montes LR, López DJ, Sot J, Bagatolli LA, Stonehouse MJ, Vasil ML, Wu BX, Hannun YA, Goñi FM, Alonso A (2008) Ceramideenriched membrane domains in red blood cells and the mechanism of sphingomyelinaseinduced hot-cold hemolysis. Biochemistry 47:11222–11230.

Chapter 10 Visualization of Lipid Domain-Specific Protein Sorting in Giant Unilamellar Vesicles Martin Stöckl, Jörg Nikolaus, and Andreas Herrmann Abstract Recent studies suggest that phospholipids in the plasma membrane of mammalian cells are not homogenously distributed but may form domains either by lipid–lipid interactions or/and as consequence of lipid– protein interactions. Such lipid compartments may act as protein recruiting platforms which, for example, are essential components of cell signaling pathways. Model membrane systems with a defined lipid composition are ideally suited to study domain-specific interactions of peripheral and integral membrane proteins. Giant unilamellar vesicles (GUVs) offer the opportunity to directly visualize in parallel, both the lateral lipid domains and the interaction sites of proteins using fluorescence microscopy. The application of GUVs is exemplarly illustrated for studying domain-specific interactions of the protein a-synuclein and the domain-specific distribution of synthetic transmembrane peptides. Key words: Giant unilamellar vesicles (GUVs), Fluorescence microscopy, Lipid domain, Raft, a-Synuclein, Transmembrane domain

1. Introduction In recent years, the relevance of lipid domains in plasma membranes of eukaryotic cells serving as a platform for recruitment and enrichment of specific integral and peripheral membrane proteins has become a challenging research focus in cell biology, biochemistry, and biophysics. Those domains may play an important role in cellular uptake mechanisms, signal transduction, and vesicle and virus budding (1–4). Domains are formed by lipid–lipid interactions, but can be also triggered by protein–lipid interactions. However, specific attention has been given to domains based on lipid–lipid interactions. Interaction between cholesterol and glycosphingolipids or saturated phospholipids, in particular phosphatidylcholine, is known to generate liquid ordered (lo) V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_10, © Humana Press, a part of Springer Science+Business Media, LLC 2010

115

116

Stöckl, Nikolaus, and Herrmann

domains clearly distinct from liquid-disordered (ld) domains (5, 6). The latter are essentially formed by unsaturated phospholipids and characterized by a reduced lipid packing density and membrane thickness with respect to lo domains. Both lipid packing and membrane thickness are important factors for specific redistribution of integral and peripheral proteins to distinct lipid domains. Of course, other factors are feasible. Enrichment of phospholipids with charged headgroups in distinct lipid domains can direct peripheral proteins to those domains via electrostatic interaction (7, 8). In native cells, it is difficult to visualize lipid domains and to study domain-specific enrichment of membrane proteins in the plasma membrane. Indeed, several attempts to image lo domains, the so-called rafts, in the plasma membrane of mammalian cells suggested that they are highly dynamic and very small, in the order of 10–100 nm (9). Giant Unilamellar Vesicles (GUVs) have become an attractive tool to study lipid domains. GUVs of appropriate lipid mixtures have been shown to form lo and ld domains with a size in the micrometer scale (5, 10, 11). These domains can be visualized by various lipid-like fluorophores which preferentially enrich either in the ld or in the lo domain. Such model systems are well suited to study the association of membrane proteins with specific lipid domains. For this purpose, the lateral distribution of fluorescently tagged proteins can be compared with the lipid domain-specific distribution of lipid-like fluorophores. For the protein a-synuclein, which is associated with the development of neurodegenerative diseases as Parkinson’s disease, it has been suggested that interaction with membranes may play a relevant role in pathogenesis. However, a controversy about the membrane binding site of a-synuclein exists (12, 13). An approach based on lipid domainforming GUVs allows direct visualization of potential binding sites of fluorescent-tagged proteins (8). Using the GUV model system, it is also possible to study the lateral distribution of transmembrane domains of integral membrane proteins. Exemplarily, the distribution of the so-called LV peptides, which were designed as a low-complexity model that mimics fusion protein transmembrane domains, will be illustrated here (14).

2. Materials 2.1. Preparation of GUVs on Indium Tin Oxide (ITO) Slides

1. ITO slides are glass slides (50 × 60 mm; 1.1 mm thickness) which are coated on one side with ITO (R £ 20 W/sq) from Präzisions Glas & Optik GmbH (Iserlohn, Germany). 2. Chloroform, methanol, and trifluoroethanol were of spectroscopic grade (Uvasol; Merck KGaA, Darmstadt, Germany).

Visualization of Lipid Domain-Specific Protein Sorting in Giant Unilamellar Vesicles

117

Fig. 1. Chamber design based on ITO-coated glass slides

3. Lipids and cholesterol were obtained from Avanti Polar Lipids (Alabaster, AL). 4. Polytetrafluoroethene (PTFE) spacers can be cut from several layers of PTFE film for surface protection. Alternatively, a silicone rubber spacer (thickness: 1 mm) can be used. For a sketch of the chamber assembly, see Fig. 1. PVC-based nonmelting plasticine (FimoTM; Eberhard Faber GmbH, Germany) was used for sealing chambers. 5. Swelling buffer: 250 mM sucrose and as a preservative 15 mM NaN3 in double deionized water. Measure osmolality of buffer (approx. 280 mOsm/kg). (see Note 1) Remove particles by means of filtration (0.22 µm pores). Store at 4°C until use. Note that NaN3 is highly toxic. 2.2. Preparation of GUVs on Titanium Slides

1. Titanium chambers (33 × 50 mm; 2 mm thickness) were made from pure titanium slides from Alfa Aesar GmbH (Karlsruhe, Germany). For a sketch of the chamber assembly see Fig. 2. 2. Parafilm is a product of Pechiney plastic packaging (Chicago, IL). 3. Swelling buffer: 250 mM sucrose and as a preservative 15 mM NaN3 in double deionized water. Measure osmolality of buffer (approx. 280 mOsm/kg) (see Note 1). Remove particles by means of filtration (0.22 µm pores). Store at 4°C until use. Note that NaN3 is highly toxic.

118

Stöckl, Nikolaus, and Herrmann

Fig. 2. Chamber design based on titanium slides

4. PVC based non melting plasticine (FimoTM; Eberhard Faber GmbH, Germany) is used for sealing of filling holes. 2.3. Preparation of GUV Harboring Transmembrane Peptides

2.4. Membrane Binding of aSynuclein Evaluated by Laser Scanning Confocal Microscopy

1. Materials for GUV preparation on ITO or titanium slides correspond to the materials described in Subheading 2.1 (ITO slides) or Subheading 2.2 (titanium slides), respectively. 2. The peptide (amino acid sequence Rh-KKKKWLLVLLVLLVLLVLL VLKKKK-Rh) was synthesized by Boc chemistry (PSL, Heidelberg, Germany by courtesy of Prof. Dieter Langosch, TU Munich). Rhodamine-tagged peptides were made by coupling of a Lys derivative (Fmoc-Lys(Dde)-OH) to the C- and N-termini during synthesis. Reaction of the peptide with 5-(and-6)carboxytetramethylrhodamine succinimidylester (TMR-SE) yielded TMR-labeled peptides (see Notes 2 and 3). 1. Tetramethylrhodamine-6-maleimide was obtained from Molecular Probes (Eugene, OR). 2. Microscopy buffer: 280 mM glucose, 5.8 mM NaH2PO4, and 5.8 mM Na2HPO4 in double deionized water. Measure osmolality of buffer (approx. 300 mOsm/kg). If necessary, adjust by dilution with double deionized water. Remove particles by means of filtration (0.22 µm pores). Store at 4°C until use. 3. Glass coverslips, 24 × 60 mm (thickness: No. 1), were obtained from Carl Roth GmbH (Karlsruhe, Germany). Coverslips were used without prior cleaning.

Visualization of Lipid Domain-Specific Protein Sorting in Giant Unilamellar Vesicles

119

3. Methods Using appropriate lipid mixtures, GUVs with laterally segregated lipid domains are readily accessible by the electroformation technique using ITO-coated glass slides developed by Angelova and Dimitrov (15, 16). However, in a recent work it has been shown that the ITO may catalyze the oxidation of unsaturated lipids (17). Chambers made from titanium sheets can be used to circumvent that issue. Yet, no differences were found using the described lipid mixtures. For visualization of the domain structure, specific fluorescent lipid analogues such as C6-NBD-PC (1-palmitoyl-2-[6-[(7-nitro-2-1,3-benzoxadiazol4-yl)amino]hexanoyl]-sn-glycero-3-phosphocholine) are included in the lipid mixture. The ld domains in the vesicles show bright green fluorescence as C6-NBD-PC preferentially distributes to those domains, while the lo domains, from which C6-NBD-PC is excluded, appear almost completely dark in fluorescence microscopy. After incubation of a GUV suspension with the fluorescent-labeled proteins, protein–membrane interactions can be qualitatively and quantitatively investigated by measuring the fluorescence level at the vesicle membrane. For the investigation of transmembrane peptides, these hydrophobic peptides can be directly mixed to the lipid mixture before vesicle preparation. 3.1. Preparation of GUVs on ITO Slides

1. Clean two ITO slides (Fig. 1) in three successive steps by washing with detergent, deionized water, and finally ethanol. Rub down with lint-free paper. 2. Prepare lipid stocks by dissolving powdered lipids in chloroform and dilute to a suitable concentration (5 mM). If a turbid solution results (lipids are not completely dissolved), add methanol until the lipid stock is clear. To minimize solvent evaporation, stocks should be stored in tightly sealed glass vials at −20°C until use. 3. For GUV preparation, mix on ice 100 nmol of total lipids in a volume of 50 µl chloroform in a small glass vessel. For detection of domains by fluorescence microscopy, add 1 nmol of C6-NBD-PC. Keep the stocks on ice all the time. See Note 4 for lipid mixtures. 4. Place the slides on a heating stage and heat to approx. 50°C. Pipette 25 µl of the lipid mixture in small droplets on each of the conducting side of the slides. Let the solvent evaporate completely. Check the quality of spotted lipid films (see Note 5). 5. Place the spotted slides in an exsiccator and apply vacuum (<10 mbar) for at least 1 h to evaporate remaining traces of solvent.

120

Stöckl, Nikolaus, and Herrmann

6. Assemble the GUV chamber consisting of the two lipidcoated ITO slides and an additional PTFE or silicone rubber spacer (1 mm thickness). The chamber is held together by two clamps. If PTFE spacers are used, the chamber has to be sealed with nonmelting modeling clay to prevent leakage. 7. Preheat the required amount of swelling buffer (approx. 1 ml per chamber) to 50°C. Fill the whole chamber with the heated buffer using a syringe and needle. Seal the chamber opening with modeling clay. 8. For electroformation of GUVs, heat the chamber to 50°C and apply a sinusoidal alternating voltage (10 Hz) which is stepwise increased from 0.02 to 1.1 V over 48 min and hold this voltage for at least 100 min. To detach the vesicles, incubate for further 30 min with an increased voltage of 1.3 V at a frequency of 4 Hz. 9. Remove the vesicle suspension from the chamber with a syringe and needle. GUVs are stable for a few days when stored at room temperature in the dark. 3.2. Preparation of GUVs on Titanium Slides

1. For cleaning of titanium chambers and preparation of lipid stocks see steps 1–3 of the protocol for GUV preparation on ITO slides. 2. To achieve an insulated layer between the bottom and top titanium slides, heat the bottom slide to approx. 50°C and surface-fuse one layer of Parafilm to the rim of the slide. Cut away excess Parafilm from the edges and the cavity of the slide with a razor blade. 3. Put the slides on a heating stage and heat to approx. 50°C. Pipette 25 µl of the lipid mixture in small droplets on each of the slides. Let the solvent evaporate completely. 4. Put the spotted slides in an exsiccator and apply vacuum (<10 mbar) for at least 1 h to evaporate remaining traces of solvent. 5. For chamber assembly, heat the bottom slide until the Parafilm layer melts and becomes sticky. Press on the top slide firmly to seal the chamber. After cooling to room temperature, insulation should be checked by means of measuring the resistance between the slides with a multimeter. The resistance should be >200 kW. 6. Preheat the required amount of swelling buffer (approx. 1 ml per chamber) to 50°C. Fill the whole chamber with the heated buffer with a pipette. Seal the filling holes with modeling clay. 7. Electroformation is equivalent to the procedure described for ITO slides. GUV suspension again is removed by pipetting.

Visualization of Lipid Domain-Specific Protein Sorting in Giant Unilamellar Vesicles

3.3. Preparation of GUVs Harboring Transmembrane Peptides and Visualization of the Lateral Distribution of the Peptides by Confocal Laser Scanning Microscopy

121

1. GUVs harboring transmembrane peptides can be generated using either ITO (see Subheading 3.1) or titanium slides (see Subheading 3.2) following the respective protocol. 2. For preparation of GUV, add 1 nmol of the peptide dissolved in trifluoroethanol and 1 nmol of C6-NBD-PC for the visualization of lipid domains to 100 nmol of total lipids including cholesterol in a total volume of 50 µl chloroform in a small glass vessel. 3. Continue GUV preparation by spotting small droplets of the lipid/peptide mixture according to step 4 (ITO) or step 3 (titanium) of the protocol, respectively (see Note 5). 4. To investigate the lateral distribution of the peptides, suspensions of GUVs containing peptide and C6-NBD-PC are added to microscopy buffer at a ratio of 1:1 to 1:3. Prior to microscopy, GUVs are given some minutes to settle down on a cover slip because of the greater specific density. 5. For microscopy, follow steps 4–6 of the Subheading 3.4 about laser scanning confocal microscopy. NBD fluorescence – enriched in the ld domains – together with the rhodamine fluorescence of the peptides allows a direct visualization of the lateral distribution of the peptides (see Fig. 3). See Note 10 for lipid abbreviations.

Fig. 3. Lateral distribution of transmembrane peptides in GUVs of two lipid mixtures. TMR-labeled synthetic transmembrane peptide (1 mol%) and C6-NBD-PC (1 mol%) were incorporated into GUVs prepared from a nonraft lipid mixture of POPC/DOPE/DOPS (3/1/1 – molar ratio) (left column) and in lipid domain-forming GUVs made from DOPC/SSM/Chol (1/1/1 – molar ratio) (middle and right column). Confocal images of the equatorial plane of the GUVs show C6-NBD-PC fluorescence (top) and fluorescence of TMR-labeled peptide (bottom). C6-NBD-PC as a marker for the ld domain and the peptide are colocalized. Images were taken at 25°C. The scale bar corresponds to 10 mm

122

Stöckl, Nikolaus, and Herrmann

3.4. Membrane Binding of a-Synuclein Evaluated by Laser Scanning Confocal Microscopy

1. Perform covalent labeling of protein with an appropriate reactive fluorophore. Fluorescence of label should not interfere with fluorescence properties of lipid analogues used for domain visualization. Here, a-synuclein was labeled with tetramethylrhodamine-6-maleimide (TMR) (protein labeling was done according to the manufacturer’s protocol). 2. In a reaction tube, mix 75 µl of microscopy buffer (see Note 6) with labeled protein to yield the desired protein concentration in a total volume of 100 µl. Add 25 µl GUV suspension and mix well by pipetting. In case of strong membrane affinity, protein concentrations in the range of tenth of nanomoles per liter are sufficient. 3. For microscopy, pipette the sample on a coverslip (see Note 7). Wait a few minutes for the vesicles to settle down due to their greater specific density. 4. Using a laser scanning confocal microscope, the NBD lipid analogue is ideally excited with the 488-nm line of an Ar-ion laser; for excitation of the TMR fluorescence, a wavelength of 543 nm (He–Ne laser) or 559 nm (diode laser) is suitable (see Note 8). 5. Images of the equatorial focal plane of a single GUV allow a straightforward identification of the lipid domains. In the case of C6-NBD-PC, the ld domains show a bright green fluorescence due to the preferential enrichment of the probe. In contrast, lo or gel phase domains appear as dark regions, since the C6-NBD-PC is excluded (see Fig. 4 (DPPG), Fig. 5 and Note 9). See Note 10 for lipid abbreviations. 6. Binding of the fluorescent-labeled protein to the membrane and – if applicable – specific lipid domains can be directly deduced from the images by comparison with the C6-NBD-PC fluorescence (see Figs. 4 and 5). See Note 10 for lipid abbreviations. 7. For quantification of the amount of membrane bound protein, intensity profiles through the equatorial section of single vesicles can be used. The peak values of TMR fluorescence at the membrane are quite reliable for quantification of the bound protein (see Fig. 6 and Note 9).

4. Notes 1. The amount of sucrose in the swelling buffer is not predefined, but depends on the osmolality of the microscopy buffer. The osmolality of the swelling buffer should be approx. 20 mOsm/ kg lower or at least equal to that of the microscopy buffer, since otherwise the stability of the GUVs can be impaired.

Visualization of Lipid Domain-Specific Protein Sorting in Giant Unilamellar Vesicles

123

Fig. 4. Binding of a-synuclein requires negatively charged lipids in liquid disordered phase. Laser scanning microscopy images showing equatorial sections of GUVs incubated with fluorescent labeled a-synuclein (0.1 µM). All GUVs were labeled with 1 mol% C6-NBD-PC. DOPC: DOPC only; All other: 70 mol% DOPC and 30 mol% of indicated anionic lipid. Top image – fluorescence of C6-NBD-PC; Bottom image – fluorescence of TMR-labeled a-synuclein. All images were obtained at 25°C. Some GUVs contain smaller vesicles or lipid debris. The scale bar corresponds to 10 mm

2. The length of the hydrophobic stretch of the transmembrane peptide varies with the number of hydrophobic amino acids in the sequence of the peptide. Choosing lipids with corresponding acyl chain length may help to avoid a mismatch situation in which hydrophobic side chains may become exposed to polar environment, which is energetically unfavorable. 3. The presence of terminal lysines is a common approach to enhance the peptide solubility and to promote a transmembrane insertion into the membrane (18). 4. The lipid mixtures used for GUV preparation can be adapted to meet the requirements of the experiment. GUVs with membranes being solely in the liquid disordered state can be formed from unsaturated (e.g., DOPC) or partly unsaturated (e.g., POPC) phosphatidylcholines (PCs) with different acyl chain lengths. Alternatively, also PC preparations of natural sources

124

Stöckl, Nikolaus, and Herrmann

Fig. 5. Binding of a-synuclein requires negatively charged lipids in liquid disordered phase. Laser scanning microscopy images showing equatorial sections of GUVs incubated with fluorescent labeled a-synuclein (0.1 µM). All GUVs were labeled with 1 mol% C6-NBD-PC. Raft: DOPC/SSM/Chol = 33/33/33; Raft (DOPS): DOPC/SSM/Chol/DOPS = 25/25/25/25; Raft (DSPS): DOPC/SSM/Chol/DSPS = 25/25/25/25. Top image – fluorescence of C6-NBD-PC; Bottom image – fluorescence of TMR-labeled a-synuclein. All images were obtained at 25°C. Some GUVs contain smaller vesicles or lipid debris. The scale bar corresponds to 10 mm

Fig. 6. Quantitative analysis of labeled a-synuclein binding to GUV (DOPC/DOPS = 70/30). (a) In order to measure the fluorescence intensity at the vesicle membrane, a line profile in the equatorial plane of the GUV is selected. (b) The amount of labeled protein bound at the membrane can be measured at the fluorescence peaks found in the profile. (c) Mean intensities of fluorescence at the GUV membrane for different a-synuclein concentrations. Mean values measured from six different GUVs. Bars correspond to the standard error of the mean. The data were fitted according to the scaled particle theory (19)

Visualization of Lipid Domain-Specific Protein Sorting in Giant Unilamellar Vesicles

125

(e.g., egg yolk) can be used. To study the effect of anionic lipids (e.g., phosphatidylserine (PS) or phosphatidylglycerol (PG)), rather large fractions of these lipids can be included in the lipid mixture (e.g., 70 mol% of DOPS with 30 mol% DOPC still yield GUVs). As long as these lipids are unsaturated, no phase separation occurs. For GUVs with coexisting ld and lo domains, use a ternary lipid mixtures consisting of unsaturated PC in combination with saturated PC or sphingomyelin and cholesterol. Especially, mixtures of DOPC/SSM/ Chol = 1/1/1 and DOPC/DPPC/Chol = 1/1/1 are often used, and the phase behavior is well studied. An overview (by far not complete) of possible lipid mixtures consisting of zwitterionic lipids has been published, together with the respective phase transition temperatures, by Veatch and Keller (5). 5. The lipid spots on ITO slides should show violet color and concentric ring-like structures. If reproducibly small yellow lumps of lipids on the slides are formed, evaporate the chloroform of the lipid (lipid/peptide) mixture in a nitrogen stream and redissolve the lipids in 50 µl trifluoroethanol. Then, pipette 25 µl of the lipid mixture in one large drop on each of the ITO slides before heating, and subsequently place the spotted slides on a heating stage at approx. 50°C to evaporate the solvent and to form a homogenous lipid film. Normally, this problem can also be solved by the use of titanium chambers instead of ITO slides. 6. The composition of the microscopy buffer can be adjusted to the requirements of the experiment. Instead of glucose, any ionic or nonionic substance can be used as long as it does not disturb the membrane integrity. 7. Due to solvent evaporation, GUVs will be stable only for a limited time. For prolonged periods, instead of coverslips cell culture dishes with a glass bottom can be used. If GUVs tend to spread quickly by forming a flat bilayer at the glass surface, the use of a polylysine-coated support can be advantageous. 8. Microscopy can also be done with an epifluorescence microscope, although a high background from unbound protein or label may result. In each case, the experimental setup should be checked for spectral crosstalk with samples containing only one of the fluorescent dyes. For observation, at least a 40× air objective is adequate, although a 60× oil objective with a high numerical aperture will greatly enhance fluorescence intensity and resolution. 9. As labeling of proteins can significantly change their properties, it must be precluded that the membrane binding of the protein is changed by the introduced label. Therefore, important controls include the incubation of the vesicles with

126

Stöckl, Nikolaus, and Herrmann

free label to prove that it does not bind to the membrane by itself. Additionally, competition experiments with equal amounts of unlabeled and labeled protein should show a reduction of membrane fluorescence of 50% if the binding effectiveness is not changed by the label. 10. Lipids were abbreviated using the following scheme: DO, di-oleoyl-; DP, di-palmitoyl-; DS, di-stearoyl-; PO, 1-palmitoyl2-oleoyl-; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PG, phosphatidylglycerol; PS, phosphatidylserine; SSM, stearoyl-sphingomyelin; Chol, cholesterol. References 1. Allen JA, Halverson-Tamboli RA, Rasenick MM (2007) Lipid raft microdomains and neurotransmitter signalling. Nat Rev Neurosci 8:128–140 2. Hanzal-Bayer MF, Hancock JF (2007) Lipid rafts and membrane traffic. FEBS Lett 581:2098–2104 3. Lajoie P, Nabi IR (2007) Regulation of raftdependent endocytosis. J Cell Mol Med 11:644–653 4. Nayak DP, Hui EK, Barman S (2004) Assembly and budding of influenza virus. Virus Res 106:147–165 5. Veatch SL, Keller SL (2003) Separation of liquid phases in giant vesicles of ternary mixtures of phospholipids and cholesterol. Biophys J 85:3074–3083 6. Radhakrishnan A, McConnell H (2005) Condensed complexes in vesicles containing cholesterol and phospholipids. Proc Natl Acad Sci U S A 102:12662–12666 7. Lim LH, Pervaiz S (2007) Annexin 1: the new face of an old molecule. Faseb J 21:968–975 8. Stockl M, Fischer P, Wanker E, Herrmann A (2008) Alpha-synuclein selectively binds to anionic phospholipids embedded in liquid-disordered domains. J Mol Biol 375:1394–1404 9. Sengupta P, Baird B, Holowka D (2007) Lipid rafts, fluid/fluid phase separation, and their relevance to plasma membrane structure and function. Semin Cell Dev Biol 18:583–590 10. Korlach J, Schwille P, Webb WW, Feigenson GW (1999) Characterization of lipid bilayer phases by confocal microscopy and fluorescence correlation spectroscopy. Proc Natl Acad Sci U S A 96:8461–8466 11. Dietrich C, Bagatolli LA, Volovyk ZN, Thompson NL, Levi M, Jacobson K, Gratton E

12.

13.

14.

15. 16.

17.

18.

19.

(2001) Lipid rafts reconstituted in model membranes. Biophys J 80:1417–1428 Fortin DL, Troyer MD, Nakamura K, Kubo S, Anthony MD, Edwards RH (2004) Lipid rafts mediate the synaptic localization of alpha-synuclein. J Neurosci 24:6715–6723 Kubo S, Nemani VM, Chalkley RJ, Anthony MD, Hattori N, Mizuno Y, Edwards RH, Fortin DL (2005) A combinatorial code for the interaction of {alpha}-synuclein with membranes. J Biol Chem 280:31664–31672 Ollesch J, Poschner BC, Nikolaus J, Hofmann MW, Herrmann A, Gerwert K, Langosch D (2008) Secondary structure and distribution of fusogenic LV-peptides in lipid membranes. Eur Biophys J 37:435–445 Angelova MI, Dimitrov DS (1986) Liposome electroformation. Faraday Discuss Chem Soc 81:303–311 Angelova MI, Soleau S, Meleard P, Faucon JF, Bothorel P (1992) Preparation of giant vesicles by external AC electric fields. Kinetics and applications. Prog Colloid Polym Sci 89:127–131 Ayuyan AG, Cohen FS (2006) Lipid peroxides promote large rafts: effects of excitation of probes in fluorescence microscopy and electrochemical reactions during vesicle formation. Biophys J 91:2172–2183 Hesselink RW, Koehorst RB, Nazarov PV, Hemminga MA (2005) Membrane-bound peptides mimicking transmembrane Vph1p helix 7 of yeast V-ATPase: a spectroscopic and polarity mismatch study. Biochim Biophys Acta 1716:137–145 Chatelier RC, Minton AP (1996) Adsorption of globular proteins on locally planar surfaces: models for the effect of excluded surface area and aggregation of adsorbed protein on adsorption equilibria. Biophys J 71:2367–2374

Chapter 11 Biosynthesis of Proteins Inside Liposomes Pasquale Stano, Yutetsu Kuruma, Tereza Pereira de Souza, and Pier Luigi Luisi Abstract Protein expression is the most complex metabolic reaction that has been encapsulated in liposomes, mainly as an intermediate step toward the synthesis of minimal semisynthetic cells. Although there are different experimental approaches to achieving the synthesis of proteins inside liposomes and it is therefore not possible to give a standard recipe, all methods follow a general strategy, which is briefly discussed. On this basis, we provide general indications for designing and realizing protein-expressing liposomes. Our approach for the green fluorescent protein expression inside 200-nm extruded vesicles is described in detail. Key words: Cell-free protein expression, PURESYSTEM, Enhanced green fluorescent protein (EGFP), Minimal cell, Synthetic biology

1. Introduction 1.1. Background

The biosynthesis of functional proteins is, to date, the most complex metabolic network that has been synthetically reconstructed inside lipid vesicles (liposome). The interest and the relevance for such kind of work concern several fundamental aspects of biology and important applications in biotechnology. First, it represents the most advanced step toward the bottomup construction of a semisynthetic minimal cell (1), which is an ambitious goal aiming to understand the very nature of cellular life. This research started about 15 years ago by carrying out molecular biology reactions of increasing complexity inside liposomes (for a review, see ref. (1)). Its final target is the construction of molecular systems that exhibit living-like properties, guided by the theory of autopoiesis (2, 3) also considering that (minimal)

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_11, © Humana Press, a part of Springer Science+Business Media, LLC 2010

127

128

Stano et al.

life emerges as a system property as soon as a certain degree of molecular complexity and spatial/functional organizations are reached (4). The theoretical analysis of the minimal number of molecular components required to construct cellular life states that about 200 genes constitute the “minimal genome,” i.e., the genome required for cell survival in a highly permissive environment. According to this view, the environment provides the cell with several low-molecular-weight compounds, whereas the cell synthesizes by itself the characteristic and functional macromolecules coded by the minimal genome. Based on this analysis, it is possible to conceive an experimental research program for the

Fig. 1. Protein synthesis inside liposomes. The whole transcription–translation (T&T) kit, inclusive of an energy regeneration system, is entrapped inside liposomes. The T&T kit, composed of macromolecules (translation factors, ribosomes, tRNAs, aa-tRNA synthases, kinases), and low-molecular-weight compounds (nucleotides, amino acids, salts, phosphate donors) effect the synthesis of functional proteins. Unentrapped components are removed, or their reaction is inhibited by addition of ribonuclease, protease, or EDTA. In advanced models, membrane pores (a-hemolysin (10)) are inserted in order to exchange low-molecular-weight compounds, without releasing macromolecules from the liposome. Drawing adapted from (20), with modifications

Biosynthesis of Proteins Inside Liposomes

129

construction of minimal cells of increasing complexity: from simple liposome bioreactors hosting a single enzyme to very complex systems as the whole transcription–translation (T&T) machinery inside liposomes (Fig. 1). Along the roadmap to living (although limping) minimal cells, which are models of primitive cells (1), the expression of one (or more) functional protein(s) inside liposomes has been reported by us and several other groups (5–15). In addition to basic research, the encapsulation of entire metabolic networks inside liposomes has great biotechnological potentiality as therapeutic of diagnostic agents, as discussed by Pohorille and Deamer (16), or as tools for genome replication, modified RNA synthesis, drug discovery, or generation of new microorganism, within the context of the emerging discipline of synthetic biology (17). 1.2. Experimental Strategy

Today, the biosynthesis of proteins inside liposomes is achieved by combining cell-free in vitro protein expression and liposome technology. It should be noted that there is not yet a standard approach to carrying out protein biosynthesis inside liposomes. Different researchers have used different experimental setups, which differ with respect to (i) size and morphology of liposomes, (ii) liposome composition and lipid concentration, (iii) liposome preparation, (iv) nature of cell-free T&T kits and of the external inhibitors, (v) nature of expressed protein(s), and (vi) protein detection method. In addition, only in one case it has been reported that the protein-expressing liposome can exchange matter with the environment (10), realizing a thermodynamically open system; in all other cases, liposomes act essentially as a nonpermeable compartment, i.e., a closed system (5–9, 11–15). The general strategy can be summarized as shown in Fig. 2. The gene of interest downstream to a proper promoter (generally T7, but SP6 has been also used (9)) is mixed together with a T&T kit, composed of cell extracts (18) (homemade or commercially available), or completely reconstituted from purified enzymes, ribosomes, and tRNAs (19, 20). The mixture is kept at low temperature (in ice) in order to prevent the beginning of the reaction, and then liposomes are formed and processed according to specific requirements. Soon after, an inhibitor is added to the liposome suspension. The inhibitor should not permeate the liposome membrane. Alternatively, the external medium may be exchanged with an isotonic buffer. Following proper incubation at 37°C, the synthesis of the protein(s) inside liposomes is then verified analytically (when possible, in real time). Table 1 summarizes the diverse approaches, according to the six (i–vi) points listed above. Vesicles of very different size and morphology have been successfully employed to express proteins (mainly green fluorescent proteins, GFPs) in their inner aqueous

130

Stano et al.

Fig. 2. General strategy for protein expression inside liposomes. Liposomes are formed in a solution containing a transcription–translation (T&T) kit, composed of cell extracts or reconstituted by purified components (i.e., PURESYSTEM), and a gene of interest, under T7 promoter. The liposome suspension is formed and processed at low temperature to prevent the beginning of the protein synthesis. Then ribonuclease (or protease, or EDTA) is added to the sample in order to inhibit the synthesis of protein by external unentrapped molecules (route a). The inhibitors should not permeate the liposome membrane. Alternatively, the external solution is exchanged by size exclusion chromatography (route b). Compartmentalized protein synthesis will occur within 2–3 h incubation at 37°C

core. Uni- or oligolamellar vesicles with diameter between 200 and 400 nm can be obtained by the extrusion method (21) or by the ethanol injection method (22). Large multilamellar vesicles are generally obtained by hydration of a lipid film or of a liposome cake (prepared by freeze-drying of preformed liposomes) (23), whereas giant vesicles (GVs) can be generated by natural swelling (24) or by the new method of w/o emulsion inversion, developed by Pautot et al. (25) (notice that there are no report on protein expression in GVs formed by the electroswelling method). In order to carry out protein biosynthesis inside vesicles, homemade or commercially available cell extracts have been largely used. After the initial report of Ueda and coworkers (19, 20) and the successive introduction in the market of purified T&T components (i.e., the PURESYSTEM, sold by the Japanese company Post Genome Institute- Co., Ltd., Tokyo), it is foreseeable that this totally reconstituted kit will prevail over the E. coli cell extracts, which actually can be seen as a “black box” of not well-defined molecular composition. In fact, the concentrations of all molecular species in the

VET400

MLV, GV

LUV

GV

MLV, GV

GV

MLV, GV

MLV

1

2

3

4

5

6

7

8

50–200 mM POPC

48 mM POPC, PLPC, SOPC, SLPC, cholesterol, DSPEPEG5000

EggPC

15 mM EggPC, CHOL, DSPE-PEG5000 (1.5:1:0.08)

11 mM DOPC, DOPG (10:1)

1.5 mM POPC

20 mM EggPC, CHOL, DSPE-PEG5000 (1.5:1:0.08)

66.7 mM POPC

(i) Size and (ii) Composition and Entry morphology concentration

PURE SYSTEM; 5.7 mg/ mL RNase

PURE SYSTEM; 7.3 mg/ mL RNase

Liposome cake rehydration

Liposome cake rehydration

Cell extract (Roche); no external inhibitor

Homemade; 12 mg/mL RNase

Cell extracts (Ambion); 50 mg/mL proteinase K

Cell extracts (Promega); 33 mM EDTA

Homemade; 12 mg/mL RNase

Homemade; 35 mM EDTA

(iv) T&T and inhibitors

w/o emulsion droplets inversion

Liposome cake rehydration

Natural swelling

Ethanol injection

Film hydration

Film hydration, FT, Extrusion

(iii) Preparation

Table 1 Summary of protein biosynthesis inside liposomes

EGFP

EGFP

EGFP and alphahemolysin

GFPmut1 and T7RNApol

rsGFP

EGFP

GFPmut1

Poly(Phe)

(v) Expressed protein(s)

Fluorescence: batch fluorimetry and confocal laser microscopy

Fluorescence: flow cytometry

(12)

(11)

(10)

(9)

(8)

(7)

(6)

(5)

Ref.

(continued)

Fluorescence: laser scanning microscopy

Radiolabeling and fluorescence: SDS-PAGE and flow cytometry

Fluorescence: laser scanning microscopy

Fluorescence: batch fluorimetry

Fluorescence: flow cytometry and laser scanning microscopy

Radiolabeling: SDSPAGE

(vi) Detection method

Biosynthesis of Proteins Inside Liposomes 131

PURE SYSTEM; 20 mg/mL RNase

Radiolabeling: SDSPAGE

Fluorescence: batch fluorometry

EGFP

GPAT and LPAAT

Radiolabeling: SDSPAGE

(vi) Detection method

GPAT and LPAAT

(v) Expressed protein(s)

(15)

(14)

(13)

Ref.

MLV multilamellar vesicles; GV giant vesicles; VET400 vesicles by extrusion techniques, with diameter 400 nm; LUV large unilamellar vesicles (100–200 nm); LUV-EI large unilamellar vesicles obtained by the ethanol injection method; POPC 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphatidylcholine; egg-PC egg yolk phosphatidylcholine; CHOL cholesterol; DSPE-PEG5000 1,2-distearoyl-sn-glycero-3-phosphatidylethanolamine-poly(ethylenglycol) 5000; DOPC 1,2-dioleoyl-sn-glycero-3-phosphatidylcholine; DOPG 1,2-dioleoyl-sn-glycero-3-phosphatidylglycerol; PLPC 1-palmitoyl-2-linoleoyl-sn-phosphatidylcholine; SOPC 1-stearoyl-2-oleoyl-sn-phosphatidylcholine; SLPC 1-stearoyl-2linoleoyl-sn-phosphatidylcholine; POPE 1-stearoyl-2-oleoyl-sn-phosphatidylethanolamine; POPG 1-stearoyl-2-oleoyl-sn-phosphatidylglycerol; FT freeze-and-thaw; w/o water-in-oil; T&T transcription and translation kit; EDTA ethylendiaminotetraacetic acid; RNase ribonuclease; SEC size exclusion chromatography; poly(Phe) poly(phenylalanine); GFPmut1 green fluorescent protein mutant 1; EGFP enhanced green fluorescent protein; rsGFP red-shifted green fluorescent protein; T7RNApol T7 RNA polymerase; GPAT glycerol-3phosphate acyltransferase; LPAAT lysophosphatidic acid acyltransferase; SDS-PAGE sodium dodecylsulfate-polyacrylamide gel electrophoresis

200 mM POPC, POPE, POPG, cardiolipin (50.8:35.6:11.5:2.1)

Film hydration

MLV

11

Cell extracts (Promega) and PURE SYSTEM; 180 mg/mL RNase or SEC

VET200 and 8.5–15 mM LUV-EI POPC

10

PURE SYSTEM; 20 mg/mL RNase

(iv) T&T and inhibitors

Ethanol injection, extrusion

Film hydration

MLV

9

200 mM POPC

(iii) Preparation

(i) Size and (ii) Composition and Entry morphology concentration

Table 1 (continued)

132 Stano et al.

Biosynthesis of Proteins Inside Liposomes

133

PURESYSTEM are known (see Note 1) and individually tunable, at least in principle. Protein synthesis that may occur outside vesicles can be inhibited by the addition of ethylenediaminetetraacetic acid (EDTA) (Mg(II) chelator), or ribonuclease (RNase A, digesting the mRNA), or proteinase K (digesting the enzymes needed to accomplish the T&T processes). Finally, almost all studies refer to the synthesis of GFPs, which is a model protein with easy detection. Notice that GFP is fluorescent due to an intramolecular cyclocondensation between three residues (Ser 65, Tyr 66, Gly 67) and consequent oxidation to give a p-hydroxybenzylideneimidazolinone (26). It is recognized that the appearance of its fluorescence implies a correct fold, so that it can be said that the produced protein is “functional” (26). In addition to GFPs, some studies have focused on the expression of T7 RNA polymerase (Table 1, entry 5) in order to create a two-stage cascading expression network, or of a-hemolysin (Table 1, entry 6), in order to make the liposome membrane porous. All these proteins (GFP, RNA polymerase, a-hemolysin) are water soluble. Only recently, Kuruma et al. (13, 15) reported on the synthesis of membrane proteins inside liposomes (Table 1, entries 9 and 11). 1.3. Choice of the Conditions

From the foregoing discussion, it is evident that there does not exist a unique and specific protocol to carry out protein expression inside liposomes. The choice of the different variables (i–vi) very much depends on the aim of the experiment; moreover, some analytical techniques are possible only for specific cases (e.g., fluorescence-based techniques can be used only when GFPs are expressed). Here we shortly report some general guidelines. (i) Vesicle size and morphology. Unilamellar vesicles should be preferred whenever is possible. Working with unilamellar GVs (e.g., diameter ³10 mm), for example, is advantageous, since it is possible to follow directly the evolution of the system, for example, at morphological level or by employing some fluorescent reporters. On the other hand, GVs must be observed one by one, or in small groups, and the advantage of large number averaging is lost. On the contrary, smaller vesicles (100–200 nm), especially if homogeneously sized, provide a good experimental model, and their preparations are reproducible but suffer the disadvantage of a small internal volume and therefore need highly sensitive methods for assessing the yield of the synthesized protein. Intermediate sized vesicles, e.g., ca. 500–1,000 nm diameter vesicles, are generally oligoor multilamellar and difficult to produce. To the best of our knowledge, there are no standard methods to produce unilamellar vesicles with those sizes. Multilamellar vesicles, although convenient in some applications, are to be considered unsuitable to carry out protein expression in a controlled way for a liposome system.

134

Stano et al.

(ii) Liposome composition and lipid concentration. It is interesting to notice that protein expression has been carried out mainly in phoshatidylcholine vesicles. This lipid is chemically quite inert (for example, negatively charged lipids which may interact with Mg2+ ions, or positively charged lipids which strongly interact with nucleic acids and are reported to strongly reduce the protein synthesis (5, 27)), readily forms stable vesicles, and ensures good entrapping yields. DOPC and POPC have low phase-transition temperatures −18.3 and −2.5°C in water, respectively, (28), so that they can be used at a low (+4°C) temperature. Liposome composition is an important issue in the case of expressing membrane proteins (15, 29) and when liposomes with special performances are required. (iii) Liposome preparation. This is of course related to point (i). Since enzymes and ribosomes have to be entrapped inside liposomes, preparation protocols, which require the application of harsh conditions or high detergent concentration may lead to poor results. The hydration of lipid film or rehydration of liposome cake (obtained by freeze-drying previously formed liposomes) is the safest procedures. Ethanol injection method is also applicable, but yields are partially reduced because of the presence of alcohol in the final suspension (methanol can also be used: the protein yield is slightly higher, but the liposome size distribution is broader). Extrusion also leads to a reduction of yield (14), whereas mild (bath) sonication and freeze-andthaw are apparently well tolerated by T&T kits. A systematic study on the effect of liposome manipulation on the protein synthesis inside liposomes is still missing, so that the conclusions sketched above must be considered with caution. (iv) Nature of cell-free T&T kits and of the external inhibitors. For the expression of prokaryotes protein, the choice is essentially between cell extracts (homemade or bought from several companies) and the PURESYSTEM. Cell-free kits based on cell extracts generally provide higher yields than PURESYSTEM; their composition, however, is unknown and cannot be considered really functional for the constructive paradigm of synthetic biology. PURESYSTEM offers the additional advantage of easy purification of expressed protein, due to the fact that its components are all His-tagged, and the target protein can be easily isolated by selective removal of His-tagged components (19). Cell extracts are also generally viscous and may affect the liposome preparation steps. RNase A has been widely employed to avoid the occurrence of the reaction outside the liposomes (6, 9, 12), but its effectiveness must be checked in control experiments to take into account the effect of liposomes. Proteinase K has also been used (8). Protein expression can be inhibited by EDTA

Biosynthesis of Proteins Inside Liposomes

135

(>30 mM). EDTA does not suffer limitations as in the case of macromolecular inhibitors (RNase A and proteinase K); however, it may provide osmotic stress to liposomes, and its use must be checked with respect to liposome size, morphology, and manipulation/processing. The most elegant way to eliminate external synthesis is to exchange the external solution with an isotonic buffer by size exclusion chromatography (14), see Fig. 2b. (v) Nature of expressed protein(s). This parameter depends on the aim of the experiment. Water-soluble or membrane-associated proteins require different strategies, mainly related to the composition of liposome membrane. In fact, when a membrane protein must be expressed inside vesicles, great care must be taken for the choice of lipids, which will affect the membrane/protein interaction and also the protein activity. Liposome membrane, in a sense, should be considered in these cases as a reactant. When the research aim is not the expression of a specific protein, and the experiment is focused on other aspects of compartmentalized reactions, an easily detectable water-soluble protein must be chosen. (vi) Protein detection method. The common strategy here follows the classical methods developed for cell-free protein expression. The most general detection method involves the incorporation of labeled amino acids in the protein sequence. Very common is the use of [35S]-methionine; when required, however, other 35S-, 14C-, or 3H-labeled amino acids can be employed (see Note 2). The freshly synthesized protein, in all these methods, is concentrated by precipitation (trichloroacetic acid (TCA) or acetone precipitation, see Note 3) and then analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). In the case of radiolabeling, the dried gel is then exposed to autoradiographic film or detected with a phosphor imager. Alternatively to the incorporation of labeling amino acids, the protein can be detected by Western blotting (primary and secondary antibodies are required). Here the synthesized protein is analyzed by SDSPAGE and then transferred to the membrane for primary and secondary antibody binding and immunodetection. These methods are very useful, since they can be opportunely applied to every protein, regardless from their sequence, structure, and function. In particular cases, however, it is possible to directly detect the expressed protein by probing its “activity.” This is possible when the target protein is an enzyme or an intrinsically fluorescent protein, such as in the case of the family of GFPs. If the protein activity is successfully measured, this will provide strong evidence for the proper synthesis and folding of the synthesized protein. The enzymatic assay must

136

Stano et al.

be properly designed for each case, also considering the compartmentalized nature of the system, and therefore the problem of membrane permeability. This will affect the way of substrate delivery. When GFPs are expressed, on the other hand, the product analysis becomes easier. For this reason, the first studies on the biosynthesis of proteins inside liposomes have been carried out with GFPs as target proteins. In fact, GFPs can be revealed by following their intrinsic fluorescence. GFP fluorescence can be measured by single-wavelength measurements (5, 14), emission spectra measurements (5, 14), laser scanning fluorescence microscopy (6, 8, 10, 12, 14), and flow cytometry (6, 9, 11). Many of these methods can be used to monitor fluorescence signals in real time, providing a facile way to follow the kinetics of protein synthesis. A serious problem in fluorescent measurements of GFP inside liposomes stems from sample turbidity, which gives rise to a scattering-based inner filter effect (IFE) (see Note 4). It is clear that the expression of proteins inside liposomes is still a challenging task and that it is difficult to provide a standard recipe for all cases. Here we describe in detail a protocol that has been successfully developed in our laboratory concerning the expression of enhanced GFP (EGFP) in extruded vesicles. The production of EGFP can be followed flurometrically; however, we also provide the classical protocol to verify its production by SDS-PAGE.

2. Materials 2.1. Cell-Free Protein Expression Kit, Buffer and Reagents

1. Plasmid pWM-T7-EGFP (3,026 bp, see Note 5) (BioTecton, Zurich, Switzerland), encoding egfp gene under T7 promoter dissolved in pure nuclease-free water (1 mg/mL) and stored in proper aliquots at −20°C. 2. PURESYSTEM Classic II kit (Post Genome Institute Co., Ltd., Tokyo, Japan) (see Note 6), must be properly divided in aliquots and stored at −80°C. Avoid freeze–thaw cycles. Do not mix solution A and solution B before use. 3. l-[35S]-Methionine, 1,000 Ci/mmol (GE Healthcare, code SJ1015) (this chemical is needed only when the synthesized protein must be analyzed by SDS-PAGE and autoradiography). 4. Nuclease-free water (produced by Milli-Q apparatus, Millipore). 5. Recombinant EGFP (BD Bioscience, Switzerland), stored at −20°C in the dark.

Biosynthesis of Proteins Inside Liposomes

137

6. “Multicore”-like sterile buffer (MC): 125 mM potassium acetate, 25 mM Tris acetate, pH 7.5. Store at 4°C. 7. RNase A, 2 mg/mL (Sigma-Aldrich), in MC buffer; freshly prepared and kept in ice. 8. Sterile Eppendorf-like 1.5-mL vials and 200 mL PCR-like vials (see Note 7). 2.2. Liposome Preparation, Manipulation, and Processing

1. Synthetic POPC, i.e., 1-palmitoyl-2-oleoyl-sn-glycero-3phosphatidylcholine, (Avanti Polar Lipids, Inc., Alabaster, USA) is dissolved in absolute ethanol (500 mM) and stored in aliquots under nitrogen atmosphere at −20°C. 2. Two sterile Teflon-covered magnetic stirring bars; length 5 mm, diameter 2 mm (Sigma-Aldrich, Code Z328839). Wash extensively with ethanol and dry before use. 3. Microsyringe, 10 mL, with a narrow needle (Hamilton). 4. Extruder “Liposo-fast” (Avestin Inc., Ottawa, Canada), equipped by two 250-mL gastight microsyringes (Hamilton). Its components contain glass, Teflon, viton (O-rings), and stainless steel. They are easily cleaned and the entire instrument is autoclavable. 5. Three drain discs, 13 mm (Osmonics, Inc., Article Nr. C32WP01300). 6. Two polycarbonate membranes, 200 nm pore size (Nuclepore Track-Etched Membranes, Whatman, Maidstone, UK).

2.3. S DS-PAGE

1. Stacking buffer (4×): 0.5 M Tris–HCl, pH 6.8, 0.4% (w/v) SDS. Store at room temperature. 2. Separating buffer (3.75×): 1.5 M Tris–HCl, pH 8.8, 0.4% SDS. Store at room temperature. 3. SDS running buffer (10×): 250 mM Tris, 1.92 M glycine, 1% SDS. Store at room temperature. 4. Staining buffer: 0.5% (w/v) Coomassie Brilliant blue R-250 (CBB), 50% (v/v) methanol, 10% (v/v) acetic acid. Store at room temperature. 5. Destaining (fixing) buffer: 50% methanol, 10% acetic Acid. Store at room temperature. 6. SDS loading buffer: 125 mM Tris–HCl. pH 6.8, 20% (v/v) glycerol, 4.6% SDS, 0.006% (w/v) Bromophenol Blue (BPB), 10% (v/v) b-mercaptoethanol (add before use). Store at room temperature. 7. Bisacrylamide/30% acrylamide: 30% (w/v) acrylamide, 0.8% (w/v) N,N¢-methylenebisacrylamide. Store at 4°C. 8. Ammonium peroxodisulfate (APS) (10% (w/v)). Store at 4°C.

138

Stano et al.

9. N,N,N¢,N¢-Tetramethylethylenediamine (TEMED). Store at 4°C. 10. Acetone, diethyl ether (analytical grade). Important: All chemicals should be dissolved in distilled water. Use gloves when acrylamide is handled, because it is a neurotoxin when unpolymerized and so care should be taken not to receive exposure.

3. Methods 3.1. Preparation of the Extruder

1. Assemble the extruder by placing three drain discs (see Note 8) and two polycarbonate membranes (200 nm pore size) between the two Teflon blocks, so that the two membranes are sandwiched between the three drain discs (i.e., drain disc – membrane – drain disc – membrane – drain disc). Close the extruder by tightening the two external screws. 2. Wash the extruder by passing through the membranes 200 mL MC buffer. 3. Keep the assembled extruder in the cold room at least 30 min before the experiment (see Note 9).

3.2. Preparation of Reaction Mixture and Liposome Formation

All operations described below must be done in the cold room (4°C). All equipmentfor liposome formation/manipulation/processing must be kept in the cold room at least 30 min before the beginning of experiments. 1. Thaw all components required for the transcription–translation reaction in ice (solution A and B of PURESYSTEM, pWMT7-EGFP plasmid). Nuclease-free water, “multi-core”-like buffer, POPC solution in ethanol, and RNase A solution must also be kept in ice. 2. Gently mix 100 mL solution A, 40 mL solution B, and 50 mL nuclease-free water in a sterile Eppendorf-like vial, and keep the resulting mixture in ice. This mixture will provide the positive control sample (protein synthesis inside and outside liposomes) and the test sample (protein synthesis only inside liposomes). Negative control samples (no protein synthesis) are similarly prepared, but 20 mL of RNase A solution is also added at this stage. 3. Add two Teflon-covered followers to the mixture prepared (point 2). 4. Add 4 mL plasmid pWM-T7-EGFP (1 mg/mL) to the mixture (point 3).

Biosynthesis of Proteins Inside Liposomes

139

5. Put the Eppendorf-like vial containing 194 mL of T&T/ gene mixture in a small water/ice bath over a magnetic stirrer and stir at about 500 rpm; quickly inject 1.5 mL POPC solution; charge the microsyringe again and inject three further 1.5 mL fractions, for a total of four injections (total 6 mL). To avoid excess shearing force, decrease the stirring rate between injections. The suspension should appear turbid, by the presence of 15 mM POPC (final concentration). The ethanol content is 3% v/v. 6. Put the liposome suspension in ice. 7. Repeat the procedure (points 3–6) for the negative control sample. When the EGFP synthesis has to detected on the SDS-PAGE gel, 4 mL of l-[35S]methionine (1,000 Ci/mmol) must be supplied to the PURESYSTEM reaction mixture (total reaction volume 200 mL), and the amount of pure water decreased to 46 mL. 3.3. Liposome Processing

1. Mount a needle on the 250-mL gastight microsyringe of the extruder and fill it with the liposome suspension. Remove the needle. 2. Insert the syringe into the extruder inlet and extrude the liposomes by passing them 11 times back and forth through the two polycarbonate membranes (see Note 10). Avoid prolonged contact between hands and syringes, in order to avoid heat transfer. 3. Collect the extruded vesicles and make 50 mL aliquots inside PCR-type 200-mL vials. Put the aliquots in ice. (see Note 11). 4. Add 5 mL of cold MC buffer to the positive control samples (protein synthesis inside and outside liposomes). Keep the samples in ice. 5. Add 5 mL of cold RNase A solution to the other fractions (final RNase A concentration 0.18 mg/mL) and mix gently; these will be the test samples (protein synthesis only inside vesicles). Keep the samples in ice. 6. Repeat the procedure (points 1–3) for negative control samples, but make 55 mL aliquots. Keep the samples in ice.

3.4. Real-Time Fluorescence Analysis of EGFP-Producing Liposomes

1. Real-time (RT) PCR instruments can be used as sensitive fluorimeters if the temperature is set constant (37°C). We have successfully employed a RT-PCR Corbett Rotor-Gene 6000. Such an instrument improves reproducibility, enhances sensibility, allows the use of small sample volumes (25 or 50 µL), and provides homogeneous heating. Instrumental settings: fluorescence channel: green (excitation 470 nm, emission 510 nm); gain 5, 7, or 10; cycle setting: 120 s at 37.0 (±0.5)°C; 400 rpm.

140

Stano et al.

Fig. 3. Analysis of produced EGFP by (a) real-time fluorimetric assay and (b) SDS-PAGE. In (a) the fluorescence versus time curves show that positive control sample (EGFP expression inside and outside liposomes) reaches high fluorescence values, whereas negative control sample (RNase A added before the liposome formation) has low fluorescent values that do not increase in time. The test sample, on the other hand, can express EGFP inside liposomes, since RNase A has been added after liposome formation. In this example, the relative yield of inside-liposome production is about 3% ( 28.5, 1.9, and 0.8 being the values of fluorescence for the three samples). In (b), the classical pattern of SDS-PAGE analysis is shown (M MW markers; P positive control; T test sample; N negative control). The marker has been added for the sake of clarity (after radiolabeling, only expressed proteins can be visualized, and correspondence with markers for MW check should be done in a separate experiment)

2. Insert the 55-mL samples (positive controls, test samples, negative controls) in the RT-PCR machine and record fluorescence for 3–5 h. This will provide the fluorescence versus time curves for each sample. Remove the sample from ice immediately before starting fluorescence recording. 3. Read the asymptotic fluorescence value for the samples, and average values for samples run in duplicate. Calculate the EGFP yield inside liposomes as follows: EGFP relative yield (%) = 100 × (Ftest − Fneg)/(Fpos − Fneg). Typical relative yields range from 1 to 10% (see Fig. 3a). 3.5.Calibration Line for Fluorescence Analysis

1. A calibration line is needed to quantify the production of EGFP in terms of absolute yield (EGFP amount) (for an alternative method, see Note 12). 2. Create five samples (55 mL) by mixing solution A, solution B, water, and non-EGFP coding plasmid of same size as pWM-T7EGFP, preformed 200 nm extruded POPC liposomes, RNase if required, and different amounts of recombinant EGFP (for example: 0, 0.1, 0.3, 0.6, 0.8 mg/mL), so that the composition of the samples is qualitatively and quantitatively similar to real

Biosynthesis of Proteins Inside Liposomes

141

experimental samples, and measure their fluorescence on the RT-PCR machine for 10 min. Five straight lines (fluorescence vs. time) should be obtained. Use the fluorescence values to construct a calibration line (fluorescence vs. [EGFP]). 3. Use the calibration line to measure the amount of EGFP produced in the liposome-based experiment. 3.6. SDS-PAGE Analysis

1. In this study, Bio-Rad Mini-PROTEAN Tetra Cell kit is used for the SDS-PAGE analysis, since this kit has been commonly used in molecular biology laboratories. Before composing the kit, two mini glasses are first cleaned up by distilled water and then rinsed by 95% ethanol. After completely dry, the glasses are assembled by using the accessory platform. 2. Mix the separating gel (lower gel) solution as follows. For 10 mL of 12% gel: 4 mL of 30% acrylamide/bisacrylamide, 2.5 mL of 4× separating gel buffer, and 100 mL of 10% APS are mixed, and then distilled water is added up to 10 mL. Ten microliters of TEMED is also supplied and mixed just before pouring into the glasses. The percentage of gel can be varied by adjusting the volumes of acrylamide solution and water. 3. Pour the gel solution into the slit of the composed glasses leaving space for the stacking gel. Isopropanol or distilled water is overlaid on the poured solution to avoid drying the solution surface and to allow formation of a horizontal surface. The gel should polymerize in about 30 min. 4. Pour off the isopropanol or water completely. 5. Mix the stacking gel solution (upper gel) as follows. For 3 mL of 5% gel: 0.5 mL of 30% acrylamide/bisacrylamide, 0.8 mL of 4× stacking gel buffer, and 30 mL of 10% APS are mixed, and then distilled water is added up to 3 mL. Three microliters of TEMED is also supplied and mixed just before pouring into the glasses. 6. Pour the gel solution over the separating gel, and insert the comb. The stacking gel should polymerize within 30 min. 7. Prepare the running buffer by diluting 50 mL of 10× running buffer with 450 mL of water. The diluted buffer should be stirred by a magnetic stirrer bar in a beaker. 8. Once the stacking gel has polymerized, insert the gel assembly unit to the electrophoresis apparatus and supply the running buffer. Carefully remove the comb within the running buffer and wash the wells by pipetting. Check that there is no air at the bottom of the gel. When bubbles are present, remove them by a syringe fitted with thin distorted needle. 9. Before loading the samples (positive control, test sample, and negative control), a pretreatment must be carried out in

142

Stano et al.

order to avoid the jumbling of the gel pattern due to lipid components. For this, proteins must be precipitated by acetone (see Note 3). Acetone (200 mL) is added to liposomecontaining samples (50 mL) and incubated on ice for 30 min. Then, the samples are centrifuged at 18000 × g for 10 min. The resulting supernatants (acetone) are removed by pipetting, and additional acetone (200 mL) is added again to the precipitates. The suspensions are vortexed and the suspended precipitates are finely dispersed by using bath sonication, and then centrifuged (18000 × g for 10 min). After removal of the supernatants (acetone), 100 mL of diethylether is added to the precipitates. The precipitates are dispersed again by bath sonication and centrifuged (18000 × g for 10 min). After removing the supernatants (diethylether), the resulting precipitates are completely dried by a centrifuge-evaporator for 5 min. Finally, the dried precipitates are dissolved in SDS loading buffer with the appropriate volume (max 20 mL). Through these processes, lipid components can be removed, and the produced radiolabeled [35S]-protein can be safely analyzed by SDS-PAGE. The dissolved samples should be boiled at 95°C for 5 min just before loading into the gel. 10. Carefully load the samples into the wells, which include one well for the prestained molecular weight marker. 11. Complete the assembly of the apparatus and connect to a power supply. Apply a voltage (less than 100 V) until the dye of sample enters the separating gel. After the sample enters the separating gel, the voltage can be increased up to 200 V. Stop the gel running when the dye front arrives at the bottom of gel, and carefully remove the gel from the glasses. 12. If the sample is nonradiolabeled, the gel is processed to stain by staining buffer, and then destained by destaining buffer with a piece of paper towel. If the sample is radiolabeled, incubate the gel with fixing buffer. All these steps are done for 30 min with gentle shaking. 13. Take a photograph of the stained gel (nonradiolabelled). 14. For the radiolabeled gel, the gel is dried up on a filter paper (Whatman) by using a Bio-Rad Gel Dryer, for 30 min at 80°C. 15. The dried gel can be analyzed by an imaging system, such as GS-525 Molecular Imager System (BioRad). Expose the gel to the storage phosphor imaging screen for at least 12 h (this time must be optimized according to the yield of the expressed protein), and then analyze the screen by the IR-laser-scanning acquisition module. A typical pattern is shown in Fig. 3a.

Biosynthesis of Proteins Inside Liposomes

143

4. Notes 1. The PURESYSTEM is composed by 10 translation factors (IF1, IF2, IF3, EF-G, EF-Tu, EF-Ts, RF1, RF2, RF3, RRF), 20 aminoacyl-tRNA synthases, T7 RNA polymerase, ribosomes, methionyl-tRNA formyltransferase, creatine kinase, myokinase, nucleoside diphosphate kinase, pyrophosphatase, and a mix of tRNAs (37 macromolecular compounds + tRNAs). In addition, low-molecular-weight compounds are present (ATP, GTP, CTP, UTP, DTT, spermidine, creatine phosphate, 20 amino acids, 10-formyl-5,6,7,8-tetrahydrofolic acid, potassium glutamate, and magnesium acetate) (19, 20). 2. Alternative to radiolabeling, biotinylated amino acids (30) or fluorescently labeled amino acids (31) can be used. In the first case, after SDS-PAGE and blotting, streptavidin-linked enzymes (phosphatase, peroxidase) allow colorimetric or chemiluminescent detection. In the second case, direct visualization of the fluorescent bands in the gel/membrane is possible. 3. Acetone precipitation (followed by diethyl ether rinse) should be preferred to TCA precipitation (12, 13, 15). In fact, acetone effectively solubilizes lipids, which may interfere in SDS-PAGE analysis. TCA precipitation is the classic choice in liposome-free cell-free synthesis. 4. Turbidity-based IFE is not very well described in literature, in contrast to absorbance-based IFE. IFE leads to a reduction of fluorescence signal due to absorption or scattering of the excitation and emitted radiation. In some extreme cases, the IFE may remove completely the fluorescence signal, especially the signal is not very strong. It follows that the quantification of a fluorescent molecule may be difficult in the presence of IFE. To solve this problem, two approaches can be followed. In the first approach, liposomes are solubilized as micelles by addition of a detergent (e.g., sodium cholate). This treatment reduces IFE by reducing the size of particles present in the solution, and therefore by reducing the scattered light. Of course, the effect of such a detergent on the protein fluorescence must be checked in preliminary experiments. In the second approach, a proper calibration line (in the presence of liposomes) must be constructed. The liposomes used to record the calibration line should be similar in average size and dispersity to the liposomes obtained in the real experiment. 5. Detail of the T7 (underlined), linker (italics), and EGFP sequences of pWM-T7-EGFP plasmid: TAATAC GACTCA CTATAG – GGAGAC CACAAC GGTTTC CCTCTA

144

Stano et al.

GAAATA ATTTTG TTTAAC TTTAAG AAGGAG ATATAC AT – EGFP sequence. Plasmid purity should be carefully checked, as well as the quality of water. Nuclease contamination, e.g., deriving from the plasmid preparation, should be avoided. 6. http://www.postgenome.jp/english/index.html 7. PCR-like vials should be colorless and possibly made by thin plastic, in order to have higher transparency. 8. Cut three drain discs so that their diameter fits well with the head area (limited by the O-rings) on the extruder Teflon blocks. 9. In some cases, low temperature causes shrinkage of the pistons’ tips, giving rise to leakage during extrusion. A preliminary check of the extruder in the cold room is suggested. 10. A possible problem at this stage is leakage from the extruder, due to imperfect assembly of the extruder or to the wear of mechanical parts. Check the extruder (in the cold room) in a preliminary experiment by extruding liposomes at similar concentration. Very small volumes (e.g., 50 mL) cannot be effectively extruded. 11. Label the vials on the top. 12. Use the standard addition method to quantify EGFP in a complex matrix. In this method, small aliquots of pure EGFP (at a known concentration) are added to the sample where EGFP of unknown concentration has to be determined. The addition(s) should provide an increase of EGFP concentration of the same order of magnitude as the unknown EGFP concentration (e.g., if synthesized EGFP is present at an estimated concentration of 0.05 mg/mL, add EGFP aliquots so that the total estimated EGFP concentration becomes 0.1 mg/ mL, 0.2 mg/mL, etc.). Record the fluorescence values and make a plot of recorded fluorescence versus the added EGFP concentration. The x-intercept (a negative number) is the concentration of EGFP in the sample. This method takes into account the interference of the matrix (liposomes, T&T kit, etc.) with the fluorescence signal of EGFP, and do not need a calibration line. Further information can be found in classical analytical chemistry books, for example (32).

Acknowledgements This work was carried out within the SYNTHCELLS project (Approaches to the Bioengineering of Synthetic Minimal Cells, EU Grant #FP6–043359) and further supported by the Human Frontiers Science Program.

Biosynthesis of Proteins Inside Liposomes

145

References 1. Luisi PL, Ferri F, Stano P (2006) Approaches to semi-synthetic minimal cells: a review. Naturwissenschaften 93:1–13 2. Varela FJ, Maturana HR (1974) Autopoiesis: the organization of living systems, its characterization and a model. Biosystems 5:187–196 3. Luisi PL (2003) Autopoiesis: a review and a reappraisal. Naturwissenschaften 90:49–59 4. Luisi PL (2006) The emergence of life: from chemical origin to synthetic biology. Cambridge University Press, Cambridge 5. Oberholzer T, Nierhaus KH, Luisi PL (1999) Protein expression in liposomes. Biochem Biophys Res Comm 261:238–241 6. Yu W, Sato K, Wakabayashi M, Nakatshi T, Ko-Mitamura EP, Shima Y, Urabe I, Yomo T (2001) Synthesis of functional protein in liposome. J Biosci Bioeng 92:590–593 7. Oberholzer T, Luisi PL (2002) The use of liposomes for constructing cell models. J Biol Phys 28:733–744 8. Nomura SM, Tsumoto K, Hamada T, Akiyoshi K, Nakatani Y, Yoshikawa K (2003) Gene expression within cell-sized lipid vesicles. ChemBioChem 4:1172–1175 9. Ishikawa K, Sato K, Shima Y, Urabe I, Yomo T (2004) Expression of a cascading genetic network within liposomes. FEBS Lett 576:387–390 10. Noireaux V, Libchaber A (2004) A vesicle bioreactor as a step toward an artificial cell assembly. Proc Natl Acad Sci USA 101:17669–17674 11. Sunami T, Sato K, Matsuura T, Tsukada K, Urabe I, Yomo T (2006) Femtoliter compartment in liposomes for in vitro selection of proteins. Anal Biochem 357:128–136 12. Murtas G, Kuruma Y, Bianchini P, Diaspro A, Luisi PL (2007) Protein synthesis in liposomes with a minimal set of enzymes. Biochem Biophys Res Comm 363:12–17 13. Kuruma Y (2007) Question 7: Biosynthesis of phosphatidic acid in liposome compartments – Toward the self-reproduction of minimal cells. Orig Life Evol Biosph 37:409–413 14. Souza T, Stano P, Luisi PL (2009) The minimal size of liposome-based model cells brings about a remarkably enhanced entrapment and protein synthesis. ChemBioChem 10: 1056–1063 15. Kuruma Y, Stano P, Ueda T, Luisi PL (2009) A synthetic biology approach to the construction of membrane proteins in semi-synthetic minimal cells. Biochim. Biophys. Acta 1788: 567–574 16. Pohorille A, Deamer D (2002) Artificial cells: prospects for biotechnology. Trends Biotechnol 20:123–128

17. Forster AC, Church GM (2006) Towards synthesis of a minimal cell. Mol Syst Biol 2:45 18. Zubay G (1973) In vitro protein systems synthesis of in microbial systems. Annu Rev Genet 7:267–287 19. Shimizu Y, Inoue A, Tomari Y, Suzuki T, Yokogawa T, Nishikawa K, Ueda T (2001) Cell-free translation reconstituted with purified components. Nature Biotechnol 19:751–755 20. Shimizu Y, Kanamori T, Ueda T (2005) Protein synthesis by pure translation systems. Methods 36:299–304 21. Hope MJ, Bally MB, Webb G, Cullis PR (1985) Production of large unilamellar vesicles by a rapid extrusion procedure. Characterization of size distribution, trapped volume and ability to maintain a membrane potential. Biochim Biophys Acta 812:55–65 22. Batzri S, Korn ED (1973) Single bilayer liposomes prepared without sonication. Biochim Biophys Acta 298:1015–1019 23. Kikuchi H, Suzuki N, Ebihara K, Morita H, Ishii Y, Kikuchi A, Sugaya S, Serikawa T, Tanaka K (1999) Gene delivery using liposome technology. J Control Release 62:269–297 24. Nomura S-IM, Yoshikawa Y, Yoshikawa K, Dannenmuller O, Chasserot-Golaz S, Ourisson G, Nakatani Y (2001) Towards proto-cells: “primitive” lipid vesicles encapsulating giant DNA and its histone complex. ChemBioChem 6:457–459 25. Pautot S, Frisken BJ, Weitz DA (2003) Production of unilamellar vesicles using an inverted emulsion. Langmuir 19:2870–2879 26. Tsien RY (1998) The green fluorescent protein. Annu Rev Biochem 67:509–544 27. Tachibana R, Harashima H, Ishida T, Shinohara Y, Hino M, Terada H, Baba Y, Kiwada H (2002) Effect of cationic liposomes in an in vitro transcription and translation system. Biol Pharm Bull 25:529–531 28. Cevc G (ed) (1993) Phospholipids handbook. Marcel Dekker Inc, New York 29. Lee AG (2004) How lipids affect the activities of integral membrane proteins. Biochim Biophys Acta 1666:62–87 30. Kurzchalia TV, Wiedmann M, Breter H, Zimmermann W, Bauschke E, Rapoport TA (1988) tRNA-mediated labeling of proteins with biotin. A nonradioactive method for the detection of cell-free ranslation products. Eur J Biochem 172:663–668 31. PROMEGA Product # L5001, technical NOTES, # TB285 32. Harris DH (2003) Quantitative chemical analysis. W. H. Freeman and Co., New York

Chapter 12 Study of Respiratory Cytochromes in Liposomes Iseli L. Nantes, Cintia Kawai, Felipe S. Pessoto, and Katia C.U. Mugnol Abstract Important findings regarding the structure and function of respiratory cytochromes have been made from the study of these hemeproteins associated to liposomes. These studies contributed to the comprehension of the biological role of these proteins in the electron transfer process, the regulatory mechanisms, the energy transduction mechanisms, the protein sites that interact with mitochondrial membranes and the role played by the non-redox subunits present in the protein complexes of the respiratory chain of eukaryotes. In this chapter, the protocols developed to study cytochrome bc1 activity in liposomes and the binding of cytochrome c to lipid bilayers is presented . The former protocol was developed to study the mechanism of energy transduction related to the topology of the components of bc1 complex in the mitochondrial membrane. These studies were done with purified cytochrome bc1 complexes reconstituted into potassium-loaded vesicles. The latter protocol was developed to study the influence of pH, DpH, and DY on the interaction of cytochrome c with liposomes that mimic the inner mitochondrial membrane. Key words: Cytochromes bc1, Eukaryotic cytochrome c, Respiratory chain, Energy transduction, Liposomes

1. Introduction 1.1. The bc1 Complex

Several studies of bc1 complex incorporated into liposomes can be found in the literature. These studies are concerned with the mechanism of proton translocation in the cytochrome bc1 complex and the function of individual subunits of the enzyme in the energy transduction process. Two different mechanisms have been proposed to explain the coupling of the electron transfer in the bc1 complex with the proton translocation from the matrix to the intermembrane space: the redox loop and the proton pump mechanisms. The redox loop mechanism explains the coupling by the

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_12, © Humana Press, a part of Springer Science+Business Media, LLC 2010

147

148

Nantes et al.

existence of arranged redox centers that, after reduction, accept protons from the matrix and, after reoxidation, release the protons to the intermembrane space. According to the proton pump mechanism, the reduction and re-oxidation of protein redox centers induce changes in the pKa of their amino acid side chains and their alternate exposure to the internal and external side of the membrane (1–3). The electron transfer pathway through bc1 complex is linked with the coenzyme Q cycle. The coenzyme molecules solved inside the membrane lipid fraction are completely reduced to QH2 by accepting electrons from complex I or II and the high potential b562 of complex III concomitant with the uptake of two protons from the mitochondrial matrix. In the re-oxidation process, QH2 donates one electron to cytochrome c1 via the iron sulfur protein (ISP) center, and one electron to heme b566 that recycles it back to oxidized coenzyme Q via heme b562. Concomitant with the re-oxidation, coenzyme Q releases two protons in the intermembrane space (Scheme 1). According to the originally proposed coenzyme Q cycle (4), in the bc1 complex, approximately 40% of the transmembrane potential is generated by the vectorial transmembrane electron transfer from cytochrome b566 to cytochrome b562 and the remaining 60% has been attributed to the electron transfer from cytochrome b562 to Q localized in the i center of the bc1 complex (5) and to

O

CH3

H3C O

reduced complex I or II

oxidized complex I or II

e–

H3C O

H

10

Fe3+cytc1

e–

OH CH3

H3C O 10

O

CH3

e– reclycled electron

2H+ from the matrix

e–

CH3

H3C O

CH3

Fe3+b562

Fe3+b566 O

H

Fe2+b562

e–

e–

H3C O

10

O

Fe2+b566

Fe2+cytc1

CH3

H3C O

CH3

O

O

H3C O

H3C O

H 2H+ to the intermembrane space

10

OH

H

CH3

e– transferred electron Scheme 1. Coenzyme Q cycle related to bc1 electron transport and the generation of transmembrane potential according to the redox loop mechanism

Study of Respiratory Cytochromes in Liposomes

149

the transfer of protons associated with the oxidation and reduction of Q (6, 7). Several studies concerning the mechanism that couples the electron transfer to a transmembrane proton electrochemical potential have been done by using bc1 complex incorporated into liposomes and involve the measurement of the catalytic and the back reaction, i.e., the electron transfer from a component with a more positive redox potential to a component with a more negative potential driven by an electrochemical potential gradient of protons. The study of electron transfer within the purified cytochrome bc1 complex reconstituted in liposomes has several advantages relative to the study with submitochondrial particles because, in the model system, the influence of other components of the electron transfer chain is eliminated, and the effect of Q can be studied by using the bc1 complexes depleted and replenished of Q. The mechanism of energy transduction related to the topology of the components of bc1 complex in the mitochondrial membrane has been studied in purified cytochrome bc1 complexes reconstituted into K+-loaded vesicles. In this model system, the addition of valinomycin to the K+-loaded proteoliposomes in a medium that does not have external KCl promotes the back electron transfer from ascorbate-reduced cytochrome c1 to ISP which, in turn, reduces Q to QH2. The reduced coenzyme then reduces the high potential heme b562. The thermodinamically unfavorable electron transfer from ISP to Q and from the high potential b562 to low potential b566 is coupled with proton uptake from the internal side of the membrane. Thus, in this model, the reversed transport of electrons is closely coupled to the reverse flow of protons related to the Q cycle (8). However, the demonstration that the reduction potential of the protein redox centers are responsive to conformational changes led to the postulation that there are least two redox-linked ionizable groups (9–12) in the b subunit, one closer to the outside and another closer to the inside. These groups have pKa values that vary according to the redox states of both hemes b and, possibly, to the redox states of quinone. More recently, a model compromising between pure “redox-loop” and pure “protonpump” mechanisms has been proposed . According to this model, in the respiratory chain, the one-electron reduction of oxidized coenzyme Q by complex I or complex II leads to the formation the semiquinone intermediate Q•−. This intermediate accepts one electron from b562 and is converted to the reduced form. The reduction of Q•− is accompanied by the uptake of one proton from the matrix and one proton from a redox-linked ionizable group. QH2 is reconverted to semiquinone by donating one electron to b566 and one proton to a redox-linked ionizable group while the second proton is delivered in the positive side of the membrane. The second electron is transferred to ISP and from that to cytochrome c1. The conformational changes promoted by the electron

150

Nantes et al. oxidized complex I or II O

e–

O

CH3

H3C O

H

10

H3C O

CH3

O Fe3+cytc1

H+

OH

CH3

CH3

H3C O H

H3C O

CH3

H+

e– reclycled electron

GIG

e– transferred electron GIG generic ionizable group

H+

GIGH

H+

e–

O

10

H+

Fe3+b562

Fe3+b566

O

GIG

e–

e–

H3C O

CH3

Fe2+b562

e– Fe2+cytc1

H

10

O

Fe2+b566

H3C O

CH3

H3C O

H3C O

H+

reduced complex I or II

10

OH

GIG

matrix H

CH3

GIGH

intermembrane space

H+

proton from and to GIG

Scheme 2. Coenzyme Q cycle related to bc1 electron transport and the generation of transmembrane potential according to the redox cycle associated to the proton pump mechanism

transfer to b subunit lead to the uptake of a proton from the matrix side by the redox-linked ionizable group concomitant with the electron transfer from b566 to b562 (Scheme 2). 1.2. Respiratory Cytochrome c

Cytochrome c is a mobile electron carrier in the respiratory chain that exhibits also the ability to detach from the inner mitochondrial membrane to integrate the apoptosome in the cytosol and promote apoptosis (13, 14). The cellular oxidative stress and loss of mitochondrial membrane potential promoted as Ca2+-induced permeability transition or photodynamic therapy (PDT), are known inducers of cytochrome c release and apoptosis (15–24). Therefore, the nature and specificity of the interaction of cytochrome c with model membranes have been extensively studied (25–28). The association of cytochrome c with phospholipid membranes involves both electrostatic and hydrophobic interactions (29–31). In model systems, at pH 7.4, the interaction of cytochrome c with acidic phospholipids occurs via site A (19, 32, 33), an electrostatically interacting site constituted of basic residues in cytochrome c, probably Lys72 and Lys73.

Study of Respiratory Cytochromes in Liposomes

151

The interaction of cytochrome c with the phospholipid head group leads to the accommodation of one phospholipid acyl chain within a hydrophobic channel of the protein lined by hydrophobic amino acid residues from the surface of the protein to the heme crevice. The insertion of a phospholipid acyl chain in the cytochrome c structure leads to perturbation in the heme iron spin state detected by electron paramagnetic resonance (EPR) (32). These studies have shown a clear correlation between the nature of the lipid acyl chain and the spin states of cytochrome c interacting with different types of lipid membranes. However, on the external surface of the inner mitochondrial membrane, the clusters of positively charged amino acid side chains (site A) are important for the interaction of the cytochrome c redutase and oxidase (13, 14, 34). More recently, the interaction of cytochrome c with mitochondrial mimetic vesicles of PCPECL (1,2-dipalmitoyl-sn-glycero-3-phosphocholine, 1,2-dipalmitoyl-snglycero-3-phosphoethanolamine and heart cardiolipin) was investigated over the 7.4–6.2 pH range by means of turbidimetry and photon correlation spectroscopy. In the presence of cytochrome c, the decrease of pH induced vesicle fusion dependent of protein ionizable groups with a pKa(app) around 7.0. The carbethoxylation of these groups by diethylpirocarbonate impaired cytochrome c-induced vesicle fusion, remaining cytochrome c association to vesicles unaffected. Matrix-Assisted Laser Desorption/Ionization (MALDI) Time-of-Flight (ToF) analysis revealed that Lys22, Lys27, His33 and Lys87 cytochrome c residues were the main targets for carbethoxylation performed at low pH values (<7.5). This site was denominated site L because the high lysine content. The structural analysis of cytochrome c structure shows that these amino acid residues belong to clusters of positively charged amino acids responsible for the lowering of the pKa. Subsequent unpublished studies have revealed that the association of cytochrome c with PCPECL liposomes via site L and monitored by turbidimetry is favored by electric potential difference provided by DpH and DY. Thus, the turbidimetry analysis of these vesicles in the presence of cytochrome c can be used to determine the existence and loss of electric potential difference in these vesicles. Besides, turbidimetry, z potential analysis is also an adequate technique to detect the binding and vesicle fusion promoted by cytochrome c.

2. Materials 2.1. Protocols to Study the Back Electron Transfer in the bc1 Complex

1. 10 l 0.1 M phosphate buffer pH 7.4. 2. 1,300 ml 0.02 M phosphate buffer, pH 7.4. 3. 250 ml borate–phosphate buffer: 0.1 M Na2HPO4 and H3BO3. 4. 800 ml 0.01 M KH2PO4.

152

Nantes et al.

2.1.1. Purification of bc1 Complex from beef heart muscle

1. 50 mM phosphate buffer, pH 7.4, containing 0.25 M sucrose.

2.1.2. Separation of Cytochrome bc1 Complex from SuccinateCytochrome c Reductase

4. 50 mM Tris–HCl buffer, pH 8.0, containing 0.67 M sucrose.

2. 50 mM phosphate–borate buffer, pH 7.8. 3. 50 mM phosphate buffer, pH 7.4. 5. 200 mM stock solution of succinate was prepared in an argon atmosphere in the presence of 10 mM dithiothreitol. 6. Potassium Deoxycholate. 7. Ammonium acetate solution, which was prepared according to Hatefi and Rieske (35) by dissolving 454 g of solid ammonium acetate in 613 ml of H2O. 8. 18.5% ammonium acetate saturation (20 ml of 50% saturated ammonium acetate solution to 100 ml solution). 9. 2 M urea. 10. Sodium cholate.

2.1.3. K+-Loaded Cytochrome bc1 Vesicles

1. 500 mM Stock solution cytochrome c (horse heart, type III) was prepared in deionized water (see Note 1) and the concentration was calculated using molar absorptivity e409 = 106.1 × 103 M/cm (36) and stored at −20°C. 2. Stock solution of valinomycinin ethanol. 3. Stock solution of nigericin in ethanol. 4. Stock solution of antimycin A in ethanol. 5. 48 mM stock solution of safranine in deionized water. 6. 50 mM stock solution of sodium ascorbate in deionized water (see Note 2). For a storage during hours the solution should be saturated with argon (see Note 3). 7. Asolectin (soybean phospholipids) and purified partially according to Sone et al. (37). 8. 10 mM 3-(N-mopholino)propanesulfonic acid (MOPS), pH 7.4, prepared by titration of the solution with NaOH solution. 9. 100 mM KCl. 10. 100 mM NaCl.

2.2. Protocols to Study the Interaction of Respiratory Cytochrome c Sites A and L with Lipid Bilayers (38)

1. Stock solution cytochrome c (horse heart, type III) was prepared in deionized water and the concentration was calculated using molar absorptivity e409 = 106.1 × 103 M/cm (39) and stored at −20°C (see Note 6). 2. HCl, NaOH, NaCl, HClO4, KH2PO4 and ammonium molybdate used to prepare stock solutions at deionized water (conductivity = 18 ohms). 3. HEPES (4-(2-Hydroxyethyl)piperazine-1-ethanesulfonic acid) buffer prepared at different pH values by the titration of the solution with NaOH and HCl.

Study of Respiratory Cytochromes in Liposomes

153

4. Tris base 2-Amino-2-(hydroxymethyl)-1,2-propanediol buffer solution prepared by the titration of the solution with NaOH and HCl. 5. Acetic acid used to prepare acetate buffer. 6. Ascorbic acid and d-glucose used to prepare stock solutions at deionized water (conductivity = 18 ohms). 7. Chloroform used to solve phospholipids stock solutions. 8. Egg phosphatidylcholine(PC), Lipids 1,2-dioleoyl-sn-glycero-3phosphoethanolamine (DOPE) and 1,1¢,2,2¢tetraoleoyl cardiolipin (CL). To prepare the lipid stock solution, each lipid type is weighed and dissolved in chloroform in order to obtain a concentration of 20 mg/ml stored at −20°C. The correct concentration was determined by phosphate dosage. Total phospholipid content is determined by phosphate analysis, using the method of Bartlett (40), with some modifications. 2 ml of each lipid is dried in a pyrex tube and heated in hot block at 120°C for 10 min. In the following, 0.4 ml HClO4 70% is added in a tube of HClO4 and heated at 180°C for 1 h. After this step, the tubes are removed from the hot block and cooled to room temperature. 1.0 ml of water and 0.4 ml of 1.252% ammonium molybdate are added inside each tube and the samples are vortexed. 0.4 ml of fresh ascorbic acid 3% is added, mixed by vortex and then incubated at 90°C in water bath for 10 min. The samples are then cooled to room temperature and absorbance at 797 nm is recorded. Concentrations are determined using a calibration curve obtained with KH2PO4 standard solutions from 0 to 100 nmols. 9. Polycarbonate membrane filter with 100 and 1,000 nm pore. 10. Spectrapor dialyze tubing of 3,500 molecular weight cutoff. 11. PyPC (2-(10-(1-pyrene) decanoyl) phosphatidylcholine).

3. Methods 3.1. Purification of bc1 Complex from Beef Heart Muscle

Step 1: Extraction of succinate-cytochrome c reductase from beef heart muscle 1. The preparation of heart muscle is based on Keilin and Hartree method with modifications done by King (1961) (41). Heart muscle is cleaned of fat and connective tissues, and minced in a power-driven meat grinder. One kilogram of mince is washed with 20 l of tap water in a plastic bucket and efficient mechanical stirring is carried out for 20 min (see Notes 3 and 4). In the paper published in 1961 by King, the author describes the model of the mechanical stirrer, equipped with a three-blade stainless steel propeller of approximately 6 in. across (see Notes 4 and 5).

154

Nantes et al.

2. The mince is pressed and squeezed either by hand in cheese cloth, or by a hydraulic press in canvas (37). Usually, six washings result in a mince as a light yellow, but not pink, preparation. The mince is then used either directly for grinding, or further mixed with 10 l of 0.1 M phosphate buffer, pH 7.4, and stirred mechanically for approximately 1 h. It is finally washed with 20 l of deionized water. 3. Five hundred grams of the hard-pressed mince thus obtained are ground in a mechanical mortar for 2 h with 700 ml of 0.02 M phosphate buffer, pH 7.4 and 200 g of sand (37). 4. The mixture is diluted with another 600 ml of buffer and then centrifuged for 30 min at 600 × g. 5. The upper turbid layer is recentrifuged at 6,000 × g for 90 min. 6. The pellet is dispersed in 250 ml of borate–phosphate buffer, with a Potter-Elvehjem type homogenizer. All operations described are done at room temperature. 7. The turbid supernatant fluid from the first centrifugation is cooled to approximately 2°C by the addition of crushed ice and brought to pH 5.5 with 0.1N acetic acid. 8. The mixture is centrifuged at 600 × g for 15 min in a refrigerated centrifuge. 9. The clear supernatant fraction is discarded and the precipitate is washed with 800 ml 0.01 M KH2PO4 and centrifuged again. The residue is finally suspended in borate–phosphate buffer. The heart muscle preparation is not stable at room temperature in mediums with pH values lower than 6.8. Step 2: Separation of cytochrome bc1 complex and succinate dehydrogenase from the succinate-cytochrome c reductase extracted from beef heart muscle Separation of the bc1 complex and succinate dehydrogenase from succinate-cytochrome c reductase can be achieved essentially according to the method previously described (42) except that a slightly lower pH and a more strictly anaerobic condition should be used in order to get a better recovery of succinate dehydrogenase. 1. Succinate-cytochrome reductase as prepared was dialyzed against 50 mM phosphate buffer, pH 7.4, containing 0.25 M sucrose overnight with two changes of buffer and frozen at −70°C until be used. 2. Succinate-cytochrome reductase, 10 mg/ml, was centrifuged at 170,000 × g for 45 min and the collected precipitate was resuspended in 50 mM phosphate–borate buffer, pH 7.8, to a protein concentration of 15 mg/ml. 3. Then it was incubated with 20 mM succinate in an argon atmosphere in the presence of 1 mM dithiothreitol for 30 min before the pH of the solution was brought to 10 by anaerobic addition of NaOH.

Study of Respiratory Cytochromes in Liposomes

155

4. The mixture was further incubated at 0°C for 20 min before being centrifuged at 170,000 × g for 45 min. 5. Succinate dehydrogenase was collected from the supernatant solution and the collected precipitate containing the cytochrome bc1 complex was resuspended and homogenized in 50 mM phosphate buffer, pH 7.4 and then brought to pH 10.3 once more by the addition of 0.1N NaOH. 6. After incubation at 0°C for 20 min, the mixture was centrifuged at 170,000 × g for 40 min. 7. The bc1 complex was collected in the precipitate, and the supernatant solution, containing a small amount of succinate dehydrogenase, was discarded. 8. The precipitate was suspended in 50 mM Tris–HCl buffer, pH 8.0, containing 0.67 M sucrose and kept at –70°C for the isolation of ubiquinone-binding protein and bc1 III complex (ubiquinol-cytochrome c reductase). Step 3: Separation of ubiquinone-binding protein and bc1 complex from ubiquinol-cytochrome c reductase (complex III) 1. The frozen bc1 complex is thawed and protein concentration adjusted to 20 mg/ml with 50 mM Tris–HCl buffer, pH 8.0, containing 0.67 M sucrose. 2. The suspension is solubilized with potassium deoxycholate at a concentration of 0.5 mg per mg of protein, and fractionated with ammonium acetate solution. 3. Ammonium acetate solution is added to the deoxycholatesolubilized bc1 complex, 11 ml of 50% saturated ammonium acetate solution per 100 ml solution, to give 5% ammonium acetate saturation. 4. The mixture is stirred at 0°C for 20 min and then centrifuged at 50,000 × g for 30 min. The supernatant is collected and brought to 10% ammonium acetate saturation (12 ml of ammonium acetate solution to 100 ml of solution). 5. The pellet, which contains ubiquinone-binding protein, is collected by centrifuging at 50,000 × g for 30 min and dissolved in 50 mM Tris–HCl buffer, pH 8.0, containing 0.67 M sucrose. 6. The crude ubiquinone-binding protein preparation can be frozen at −70°C for several weeks without significant loss of activity. 7. The supernatant solution is collected and brought to 18.5% ammonium acetate saturation (see Subheading 2). 8. After stirring for 20 min, the solution is centrifuged at 170,000 × g for 30 min. 9. The pellet which contains purified ubiquinol-cytochrome reductase is dissolved in 50 mM Tris–HCl pH 8.0, containing 0.67 M sucrose, diluted to a protein concentration of approximately 20 mg/ml, and frozen at −70°C.

156

Nantes et al.

10. The ubiquinol-cytochrome c reductase thus obtained can be further purified by ammonium sulfate fractionation in the presence of 2 M urea and 0.5% sodium cholate at a protein concentration about 10 mg/ml. 11. The highly purified ubiquinol-cytochrome c reductase is recovered in the precipitate formed between 55% and 70% ammonium sulfate saturation. 12. Succinate dehydrogenase, ubiquinone-binding protein and bc1 complex can be identified by their enzymatic activities as described by (43). 13. By following the above procedure, a SDS–polyacrylamide gel electrophoresis, of the purified sample shows that the bc1 is free from contamination by succinate dehydrogenase protein. 3.2. Preparation of Vesicles

Vesicles containing cytochrome bc1 vesicles were prepared by the cholate dialysis method (44–46) as follows in presence of potassium buffer. 1. Asolecitin dissolved in chloroform was mixed in the pyrex tube and then evaporated with N2 gas in order to obtain film and kept under reduced pressure for at least 2 h. 2. The lipid film was resuspended at concentration of 40 mg/ml in 10 mM MOPS/NaOH, pH 7.4, 100 mM KCl and 1.5% sodium cholate and then sonicated for 10–15 min at 0˚C.

3.3. Reconstitution of Proteoliposomes

1. The purified bc1 complex isolated from beef heart muscle according to the method describe by Yu and Yu, 1980 (40), was added for final concentration of 5 mM, incubated for 5 min at 4°C. 2. The solution was dialyzed against 400 times volume in 10 mM MOPS/NaOH, pH 7.4 plus 100 mM KCl at 4°C for 24 h with two changes of outer buffered solution. 3. In order to remove the external potassium, the proteoliposomes were dialyzed again against 10 mM MOPS/NaOH, pH 7.4 plus 100 mM NaCl at 4°C for two days with four changes of the dialysis medium.

3.4. Reduction of Cytochrome c Coupled to ValinomycinInduced K+ Diffusion Potential (8)

1. Reduction of cytochrome c1 was done by spectroscopic measurements can be carried out in a UV-visible spectrophotometer equipped with a thermostated cell holder. 2. The catalytic activity of the reconstituted bc1 complex is assayed at 20°C in a medium containing 10 mM MOPS/ NaOH, pH 7.4 and 100 mM KCl, 50 mM cytochrome c, and 25 mM Q2H2. 3. The initial rate of cytochrome c reduction is corrected against non-enzymatic reduction of cytochrome c by ubiquinol in the absence of bc1 vesicles. Proteoliposomes reconstituted

Study of Respiratory Cytochromes in Liposomes

157

with the bc1 complex exhibit the oxidation control ratio ranging from 7 to 9 that is determined as the ratio of the enzyme activity in the presence and absence of 0.2 mg of valinomycin plus 2 mg of nigericin. 4. The ability of K+ diffusion potential can be checked spectrophotometrically using safranine as an optical probe for membrane potential (47) as follows. A 5 ml portion of proteoliposomes containing the bc1 complex is added to 1 ml of a medium containing 4.8 mM safranine, 10 mM MOPS/ NaOH, pH 7.4 and 100 mM NaCl. The reaction is initiated by the addition of valinomycin, and the change in the membrane potential is measured by the changes in the absorbance of safranine at 525 nm (48) or at 530–578 nm (4). 5. The reduction of cytochrome c and c1 is determined spectrometrically using the following extinction coefficients: De550 nm = 19.2 mM/cm for cytochrome c and De553−539 nm (reduced−oxidaded) = 17.5 mM/cm for cytochrome c1. (reduced−oxidaded) 6. The response of the reduction level of cytochrome b in the reconstituted bc1 complex to the valinomycin-induced K+ diffusion potential can be monitored spectrophotometrically. In these experiments, cytochrome c1 and iron–sulfur protein in the bc1 complex are reduced with 5 mM sodium ascorbate in a medium in which there is no external KCl. In the following, valinomicin is added and this procedure is accompanied by a rapid increase in absorbance at 562 nm relative to 575 nm, the wavelength chose to monitor the redox state of cytochrome b components. After approximately 3 min, it is observed the maximum reduction level of reduced cytochrome b. 7. The rapid reoxidation of b components can be observed after addition of nigericin that collapses the K+ gradient. 8. The dependence of cytochrome b reduction on the magnitude of the diffusion potential (DY) formed across liposomal membranes induced by valinomycin can be monitored by changing the concentration of KCl in the reaction medium. The DY can be calculated from the Nernst equation DY (mV) = 60 log([K+]in/[K+]out) and assuming that the internal KCl concentration is 100 mM. Table 1 shows the results obtained by measuring the cytochrome b reduction as a function of the membrane potential. Cytochrome b562 was not reduced at potentials below 30 mV, and the reduction and the maximal reduction is attained at more than 210 mV (the initial concentration of external KCl was less than 30 mM). The data were obtained from ref. (8). 9. According to the reported method (49) ubiquinone-10 content of the purified bc1 complex can be determined spectrophotometrically using an extinction coefficient of 12.25 mM/cm of the difference between the oxidized and reduced quinine at 275 nm (50).

158

Nantes et al.

Table 1 Results obtained by measuring the cytochrome b reduction as a function of the membrane potential Potential (mV)

A(562–575)

0

0

30

0

60

0.0005

87

0.0016

120

0.0029

150

0.0033

180

0.0035

205

0.0039

240

0.00375

10. The total protein content can be estimated according to Lowry et al. (51). 3.5. Study the Interaction of Respiratory Cytochrome c Sites A and L with Lipid Bilayers [52]

Step 1: Preparation of PCPECL liposomes 1. Aliquots of lipids dissolved in chloroform and calculated to a final concentration of 5 mM in a final volume of 2 ml, are mixed in the pyrex tube and the solvent evaporated by N2 or argon gas stirring (see Note 7). The aliquot of lipid stock to be added in the pyrex is calculated according to the molar percentage of each lipid in the mixture: 50% PC, 30% PE and 20% CL. 2. After complete evaporation of the solvent, the lipid film was kept under reduced pressure for at least 2 h to eliminate residual solvent. The films can be used immediately after the preparation or stored at −20°C until use (see Notes 7 and 8). 3. The liposomes are prepared by adding 1 ml cold buffer followed by vortex stirring. This procedure renders multilamelar liposomes. Unilamelar liposomes can be obtained by extrusion of hydrated lipid dispersions in a mini-extruder. Samples are subjected at least to 11 passes through one polycarbonate filters (100 or 1,000 nm pore size) installed in tandem (53). After this procedure one more milliliter of buffer is added in the lipid pyrex tube and the previous step is repeated using the same filters. 4. The preparation of liposomes with DpH by varying inside and outside pH at a range of 6.2–8.2 is done by the hydration of the lipid film with 5 mM HEPES buffer.

Study of Respiratory Cytochromes in Liposomes

159

5. For inside pH variations (acidic outside) the lipid films are individually hydrated by 5 mM HEPES buffer at the pH values from 6.2 to 8.2. After the procedures described in the item 3 above, the external pH is corrected to 6.2 by HCl titration. 6. For outside pH variations (alkaline inside), one lipid film is hydrated by 5 mM HEPES buffer pH 8.2. After the procedures described in the item 3 above, the external pH is corrected with NaOH solution to the desired pH (from 6.4 to 8.2). Due to the high permeability of lipid bilayers to protons (54), the liposomes should be used for experimental procedures immediately after the preparation. 7. To obtain liposomes with DY, the lipid film is hydrated by 5 mM HEPES buffer pH 6.2 containing NaCl at a concentration range of 50–150 mM. The NaCl outside liposomes is eliminated by exhaustive dialysis against 500 times volume of 5 mM HEPES in presence of d-glucose with equivalent osmolarity (Os/kg). 150 mM NaCl (~0.277 Os/kg) = 266 mM d-glucose (55). The dialysis is performed using Spectrapor tubing (3,500 molecular weight cutoff) previously hydrated for 5 min. Step 2: Binding of cytochrome c to PCPECL liposomes and vesicle fusion 1. Determination of cytochrome c binding to PCPECL liposomes and vesicle fusion by PyPC (2-(10-(1-pyrene) decanoyl) phosphatidylcholine) fluorescence. This technique permits to detect the binding of cytochrome c to PyPC-labeled vesicles because the hemeprotein quenches the fluorescence of pyrene. The vesicle fusion is monitored by the decrease in the excimer/ monomer (E/M) ratio of PyPC (56, 57) present in PCPECL vesicles, which were mixed with unlabeled vesicles in the presence of protein. In conditions, in which the binding occurs predominantly via site A, there is no vesicle fusion and only partial quenching of pyrene fluorescence is observed without significant changes in the E/M ratio. In conditions in which sites A and L of cytochrome c are protonated, the protein binding is accompanied by vesicle fusion and, besides pyrene fluorescence quenching, changes in the E/M ratio are also detected. (a) The preparation of PyPC-labeled liposomes is done as follows. Aliquots of small unilamellar PCPECL (50/30/20% mol) vesicles containing PyPC 7% mol are added to identical unlabelled PCPECL vesicles to a final ratio of 1/10 labeled/unlabeled vesicles. Fluorescence emission spectra are recorded at 30°C using a fluorescence spectrometer. (b) To determine the binding of cytochrome c and vesicle fusion, aliquots of cytochrome c to a final concentration of 0.9 mM are added to 50 mM PCPECL liposomes in 10 mM HEPES pH 6.2 or 7.4 (3 ml) and the fluorescence

160

Nantes et al. 2100

2100

Relative Fluorescence

Monomer

pH 7.4

pH = 6.2 1400

1400

700

0

700

Excimer

400

450

500

400

450

500

0

Wavelength (nm)

Fig. 1. Effect of cytochrome c on the fluorescence emission spectra of PCPECL vesicles containing 7 mol% PyPC. Vesicles (0.05 mM PCPECL; labeled/unlabeled 1:10) incubated at 30 °C in 10 mM HEPES buffer with 0.9 mM cyt c for 378 s at pH 6.2 and pH 7.4. Thin solid line corresponds to the initial time, before addition of cytochrome c and thick solid line corresponds to the final incubation time

is monitored during 10 min with the following parameters: excitation wavelength at 344 nm, scan range = 350 to 550 nm and emission slits = 5 nm. The E/M ratio is calculated by using the fluorescence intensity at 459 nm (excimer) and at 398 nm (monomer). Figure 1 shows an example of the results obtained in conditions in which cytochrome c binds to lipid bilayer (pH = 7.4) and in conditions in which the binding was favored and accompanied by vesicles fusion (pH = 6.2). 3.6. Determination of Cytochrome c Binding to PCPECL Liposomes and Vesicle Fusion by Turbidimetry

1. Turbidity measurements are conducted in a spectrophotometer using quartz cuvettes of 1 cm light path and a slit of 0.5 nm. 2. The kinetics of vesicle fusion is monitored continuously by the increase of the turbidity at 480 nm as a function of time. Vesicles alone, without cytochrome c are used as blank. 3. A typical kinetics is carried out with 1 ml of 0.25 mM liposomes in HEPES buffered water pH 6.2 in the presence of 4.0 mM cytochrome c, at 30°C, during 10 min with 1 s time interval. 4. Because the protonation of cytochrome c site L, the cytochrome c-promoted PCPECL vesicle fusion is favored at acidic pH. Transmembrane potential provided by DpH and DY acts synergistically with external acidic pH to favor the binding of cytochrome c to lipid bilayers. Therefore, the comparative capacity of cytochrome c to fuse negatively charged vesicles is an indicator of the presence of transmembrane potential. An example of these effects is shown in Fig. 2.

Study of Respiratory Cytochromes in Liposomes

161

Turbidity (480 nm)

0,8 pHi = 7.4, pHo = 6.2 pHi = pHo = 6.2 + 50 mM NaCl(in) pHi = pHo = 6.2

0,6

0,4

0,2

0,0

0

100

200

300 400 Time (s)

500

600

Fig. 2. Effect of DpH and DY on the turbidity of the medium containing PCPECL vesicles and cytochrome c. Experiments were carried out at 30°C using 0.25 mM lipids containing 50% DPPC, 30% DPPE, and 20% CL in 10 mM HEPES buffer in the conditions indicated in the Figure

0.6

Turbidity (480 nm)

0.5 0.4 0.3 0.2 0.1 0.0 6.0

6.4

6.8 7.2 Outside pH

7.6

8.0

Fig. 3. Effect of outer pH (inside 7.4) on the turbidity at 480 nm of the medium containing PCPECL vesicles. The turbidity was determined 10 min after incubation at 30°C. Experiments were carried out at with 0.25 mM lipids containing 50% DPPC, 30% DPPE, and 20% CL or phosphatidylglycerol

5. The data on turbidity increase (Dturbidity at 480 nm) as a function of external pH (inside pH = 7.4) render a sigmoid curve that is well fitted by Eq. 1 and Fig. 3.

∆turbidity =

lim1 10(pH-pk a) 10(pH-pk a) + 1

(1)

162

Nantes et al.

3.7. Determination of Cytochrome c Binding to PC/PE/CL Liposomes by Mean Zeta-Average Diameters (Dz) and Zeta-Potentials (z)

1. At 25°C, Dz and z are determined by dynamic light scattering and laser-Doppler microelectrophoresis, respectively, using a ZetaPlus-ZetaPotential Analyzer (Brookhaven Instruments Corporation, Holtsville, NY) equipped with a 570 nm laser and dynamic light-scattering at 90° for particle sizing (58). Particle size (Dz) and zeta-potential (z) are the mean values calculated from at least 10 independent measurements with standard deviations lower than 5% from the mean. z is determined, in pure water, from electrophoretic mobility m and the Smoluchowski’s equation: z = mh/e, where h is the medium viscosity and e the medium dielectric constant. 2. For the mean zeta-average diameters (Dz) and zeta-potentials (z) measurements the final lipid concentration can be varied from 0.25 to 1.0 mM. The binding of cytochrome c to negatively charged vesicles is accompanied by increase of potential and in the occurrence of fusion by increase of vesicle mean diameter. In conditions in which the binding of cytochrome c to the negatively charged PCPECL vesicles is favored, the z potential increase to less-negative values and in conditions in which the vesicle fusion is also favored the mean diameter of the vesicles is significantly increased.

4. Notes 1. All solutions necessary for the protocols described in this chapter should be prepared in water that has a resistivity of 18.2 MW cm. This standard is referred to as “deionized water” in this chapter. 2. Ascorbate is susceptible to oxidation and only fresh solution should be used for the reduction of cytochrome c. For storage during hours, the solution should be saturated with argon. 3. The saturation of solutions with argon can be done by purging the gas directly inside the solution during at least 10 min followed by closing the tube or cuvette with a lid septum. 4. The mechanical stirring of tissues (heart muscle in this chapter) should be done under refrigeration that could be warranted by an ice bath. 5. For a better yield and to avoid proteolytic fragmentation of the proteins of the bc1 complex during the extraction procedure, the following inhibitors of proteases should be added to the solutions: 0.1 M aminocaproic acid, 6.5 mM benzamidine, 5.5 mM iodoacetamide and 0.1 mM phenylmethylsuphonyl fluoride (PMSF).

Study of Respiratory Cytochromes in Liposomes

163

6. The UV spectrometry should be done by using quartz cuvettes because glass cuvettes impair the passage UV light. 7. The preparation of liposomes could be done by using extrusor, tip or bath sonicator. If the bath sonicator is used, the temperature should be set to 50°C. The formation of unilamelar liposomes can be observed by the decrease of the sample turbidity. If the extrusor is used, it can be heated to facilitate the traffic of the solution. 8. In the preparation of lipid films, the evaporation of chloroform should be done under direct jet of argon or nitrogen concurrently with the manual rotation of the tube maintained in an inclined position relative to the needle. References 1. Trumpower BL (1990) The protonmotive Q cycle. J Biol Chem 265:11409–11412 2. Boyer PD (1993) The binding change mechanism for ATP synthase. Some probabilities and possibilities. Biochim Biophys Acta 1440:215–250 3. Erecinska M, Wilson DF (1982) Regulation of cellular energy metabolism. J Membr Biol 70:1–14 4. Mitchell P (1976) Possible molecular mechanisms of the protonmotive function of cytochrome systems. J Theor Biol 62:327–367 5. Robertson DE, Dutton PL (1988) The nature and magnitude of the charge-separation reactions of ubiquinol cytochrome c2 oxidoreductase. Biochim Biophys Acta 935: 273–291 6. Drachev LA, Kaurov BS, Mamedov MD, Mulkidjanian AY, Semonev AY, Shinkarev VP, Skulachev VP, Verkhovsky MI (1989) Biochim Biophys Acta 973:189–197 7. Konstantinov A, Kunz WS, Kamensky YA (1981) In: Skulachev VP, Hinkle PC (eds) Chemiosmotic proton circuits in biological membranes, Addison-Wesley pp 123–146 8. Miki T, Mikin M, Orii Y (1994) Membrane potential-linked reversed electron transfer in the Beef Heart Cytochrome bcl complex reconstituted into potassium-loaded phospholipid vesicles. J Biol Chem 269:1827–1833 9. West IC, Mitchell P, Rich PR (1988) Electron conduction between b cytochromes of the mitochondrial respiratory chain in the presence of antimycin plus myxothiazol. Biochim Biophys Acta 933:35–41 10. Konstantinov AA, Popova E (1988) In: Papa S (ed) Cytochrome systems: molecular biology and bioenergetics, Plenum Publishing Corp., New York, pp 751–765

11. Kamensky YuA, Artzabanov VYu, Shevchenko DV, Konstantinov AA (1979) Effect of antimycin A on redox-dependent protonation of the b cytochromes of the mitochondrial respiratory chain. Dokl Acad Nauk SSSR 249:994–997 12. Von Jagow G, Link TA, Ohnishi T (1986) Organization and function of cytochrome b and ubiquinone in the cristae membrane of beef heart mitochondria. J Bioenerg Biomembr 18:157–179 13. Kluck RM, BossyWetzel E, Green DR, Newmeyer DD (1997) Science 275:1132–1136 14. Yang J, Liu XS, Bhalla K, Kim CN, Ibrado AM, Cai J, Peng T, Jones DP, Wang X (1997) Science 275:1129–1132 15. Crompton M (1999) The mitochondrial permeability transition pore and its role in cell death. Biochem J 341:233–249 16. Lenaz G (1998) Quinone specificity of complex. Biochim Biophys Acta 1364:207–221 17. De Giorgi F, Lartigue L, Bauer MK, Schubert A, Grimm S, Hanson GT, Remington SJ, Youle RJ, Ichas F (2002) The permeability transition pore signals apoptosis by directing Bax translocation and multimerization. FASEB J 16:607–609 18. Matroule JY, Carthy CM, Granville DJ, Jolois O, Hunt DW, Piette J (2001) Mechanism of colon cancer cell apoptosis mediated by pyropheophorbide-a methylester photosensitization. Oncogene 20:4070–4084 19. Sugawara T, Lewen A, Gasche Y, Yu F, Chan PH (2002) Overexpression of SOD1 protects vulnerable motor neurons after spinal cord injury by attenuating mitochondrial cytochrome c release. FASEB J 16:1997–1999 20. Kowaltowski AJ, Castilho RF, Vercesi AE (2001) Mitochondrial permeability transition and oxidative stress. FEBS Lett 495:12–15

164

Nantes et al.

21. Kowaltowski AJ, Vercesi AE, Fiskum G (2000) Bcl-2 prevents mitochondrial permeability transition and cytochrome c release via maintenance of reduced pyridine nucleotides. Cell Death Differ 7:903–910 22. Kowaltowski AJ, Vercesi AE (1999) Mitochondrial damage induced by conditions of oxidative stress. Free Radic Biol Med 26:463–471 23. Barros MH, Netto LE, Kowaltowski AJ (2003) H2O2 generation in Saccharomyces cerevisiae respiratory pet mutants: effect of cytochrome c. Free Radic Biol Med 35:179–188 24. Kowaltowski AJ, Turin J, Indig GL, Vercesi AE (1999) Mitochondrial effects of triarylmethane dyes. J Bioenerg Biomembr 31:581–590 25. Wuthrich K, Aviram I, Schejter A (1971) Structural studies of modified cytochromes c by nuclear magnetic resonance spectroscopy. Biochim Biophys Acta 253:98–103 26. Nicholls P (1974) Cytochrome c binding to enzymes and membranes. Biochim Biophys Acta 346:261–310 27. Pinheiro TJ (1994) The interaction of horse heart cytochrome c with phospholipid bilayers. Structural and dynamic effects. Biochimie 76:489–500 28. Nantes IL, Zucchi MR, Nascimento OR, Faljoni-Alario A (2001) Effect of heme iron valence state on the conformation of cytochrome c and its association with membrane interfaces. A CD and EPR investigation. J Biol Chem 276:153–158 29. Rytömaa M, Kinnunen PK (1995) Reversibility of the binding of cytochrome c to liposomes. Implications for lipid–protein interactions. J Biol Chem 270:3197–3202 30. Tuominen EK, Wallace CJ, Kinnunen PK (2002) Phospholipid–cytochrome c interaction: evidence for the extended lipid anchorage. J Biol Chem 277:8822–8826 31. Zucchi MR, Nascimento OR, Faljoni-Alário A, Prieto T, Nantes IL (2003) Modulation of cytochrome c spin states by lipid acyl chains: a continuous-wave electron paramagnetic resonance (CW-EPR) study of haem iron. Biochem J 370:671–678 32. Rytömaa M, Mustonen P, Kinnunen PKJ (1992) Reversible, nonionic, and pH-dependent association of cytochrome c with cardiolipinphosphatidylcholine liposomes. J Biol Chem 267:22243–22248 33. Rytömaa M, Kinnunen PKJ (1994) Evidence for two distinct acidic phospholipid-binding sites in cytochrome c. J Biol Chem 269:1770–1774 34. Pelletier H, Kraut J (1992) Crystal structure of a complex between electron transfer partners, cytochrome c peroxidase and cytochrome c. Science 258:1748–1755

35. Hatefi Y, Rieske JS (1967) Methods Enzymol 10:225–230 36. Margoliash E, Frohwirt N (1959) Spectrum of horse-heart cytochrome c. Biochem J 71:570–572 37. Sone N, Yoshida M, Hirata H, Kagawa Y (1977) J Biochem (Tokyo) 81:519–528 38. Kawai C, Prado FM, Nunes GLC, Di Mascio P, Carmona-Ribeiro AM, Nantes IL (2005) pH-dependent interaction of cytochrome c with mitochondrial mimetic membranes. The role of an array of positively charged amino acids. J Biol Chem 280:34709–34717 39. Margoliash E, Frohwirt N (1959) Spectrum of horse-heart cytochrome c. Biochem J 71:570–572 40. Bartlett GR (1959) Phosphorus assay in the column chromatography. J Biol Chem 234:466–468 41. King TE (1961) J Biol Chem 236:2342–2346 42. Yu CA, Yu L, King TE (1974) J Biol Chem 249:4905–4910 43. Yu CA, Yu L (1980) Resolution and reconstitution of succinate-cytochrome c reductase: preparations and properties of high purity succinate dehydrogenase and ubiquinol-cytochrome c reductase. Biochim Biophys Acta 591:409–420 44. Leung KH, Hinkle PC (1975) Reconstitution of ion transport and respiratory control in vesicles formed from reduced coenzyme Q-cytochrome c reductase and phospholipids. J Biol Chem 250:8467–8471 45. Beattie DS, Villalobo A (1982) Energy transduction by the reconstituted b-c1 complex from yeast mitochondria. Inhibitory effects of dicyclohexylcarbodiimide. J Biol Chem 257:14745–14752 46. Miki T, Orii Y, Mukohata Y (1987) A mechanism of respiratory control: studies with proteoliposomes containing cytochrome oxidase and bacteriorhodopsin. J Biochem 102: 199–209 47. Gutweniger H, Massari S, Beltrame M, Colonna R, Veronesse P, Ziche B (1977) Biochim Biophys Acta 459:216–224 48. Miki T, Orii Y (1986) Cytochrome c peroxidase activity of bovine heart cytochrome oxidase incorporated in liposomes and generation of membrane potential. J Biochem 100: 735–745 49. Redfeam ER (1967) Isolation and determination of ubiquinone. Methods Enzymol 10:381–384 50. Tsai A-L, Olson JS, And Palmer G (1987) The kinetics of reoxidation of yeast complex III. An evaluation of the Q-cycle. J Biol Chem 262:8677–8684

Study of Respiratory Cytochromes in Liposomes 51. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951) Protein measurement with the Folin phenol reagent. J Biol Chem 193:265–275 52. Kawai C, Prado FM, Nunes GLC, Di Mascio P, Carmona-Ribeiro AM, Nantes IL (2005) pH-dependent interaction of cytochrome c with mitochondrial mimetic membranes. The role of an array of positively charged amino acids. J Biol Chem 280:34709–34717 53. Hunter DG, Frisken BJ (1998) Effect of extrusion pressure and lipid properties on the size and polydispersity of lipid vesicles. Biophys J 74:2996–3002 54. Nichols JW, Deamer DW (1980) Net protonhydroxyl permeability of large unilamellar liposomes measured by an acid–base titration technique. Proc Natl Acad Sciences (USA) 77:2038–2042

165

55. Handbook of Chemistry and Physics (1983) part II 56. Schenkman S, Araujo PS, Dijkman R, Qina FH, Chaimovich H (1981) Effects of temperature and lipid composition on the serum albumin-induced aggregation and fusion of small unilamellar vesicles. Biochm Biophy Acta 649:633–641 57. Carmona-Ribeiro AM, Yoshida LS, Chaimovich H (1985) Salt effects on the stability of dioctadecyldimethylammonium chloride and sodium dihexadecyl phosphate vesicles. J Phys Chem 89:2928–2933 58. Grabowski EF, Morrison ID (1983) Particle size distributions from analyses of quase-elastic lightscattering data. In: Dahneke B (ed) Measurements of suspended particles by quase-elastic light-scattering, Wiley-Interscience, NY, p 7

Chapter 13 Use of Liposomes to Evaluate the Role of Membrane Interactions on Antioxidant Activity Salette Reis, Marlene Lúcio, Marcela Segundo, and José L.F.C. Lima Abstract Cellular membranes, which contain abundant phospholipids, such as phosphatidylcholine, are major targets subjected to the damage caused by free radicals. Cellular damage due to lipid oxidation is strongly associated with ageing, carcinogenesis and other diseases. In addition, lipid oxidation is an important deteriorative reaction in the processing and storage of lipid-containing foods. Liposomes have been used extensively as biological models for in vitro lipid oxidation studies. The resemblance between the liposomal and membrane bilayer core makes liposomes a very useful tool to investigate the significance of the antioxidant-membrane interactions for antioxidant activity. The antioxidant activity of a compound is strongly influenced by numerous factors including the nature of the lipid substrate, the hydrophilic– lipophilic balance of the antioxidant, the physical and chemical environments of the lipids, and various other interfacial interactions. Thus, compounds that are effective antioxidants in one model system or food matrix may be unsuitable in other systems. This chapter describes fluorescent probes-based methods commonly used for testing antioxidant activity in liposomes and stresses the need to combine antioxidant assays and drug-membrane interaction studies to get a better description of the antioxidants’ profile considering their location in lipid bilayer and their effect on membrane fluidity and consequently provide additional information to that obtained currently from assays performed in aqueous buffer media. Key words: Liposomes, Radical inducers, Peroxyl radicals, Hydroxyl radicals, Antioxidant activity, In vitro assays, Interactions with membranes

1. Introduction Reactive oxygen species (ROS), and reactive nitrogen species (RNS), are generated in living organisms through many pathways. Accumulation of reactive species in aerobic organisms is thought to cause oxidative damage of biological macromolecules such as DNA, proteins, and membrane lipids (1, 2). V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_13, © Humana Press, a part of Springer Science+Business Media, LLC 2010

167

168

Reis et al.

In biological systems, the oxidation of membrane and lipoprotein lipids in circulation plays a central role in many pathologic processes (3–6), especially in the context of oxidative stress-related pathologies, including cardiovascular and neurodegenerative diseases (7–9). Furthermore, oxidation of lipids in food is also of great concern, since it leads to fatty acid decomposition and development of undesirable rancid odors and flavors, with a decrease in nutritional value and food safety (10–12). Therefore, there is an increasing interest in the efficacy of antioxidant activity of the many naturally occurring molecules in food and biological systems to minimize potential oxidative damage in vivo and/or to retard the oxidation of easily oxidizable materials in foods. While the antioxidant activity of many synthetic and naturally occurring compounds has been studied in solution and such results are extremely useful to provide an index for the intrinsic antioxidative activity, they bear little relevance to the multicellular systems, in which the tested compounds have to be metabolised and transported to the “reaction sites”. According to Halliwell and Gutteridge (13), the mechanisms of antioxidant action can include: (a) suppressing reactive species formation either by inhibition of enzymes or by chelating trace elements involved in free radical production; (b) scavenging reactive oxygen species; and (c) up-regulating or protecting antioxidant defence. Methods, which involve homogeneous solutions only assess free radicalscavenging activity. However, biological systems are inherently quite heterogeneous (6). There are both hydrophilic and hydrophobic domains in these systems which cannot be reproduced in homogeneous solutions. Consequently, the effectiveness of antioxidants in food or biological systems may depend not only on their chemical interactions with components of the oxidation pathway, but also on their location and partitioning into interfacial regions of the lipid membranes (14). It can thus be inferred that when the antioxidant of biological material is to be evaluated, the type and polarity of the system used as the substrate significantly affects the activity of the antioxidant in study (15, 16) and therefore, biological membranes may be regarded as more relevant substrates for evaluating antioxidant activity in food and in biological systems (17, 18). Examples of lipid-containing substrates for the study of antioxidants include the use of low-density lipoproteins (LDL) in the in vitro LDL oxidation assay, which has become popular for assessing the antioxidant activity of wines (19). Other lipid-containing substrates, such as erythrocytes (20), primary hepatocytes (21), microsomal membrane preparations (22–25) can also be used to evaluate the in vitro antioxidant activity of foods and phenolic compounds.

Role of Membrane Interactions on Antioxidant Activity

169

In view of the complexity of biological membranes, much of the work described in the literature was performed in simplified model systems able to simulate a membrane environment (26). Model lipid oxidation systems included micelles (27, 28) and mostly lipid vesicles (liposomes). Liposomes, in particular, are relevant to the field of lipid oxidation studies for at least two reasons: (1) they constitute a simple and convenient system where the oxidation process can be reproduced with different levels of complexity allowing the analysis of water- and lipid-soluble antioxidants in the same system (29, 30) and (2) liposomes are currently being used for diagnostic and therapeutic purposes and lipid oxidation should be evaluated since it may impair their stability (29). Several studies indicated that a clear correlation exists between the study of antioxidant potency of various compounds in liposomes used as membrane models and the protective effects of these compounds against oxidation in biological systems (6). However, it is important to note that the possibility of gaining biologically relevant information on the potency of antioxidants in liposomes as membrane model systems cannot be assured without a thorough analysis of the results obtained and experimental conditions used. As a matter of fact, antioxidant effects depend on several factors, including the inducer of oxidation and its concentration; the composition, charge surface and fluidity of the liposomes. Ultimately antioxidant effects depend on the interactions of the compounds in study with the lipid membranes. Regarding these concerns, this chapter describes the use of fluorescent probes commonly used for testing antioxidant activity in liposomes and stresses the need to combine the antioxidant assays and drug-membrane interaction studies to get a better description of the antioxidants’ profile and provide additional information to that obtained currently from assays performed in aqueous buffer media. 1.1. Methods for Assessing Antioxidant Activity

The antioxidant activities of many compounds and biological fluids and extracts have been evaluated in liposome systems by various methods. In general, the oxidation is carried out in the presence and absence of an inducer and the extent of suppression of the rate of oxidation and duration of inhibition produced by antioxidant are measured. Some of the most commonly used methods for testing antioxidant activity measure the extent of oxidation from the oxygen uptake, peroxide formation, conjugated diene formation, formation of a coloured compound between TBA and malonaldehyde, chemiluminescence emission, and disappearance of fluorescence or ESR signal from a reference probe (31). Among the mentioned methods, the assays using molecular probes appear as more convenient avoiding the experimental complexity and the limitations of directly monitoring reaction kinetics of the inhibited autoxidation of lipids in the liposome model used.

170

Reis et al.

In general, the assays using probes have the following components: (a) a thermal radical generator to give a steady flux of peroxyl radicals in air-saturated solution; (b) a molecular probe (UV or fluorescence) for monitoring reaction progress; (c) antioxidant that competes with probes (substrates in this case) for the radicals and inhibits or retards the probe oxidation and (d) reaction kinetic parameters collected for antioxidant capacity quantification. The difference between the methods using probes lies mostly in the type of inducers used to initiate oxidation and the type of probes used to follow the oxidation progress which, in turn, depend on the aim of the study. Therefore, this chapter purposes the evaluation of antioxidant activity of any compounds using three different fluorescent probes specifically diphenylhexatriene propionic acid (DPH-PA); hexadecanoyl aminofluorescein (HDAF) and fluorescein. The degree of lipid oxidation is indirectly monitored by the oxidation of the probes with a consequent decay in their fluorescence intensity. The probes have different lipophilic properties, and consequently different locations (32–35) and thus, the antioxidant capacity of compounds can be related to the molecular distribution between aqueous and lipid media, and to other components of the interaction of the compounds with lipid membranes, namely, the effect in membrane surface charge; and the effect in membrane physical properties (Fig. 1). From the observation of Fig. 1 it becomes evident that the choice of a certain inducer depends on the aim of the study. If the researcher intends to study lipophilic antioxidants (e.g. a-tocopherol) or follow the oxidation of a lipophilic probe (e.g. DPH-PA) it is reasonable to choose a liposoluble radical initiator. In this way, AMVN is a synthetic azo-compound that dissociates spontaneously to form carbon-centered free radicals in the hydrophobic core of the bilayer which, in turn, produce peroxyl radicals when oxygen is present (6, 36). If the researcher intends to study hydrophilic compounds or follow the oxidation of hydrophilic probe (e.g. fluorescein) then it is preferable to choose a hydrophilic inducer like the commonly used water-soluble radical initiator, AAPH, which generates radicals in the aqueous phase of the liposomes. Metal ions are also used as hydrophilic inducers. However, some water-soluble antioxidants including ascorbic acid can become pro-oxidant in the presence of metal ions, depending on the antioxidant: metal ratio, so it is not considered desirable to use this method of generating radicals for the assessment of natural extracts where the concentration of different antioxidants will vary widely (17). If the studied compounds are soluble in the aqueous media but locate in the polar surface region of the lipid bilayers (e.g. trolox) or the oxidation is followed by a probe located both in polar and hydrophobic regions of the bilayer, it might be helpful

Role of Membrane Interactions on Antioxidant Activity

171

FLUORESCEIN HO

O

O

COO

HDAF HO

O

O

ROO

–

AAPH

Interactions with membranes: Effect in membrane surface charge Membrane partition Membrane location Effect in membrane physical properties

–

COO

DPH-PA –

AQUEOUS MEDIA

COC

HN O

LIPID MEDIA

CH3

AMVN

ROO•

Fig. 1. Schematic illustration of the principle of the assays used to evaluate the antioxidant activity against peroxyl radical (ROO•) showing the location of the fluorescent probes fluorescein, hexadecanoyl aminofluorescein (HDAF) and diphenylhexatriene propionic acid (DPH-PA) in a liposome-aqueous media; the generation of the radical in the aqueous and liposome media, and the lipid-water partition of the antioxidant studied

to perform a comparative assay with both liposoluble and hidrosoluble radical initiators. 1.2. Methods for Assessing Membrane Interactions

As it has been previously said, the efficacy of the antioxidants in vivo is determined not only by the reactivity toward a radical, but also by their interaction with membrane lipid bilayers. Several aspects of this interaction are explored in this chapter and include: location and partitioning of the antioxidant into interfacial regions in liposome systems; as well as the antioxidant mobility at the microenvironment determined by membrane physical changes and also interactions between specific membrane surface charges and antioxidants of different charge types (31, 37–39) (Fig. 1).

1.2.1. Assessing Location and Partitioning of the Antioxidant into Interfacial Regions in Liposome Systems

The rate of scavenging depends on the chemical structure of the inhibitor, but it may also be promoted by concentrating the scavenger in the neighbourhood of the target lipid. It is for the latter reason that the precise positioning, orientation and partition of many of the currently available antioxidant compounds within

172

Reis et al.

the membrane, is also an important factor contributing to their scavenger effectiveness so when both antioxidant and free radical coincide in the same place, the antioxidant effect is greater (38). For a similar reason, the partition coefficient has also great importance because it determines the accessibility and interaction of antioxidants to the lipid radicals within the liposome (15). The type of radicals also determines the rate of scavenging, which will, in turn, be dependent on the antioxidant location. Peroxyl radicals are known to be highly polarized. As a consequence, such radicals formed initially in the hydrophobic region of membranes are expected to diffuse rapidly to the polar head group region near the aqueous phase where a water-soluble antioxidant could terminate them – diffusion-trapping mechanism (37). It is likely that compounds that are mostly located near the surface of membranes, where aqueous peroxyl radicals are easily trapped and more readily accessible, have better antioxidative effect in liposomes. On the other hand,•OH radicals, although are generated in the aqueous phase, diffuse easily across the bilayer and react with acyl chains causing their oxidation. In this case, the efficiency of an antioxidant to interrupt the oxidative sequence will be greater when located at the hydrophobic region of the bilayer, as the acyl chains are (38). The partition coefficient (Kp) can be determined using derivative spectroscopy and fluorescence-quenching techniques, which permit evaluating and comparing the extent of penetration and/ or interaction of antioxidants with membrane phospholipids. Moreover, fluorescence-quenching studies and zeta potential measurements can be used to preview the location of the antioxidants within the bilayer (40). 1.2.2. Assessing Membrane Physical Changes Induced by the Antioxidant

Antioxidants can act either by scavenging free radicals or by modifying their propagation across cell membranes. Several authors have demonstrated that free radical scavenger antioxidants could interact more efficiently with lipid radicals in a disordered lipid bilayer (41, 42). Then, the fluidizing effect of compounds could favor the known antioxidant capability and scavenging characteristics of these compounds (41, 42). Moreover, an agent effective against free radical oxidation may also act by altering the cell membrane in such a way that the free radical chain is unable to propagate efficiently (39). Modifications of membrane biophysical properties elicited by the antioxidants under study can be evaluated by fluorescence measurements of anisotropy (40).

1.2.3. Assessing Interactions Between Membrane Surface and Antioxidant Charge

It is important to consider the possible effect of conditions (e.g. pH) and the polarity of antioxidants as well as the charge type of the target membranes when employing antioxidants for medicinal purposes (37). According to this, zeta-potential measurements

Role of Membrane Interactions on Antioxidant Activity

173

can be used to evaluate changes in membrane surface and thus obtain more information about the antioxidant’s binding behaviour (40).

2. Materials 2.1. Preparation of Liposomes

1. Egg l-a-phosphatidylcholine (EPC) (Sigma, St. Louis, MO) is stored at −20°C. 2. Hydration buffer: 10 mM Hepes, pH 7.4, 0.1 M NaCl. Store at room temperature (see Notes 1 and 2). 3. Chloroform and methanol. 4. Stainless-steel extruder (Lipex Biomembranes, Vancouver, BC, Canada), equipped with a circulating water bath. 5. Polycarbonate filters of three different sizes: 400, 250, and 100 nm (Nucleopore, Pleasanton, CA). 6. Malvern ZetaSizer 5000 (Malvern Instruments, Malvern, Worcestershire, UK) for quasielastic light scattering analysis.

2.2. Methods for Assessing Antioxidant Activity 2.2.1. Studies Using Fluorescein as a Fluorescent Probe in Aqueous and Liposome Media

1. Phosphate buffer: 75 mM KH2PO4, 0.1 M NaCl. Adjust pH to 7.4 with 10% KOH. Store at room temperature (see Note 2). 2. Fluorescein stock solution is prepared by dissolution of 3.77 mg fluorescein sodium salt (Aldrich, St. Louis, MO) in 25.00 mL of phosphate buffer to achieve a concentration of 0.42 mM and stored in dark conditions at 4°C. The fluorescein stock solution at such conditions can last several months. A fresh fluorescein working solution is prepared daily at 480 nM (studies in liposome media) and at 576 nM (studies in aqueous media) by dilution of the stock solution in the phosphate buffer (see Note 3). 3. The radical inducer 2, 2¢-azobis(2-amidinopropane) dihydrochloride (AAPH) (Fluka, St. Louis, MO) is completely dissolved in the phosphate buffer at 188 mM (510 mg of AAPH in 10.00 mL of buffer when studies are performed in liposome media) and at 36 mM (98 mg of AAPH in 10.00 mL of buffer when studies are performed in aqueous media) and is kept in an ice bath (see Notes 4 and 5). 4. Standard preparation: 0.5 g of Trolox (Carboxy-2, 5, 7, 8-tetramethyl-6-chromanol) (Fluka, St. Louis, MO) is dissolved in 10.00 mL of phosphate buffer (pH 7.4) to give a 200 mM stock solution. The stock solution is diluted with the same phosphate buffer to trolox working solutions in the range of 0–100 mM (see Note 6). 5. Sample preparation: pure compounds are directly dissolved in phosphate buffer (pH 7.4) to give a 200 mM stock solution.

174

Reis et al.

The stock solution is diluted with the same phosphate buffer to samples working solutions in the range of 0–100 mM. 6. 96-well polystyrene microplates International Inc, Bridgeport, NJ).

and

7. A FL600 microplate fluorescence Instruments, Inc., Winooski, VT). 2.2.2. Studies Using 5-Hexadecanoyl Aminofluorescein as a Fluorescent Probe

covers reader

(VWR (Bio-Tek

1. Phosphate buffer: 75 mM KH2PO4, 0.1 M NaCl. Adjust pH to 7.4 with 10% KOH. Store at room temperature (see Note 2). 2. HDAF stock solution is prepared by dissolution of 1.46 mg of 5-N-hexadecanoyl-aminofluorescein (HDAF; Aldrich, St. Louis, MO) in 10.00 mL of dimethylsulphoxide (DMSO) to achieve a concentration of 0.293 mM and stored in dark conditions at −20°C. 3. The radical inducer 2, 2¢-azobis(2-amidinopropane) dihydrochloride (AAPH) (Fluka, St. Louis, MO) is completely dissolved in the phosphate buffer at 545.5 mM (1.48 g of AAPH in 10.00 mL of buffer) and is kept in an ice bath (see Notes 4 and 5). 4. Standard preparation: 1 g of Trolox (Carboxy-2, 5, 7, 8-tetramethyl-6-chromanol) (Fluka, St. Louis, MO) is dissolved in 10.00 mL of phosphate buffer (pH 7.4) to give a 400 mM stock solution. The stock solution is diluted with the same phosphate buffer to trolox working solutions in the range of 0–200 mM. 5. Sample preparation: pure compounds are directly dissolved in phosphate buffer (pH 7.4) to give a 400 mM stock solution. The stock solution is diluted with the same phosphate buffer to samples working solutions in the range of 0–200 mM. 6. 96-well polystyrene microplates International Inc, Bridgeport, NJ).

and

covers

(VWR

7. A FL600 microplate fluorescence reader (Bio-Tek Instruments, Inc., Winooski, VT). 2.2.3. Studies Using Diphenylhexatriene Propionic Acid as a Fluorescent Probe

1. Hepes buffer: 10 mM Hepes, pH 7.4, 0.1 M NaCl. Store at room temperature (see Note 2). 2. DPH-PA stock solution is prepared by dissolution of 2.92 mg of 3-(p-(6-phenyl)-1,3,5-hexatrienyl)-phenylpropionic acid (DPH-PA; Invitrogen Corporation, Carlsbad, California, USA) in 5.00 mL of tetrahydrofuran (THF) to achieve a concentration of 1.92 mM and stored in dark conditions at −20°C (see Note 7). 3. The radical inducer 2, 2¢-azobis(2-amidinopropane) dihydrochloride (AAPH) (Fluka, St. Louis, MO) is completely dissolved in the phosphate buffer at 67.5 mM (183 mg of AAPH in 10.00 mL of buffer and is kept in an ice bath (see Notes 4 and 5)).

175

Role of Membrane Interactions on Antioxidant Activity

4. Standard preparation: 1 g of Trolox (Carboxy-2, 5, 7, 8tetramethyl-6-chromanol) (Fluka, St. Louis, MO) is dissolved in 10.00 mL of phosphate buffer (pH 7.4) to give a 400 mM stock solution. The stock solution is diluted with the same phosphate buffer to trolox working solutions in the range of 0–200 mM. 5. Sample preparation: pure compounds were directly dissolved in phosphate buffer (pH 7.4) to give a 1.35 mM stock solution. The stock solution was diluted with the same phosphate buffer to samples working solutions in the range of 0–675 mM. 6. Steady-state spectrofluorimeter (Perkin-Elmer LS 50B). 2.3. Methods for Assessing Membrane Interactions

1. Materials for liposome preparation (see Subheading 2.1) 2. Samples: pure compounds are directly dissolved in Hepes buffer (pH 7.4). 3. DPH stock solution is prepared by dissolution of 5.8 mg of 1,6-diphenyl-1,3,5-hexatriene (DPH; Invitrogen Corporation, Carlsbad, California, USA) in 5.00 mL of tetrahydrofuran (THF) to achieve a concentration of 5 mM and stored in dark conditions at −20°C (see Note 7). 4. UV-VIS spectrophotometer (Perkin-Elmer Lambda 45, Waltham, Massachusetts, USA). 5. Steady-state spectrofluorimeter Waltham, Massachusetts, USA).

(Perkin-Elmer

LS

50B,

6. Malvern ZetaSizer 5000 (Malvern Instruments, Malvern, Worcestershire, UK) for quasielastic light scattering analysis.

3. Methods 3.1. Preparation of Liposomes

1. Liposomes suspension is prepared by dissolving EPC in chloroform/methanol (9:1) (10–20 mg lipid per mL of organic solvent) in a round-bottomed flask. 2. Solvent is removed under reduced pressure on a rotary evaporator with the water bath set at 30°C. N2 is introduced to re-establish atmospheric pressure and the flask covered in aluminium foil. A pump is then used to keep the flask under a vacuum of <0.5 mmHg for at least 3 h. After evaporation, the vacuum is released while introducing N2. A dry thin lipid film is obtained in the round-bottomed flask walls. 3. Hydration of the thin lipid film is made by addition of suitable hydration buffer (see Subheading 2.1) for approximately 30 min at 25°C (see Note 8). The proportion lipid: buffer should be 2.1 mg/mL, which gives a lipid concentration of 3 mM (if the relative molecular mass of EPC is taken as 700).

176

Reis et al.

4. The flask is vortexed at room temperature (see Note 8) for 20 min, producing a homogeneous white suspension of multilamellar liposomes (MLV). The multilamellar suspension is then shaken while being sonicated for 30 s in an ultrasonic bath. This ensures complete recovery of the lipid from the flask walls (see Note 9). 5. The MLV suspension is transferred to an extrusion device maintained at 25°C by a circulating water bath to form large unilamellar vesicles (LUV) (see Note 8). Prior to extrusion through the final pore size, MLV suspensions are disrupted by pre-extruding the suspension through a larger pore size (typically, five times through 400 nm pore size, then five times through 250 nm pore size) (see Notes 10 and 11). Finally, extrude the lipid suspension ten times through 100 nm pore size (see Note 12). 6. Size distribution of the extruded vesicles should be determined by quasielastic light scattering analysis using a heliumneon laser (633 nm) as a source of incident light, operating at a scattering angle of 90° and a temperature of 25°C. Extrusion through filters with 100 nm pores typically yields large, unilamellar vesicles (LUV) with a mean diameter of 120–140 nm. Mean particle size depends on lipid composition but is quite reproducible from batch to batch. 3.2. Methods for Assessing Antioxidant Activity

The methods for assessing antioxidant activity described in this chapter are based on the decay in probes fluorescence intensity monitored over time. In general, samples, controls, and standard (Trolox of four or five different concentrations for construction of a standard curve) are mixed with the fluorescent probe in aqueous or liposome media and incubated at constant temperature (37°C) before AAPH solution is added to initiate the reaction. As the reaction progresses, the probe is oxidized and fluorescence intensity decreases. In the presence of antioxidant, the fluorescence decay is inhibited. Data obtained are converted to relative fluorescence values, by dividing the fluorescence intensity at a given time by the fluorescence intensity at 0 min. The ability of the compounds studied to act as antioxidants is analysed by determining their IC50 values, which are defined as the concentration (in mM) of each compound required to obtain a ratio defined by (1) equal to 50%: AUCantioxidant/trolox − AUCblank AUCblank

× 100

(1)

where AUC is the area under the curve obtained for the sample and blank assays.

Role of Membrane Interactions on Antioxidant Activity 3.2.1. Studies Using Fluorescein as a Fluorescent Probe in Aqueous and Liposome Media

177

1. For the assays in aqueous media, a 96 well plate is prepared (maximum well volume = 320 mL) with a reaction mixture (300 mL) containing the following volumes of reagents in potassium phosphate buffer (75 mM, I = 0.1 M, pH = 7.4): 25 mL of fluorescein working solution (see Subheading 2.1, item 1); 150 mL of samples or standard (trolox) working solutions. Blanks are made with the radical initiator and without the sample or standard replacing sample or standard by 150 mL of potassium phosphate buffer (see Note 13). Controls are made without the radical initiator replacing it by potassium phosphate buffer: 25 mL of fluorescein working solution; 150 mL of samples or standard (trolox) working solutions and 125 mL of potassium phosphate buffer (75 mM, I = 0.1 M, pH = 7.4) instead of AAPH solution (see Note 14). For the assays in liposome media, a 96 well plate is prepared (maximum well volume = 320 mL) with a reaction mixture (300 mL) containing the following volumes of reagents in potassium phosphate buffer (75 mM, I = 0.1 M, pH = 7.4): 80 mL of LUV suspension (see Subheading 3.1); 30 mL of fluorescein working solution (see Subheading 2.1, item 1); 150 mL of samples or standard (trolox) working solutions. Blanks are made replacing sample by 150 mL of potassium phosphate buffer (see Note 13). Controls are made without the radical initiator replacing it by potassium phosphate buffer: 80 mL of LUV; 30 mL of fluorescein-working solution; 150 mL of samples or standard (trolox) working solutions and 40 mL of potassium phosphate buffer (75 mM, I = 0.1 M, pH = 7.4) instead of AAPH solution (see Note 14). 2. The plate was covered with a lid and incubated in the preheated (37°C) microplate reader (Synergy HT, Bio-Tek, Winooski, VT, USA) for 10 min with a 3 min shaking during this time. 3. Incubation is followed by the addition of radical initiator solution (see Subheading 2.1, item 1); to all wells (samples, standard trolox and blanks) except for the wells containing controls: 125 mL of AAPH solution (in assays in aqueous media) or 40 mL of AAPH solution (in assays in liposome media). Thus, the final concentration for each well in aqueous media is: 48 nM of fluorescein; increasing concentrations of samples or standard trolox (0–50 mM) and 15 mM of AAPH. In liposome media, the final concentrations for each well is: 800 mM of LUV suspension; 48 nM of fluorescein; increasing concentrations of samples or standard trolox (0–50 mM) and 25 mM of AAPH. 4. After addition of AAPH, the plate is immediately transferred to the plate reader (see Note 15), and the fluorescence intensity was

178

Reis et al.

measured every minute for 180 min at previous conditions (pH 7.4, 37°C) with fluorescence filters for an excitation wavelength of 485 ± 20 nm and an emission wavelength of 528 ± 20 nm. The plate reader was controlled by software KC4 3.0. 3.2.2. Studies Using 5-Hexadecanoyl Aminofluorescein as a Fluorescent Probe

1. Liposomes labelled with HDAF probe are prepared in a round-bottomed flask by dissolving EPC in chloroform/ methanol (9:1) (10–20 mg lipid per mL of organic solvent) and adding an aliquot of the probe to the organic phase so that the ratio lipid to probe is 6,000:1 (see Note 16). Typically, 21 mg of EPC and 20 mL of HDAF stock solution are added to the chloroform/methanol (9:1) mixture and dried together under reduced pressure on a rotary evaporator with the water bath set at 30°C as previously described in preparation of liposomes method (see Subheading 3.1). 2. Hydration of the thin lipid film is made by addition of potassium phosphate buffer (75 mM, I = 0.1 M, pH = 7.4) as hydration buffer (see Subheading 2.1, item 1) for approximately 30 min at 25°C (see Note 8). The proportion lipid: buffer should be 2.1 mg/mL, which gives a lipid concentration of 3 mM (if the relative molecular mass of EPC is taken as 700) and a probe concentration of 0.5 mM. After labelling the lipid with the probe and hydration of the thin lipid film, the preparation of LUV follows the procedure previously described in the preparation of liposomes method (see Subheading 3.1). 3. A 96 well plate is prepared (maximum well volume = 320 mL) with a reaction mixture (300 mL) containing: 80 mL of HDAF labelled LUV and 110 mL of samples or standard (trolox) working solutions. Blanks are made with the radical initiator and without the sample or standard replacing sample or standard by 110 mL of potassium phosphate buffer (see Note 13). Controls are made without the radical initiator replacing it by potassium phosphate buffer: 80 mL of HDAF labelled LUV; 110 mL of samples or standard (trolox) working solutions and 110 mL of potassium phosphate buffer (75 mM, I = 0.1 M, pH = 7.4) instead of AAPH solution (see Note 14). 4. The plate is covered with a lid and incubated in the preheated (37°C) microplate reader (Synergy HT, Bio-Tek, Winooski, VT, USA) for 10 min with a 3 min shaking during this time. 5. Incubation is followed by the addition of 110 mL of radical initiator solution (see Subheading 2.1, item 1); to all wells (samples, standard trolox and blanks) except for the wells containing controls. Thus, the final concentrations for each well in aqueous media are: 800 mM of LUV suspension 0.13 mM of HDAF; increasing concentrations of samples or standard trolox (0–200 mM) and 200 mM of AAPH.

Role of Membrane Interactions on Antioxidant Activity

179

6. After addition of AAPH, the plate is immediately transferred to the plate reader (see Note 15), and the fluorescence intensity was measured every minute for 360 min at previous conditions (pH 7.4, 37°C) with fluorescence filters for an excitation wavelength of 485 ± 20 nm and an emission wavelength of 521 ± 20 nm. The plate reader was controlled by software KC4 3.0. 3.2.3 Studies Using Diphenylhexatriene Propionic Acid (DPH-PA) as a Fluorescent Probe

1. Liposomes labelled with DPH-PA probe are prepared in a round-bottomed flask by dissolving EPC in chloroform/ methanol (9:1) (10–20 mg lipid per mL of organic solvent) and adding an aliquot of the probe to the organic phase so that the ratio lipid to probe was 300:1 (see Note 16). Typically, 10 mg of EPC and 25 mL of DPH-PA stock solution are added to the chloroform/methanol (9:1) mixture and dried together under reduced pressure on a rotary evaporator with the water bath set at 30°C as previously described in the preparation of liposomes method (see Subheading 3.1). 2. Hydration of the thin lipid film is made by addition of Hepes buffer (10 mM, I = 0.1 M, pH = 7.4) as hydration buffer (see Subheading 2.1, item 1) for approximately 30 min at 25°C (see Note 8). The proportion lipid: buffer should be 1.05 mg/mL. After labelling the lipid with the probe and hydration of the thin lipid film, the preparation of LUV follows the procedure previously described in the preparation of liposomes method (see Subheading 3.1). 3. An aliquot of 1 mL of the LUV suspension, containing 1.5 mM of EPC and 5 mM of DPH-PA, is incubated in a fluorimetric cuvette placed in a thermostatted holder (37°C) for 5 min with continuous stirring and away from light. An aliquot (400 mL) of samples or standard (trolox) working solutions is added to the mixture and incubated for 5 min. Blanks are made with the radical initiator and without the sample or standard replacing sample or standard by 400 mL of Hepes buffer (see Note 13). Controls are made without the radical initiator replacing it by Hepes buffer: 1 mL of DPH-PA labelled LUV; 400 mL of samples or standard (trolox) working solutions and 400 mL of Hepes buffer (10 mM, I = 0.1 M, pH = 7.4) instead of AAPH solution (see Note 14). 4. Incubation is followed by the addition of 400 mL of radical initiator solution (see Subheading 2.1, item 1); to the cuvette (in the assays of samples, standard trolox and blanks) except for the control assays (see Note 17). Thus, the final concentrations for each assay are: 800 mM of LUV suspension 2.7 mM of DPH-PA; increasing concentrations of trolox and samples (0–150 mM) and 15 mM of AAPH.

180

Reis et al.

5. After addition of AAPH, the fluorescence measurements are immediately started (see Note 15), and the fluorescence intensity is monitored every minute for 60 min at previous conditions (pH 7.4, 37°C) using a steady-state spectrofluorimeter (Perkin-Elmer LS 50B) at excitation and emission wavelengths of 399 and 435 nm, respectively. 3.3. Methods for Assessing Membrane Interactions 3.3.1. Determination of Lipid/Water Partition Coefficient by Derivative Spectrophotometry

The extent of interaction of a solute with a microheterogeneous system is evaluated in a quantitative way from its partition coefficient. The octanol/water biphasic system is traditionally used to evaluate partition coefficients, which are then extrapolated to biomembrane/aqueous phase systems, this being an oversimplification. Instead, it is advisable to use liposomes as biomembranes models and quantify the concentration of the solute in one of the phases (aqueous or lipidic), or both. UV–Vis absorption spectrophotometric techniques are usually based on the change of an absorption parameter upon incorporation of compounds into membranes, which permits the determination of partition coefficient values from the following equation (42, 43): Abs f = Abs a +

(Absm − Absa )K p (L ) Vϕ 1 + K p (L ) V ϕ

(2)

where Absa is the absorbance of the compound in buffer solution without lipid; Absf is the absorbance of the compound in liposome suspension with different lipid concentrations; Absm is the absorbance calculated assuming that all compound is membrane bound; [L] is the lipid concentration and Vf is the lipid molar volume (for EPC, Vf = 0.688 L/mol and the mean molecular weight is 700 (44)). Despite the overall simplicity of (2), its practical application is usually limited to systems with low light scattering background signals. When the direct application of this spectrophotometric method is prevented by high background signals, caused by the presence of liposomes or cells, the problem can be minimized by the use of second derivative spectrophotometry with Absf replaced by Df = (dnAbs)/(dln). The second derivative spectra are thus calculated from the recorded absorption spectra of a compound in the presence of increasing lipid concentration after blank subtraction. Then, the partition coefficient is calculated by fitting (2) to the experimental data (Df versus [L]), using a non-linear regression method, where the adjustable parameter is Kp (44). Second derivative spectrophotometry technique has the additional advantage of analysing the signals originated from both phases (aqueous and lipidic) with no demand for their physical separation, which may be laborious and may result in equilibria perturbation.

Role of Membrane Interactions on Antioxidant Activity

181

The procedure for the determination of lipid/water partition coefficient of an antioxidant by derivative spectrophotometry, is the following: 1. LUV suspension with a final concentration of 1.5 mM is prepared by the previously described thin lipid film hydration method (see Subheading 3.1) 2. Samples (total volume of 1,800 mL) are prepared in 2 mL Eppendorfs containing in Hepes buffer (Hepes: 10 mM, I = 0.1 M, pH 7.4): 80 mL of compound in study and increasing volumes (0–1,200 mL) of LUV suspension. This gives a series of samples containing a fixed concentration of studied compound (see Note 18) and increasing concentrations of LUV suspension (0–1,000 mM). 3. Corresponding reference suspensions are prepared identically as samples, but without compounds in study which are replaced by Hepes buffer (10 mM, I = 0.1 M, pH 7.4). 4. The resulting suspensions of samples and references are vortexed and incubated in the dark for 1 h. 5. The absorbance spectra of each sample and reference are recorded in a UV spectrometer equipped with a constant temperature cell holder at 25.0 ± 0.1°C in 1-cm path length cuvettes with adequate slits. 3.3.2. Assessing Membrane Location by Fluorescence Quenching

Fluorescence-quenching studies can be used to reveal the membrane location of compounds, which will behave as quenchers of a fluorescent probe inserted in the lipid bilayer (45). The extent of quenching is expressed by the Stern-Volmer equation (45): I0 app − 1 = K SV [Q ] I

(3)

app

where K SV is the apparent Stern-Volmer constant, [Q] is the total quencher concentration, I0 and I are the fluorescence intensity values in the absence and in the presence of the quencher (compound which location is being studied). High values of Stern-Volmer constants mean that the location of the quenching compound is very similar to the location of the fluorescent probe. According to this, it is required that the molecular location of the probe within membranes is known with certainty. This chapter describes the use of DPH probe (see Note 19) to assess membrane location of antioxidants using the following procedure: 1. Liposomes labelled with DPH probe are prepared in a roundbottomed flask by dissolving EPC in chloroform/methanol (9:1) (10–20 mg lipid per mL of organic solvent) and adding an aliquot of the probe to the organic phase so that the ratio lipid to probe was 300:1 (see Note 16). Typically, 10 mg of EPC and 9.5 mL of DPH-PA stock solution are added to the

182

Reis et al.

chloroform/methanol (9:1) mixture and dried together under reduced pressure on a rotary evaporator with the water bath set at 30°C as previously described in the preparation of liposomes method (see Subheading 3.1). 2. Hydration of the thin lipid film is made by addition of Hepes buffer (10 mM, I = 0.1 M, pH = 7.4) as hydration buffer (see Subheading 2.1, item 1) for approximately 30 min at 25°C (see Note 8). The proportion lipid: buffer should be 1.05 mg/ mL. After labelling the lipid with the probe and hydration of the thin lipid film, the preparation of LUV follows the procedure previously described in the preparation of liposomes method (see Subheading 3.1). The DPH-labelled LUV have a final EPC concentration of 1.5 mM and DPH of 5 mM. 3. Samples (total volume of 1,800 mL) are prepared in 2 mL Eppendorfs containing: 600 mL of DPH-labelled LUV suspension and increasing volumes (0–1,200 mL) of Hepes buffer solutions (Hepes: 10 mM, I = 0.1 M, pH 7.4) of compounds in study. This gives a series of samples containing a fixed concentration of DPH-labelled LUV suspension (500 mM of lipid and 1.7 mM of DPH) and increasing concentrations of studied compounds (see Note 18). 4. Corresponding reference suspension is prepared identically as sample, but without compound in study which is replaced by Hepes buffer (10 mM, I = 0.1 M, pH 7.4). 5. The resulting suspensions of reference and samples are vortexed and incubated in the dark for 1 h. 6. The fluorescent spectra of each reference and sample are recorded in a steady-state fluorescence spectrometer equipped with a constant temperature cell holder at 25.0 ± 0.1°C in 1-cm path length cuvettes with adequate excitation and emission slits. Excitation and emission wavelength is set to 361 and 432 nm, respectively (see Note 20). 3.3.3. Assessing Membrane Physical Changes Induced by the Antioxidant

Steady-state fluorescence anisotropy is a technique that assesses the range of rotational motion of the membrane-associated fluorescent probe, in this case DPH, during the lifetime of its excited electronic state (46). Using this technique, DPH probe is automatically excited with vertically polarized light and resulting fluorescence intensities are recorded with the analyzing polarizer oriented parallel (IVV) and perpendicular (IVH) to the excitation polarizer allowing the determination of steady-state fluorescence anisotropy for labelled samples by the following equation (45): rss =

I VV − GI VH I VV + 2GI VH

(4)

Role of Membrane Interactions on Antioxidant Activity

183

where G = IHV/IHH is the grating correction factor (45).The greater the extent of probe rotation during its excited state lifetime, the smaller will be the observed fluorescence anisotropy (rss), to the extent that rss = 0 for complete DPH reorientation. The probe rotational motions are tightly coupled to the phospholipid fatty acid chain fluctuations, which provide a measure of membrane “fluidity” (45). To assess membrane physical changes induced by antioxidant, liposomes labelled with DPH probe are incubated with increasing concentrations of the antioxidant tested and the steady-state fluorescence anisotropy is measured. The procedure is the following: 1. Liposomes labelled with DPH probe, reference and samples are prepared in the same way as described for fluorescencequenching measurements method (see Subheading 3.3.2). 2. The resulting suspensions of reference and samples are vortexed and incubated in the dark for 1 h. 3. The steady-state anisotropy of each reference and sample is recorded in a steady-state fluorescence spectrometer equipped with a constant temperature cell holder at 25.0 ± 0.1°C in 1-cm path length cuvettes with adequate excitation and emission slits. Excitation and emission wavelength is set to 361 and 432 nm, respectively (see Note 20). 3.3.4. Assessing Interactions Between Membrane Surface and Antioxidant Charge

1. LUV suspension with a final concentration of 1.5 mM is prepared by the previously described thin lipid film hydration method (see Subheading 3.1). 2. Samples (total volume of 1,800 mL) are prepared in 2 mL Eppendorfs containing in Hepes buffer (10 mM, I = 0.1 M, pH 7.4): 600 mL of LUV suspension and increasing volumes (0–1,200 mL) of Hepes buffer solutions (10 mM, I = 0.1 M, pH 7.4) of compounds in study. This gives a series of samples containing a fixed concentration of LUV suspension (500 mM of EPC) and increasing concentrations of studied compounds (see Note 18). 3. Corresponding reference suspension is prepared identically as samples, but without compounds in study which are replaced by Hepes buffer (10 mM, I = 0.1 M, pH 7.4). 4. The resulting suspensions of samples and references are vortexed and incubated in the dark for 1 h. 5. The zeta-potential (z-potential) values of the membrane vesicles, with and without incorporated compound (samples and reference) are determined at pH 7.4 (Hepes buffer) at 25.0 ± 0.1°C, by quasi-elastic light scattering analysis with a 90° scattering angle. The values for the viscosity and refractive index are taken as 0.890 and 1.330 cP, respectively.

184

Reis et al.

4. Notes 1. Hepes buffer interferes with fluorescent probes fluorescein and its derivative 5-hexadecanoyl aminofluorescein (HDAF). Whenever using these probes, prepare the liposomes with phosphate buffer: 75 mM KH2PO4 (Merck, Darmstadt, Germany), 0.1 M NaCl. Adjust pH to 7.4 with 10% KOH. 2. All water used for buffer preparations is double-deionised water (conductivity smaller than 0.1 mS/cm). Buffer solutions are typically filtered using standard MF-Millipore membrane filters (0.45 mm) (Millipore, Bedford, MA). 3. The fluorescence of fluorescein is highly pH dependent and thus pH of the solutions must be carefully monitored. 4. AAPH powder should be stored at room temperature in a desiccator. Buy small bottles as it may decline in quality (oxidation induction requires higher concentrations of AAPH) after opening. 5. AAPH is an azo-compound that undergoes spontaneous decomposition producing carbon-centered radicals, which in turn produce peroxyl radicals when oxygen is present, with a rate primarily determined by temperature. Therefore, it is fundamental that the AAPH solution is kept in the ice bath before and after use. AAPH solution should be discarded within 8 h. 6. Trolox has advantages over other active antioxidants (e.g. vitamin E), which are only lipid-soluble. For example, Trolox does not have to be incorporated into the lipid membrane by solvent extraction and co-evaporation methods; it can be added directly to the intact system. This makes it convenient for studies on natural biological systems and for quantitative studies on model systems (47). However, vitamin E can also be used as a lipophilic standard compound and in these cases, vitamin E is co-dissolved with the lipids in chlorophorm: methanol during the preparation of the thin lipid film. 7. Long-term storage of stock solutions of DPH-PA and DPH in THF is not recommended because of possible peroxide formation in that solvent. 8. The lipid suspension should be kept above the phase transition temperature (Tm) of the lipid during hydration and extrusion. Attempts to extrude below the Tm will be unsuccessful as the membrane has a tendency to foul with rigid membranes which cannot pass through the pores. In the case of EPC, it is possible to work at room temperature inasmuch as the transition temperature for EPC ranges

Role of Membrane Interactions on Antioxidant Activity

185

from −15 to −7°C (29). If other lipids are used, care should be taken to work always with a controlled temperature well above the Tm of the lipid used. 9. In some cases, if the complete recovery of the lipid from the flask walls is difficult, it might be useful to include glass beads while shaking the multilamellar suspension. 10. MLV suspensions are disrupted either by several freeze-thaw cycles or by pre-extruding the suspension through a larger pore size. The later process is preferred to the freeze–thawing procedure. Indeed, liposome bilayers are damaged by internal ice formation during freezing. Damaged bilayers will reassemble due to the “hydrophobic effect” and form new liposomes possibly of a different size, which may involve fusion with other liposomes. Each freeze–thaw cycle would provide an opportunity for this process to occur, both in previously unaffected liposomes and those damaged by a previous freeze–thaw cycle. 11. Extrusion is a technique in which the lipid suspension is forced, under inert (N2) atmosphere, through a polycarbonate filter with a defined pore size to yield LUV having a diameter near the pore size of the filter used. 12. To prevent the fouling of membranes and improve the homogeneity of the size distribution of the final suspension, the use of two-stacked filters is recommended. 13. Care must be taken to ensure homogeneity of each dilution by thorough mixing at each stage of addition of solutions through repeated aspiration and dispensing volumes with micropipette. To avoid human errors, it is advisable to use an eight-channel micropipette. 14. Controls should be prepared before addition of AAPH to reaction mixtures and are important to evaluate the photostability of the fluorescent probe. Fluorescein in controls is exposed to excitation light of the assay (491 nm) in the absence of AAPH over a 180 min period and there should be no showing of any significant fluorescence intensity changes over the time of the assay meaning that fluorescein, under such conditions, is photostable. 15. AAPH is the radical initiator, and the oxidation starts immediately after its addition to reaction mixture. Therefore, this initiator must be the last reagent added and its addition must be as quick as possible following immediate start of the measurement. 16. The ratio of lipid to probe should be always greater than 100:1 to prevent changes in the structure of the liposomes (40).

186

Reis et al.

17. While stirring, the fluorescence time scan is started simultaneously with the addition of radical initiator solution with a 500 mL syringe through the hole located at the top of the sample compartment. This procedure avoids the exposure of sample to light while ensuring a better homogeneity of the initiator with the sample mixture. 18. The concentrations of studied compounds should be used in the range over which Beer’s law is obeyed. 19. DPH probe was chosen because of its advantageous spectral and structural properties. The high extinction coefficient and fluorescence quantum yield in non-aqueous solvents make it possible to use very small amounts of the probe in biological studies, thus minimizing perturbations and making DPH particularly attractive for obtaining dynamic and structural information about membrane bilayers (48). Additionally, the knowledge of molecular location of DPH makes it a commonly employed probe to explore the membrane hydrocarbon chain region without a direct and significant influence on membrane properties close to the membrane-water interface (34, 49). 20. Fluorescence values are corrected for light scattering contributions by subtraction of intensities from unlabeled samples at the same conditions. These contributions are always negligible (less than 0.5%). References 1. Halliwell B, Gutteridge JMC (1990) Role of free radicals and catalytic metal ions in human disease: an overview. In: Packer L and Glazer AN (eds) Methods in enzymology. Academic, San Diego, NY 2. Davies MJ, Dean RT (1997) The pathology of protein oxidation. In: Radical-mediated protein oxidation. Oxford Science Publications, Oxford, UK 3. Mattson MP (2004) Metal-catalyzed disruption of membrane protein and lipid signaling in the pathogenesis of neurodegenerative disorders. Ann N Y Acad Sci 1012:37–50 4. Poon HF, Calabrese V, Scapagnini G, Butterfield DA (2004) Free radicals: key to brain aging and heme oxygenase as a cellular response to oxidative stress. J Gerontol A Biol Sci Med Sci 59:478–493 5. Schroepfer GJ (2000) Oxysterols: modulators of cholesterol metabolism and other processes. Physiol Rev 80:361–554 6. Schnitzer E, Pinchuk I, Lichtenberg D (2007) Peroxidation of liposomal lipids. Eur Biophys J 36:499–515

7. van Ginkel G, Sevanian A (1994) Lipid peroxidation-induced membrane structural alterations. Methods Enzymol 233:273–288 8. Mason RP, Walter M, Mason P (1997) Effect of oxidative stress on membrane structure: small-angle X-ray diffraction analysis. Free Radic Biol Med 23:419–425 9. Spiteller G (2003) Are lipid peroxidation processes induced by changes in the cell wall structure and how are these processes connected with diseases? Med Hypotheses 60:69–83 10. Al-Ismail KM, Aburjai T (2004) Antioxidant activity of water and alcohol extracts of chamomile flowers, anise seeds and dill seeds. J Sci Food Agric 84:173–178 11. Viljanen K, Kylli P, Kivikari R, Heinonen M (2004) Inhibition of protein and lipid oxidation in liposomes by berry phenolics. J Agric Food Chem 52:7419–7424 12. Diaz M, Decker EA (2004) Antioxidant mechanisms of caseinophosphopeptides and casein hydrolysates and their application in ground beef. J Agric Food Chem 52:8208–8213

Role of Membrane Interactions on Antioxidant Activity 13. Halliwell B, Gutteridge JMC (1998) Free radicals in biology and medicine. Oxford University Press, Oxford, UK 14. Baublis A, Decker EA, Clydesdale FM (2000) Antioxidant effect of aqueous extracts from wheat based ready-to-eat breakfast cereals. Food Chem 68:1–6 15. Hassimotto NMA, Genovese MI, Lajolo FM (2005) Antioxidant activity of dietary fruits, vegetables, and commercial frozen fruit pulps. J Agric Food Chem 53:2928–2935 16. Yi O, Jovel EM, Towers GHN, Wahbe TR, Cho D (2007) Antioxidant and antimicrobial activities of native Rosa sp from British Columbia. Canada Int J Food Sci Nutr 58: 178–189 17. Roberts WG, Gordon MH (2003) Determination of the total antioxidant activity of fruits and vegetables by a liposome assay. J Agric Food Chem 51:1486–1493 1 8. Yen WJ, Chang LW, Duh PD (2005) Antioxidant activity of peanut seed testa and its antioxidative component, ethyl protocatechuate. LWT-Food Sci Technol 38:193–200 19. Frankel EN, Waterhouse AL, Teissedre PL (1995) Principal phenolic phytochemicals in selected California wines and their antioxidant activity in inhibiting oxidation of human lowdensity lipoproteins. J Agric Food Chem 43:890–894 20. Tedesco I, Russo GL, Nazzaro F, Russo M, Palumbo R (2001) Antioxidant effect of red wine anthocyanins in normal and catalaseinactive human erythrocytes. J Nutr Biochem 12:505–511 21. Morel I, Abalea V, Sergent O, Cillard P, Cillard J (1998) Involvement of phenoxyl radical intermediates in lipid antioxidant action of myricetin in iron-treated rat hepatocyte culture. Biochem Pharmacol 55: 1399–1404 22. Daglia M, Papetti A, Gregotti C, Berte F, Gazzani G (2000) In vitro antioxidant and ex vivo protective activities of green and roasted coffee. J Agric Food Chem 48:1449–1454 23. Mora A, Paya M, Rios JL, Alcaraz MJ (1990) Structure-activity-relationships of polymetho xyflavones and other flavonoids as inhibitors of nonenzymatic lipid-peroxidation. Biochem Pharmacol 40:793–797 24. Plumb GW, Chambers SJ, Lambert N, Wanigatunga S, Williamson G (1997) Influence of fruit and vegetable extracts on lipid peroxidation in microsomes containing specific cytochrome P450. Food Chem 60:161–164

187

25. van Acker SABE, van den Berg D-J, Tromp MNJL, Griffioen DH, van Bennekom WP, van der Vijgh WJF, Bast A (1996) Structural aspects of antioxidant activity of flavonoids. Free Radic Biol Med 20:331–342 26. de Beer D, Joubert E, Gelderblom WCA, Manley M (2005) Antioxidant activity of South African red and white cultivar wines and selected phenolic compounds: in vitro inhibition of microsomal lipid peroxidation. Food Chem 90:569–577 27. Roginsky V, Barsukova T (2001) Chainbreaking antioxidant capability of some beverages as determined by the clark electrode technique. J Med Food 4:219–229 28. Shi H, Noguchi N, Niki E (1999) Comparative study on dynamics of antioxidative action of alpha-tocopheryl hydroquinone, ubiquinol, and alpha-tocopherol against lipid peroxidation. Free Radic Biol Med 27:334–346 29. Goñi MF, Alonso A (1989) Studies of phospholipid peroxidation in liposomes. In: CRC handbook of free radicals and antioxioxidants in biomedicine. CRC, Boca Raton, FL 30. Murakami M, Yamaguchi T, Takamura H, Matoba T (2002) A comparative study on the various in vitro assays of active oxygen scavenging activity in foods. J Food Sci 67:539–541 31. Niki E, Noguchi N (2000) Evaluation of antioxidant capacity. What capacity is being measured by which method? Life 50:323–329 32. Fernandes E, Costa D, Toste SA, Lima JLFC, Reis S (2004) In vitro scavenging activity for reactive oxygen and nitrogen species by nonsteroidal anti-inflammatory indole, pyrrole, and oxazole derivative drugs. Free Radic Biol Med 37:1895–1905 33. Kachel K, Asuncion-Punzalan E, London E (1998) The location of fluorescent probes with charged groups in model membranes. Biochim Biophys Acta 1374:63–76 34. Kaiser RD, London E (1998) Location of diphenylhexatriene (DPH) and its derivatives within membranes: comparison of different fluorescence quenching analysis of membrane depth. Biochemistry 37:8180–8190 35. Ou B, Hampsch-Woodill M, Prior RL (2001) Development and validation of an improved oxygen radical absorbance capacity assay using fluorescein as the fluorescent probe. J Agric Food Chem 49:4619–4626 36. Broniowska KA, Kirilyuk I, Wisniewska A (2007) Spin-labelled lutein as a new antioxidant in protection against lipid peroxidation. Free Radical Res 41:1053–1060

188

Reis et al.

37. Barclay L, Vinqvist MR (1994) Membrane peroxidation: inhibiting effects of water soluble antioxidants on phospholipids of different charge types. Free Radic Biol Med 16:779–788 38. Gutiérrez ME, García AF, Africa de Mandariaga M, Sagrista ML, Casadó FJ, Mora M (2003) Interaction of tocopherols and phenolic compounds with membrane lipid components: evaluation of their antioxidant activity in a liposomal model system. Life Sci 72:2337–2360 39. Saija A, Scalese M, Lanza M, Marzullo D, Bonina F, Castelli F (1995) Flavonoids as antioxidant agents: importance of their interaction with biomembranes. Free Radic Biol Med 19:481–486 40. Lúcio M, Ferreira H, Lima JLFC, Reis S (2007) Use of liposomes to evaluate the role of membrane interactions on antioxidant activity. Anal Chim Acta 597:163–170 41. Brooks P (1998) Use and benefits of nonsteroidal anti-inflammatory drugs. Am J Med 104:9S–13S 42. Kitamura K, Imayoshi N, Goto T, Shiro H, Mano T, Nakai Y (1995) Second derivative spectrophotometric determination of partition coefficients of chlorpromazine and promazine between lecithin bilayer vesicles and water. Anal Chim Acta 304:101–106 43. Kitamura K, Imayoshi N (1992) Secondderivative spectrophotometric determination

44.

45. 46.

47.

48.

49.

of the binding constant between chlorpromazine and -cyclodextrin in aqueous solutions. Anal Sci 8:497–501 Ferreira H, Lúcio M, Castro B, Gameiro P, Lima JLFC, Reis S (2003) Partition and location of nimesulide in EPC liposomes: a spectrophotometric and fluorescence study. Anal Bioanal Chem 377:293–298 Lakowicz JR (1999) Principles of fluorescence spectroscopy. Kluwer Academic/Plenum, New York Lentz BR (1993) Use of fluorescent probes to monitor molecular order and motions within liposome bilayers. Chem Phys Lipids 64:99–116 Barclay LRC, Artz JD, Mowat JJ (1995) Partitioning and antioxidant action of the water-soluble antioxidant, trolox, between the aqueous and lipid phases of phosphatidylcholine membranes – C-14 tracer and product studies. Biochim Biophys Acta 1237:77–85 Wang S, Beechem JM, Gratton E, Glaser M (1991) Orientational distribution of 1, 6-diphenyl- 1, 3, 5-hexatriene in phospholipid vesicles as determined by global analysis of frequency domain fluorimetry data. Biochemistry 30:5565–5512 Repáková J, Holopainen JM, Morrow MR, McDonald MC, Capková P, Vattulainen I (2005) Influence of DPH on the structure and dynamics of a DPPC bilayer. Biophys J 88:3398–3410

Chapter 14 Studying Colloidal Aggregation Using Liposomes Juan Sabín, Gerardo Prieto, and Félix Sarmiento Abstract Colloidal aggregation using liposomes has been studied in this chapter. As criteria of stability, the stability factor, an extension of the DLVO theory of colloidal stability, the fractal dimension of the liposome aggregates and the different regimes of aggregation (RLCA and DLCA) and the temperature have been used. Key words: Liposomes aggregation, Liposomes stability, Stability factor, Extension of DLVO model, RLCA regime of aggregation, DLCA regime of aggregation, Fractal dimension

1. Introduction Liposomes can be characterized as associating colloids, built up of amphoteric lipid molecules that self-assemble in aqueous media into spherical, self-closed structures. They consist of one or several concentric membranes; their size ranges from 20 nm to several micrometers, whereas the thickness of the membranes is around 4 nm (1). Liposomes play an important role in the colloidal science as model membranes and are important in biological, pharmaceutical and medical research because liposomes are the most effective carriers for introducing many different types of agents into cells, and the applications for liposome-based samples and products are extremely wide. Liposome dispersions are not thermodynamically stable (2). The total free energy of a dispersed system can always be lowered by reduction in the interfacial area. This tendency to aggregate is attributable to attractive van der Waals forces. The existence of other interactions between particles determines the stability. The applications of liposomes as nanoscale containers for drugs (3, 4), vaccines (5), enzymes or genetic material (6) require control and prediction of the liposome dispersion stability. This chapter introduces some methods to study the stability in solution. V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_14, © Humana Press, a part of Springer Science+Business Media, LLC 2010

189

190

Sabín, Prieto, and Sarmiento

2. Materials 1. l-a-phosphatidylcholine from egg yolk (EYPC), 2. 1,2-dimiristoyl-sn-glycero-3-phospho-choline (DMPC) 3. 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC). 4. Calcium nitrate 5. Lanthanum nitrate. 6. Methanol 7. Chloroform 8. Polycarbonate membranes. 9. Liposome preparation (see Note 1).

3. Methods 3.1. The Stability Factor (W) as a Criterion of Stability of Liposomes

The stability factor (W) is a kinetic criterion for the stability of a colloidal system (7): W =

kr ks

(1)

where kr describes the maximum coagulation rate (rapid coagulation regime where every collision between particles is effective) and ks is the particular coagulation rate (slow coagulation regime where not every collision results in coagulation). Thus, the inverse of the stability ratio provides a measure of the effectiveness of collisions leading to aggregation or fusion. The rate constants can be obtained by nephelometry. This technique consists of measuring at a low-angle, the static light scattering (see Note 2). The total scattering intensity for a dispersion of identical primary particles with a time-varying distribution sizes is (8):

I (t , q ) = 1 + 2knst I q (0)

(2)

where Iq(0) is the initial intensity of light scattered at angle q, ns is the number of primary particles and k is the rate constant. The scattered light intensity at low angles increases linearly with time, and then an absolute coagulation rate can be obtained from the slope if the number of primary particles is known. Figure 1 shows the results of a typical nephelometry experience. From the slopes of intensity versus time of Fig. 1, W can be calculated by (1). The intersection of the linear fit in the slow

Studying Colloidal Aggregation Using Liposomes

191

Fig. 1. Scattered light intensity (arbitrary units) versus time for the study of colloidal stability of phosphatidylcholine liposomes in aqueous media at different La+3 concentrations. [From Sabín J, et al. (2005) Langmuir 21:10968–10975]

Fig. 2. Stability ratio (W) versus La3+ concentration for phosphatidylcholine liposomes in aqueous media. The obtained critical coagulation concentration is 0.3 M. [From Sabín J, et al. (2005) Langmuir 21:10968–10975]

regime and the linear fit in the rapid regime define the critical coagulation concentration (ccc) at which liposome aggregates are formed (Fig. 2).

192

Sabín, Prieto, and Sarmiento

3.2. Extension to the Classical DLVO Theory to Study Stability of Liposomes

The Derjaguin-Landau-Verwey-Overbeek (DLVO) theory of colloidal stability is, at the moment, the only quantitative theory of colloidal sciences (9, 10). In their classical approach, the interaction potential between similar colloidal particles in polar (especially, aqueous) solutions is described by two types of forces: attractive van der Waals forces and repulsive double layer forces. Although the DLVO theory has been extensively applied with outstanding success, it has its inherent limitations. The development of new techniques in the determination of the interaction potential between surfaces (atomic force microscopy, for example) has shown agreement with DLVO at separation above a few nanometers but, at smaller separations, a short-range repulsive force appears, often termed “hydration” interaction. In this sense, it can be interesting to compare the experimental results of stability of liposomes with a DLVO extended theory that has, in account, the hydration potential. The van der Waals interaction VA(x) for the case of two spherical shells of equal radius, a, and thickness, d, derived by Hamaker, is given by (11): V A (x ) = −

Aa  1 2 1  A x (x + 2d) − +  − ln  12  x + 2d x + d d  6 (x + d)2

(3)

where A is the attractive Hamaker constant, a the initial radius of the liposomes, and x the distance between the two spherical shells. The Hamaker constant can be calculated, for example, using an algorithm-based on a SIMPLEX method, which minimizes the error of the nephelometry data to the function:

V b (u) exp T du 2 (u + 2a) kBT W = ∞ VA b (u) ∫0 (u + 2a)2 exp kBT du

∫

∞

0

(4)

where kB is the Bolztmann constant, T is the absolute temperature, VT represents the total DLVO potential, and b(u) is the hydrodynamic correlation factor, related to retardation on the particles diffusion due to the friction with the medium:

b (u) =

6u 2 + 13u + 2a 2 6u 2 + 4ua

(5)

where u = (x - 2a)/a The expression for the repulsive potential, VR(x) , by area unit has the form (12, 13):

VR (x ) = 2p e 0 er (a + ∆)y o2 exp [−k (x − 2∆)]

(6)

where y° is the surface potential, which coincides with z (zeta potential, see Note 3) for high values, and k is the reciprocal Debye length:

Studying Colloidal Aggregation Using Liposomes

193

Fig. 3. DLVO potentials of phosphatidylcholine liposomes as a function of the distance between two liposomes for different La3+ concentrations. [From Sabín J, et al. (2005) Langmuir 21:10968–10975]

κ2 =

( z12 z 2 + z22 z1 )cs N A e2 e 0 e r kB T

(7)

where cs is the concentration, e the elementary charge, z i the valence of the ions, and NA the Avogadro number. Finally, D is the thickness of the Stern layer and it is taken as the hydrated radius of the adsorbed ion. In Fig. 3, it can be seen an example of the DLVO potential for different ion concentrations. The aggregation of liposomes occurs when the repulsive barrier is lower than kBT at the contact surface (hydration surface D= 0.49 nm). Figure 3 shows also that the repulsive barrier is below kBT for ion concentrations larger than 0.27 M. 3.3. Fractal Dimension as a Tool to Analyze the Stability of Liposomes

The large structures that liposomes form when they aggregate produce good models for further analysis. The fractal dimension characterizes how the monomers occupy space in the aggregates and can be calculated measuring the angular dependence of the light scattered by the clusters and following the evolution of the cluster size. The aggregation mechanism can be described by two different regimes: diffusion-limited cluster aggregation (DLCA) and reaction-limited cluster aggregation (RLCA) (see Note 4). Figure 4 shows an example of evolution of size of the clusters of different liposomes along time. As can be seen in this Fig. DMPC shows the typical transition from RLCA to the DLCA aggregation. For concentrations below 0.5 × 10−3 M, the liposomes remain stable with no changes in size. At concentrations between 0.5 × 10−3 and 10−2 M, the growing of clusters is slow and exponential, as in a typical RLCA aggregation

194

Sabín, Prieto, and Sarmiento

Fig. 4. Evolution of the size of the clusters of dimyristoylphosphocholine (DMPC), dipalmitoylphsphocholine (DPPC) and egg yolk phosphatidylcholine (EYPC) liposomes along time. The aggregation was induced after starting the measurements by addition of different concentrations of Ca2+. [From Sabín J, et al. (2007) Eur Phys J E 24:201–210]

(see Note 4). For higher concentrations, the increase follows a DLCA regime and no differences between Ca2+ concentrations can be noticed. Clusters of EYPC liposomes reach a maximum size, which depends on the Ca2+ concentration and, after this size is attained, the cluster becomes stable and no other liposomes aggregate to the clusters. DPPC liposomes show an intermediate behavior. In all cases, the fractal dimensions fell in the range from 2.1 to 1.8, corresponding to the two regimes. 3.4. Analyzing the Influence of the Temperature on Stability of Liposomes

The aggregation of liposomes is influenced by temperature. As an example, we show the time evolution of DMPC liposomes when La3+ at fixed concentration is added at different temperatures (Fig. 5).

Studying Colloidal Aggregation Using Liposomes

195

Fig. 5. Size evolution as a function of time for DMPC liposomes. The arrows indicates when the addiction of La3+ and solvent, respectively, was realized. [Results not published]

As it can be seen in this Fig. 5, at 25°C, a notable increment of the size is detected, indicating a DLCA regime of aggregation. At approximately 1,500 s, solvent is added and the clusters of liposomes are destroyed, indicating a reversible aggregation. Similar behavior was observed at 37°C although in this case the size of increment is less pronounced (RLCA regime). At 40°C, the liposomes remain stable with a slight decrease in their size. This reduction could be due to osmotic effects.

4. Notes 1. Liposomes are prepared by the thin-film hydration method. A solution of the lipids in chloroform/methanol or chloroform is evaporated in a rotary evaporator to dryness under a stream of nitrogen and the resultant lipid film hydrated with double-distilled, degassed, and deionized water. This mixture is extruded a minimum of five times through polycarbonate filters of the adequate pore size. 2. In a typical experiment, the angle detecting scattered light is around 10°, the time of 120 s, and the flow cell is rectangular with 2 mm path length. The cell must be thoroughly cleaned with chromic acid and rinsed with distilled water. Normally, the liposome stability is studied in the presence of ions.

196

Sabín, Prieto, and Sarmiento

Then, equal volumes (for example, 1 mL) of salt and liposome solutions are mixed and introduced into the cell by an automatic mixing device. The dead time is, in general, quite short (14). 3. The zeta potential is calculated from the electrophoretic mobilities, mE by means of the Henry correction of Smoluchoswski equation (15):

mE =

3m E h 1 2e 0 e r f (κa )

where e0 is the permittivity of the vacuum, er is the relative permittivity dielectric constant, a is the particle radius, k is the Debye length, and eta the viscosity of water. The function f(ka) depends on the particle shape. For, k > 1 (case of the liposomes) takes the form:

f (κa) =

3 9 75 − + 2 2 2 2κa 2κ a

Electrophoretic mobilities can be measured with the appropriate instrument. For example, in most of our experiments, a Malvern Instruments Zetamaster 5002 was used, taking the average of five measurements at stationary level. The cell used was a 5 mm × 2 mm rectangular quartz capillarity. The controlled temperature was 25°C. Following Fig. 6 shows an example of the variation of the zeta potential as a function of ion concentration.

Fig. 6. Zeta potential of phosphatidylcholine liposomes as a function of La3+ concentration. [From Sabín J, et al. (2005) Langmuir 21:10968–10975]

Studying Colloidal Aggregation Using Liposomes

197

4. In the DLCA model, the cluster´s growth is controlled by diffusion and every collision between particles results in the formation of a permanent contact (16). The fractal dimension is insensitive to the sticking probability. Computer simulations of this model yield a fractal dimension of 1.8 in three dimensions. In the RLCA model (17), the particles are added one at a time at random to sites adjacent to occupied sites. This process produces a relatively compact cluster whose density correlations are independent of distance. The growth is limited by reaction kinetics at the interface of the growing clusters and only a small fraction of particle collisions leads to permanent contact. In this case, the computer simulations yield a fractal dimension of 2.0 in three dimensions. The two models of aggregation lead to two different kinds of clusters. Clusters obtained in the DLCA regime are very open structures because the particles stick where the collisions occur and the probability for branches to collide is higher than to penetrate to the center of the cluster. On the contrary, in the RLCA regime, the particles have to collide with the cluster several times before sticking, increasing the possibility of the particle to reach the center of the cluster. The resulting structure is a compact and dense cluster. Computer simulation also reveals that the size of the clusters grows differently in the DLCA and the RLCA regimes. DLCA clusters grow faster since all collisions are effective, following a power law for the average radius of gyration, R gat 1/df , where t is time and df the fractal dimension. The growth of the RLCA clusters follows an exponential law R gae at, where a is a constant which depends on the sticking probability (P).

Acknowledgments This work was supported by grants-in-aid from the Spanish “Ministerio de Ciencia e Innovación” (Project MAT2008-04722) and the “Xunta de Galicia” (Project INCITE08PXIB20603PR).

References 1. New RRC (1990) Liposomes: a practical approach. IRL, Oxford 2. Lasic DD (1993) Liposomes. From physics to applications. Elsevier, Amsterdam 3. Gregoriadis G (1987) Liposomes as drug carriers. Wiley, New York

4. Devissaguet JP, Puisieux F (1993) Les liposomes: aspects tecnologiques, Biologiques et Pharmacologiques. Les Editions Inserm, Paris 5. Alving C, Richards RL (1983) Immunologic aspects of liposomes. In: Ostro MJ (ed) Liposomes. Marcel Dekker, New York, pp 209–287

198

Sabín, Prieto, and Sarmiento

6. Puisieux F (1985) Les liposomes: applications thérapeutiques. Tec and Doc Lavoisier, Paris 7. Hidalgo-Álvarez R, Martín A, Fernández A, Bastos D, Martínez F, de las Nieves FJ (1996) Electrokinetic properties, colloidal stability and aggregation kinetics of polymer colloids. Adv Colloid Interface Sci 67:1–118 8. Lips A, Willis E (1973) Low angle light scattering technique for the study of coagulation. J Chem Soc Faraday Trans 1 69: 1226–1236 9. Derjaguin BV, Landau LD (1941) Theory of the stability of strongly charged lyophobic sols and of the adhesion of strongly charged particles in solution of electrolytes. Acta Physicochim URSS 14:633–662 10. Verwey EJW, Overbeek JThG (1948) Theory of the stability of lyophobic colloids. The interaction of particles having an electric double layer. Elsevier, Amsterdam 11. Tadmor R (2001) The London-van der Waals interaction energy between objects of various geometries. J Phys Condens Matter 13: L195–L202

12. Vincent B, Bijsterbosch H, Lyklema J (1970) Competitive adsorption of ions and neutral molecules in the Stern layer on silver iodide and its effect on colloidal stability. J Colloid Interface Sci 37:171–178 13. Bastos D, de las Nieves FJ (1994) Colloidal stability of sulfonated polystyrene model colloids. Correlation with electrokinetic data. Colloid Polym Sci 272:592–597 14. Molina-Bolívar JA, Galisteo-González F, Hidalgo-Álvarez R (1999) Development of a high sensitivity IgG-latex immunodetection system stabilized by hydration forces. Polym Int 48:685–690 15. Hunter JR (1981) Zeta potential in colloid science: principles and application. Academic, London 16. Meakin P (1983) Formation of fractal clusters and networks by irreversible diffusion-limited aggregation. Phys Rev Lett 51:1119–1122 17. Eden M (1961) Proceedings of the Fourth Berkeley Symposium on Mathematics, Statics and Probability. In: Neyman F (ed) University of California Press, Berkeley, p. 223

Chapter 15 Assessment of Liposome–Cell Interactions Jan A.A.M. Kamps Abstract Liposome–cell interactions have been assessed for over 30 years now by an enormous variety of approaches and methods. In-depth knowledge of liposome–cell interaction is still very relevant since new concepts and applications applying liposomes are being developed every day. This chapter does not aim to give a complete overview on methods to assess liposome–cell interactions but merely gives a handle to approach the assessment of liposome–cell interaction by describing some well-established methods that also allow for modification to adapt them for your specific research questions. Key words: Endocytosis, Fluorescence microscopy, Radioactive liposome markers, Fluorescent liposome markers, Cell binding, Liposome uptake, Cellular processing of liposomes

1. Introduction Liposome–cell interaction is one of the most important features in the research field of liposomes applied as versatile drug carriers, since it largely determines the fate of the compound to be delivered by the liposome. Liposome–cell interaction includes binding, uptake, intracellular homing and processing of liposomes, liposome components and liposomal content. Interactions of liposomes with cells are complex because they involve several distinct mechanisms of interaction and are dependent on many parameters such as the biological environment, cell type and liposome characteristics. Four general mechanisms of liposome–cell interactions can be distinguished: adsorption, (receptor mediated) endocytosis, lipid exchange and fusion (1–4). The nature of the liposome–cell interactions are determined by the type of cell and by liposome characteristics such as composition, size and charge, the presence and characteristics of targeting molecules on the liposome surface and by the biological environment e.g. the presence V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_15, © Humana Press, a part of Springer Science+Business Media, LLC 2010

199

200

Kamps

of plasma or serum. On top of that, assessment of liposome–cell interactions can be even more complicated because different liposome constituents may interact differently with cells. To study liposome–cell interaction, many liposome markers are available, which can be roughly divided as water-soluble markers and lipid markers, where water-soluble markers can be encapsulated in the interior of the liposomes and lipid markers can be incorporated in the liposome bilayer. Obviously, for following the cellular fate of encapsulated material, one would choose a watersoluble marker, while for the fate of the liposome a stably incorporated and inert lipid label in the bilayer would be preferable. The choice for a specific marker depends primarily on the aim of an experiment and the laboratory facilities available. Both radioactive lipid markers and fluorescent lipid markers have been shown powerful tools to determine liposome–cell interaction and examples of both will be presented. In addition, a combination of more than one marker in a liposome allows for studying the different fates of specific liposome components (5). Different liposome markers will be addressed but it is beyond the scope of this chapter to provide full overview of markers available. The protocols provided deal with in vitro experiments with cultured cells that allow for detailed study of the interaction of liposomes with cells under different conditions, though it should be emphasized that extrapolation to in vivo events can only be made with great carefulness. In a study where we addressed the uptake of liposomes containing phosphatidylserine by liver endothelial cells, we showed massive uptake of these liposomes in primary cultures of these cells, while in vivo, liver endothelial cells did not contribute in liposome uptake at all (6). To study liposome–cell interaction, there are numerous methods available. In this chapter, we will address most common liposome–cell interactions and the methods to experimentally determine binding, uptake and intracellular processing of liposomes. The protocols given have shown in our hands to give reliable information on the nature of the liposome–cell interaction but are by far not the only or even the best methods for a certain research question. However, the methods described here can form a good starting point to study liposome–cell interaction and may offer a rational for further detailed study applying more sophisticated experimental set-ups.

2. Materials 2.1. Liposome Markers

As stated in the introduction, there are many liposomal markers available to assess liposome–cell interaction. The markers listed below we, routinely, use in our laboratory for the applications described in the methods section, but for any of them alternatives

Assessment of Liposome–Cell Interactions

201

are commercially available. These markers can be applied for most if not all liposome preparation methods (see Note 1). 1. [3H]cholesteryloleyl ether (GE Healthcare Europe, Diegem, Belgium); Stable and metabolically inert radioactive lipid marker. 2. Cholesteryl[14C]oleate (GE Healthcare Europe, Diegem, Belgium); A lipid marker that is readily hydrolyzed in lysosomes of cells. 3. 1,1¢-dioctadecyl-3,3,3¢,3¢-tetramethyl indocarbocyanine perchlorate (DiI; Molecular Probes, Eugene, OR); Stable fluorescent lipid bilayer marker (exitation: 549 nm; emission: 565 nm) (see Note 2). 4. Calcein (e.g. from Sigma–Aldrich, St. Louis MO, USA); water-soluble fluorophore that is quenched at high (~100 mM) concentration (exitation: 491 nm; emission: 511 nm). 2.2. Cell Culture (General Protocol)

1. Appropriate cell culture medium depending on the cell type (e.g. Dulbecco’s Modified Eagle’s Medium (DMEM), supplemented with 10% fetal bovine serum (FCS)). 2. Cells are grown to ~80% confluency (as determined by microscopy), typically in 12 or 24 well-culture plates or on chamber slides (Lab-Tek, NUNC, Thermo Fischer Scientific) (see Note 3). 3. Before the start of the experiment, cells are washed twice with incubation medium (e.g. DMEM without FCS).

2.3. Liposome–Cell Incubations

1. Appropriate cell culture medium with or without FCS, depending on the nature of the experiment. 2. Stock solutions of labeled liposomes and when applicable stock solutions of other agents/effectors to be added to the cells. 3. Cold (£4ºC) washing buffer, e.g. phosphate-buffered saline (10 mM phosphate buffer, pH 7.4, 150 mM NaCl). 4. (a) 0.1 M NaOH for cell Lysis prior to measurement of radioactivity. (b) Phosphate-buffered saline (PBS): 10 mM phosphate buffer, pH 7.4, 150 mM NaCl, supplemented with 5% FCS or 1% bovine serum albumin (BSA) for fluorescence measurement using fluorescence-activated cell sorting (FACS). 5. (a) Scintillation counter to determine radioactivity. (b) Fluorescence-activated cell sorter.

2.4. Intracellular Processing/ Degradation of Liposomes

1. Liposomes labeled with both [3H]cholesteryloleyl ether and cholesteryl[14C]oleate. 2. Appropriate culture medium, containing an acceptor for 14 C-oleate. Serum proteins (10% FCS) or serum albumin (1–2%) may function as acceptor for 14C-oleate.

202

Kamps

3. Phosphate-buffered saline (PBS): 10 mM phosphate buffer, pH 7.4, 150 mM NaCl 4. 0.1 M NaOH for cell Lysis prior to measurement of radioactivity. 2.5. Fluorescence Microscopy

1. Cells cultured on chamber slides (Lab-Tek, NUNC, Thermo Fischer Scientific) or on microscope cover slips that have been put into a regular 12 or 24 well-culture plate prior to cell seeding. 2. Cold (£4ºC) phosphate-buffered saline or cold culture medium without FCS. 3. Fluorescence microscope with appropriate filter sets.

3. Methods 3.1. Preparation of Labeled Liposomes (General Protocol (see Note 4))

1. lipids (e.g. phospholipids and cholesterol) from stock solutions in chloroform: methanol are mixed in an appropriate molar ratio. (a) When applicable radioactive or fluorescent lipid markers are added to this mixture in an appropriate amount (typically, trace amounts, e.g. for [3H]cholesteryloleyl ether 1 µCi/µmol lipid, for Cholesteryl[14C]oleate 0.4 µCi/µmol lipid, and for DiI 0.5–1 mol% ). 2. The lipids are dried under reduced nitrogen pressure, dissolved in cyclohexane en lyophilized. 3. The lipids are then hydrated in a Hepes buffer (10 mM N-2hydroxyethylpiperazine-N¢-2-ethanesulfonic acid (Hepes), 135 mM NaCl, pH 7.4). (a) When applicable water-soluble markers are dissolved in the hydration buffer, e.g. Calcein, typically at a concentration of ~100 mM. (see Note 5). 4. The liposomes formed can be sized by repeated extrusion through polycarbonate filters (Costar, Cambridge MA, USA), pore size typically 50 nm, using a high-pressure extruder (Lipex, Vancouver, Canada). 5. When a water-soluble marker has been encapsulated, free marker has to be separated from the liposome-associated marker (e.g. by gel permeation chromatography). 6. Of each liposome preparation, phospholipid phosphorous is determined by a phosphate assay after perchloric acid destruction (7), and in case of radioactive liposomes, specific radioactivity of the liposome preparation is calculated after scintillation counting of a liposome sample. 7. Liposomes are routinely stored at 4ºC under argon atmosphere. A tube containing liposomes is carefully gassed with

Assessment of Liposome–Cell Interactions

203

argon, allowing air to flush out, after which the tube is closed with an airtight screw cap. When argon is not available, nitrogen gas can be applied (see Note 6). 3.2. Cell Binding and Uptake of [3H] cholesteryloleyl ether Labeled Liposomes

1. Replace the culture medium by serum-free medium and incubate the cells for 1 h at 37°C in a humidified 5% CO2 / 95% air atmosphere (see Note 7). 2. Remove the serum-free medium and add new medium containing appropriate amounts of [3H]cholesteryloleyl ether-labeled liposomes and if necessary, other additions (see Note 8). 3. Incubate cells for the 3 h (see Note 9) at 37°C in a humidified 5% CO2/95% air atmosphere to determine cell association (binding and uptake) or for 3 h at 4°C to determine exclusively cell binding. 4. After the incubation the culture plates are placed on ice and washed six times with cold (£4ºC) phosphate-buffered saline (see Note 10). 5. The cells are lysed in 0.1 M NaOH for 1 h at 37°C or overnight at −20°C and subsequent thawing. 6. The cell-associated 3H radioactivity is determined by liquid scintillation counting of aliquots of the lysed cell suspension and is normalized to the amount of cellular protein as determined according to Lowry (8) using bovine serum albumin as a standard. 7. Cell binding and/or cell uptake of the liposomes is calculated from the specific activity of the liposomes (dpm/µmol lipid) and cellular radioactivity (dpm/mg of cell protein). 8. This procedure distinguishes between binding plus uptake and exclusive binding of liposomes by incubating the cells either at 37ºC or 4ºC. The difference between liposome association at 37ºC and association at 4ºC is a measure for internalization (see Note 11).

3.3. Intracellular Processing/ Degradation of Liposomes

When liposomes are double-labeled with a [3H]cholesteryl ether and the radioactive marker cholesteryl[14C]oleate which is readily hydrolyzed in the lysosomes, it is possible to make an estimation of the degradation of liposomes. Due to rapid release from the cells of [14C]oleate derived from the degraded cholesteryl[14C] oleate and the complete retention of the metabolically inert [3H] cholesteryloleyl ether, the 3H/14C ratio is a sensitive measure for intracellular degradation (9) (see Note 12). 1. Replace the culture medium by serum-free medium and incubate the cells for 1 h at 37°C in a humidified 5% CO2/95% air atmosphere.

204

Kamps

2. Remove the serum-free medium and add new medium containing appropriate amounts of [3H]cholesteryloleyl ether and cholesteryl[14C]oleate-labeled liposomes, an acceptor for [14C] oleate released by the cells, and if necessary other additions. 3. Incubate the cells as described in Subheading 3.2. 4. After the incubation the culture plates are placed on ice and washed six times with cold (£4ºC) phosphate-buffered saline. 5. The cells are lysed in 0.1 M NaOH for 1 h at 37°C or overnight at −20°C and subsequent thawing. 6. The cell-associated 3H and 14C radioactivity are determined by liquid scintillation counting of aliquots of the lysed cell suspension and are normalized to the amount of cellular protein as determined according to Lowry (8) using bovine serum albumin as a standard. 7. Liposome degradation and/or processing by the cells is calculated from the specific 3H and 14C activities of the liposomes (dpm/µmol lipid) and cellular radioactivities (dpm/ mg of cell protein). The 3H/14C ratio is then a measure for cellular degradation/processing. 3.4. Fluorescence Microscopy

1. Replace the culture medium by serum free medium and incubate the cells for 1 h at 37°C in a humidified 5% CO2/95% air atmosphere. 2. Remove the serum-free medium and add new medium containing appropriate amounts of fluorescently labeled liposomes and if necessary other additions. 3. Incubate cells for the appropriate time at 37°C in a humidiffied 5% CO2/95% air atmosphere in the dark, to determine cell binding and uptake. 4. Stop the incubation by careful washing of the cells three times with cold (£4ºC) phosphate- buffered saline and then put some cold medium without FCS on the cells. 5. Transport the chamber slides or culture dishes with cover slips on ice (keep covered from light) to the microscope and in case of chamber slides remove the chambers, cover the cells with a cover slip and immediately observe the cells under the fluorescence microscope. In case of cells cultured on a cover slip, carefully remove the cover slip from the culture well and mount it upside down on an object glass and immediately observe the cells (see Note 13).

3.5 FluorescenceActivated Cell Sorting

Association of fluorescently labeled liposomes with cells can be (semi)quantified using a fluorescence activated cell sorter (10). The incubation protocol is basically the same as with radioactive labeled liposomes (incubate in the dark!).

Assessment of Liposome–Cell Interactions

205

1. After incubation in a 24 well-culture plate, stop the incubation by washing the cells two times with cold (£4ºC) phosphatebuffered saline. 2. Harvest the cells using an appropriate method e.g. trypsin treatment and transfer the cells in to FACS tubes that (in the case of trypsine treatment) contain 200 µl FCS. Keep the tubes on ice. 3. Add 3 ml of FACS buffer and wash the cells two times by centrifugation at 1600 rpm for 5 min at 4ºC. 4. Resuspend the cells in an appropriate volume and measure cell fluorescence according to the protocol of the fluorescence-activated cell sorter.

4. Notes 1. It is important to realize when studying liposome cell interaction that most methods rely on certain markers or labeled molecules. In fact, it means that one is studying the fate of the label which is not necessarily similar to the fate of the intact liposome. Therefore, it is needed to confirm data on binding, uptake and processing by other methods and to apply relevant controls. 2. Care should be taken not to expose fluorescent markers or fluorescently labeled liposomes to light. This will decrease fluorescence intensity in time. The examples described here (DiI, Calcein) are relatively insensitive to light. 3. When incubating liposomes with cultured cells, make sure that cell conditions such as medium, atmosphere and cell confluency at the time of incubation are comparable. There can be major differences in the handling of liposomes by cells depending on their confluency, i.e. growth status. 4. Liposomes can be prepared by any method, where lipid markers can be added in appropriate (trace) amounts to lipid mixture prior to hydration of the lipids, while water-soluble markers can be added during the hydration step. 5. The characteristic that calcein is quenched at high concentrations may be used to determine whether liposome content is released intracellularly. Upon intracellular release from the liposomes of the encapsulated content, the calcein will be diluted in the cytoplasm of the cell and emit fluorescence. Double labeling with a stable liposome-associated fluorophore, e.g. DiI, allows to control and/or correcting for differences in liposome uptake.

206

Kamps

6. Stability of markers has to be taken in account when studying the interaction of liposomes with cells. With the liposome markers described in the different protocols we always have routinely performed experiments within 3 weeks of preparation of the liposomes (unless stated otherwise). Although some of the liposome markers that we use showed reproducible results with regard to liposome–cell interaction after storage of the liposomes for 6 months or more we prefer to keep to the term indicated to avoid any artifacts from long-term storage of liposomes. 7. One-hour incubation with serum-free medium prior to incubation with liposomes is performed to exclude influence of serum proteins on binding and uptake of liposomes by cultured cells. However, for some research questions and/or cell types, the presence of serum in the (pre) incubation medium may be preferable. 8. For incubation of liposomes with cultured cells, dilute stock solutions of liposomes directly before the start of the experiment in the culture medium (without FCS) used in the incubation to avoid dilution of the culture medium upon addition of liposomes to the cells. The same holds true for any other addition (inhibitors, stimuli, etc). Always make sure to control for any solvent used in your incubation. 9. Incubation of 3 h is normally sufficient to determine binding and uptake (association); however, duration of incubation depends on specific research question, liposomes and cell type. For determining binding at 4ºC, the incubation should not be longer than 3 h since the quality of the cells often decreases with longer incubations under these conditions. 10. Washing cells prior to assess binding, uptake or processing of liposomes is necessary to get rid of excess (labeled) liposomes. In our hands 5–7 rounds of careful washing suffice. However, make sure to check cells between washing steps microscopically. Some (primary) cells may be very sensitive to washing and detach from their substrate. Efficiency of washing the cells can be determined by measuring the label left in the fluid taken from the cells in the subsequent washing steps. For cells that easily detach upon washing, addition of 1% (W/V) serum albumin to the washing medium may be beneficial. 11. The difference between liposome association at 37ºC and association at 4ºC can be used as a measure for internalization. However, one has to proceed here with caution since binding at 4ºC may not be representative of binding at 37ºC because of dynamic processes such as receptor recycling which do not occur at 4ºC. Differences between the association of liposomes to cells at 37ºC and 4ºC can also be

Assessment of Liposome–Cell Interactions

207

accounted for by temperature effects on bilayer mobility which may, as was shown for immunoliposomes, affect the amount of liposomes bound to the cells, even in cases where there is no uptake of liposomes (11). 12. One should take notice that the 3H/14C ratio in the cells is not only dependent on the presence of a fatty acid acceptor in the culture medium such as albumin but also of the nature of the cells used in the experiments. Specialized cells involved in lipid metabolism may not release [14C]oleate to appreciable extents but, as in the case of hepatocytes, further metabolize it to CO2 or for the production of lipoproteins. 13. For fluorescence microscopy in live cells, it is important to establish the time that the cells retain their properties. In our hands, live cells kept at 4ºC should be observed microscopically within 1 hour. One should limit the amount of samples to be examined accordingly. This period has to be established separately for every cell type. If cells are kept too long, you may observe blebbing of cells, membrane deformation and in case of fluorescence microscopy, occurrence of fluorescence in the medium in between the cells. References 1. Pagano RE, Weinstein JN (1978) Interactions of liposomes with mammalian cells. Annu Rev Biophys Bioeng 7:435–468 2. Düzgünes N, Nir S (1999) Mechanisms and kinetics of liposome-cell interactions. Adv Drug Deliv Rev 40(1–2):3–18 3. Kamps JAAM, Scherphof GL (2003) Liposomes in biological systems. In: Torchilin VP, Weissig V (eds) Liposomes, a practical approach, 2nd edn. Oxford University Press, Oxford, pp 267–288 4. Torchilin VP (2005) Recent advances with liposomes as pharmaceutical carriers. Nat Rev Drug Discov 4(2):145–160 5. Yan X, Poelstra K, Scherphof GL, Kamps JA (2004) Role for scavenger receptor B-I in selective transfer of rhodamine-PE from liposomes to cells. Biochem Biophys Res Commun 325(3):908–914 6. Kamps JA, Morselt HW, Scherphof GL (1999) Uptake of liposomes containing phosphatidylserine by liver cells in vivo and by sinusoidal liver cells in primary culture: in vivo–in vitro

7. 8. 9.

10.

11.

differences. Biochem Biophys Res Commun 256(1):57–62 Böttcher CJF, Van Gent CM, Pries C (1961) A rapid and sensitive sub-micro phosphorus determination. Anal Chim Acta 24:203–204 Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951) Protein measurement with the Folin phenol reagent. J Biol Chem 193:265–275 Derksen JT, Morselt HW, Scherphof GL (1988) Uptake and processing of immunoglobulin-coated liposomes by subpopulations of rat liver macrophages. Biochim Biophys Acta 971(2):127–136 Kamps JA, Morselt HW, Swart PJ, Meijer DK, Scherphof GL (1997) Massive targeting of liposomes, surface-modified with anionized albumins, to hepatic endothelial cells. Proc Natl Acad Sci U S A 94(21):11681–11685 Koning GA, Morselt HW, Velinova MJ et al (1999) Selective transfer of a lipophilic prodrug of 5-fluorodeoxyuridine from immunoliposomes to colon cancer cells. Biochim Biophys Acta 1420(1–2):153–167

Chapter 16 Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells Nejat Düzgünes¸, Henrique Faneca, and Maria C. Pedroso de Lima Abstract We describe fluorescence assays for membrane fusion involving the fusion of liposomes with each other and with cultured cells, fluorescence methods to assess liposome uptake by cells and the intracellular delivery of liposome contents, and assays to evaluate liposome membrane permeability. The Tb/DPA and ANTS/DPX assays monitor the intermixing of aqueous contents of liposomes. The NBD-PE/ Rhodamine-PE assay follows the intermixing of liposomal lipids. A variation of this method is suitable for detecting the mixing of the inner monolayers of liposomes. The lipid-mixing assay is also used to study the fusion of cationic liposomes and lipoplexes with cultured cells. The intracellular delivery of liposome contents are monitored, via fluorescence microscopy or flow cytometry, by measuring the release of calcein from the liposome interior, and normalized to cell-associated liposomes quantitated with Rhodamine-PE in the membrane of the same liposomes. The release of liposome contents is monitored by the increase in fluorescence of encapsulated carboxyfluorescein, calcein, or ANTS/DPX, or by the decrease in fluorescence of encapsulated Tb/DPA. Key words: Membrane fusion, Fluorescence microscopy, Fluorescence dequenching, Flow cytometry, Resonance energy transfer

1. Introduction Many studies of liposomes have employed fluorescence techniques. These studies range from the determination of the fluidity of lipid bilayers (using fluorescence anisotropy) and the intracellular fate of endocytosed liposomes (employing fluorescence microscopy), to monitoring the fusion of liposomes with each other and the fusion of viruses with liposomes (via fluorometry). In this chapter,

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_16, © Humana Press, a part of Springer Science+Business Media, LLC 2010

209

210

Düzgünes¸, Faneca and Lima

we will concentrate on fluorescence assays for membrane fusion involving the fusion of liposomes with each other and with cultured cells, fluorescence methods to assess liposome uptake by cells and the intracellular delivery of liposome contents, and assays to evaluate liposome membrane permeability.

2. Materials 2.1. Liposome Preparation 2.1.1. Large Unilamellar Liposomes Prepared by Reverse Phase Evaporation

1. A rotary evaporator (Büchi (Flawil, Switzerland), Labconco (Kansas City, MO), or equivalent) attached to an oil-free vacuum pump (Büchi), with a vacuum gauge, and inlet tubing connected to an argon tank. 2. A bath-type sonicator (Laboratory Supply Co., Hicksville, NY). 3. Vortex mixer. 4. An additional argon tank with tubing connected to a clamped Pasteur pipette. 5. A high-pressure extrusion device (Northern Lipids, Vancouver, Canada, or custom-made) or a syringe extrusion apparatus (Avestin, Toronto, Canada, or Avanti Polar Lipids, Alabaster, AL). 6. Poylcarbonate membranes with defined pore diameter (Nuclepore brand from Whatman, Kent, UK or Florham, NJ; or Poretics/GE Osmonics brand from Fisher Scientific, Waltham, MA). 7. A high-quality glass tube of approx 1 cm diameter, with a “screw-cap” with a Teflon lining at the top, and a larger tube with a fitting enabling attachment to the rotary evaporator and into which the former tube can fit. 8. Teflon tape. 9. Phospholipids, fluorescent lipids and cholesterol (Avanti Polar Lipids, Alabaster, AL or other suppliers). Lipids are kept at −20˚C or −80˚C in glass tubes or bottles, after gently flushing argon over the lipid solution. 10. Beckman Coulter N4 Plus Submicron Particle Sizer (Fullerton, CA).

2.1.2. Large Unilamellar Liposomes Prepared by Extrusion of Multilamellar Liposomes

1. A rotary evaporator (Büchi (Flawil, Switzerland), Labconco (Kansas City, MO), or equivalent) attached to an oil-free vacuum pump (Büchi), with a vacuum gauge, and inlet tubing connected to an argon tank.

Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells

211

2. Phospholipids, fluorescent lipids and cholesterol (Avanti Polar Lipids, Alabaster, AL or other suppliers). Lipids are kept at −20˚C or −80˚C in glass tubes or bottles, after gently flushing argon over the lipid solution. 3. Vortex mixer. 4. An additional argon tank with tubing connected to a clamped Pasteur pipette 5. A high-pressure extrusion device (Northern Lipids, Vancouver, Canada, or custom-made) or a syringe extrusion apparatus (Avestin, Toronto, Canada, or Avanti Polar Lipids, Alabaster, AL). 6. Polycarbonate membranes with defined pore diameter (Nuclepore, Whatman; or Poretics/GE Osmonics, Fisher Scientific). 7. A high-quality glass tube of approx 1 cm diameter, with a “screw-cap” with a teflon lining at the top, and a larger tube with a fitting enabling attachment to the rotary evaporator and into which the former tube can fit. 8. Teflon tape. 9. Beckman Coulter N4 Plus Submicron Particle Sizer (Fullerton, CA). 2.1.3. Small Unilamellar Liposomes Prepared by Sonication

1. A rotary evaporator (Büchi, Labconco, or equivalent) attached to an oil-free vacuum pump (Büchi), with a vacuum gauge, and inlet tubing connected to an argon tank. 2. A vacuum oven at room temperature. 3. A bath-type sonicator (Laboratory Supply Co., Hicksville, NY). 4. Vortex mixer. 5. An additional argon tank with tubing connected to a clamped Pasteur pipette. 6. A high-quality glass tube of approx 1 cm diameter, with a “screw-cap” with a Teflon lining at the top, and a larger tube with a fitting enabling attachment to the rotary evaporator and into which the former tube can fit. 7. Teflon tape. 8. Phospholipids, fluorescent lipids and cholesterol (Avanti Polar Lipids, Alabaster, AL or other suppliers). Lipids are kept at −20˚C or −80˚C in glass tubes or bottles, after gently flushing argon over the lipid solution. 9. Ultracentrifuge (Beckman or equivalent). 10. Beckman Coulter N4 Plus Submicron Particle Sizer (Fullerton, CA).

212

Düzgünes¸, Faneca and Lima

2.1.4. Measurement of the Phospholipid Concentration of Liposomes

1. Heating block set at 160–170˚C, with aluminum block inserts for test tubes of 25 mm diameter. 2. Glass jar pipettor (Oxford) of 1 l capacity, vortex mixer, and electrical boiling water bath into which a metal tube rack with a handle can be inserted. 3. 10 N H2SO4 prepared by adding 100 ml concentrated H2SO4 into water to make up 360 ml of solution, placed in a polypropylene dispenser jar or equivalent container. Care should be taken to wear protective gear (gloves, goggles or face mask) and to use a chemical hood. 4. 9% H2O2 (3 ml of 30% H2O2 is diluted with distilled or purified water to 10 ml final volume). 5. H2O2indicator strips (EM Quant Peroxide Test, EM Science, Gibbstown, NJ). 6. Ammonium molybdate (0.22% in 0.25 N H2SO4; 25 ml of 10 N H2SO4 are added to distilled or purified water, 2.2 g ammonium molybdate are added, and the solution is brought up to 1 L with water). The solution is placed in the Oxford pipettor. 7. ANSA (Fiske) reagent. A 15% NaHSO3 solution is prepared; 250 mg aminonaphtholsulfonic acid and 500 mg Na2SO3 are added to the NaHSO3 solution, bringing the volume up to 100 ml. To dissolve all ingredients, the solution is heated and mixed gently on a stir-plate. Then the solution is stored in the cold. 8. Ascorbate can be used as an alternative to ANSA; 15 g ascorbic acid are dissolved in 100 ml distilled or purified water and stored in the refrigerator.

2.2. Liposome Fusion: Intermixing of Aqueous Contents 2.2.1. The Tb/DPA Assay

1. Buffer A: 100 mM NaCl, 1 mM EDTA, 10 mM TES, pH 7.4 (all chemicals can be obtained from Sigma–Aldrich). All solutions are kept at 4˚C. 2. Buffer B: 100 mM NaCl, 10 mM TES, pH 7.4. 3. A vapor pressure osmometer (Wescor, Logan, Utah). 4. Three 1 × 20 cm chromatography columns (Bio-Rad, Hercules, California) and buffer reservoirs. 5. Sephadex G-75 or G-50 (Pharmacia, Uppsala), pre-swollen in distilled water containing 0.02% sodium azide, and de-gassed, and preferably kept at room temperature. 6. Solution to be encapsulated in large unilamellar “Tb- liposomes”: 2.5 mM TbCl3, 50 mM Na citrate, 10 mM TES, pH 7.4. All solutions are kept at 4˚C. 7. Solution to be encapsulated in large unilamellar “DPAliposomes”: 50 mM Na dipicolinate, 20 mM NaCl, 10 mM TES, pH 7.4.

Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells

213

8. Solution to be encapsulated in large unilamellar “Tb/DPA liposomes”: A 1:1 mixture of the above two reagents (6 and 7). 9. Solution to be encapsulated in small unilamellar “Tb-liposomes”: 15 mM TbCl3, 150 mM Na citrate, 10 mM TES, pH 7.4. 10. Solution to be encapsulated in small unilamellar “DPAliposomes”: 150 mM Na dipicolinate, 10 mM TES, pH 7.4. 11. Solution to be encapsulated in small unilamellar “Tb/DPA liposomes”: A 1:1 mixture of the reagents in 9 and 10. 12. Polystyrene 15 ml culture tubes. 13. For assay calibration, a solution of 2 mM Na dipicolinate, 10 mM TES, pH 7.4. 14. 10% (w/v) Na cholate (Calbiochem) or 16 mM C 12E8 (octaethyleneglycol-dodecyl ether; Calbiochem). 15. Teflon-coated micro, or castle-type stir-bars. 16. Fluorometer, preferably with a thermal jacket, magnetic stirring, and a pinhole for injection of fusogens into the cuvette. 17. High-pass (>530 nm) cut-off filter (e.g. Corning 3–68). 18. Appropriate phospholipids and/or cholesterol for liposome preparation (Avanti Polar Lipids). 2.2.2. The ANTS/ DPX Assay

1. Solution to be encapsulated in large unilamellar “ANTSliposomes”: 25 mM aminonaphthalene trisulfonic acid (ANTS; molecular probes), 40 mM NaCl, 10 mM TES, pH 7.4. All solutions are kept at 4˚C. The tube containing the ANTS solution is wrapped in aluminum foil to maintain darkness. 2. Solution to be encapsulated in large unilamellar “DPXliposomes”: 90 mM p-xylylene bis(pyridinium) bromide (DPX, Molecular Probes), 10 mM TES, pH 7.4. 3. Solution to be encapsulated in large unilamellar “ANTS/ DPX-liposomes” for calibration of the assay: A 1:1 mixture of the above two solutions. 4. Buffer C: 100 mM NaCl, 0.1 mM EDTA, 10 mM TES, pH 7.4. 5. Vapor pressure osmometer. 6. Sephadex G-75 or G-50 (Pharmacia, Uppsala or Sigma, St. Louis). 7. Three 1 × 20 cm Bio-Rad columns and buffer reservoirs. 8. Fluorometer, preferably with a thermal jacket, magnetic stirring, and a pinhole for injection of fusogens into the cuvette.

214

Düzgünes¸, Faneca and Lima

9. Appropriate phospholipids and/or cholesterol for liposome preparation (Avanti Polar Lipids). 2.3. Liposome Fusion: Intermixing of Lipids 2.3.1. The NBD/Rhodamine Assay

1. N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl) phosphatidylethanolamine (Avanti Polar Lipids or Invitrogen/Molecular Probes). Lipids are kept at −20˚C or −80˚C in glass tubes or bottles, after gently flushing argon over the lipid solution. 2. N-(lissamine rhodamine B sulfonyl) phosphatidylethano lamine (Avanti Polar Lipids or Invitrogen/Molecular Probes). 3. Appropriate phospholipids and/or cholesterol for liposome preparation (Avanti Polar Lipids). 4. Medium: 100 mM NaCl, 0.1 mM EDTA, 10 mM TES, pH 7.4 (Buffer C). All solutions are kept at 4˚C. 5. Fluorometer

2.3.2. Inner Monolayer Mixing Assay

1. N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl) phosphatidylethanolamine (Avanti Polar Lipids). Lipids are kept at −20˚C or −80˚C in glass tubes or bottles, after gently flushing argon over the lipid solution. 2. N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl) (Avanti Polar Lipids)

phosphatidylserine

3. DiI(5)C18 (1,1¢-dioctadecyl-3,3,3¢,3¢-tetramethylindodicar bocyanine perchlorate (Invitrogen/Molecular Probes)). 4. Appropriate phospholipids and cholesterol for liposome preparation. 5. Sodium dithionate (stock solution of 200 mM sodium dithionate, 100 mM Tris, pH 10, made shortly before use). 6. Medium: 100 mM NaCl, 0.1 mM EDTA, 10 mM TES, pH 7.4 (Buffer C). 7. Bio-Gel A-30 spin columns (0.8 mL capacity) (Bio-Rad, Richmond, CA). 2.4. Fusion of Liposomes and Lipoplexes with Cultured Cells

1. THP-1 monocytic leukemia cells are available from the UCSF Cell Culture Facility (San Francisco, CA), and other sources including the American Type Culture Collection (Manassas, VA). 2. Culture medium: RPMI-1640 medium containing 25 mM HEPES, 2% sodium bicarbonate, 10% heat-inactivated fetal bovine serum, 100 µg/ml streptomycin and 1 U/ml penicillin (Biochrom KG, Berlin; Invitrogen, Carlsbad, CA). All media are stored at 4˚C, but warmed to 37˚C before addition to cells. 3. Phenol red-free RPMI-1640 containing 25 mM HEPES buffer, kept at 4˚C.

Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells

215

4. Trypan blue solution (Sigma). 5. The cationic lipid 1,2-dioleoyl-(trimethylammonium) propane (DOTAP). All lipids are stored in glass tubes at −20˚C or −80˚C after gently flushing argon over the lipid solution. 6. Egg phosphatidylcholine (PC) (Avanti). 7. Phosphatidylethanolamine produced by transphosphatidylation of PC (PE) (Avanti). 8. N-(lissamine rhodamine B sulfonyl) phosphatidylethanolamine (Avanti). 9. Plasmid DNA: pCMV.luc. 10. Antimycin A (Calbiochem, La Jolla, CA) stock solution of 100 µg/ml. All solutions should be kept at 4˚C. 11. NaF (Sigma) stock solution of 1 M. 12. NaN3 (Sigma) stock solution of 10% (w/v). 2.5. Intracellular Delivery of Liposome Contents

1. Egg phosphatidylcholine (PC), dioleoylphosphatidylethanolamine (DOPE), lissamine rhodamine B-phosphatidylethanolamine (egg) (Rh-PE), phosphatidylserine (PS), phosphatidylglycerol (PG), distearoylphosphatidylcholine (DSPC) and poly(ethylene glycol) (2000)-distearoylphosphatidylethanolamine (DSPE-PEG) are purchased from Avanti Polar Lipids. All lipids are stored in glass tubes at −20˚C or −80˚C after gently flushing argon over the lipid solution. 2. Cholesteryl hemisuccinate (CHEMS), calcein, MES, TES, EDTA, phorbol 12-myristate 13-acetate, Triton X-100, propidium iodide, bafilomycin A1, NaN3, NaF, and NaCl are obtained from Sigma. Antimycin A is obtained from Calbiochem (La Jolla, CA). Calcein may also be obtained from (Invitrogen/ Molecular Probes). 3. Solution to be encapsulated: 80 mM calcein (Invitrogen/ Molecular Probes), 10 mM TES, pH 8.2, 1 mM EDTA, adjusted to 300 mOsm by the addition of NaCl. 4. For dialysis: TES-buffered saline (TBS, 140 mM NaCl, 10 mM TES, pH 7.4) containing 0.1 mM EDTA. 5. Polycarbonate membranes of 100 nm pore diameter (Costar, Cambridge, MA). 6. LiposoFast extruder (Avestin, Ottawa, Canada). 7. Monocytic human THP-1 cells. 8. RPMI-1640 medium, with and without phenol red (UCSF Cell Culture Facility or Invitrogen). 9. Fetal bovine serum (UCSF Cell Culture Facility or Invitrogen).

216

Düzgünes¸, Faneca and Lima

10. Antimycin A (Calbiochem, La Jolla, CA) stock solution of 100 µg/ml. 11. NaF (Sigma) stock solution of 1 M. 12. NaN3 (Sigma) stock solution of 10% (w/v). 13. Cell dissociation buffer (Life Technologies, Inc.) 14. Phosphate-buffered saline (PBS) without calcium and magnesium. 15. Becton Dickinson FACStar Plus flow cytometer (or equivalent). 16. Lab-Tek chambered coverglasses for tissue culture (Nunc, Naperville, IL). 17. Diaphot-inverted fluorescence microscope (Nikon, Melville, NY) and fluorescence image analysis system (Photon Technology International, Birmingham, NJ). 2.6. Liposome Permeability 2.6.1. Release of Carboxyfluorescein or Calcein

1. Some carboxyfluorescein or calcein preparations may contain impurities, which may be eliminated by gel filtration of a concentrated solution in water on Sephadex LH-20. 2. Solution to be encapsulated in large unilamellar liposomes: 50 mM carboxyfluorescein or calcein (Na salt) (Molecular Probes), 10 mM TES, pH 7.4. Solutions of fluorescent probes should be kept in the dark by wrapping aluminum foil over the tube, and stored at 4˚C. 3. Solution to be encapsulated in small unilamellar liposomes: 100 mM carboxyfluorescein or calcein, 10 mM TES, pH 7.4. 4. Buffer C: 100 mM NaCl, 0.1 mM EDTA, 10 mM TES, pH 7.4 5. Triton X-100 (Sigma) stock solution (100×) at 10% (w/v) or C12E8 (Calbiochem) at 80 mM. 6. Fluorometer, preferably with a thermal jacket, magnetic stirring, and a pinhole for injection of membrane-active agents (such as peptides) into the cuvette. 7. High-pass (>530 nm) cut-off filter (e.g. Corning 3–68).

2.6.2. Release of the Tb/ DPA Complex

1. Solution to be encapsulated in large unilamellar “Tb/DPA liposomes”: 1.25 mM TbCl3, 25 mM Na citrate, 25 mM Na dipicolinate, 10 mM NaCl, 10 mM TES, pH 7.4. 2. Solution to be encapsulated in small unilamellar “Tb/DPA liposomes”: 7.5 mM TbCl3, 75 mM Na citrate, 75 mM Na dipicolinate, 10 mM TES, pH 7.4. 3. Buffer A: 100 mM NaCl, 1 mM EDTA, 10 mM TES, pH 7.4. Solutions should be stored at 4˚C. 4. Buffer B: 100 mM NaCl, 10 mM TES, pH 7.4.

Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells

217

5. Sephadex G-75 or G-50 column equilibrated with Buffer A. 6. Triton X-100 stock solution (100×) at 10% (w/v) or C12E8 (20×) at 16 mM. Detergent solutions are kept at room temperature. 7. Fluorometer, as above. 2.6.3. The Release of ANTS/DPX

1. Solution to be encapsulated in large unilamellar “ANTS/ DPX-liposomes”: 12.5 mM ANTS, 45 mM DPX, 20 mM NaCl, 10 mM TES, pH 7.4. Solutions of fluorescent probes should be kept in the dark by wrapping aluminum foil over the tube, and stored at 4˚C. 2. Buffer C: 100 mM NaCl, 0.1 mM EDTA, 10 mM TES, pH 7.4. Solutions should be stored at 4˚C. 3. Sephadex G-75 or G-50 column equilibrated with Buffer C. 4. Triton X-100 stock solution (100×) at 10% (w/v) or C12E8 (20×) at 16 mM. Detergent solutions are kept at room temperature. 5. Fluorometer, as above.

3. Methods

3.1. Liposome Preparation 3.1.1 Large Unilamellar Liposomes Prepared by Reverse Phase Evaporation

1. Pure or mixed phospholipids in chloroform are measured using gastight Hamilton syringes under a chemical hood, and dispensed into a glass tube. The tube is covered with teflon tape, and then placed inside a larger tube that fits on a rotary evaporator until the lipid dries into a thin film in vacuum. The total lipid is usually 10–20 µmol. 2. A few milliliters of diethyl ether is washed with a similar volume of distilled or purified water in a tightly capped glass tube by gentle shaking, and the mixture is allowed to separate. One milliliter of the ether (the top layer) is removed by means of a glass pipette or syringe and added to the dried phospholipid film, ensuring that the lipid dissolves completely. 3. The buffer to be encapsulated (0.34 ml) is added to the phospholipid solution in ether. A gentle stream of argon gas is flushed over the mixture using a Pasteur pipette, and the tube is sealed with teflon tape and a screw-cap. The mixture is sonicated for 2–5 min, resulting in a stable emulsion. If an emulsion is not formed it is likely that the sonicator is not at an optimal setting. In this case, either the level of the water in the bath or the power supply needs to be adjusted.

218

Düzgünes¸, Faneca and Lima

4. The cap is opened in the chemical hood, the tube is sealed again with teflon tape, and placed inside the larger glass tube that fits onto the rotary evaporator. About 1 ml of water is added to the outer tube, both to facilitate thermal contact and to minimize the evaporation of the aqueous solution in the inner tube. The outer tube is immersed in the water bath of the rotary evaporator, kept at 30°C. 5. The ether is evaporated gently in controlled vacuum, at about 350 mm Hg, under constant supervision. The tube is purged occasionally with argon gas attached to the top of the evaporator to maintain the vacuum level and to prevent excessive bubbling. The vacuum is allowed to build up when a stable gel is formed. 6. To break up the gel, the inner glass tube is removed and vortexed vigorously for 5–10 s. The tube is placed again in the outer tube and controlled rotary evaporation is resumed. This step is repeated once or twice, until an aqueous opalescent suspension is formed. An additional 0.66 ml of the encapsulation buffer is added to the suspension and rotary evaporation is continued for an additional 20 min to remove any residual ether (see Note 1). 7. To achieve a uniform size distribution, the liposome suspension is passed several times through polycarbonate membranes of 100 nm pore diameter (or other desired diameter), using a high-pressure or syringe extruder. 8. The average size and the size distribution of the liposomes are assessed by dynamic light scattering in a Beckman Coulter N4 Plus Submicron Particle Sizer (or equivalent instrument). 9. Liposomes are stored at 4˚C after gently flushing the top part of the tube with a stream of argon, and sealing the cap with parafilm. 3.1.2. Large Unilamellar Liposomes Prepared by Extrusion of Multilamellar Liposomes

1. Pure or mixed phospholipids (usually 10–20 µmol) in chloroform are measured using gastight Hamilton syringes under a chemical hood, and dispensed into a glass tube. The tube is covered with teflon tape, and placed inside a larger tube that fits on a rotary evaporator until the lipid dries into a thin film in vacuum. The evaporator is purged with argon to reduce the vacuum, and the inner tube is placed in a vacuum jar or vacuum oven, and kept under high vacuum for at least 2 h to eliminate any residual chloroform. 2. The phospholipid film is hydrated with the buffer to be encapsulated (e.g., the Tb citrate solution), the tube is purged with argon gas and covered with teflon tape, and the cap is closed tightly. The mixture is vortexed for 10 min at room temperature to form multilamellar liposomes. If the lipid has a

Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells

219

high gel to liquid-crystalline phase transition temperature, it is hydrated with buffer at a higher temperature. The vortex mixing is interrupted every 30 s to place the tube in a water bath at this high temperature for about 15 s. 3. To achieve a uniform size distribution, the liposome suspension is passed 21 times through a polycarbonate membrane of 100 nm pore diameter (or other desired diameter), using a syringe extruder. If the extrusion requires too much force, the multilamellar liposomes should first be passed through a membrane of larger pore diameter (e.g. 800 or 400 nm) (see Note 2). 4. The average size and the size distribution of the liposomes are assessed by dynamic light scattering in a Beckman Coulter N4 Plus Submicron Particle Sizer (or equivalent instrument). 5. Liposomes are stored at 4˚C after gently flushing the top part of the tube with a stream of argon, and sealing the cap with parafilm. 3.1.3. Small Unilamellar Liposomes Prepared by Sonication

1. Ten micromoles of a phospholipid mixture in chloroform are dispensed into the glass tube that is then covered with teflon tape over the top of the tube. The tube is placed in the larger glass tube that fits on the rotary evaporator, and the chloroform is allowed to evaporate, using condensation precautions to trap the solvent. When the lipid forms a dry film, the evaporator is purged with argon to reduce the vacuum. The inner tube is then placed in a vacuum jar or vacuum oven, and is kept under high vacuum for at least 2 h to eliminate any residual chloroform. 2. The dried phospholipid film is hydrated with the buffer to be encapsulated, the tube is purged with argon gas and covered with Teflon tape, and the screw-cap is closed tightly. The mixture is vortexed for 10 min at room temperature. If a lipid with a high gel to liquid-crystalline phase transition temperature is used, the lipid is hydrated with buffer at a temperature above the transition temperature, and the tube is placed in a water bath at this high temperature in between short periods of vortexing. 3. The multilamellar suspension is sonicated in a bath-type sonicator for 0.5–1 h. The level of water in the bath is adjusted such that the water surface breaks up into small droplets under sonication. The top of the liposome suspension is aligned with the level of the water such that an aerosol forms occasionally in the tube. Overheating of the bath should be avoided, by circulating water through the bath or by adding some ice and re-adjusting the level of the water. For lipids with high transition temperatures, the bath should be maintained at a few degrees above this temperature.

220

Düzgünes¸, Faneca and Lima

4. To remove any remaining large liposomes, the resulting opalescent suspension is centrifuged at 100,000g for 1 h at 4˚C, preferably in a swinging bucket rotor. The supernatant is removed without disturbing the pellet, and used as the small unilamellar liposome preparation (see Note 3). 5. The average size and the size distribution of the liposomes are assessed by dynamic light scattering in a Beckman Coulter N4 Plus Submicron Particle Sizer (or equivalent instrument). 6. Liposomes are stored at 4˚C after gently flushing the top part of the tube with a stream of argon, and sealing the cap with parafilm. 3.1.4. Measurement of the Phospholipid Concentration of Liposomes

1. Liposome samples are placed in triplicate tubes at an estimated amount of less than 0.1 µmol inorganic phosphate. Phosphate standards are pipetted in triplicate (for example 0.01, 0.05, 0.075, 0.1 µmol inorganic phosphate). Four hundred microliters of 10 N H2SO4 are added to each tube that are then heated for 30 min on the heating block. 2. The tubes are cooled at room temperature and 100 µl H2O2 are added using a pipettor or repeater pipette (e.g. Pipetman or Eppendorf). The tubes are returned to the heating block for 30 min on the heating block. The fumes are tested for the absence of H2O2, using the indicator strips. 3. The tubes are placed in a round metal rack (that will fit eventually into an electric water boiler), 4.6 ml of 0.22% ammonium molybdate reagent are added to the tubes and mixed on a vortex mixer. Two hundred microliters of ANSA (or Fiske) reagent are added to each tube and then mixed on a vortex mixer. Alternatively, 100 µl ascorbate can also be used for this step. 4. The metal rack is lowered into a boiling water bath for 7–10 min, and then cooled. The contents of the tubes are transferred to spectrophotometer cuvettes. Protective latex or vinyl gloves are recommended for this procedure. A spectrophotometer with a sipper accessory is preferable to avoid handling of acid-containing tubes. 5. The absorbance of the solutions in each cuvette is measured at 812 nm or at 660 nm (if the solution is too concentrated). The phosphate content of the sample is assessed from the standard curve (see Note 4).

3.2. Liposome Fusion: Intermixing of Aqueous Contents 3.2.1. The Tb/DPA Acid Assay

1. For large unilamellar liposomes, the osmolality of the solutions inside the liposomes should match that of the medium, and if lower, should be adjusted with the addition of small amounts of NaCl. The Tb-, DPA- and Tb/DPA-liposomes are prepared as described in Subheadings 3.1.1 or 3.1.2. (see Note 5).

Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells

221

2. The 3 Sephadex G-75 or G-50 columns are equilibrated with Buffer A. The liposomes are passed through separate columns to eliminate the solution outside the liposomes, using Buffer A for elution. It is necessary to include the 1 mM EDTA in Buffer A to prevent the binding of Tb ions to the phospholipids. The first 4 ml after the liposomes are loaded on the gel are discarded, and the next 3 ml are collected in the 15-ml culture tube. Argon gas is flushed over the liposome suspensions for 10 s without causing any significant disruption of the surface of the liquid; then the tubes are capped tightly, sealed with parafilm, and stored in ice. 3. One milliliter of the collected Tb-liposomes is placed on the third Sephadex column equilibrated with Buffer B, and collected as in step 2. Since EDTA would interfere with the formation of the Tb/DPA complex when the vesicles are lysed to obtain 100% fluorescence, this chromatography step replaces the EDTA in the medium, producing the “Tb minus EDTA liposomes.” 4. The lipid concentration of the liposomes is measured in each of the culture tubes by inorganic phosphate analysis (see Subheading 3.1.3). 5. For a fusion assay utilizing 50 µM total lipid concentration, 25 µM Tb-liposomes are mixed with 25 µM DPA-liposomes. To calibrate the fluorescence to “100%” fluorescence; i.e. the maximal fluorescene that can be obtained if all the liposomes formed a mega-liposome, 25 µM of the “Tb minus EDTA liposomes” are placed in a fluorometer cuvette containing the appropriate volume of Buffer B. The amount of buffer should be adjusted to allow for the volumes of the subsequent ingredients. Ten microliters of the 2 mM DPA stock are added to the cuvette (final DPA concentration 20 µM). To lyse the liposomes, 50 µl of either of the detergent stocks are added and the fluorescence is allowed to equilibrate. The final concentrations of the detergents should be 0.5% (w/v) for Na cholate, or 0.8 mM for C12E8. 6. The excitation wavelength of the fluorometer is set at 276– 278 nm. The emission wavelength is set to 545 nm. Intermediate slit widths are used to optimize the intensity, but to minimize scattering artifacts. The high-pass cut-off filter is placed before the emission monochromator, to minimize any contributions from light scattering. Crossed polarizers can also be used to minimize lightscattering contributions. For our experiments we have used SLM 4000, SLM 8000, and Spex Fluorolog fluoro meters. Other instruments with similar light intensity and sensitivity can be used for this assay.

222

Düzgünes¸, Faneca and Lima

7. A 1:1 mixture of the Tb-liposomes and the DPA-liposomes is prepared as a stock solution in a 15-ml polystyrene culture tube, at a final lipid concentration that is tenfold higher than that used for the actual assay. For example, a suspension of 500 µM lipid is prepared to achieve a final concentration of 50 µM in the assay. The stock solution (0.1 ml) is diluted into 0.9 ml of Buffer B; this process also dilutes the final EDTA concentration to 0.1 mM. 8. For a fluorometer with a strip-chart recorder, the low fluorescence of this suspension is set to 0% Fmax, using the offset function of the fluorometer. Using the gain function, necessary adjustments are made to the 100% level with the cuvette containing the liposomes lysed in the presence of the DPA. The calibration procedure is repeated after a set of measurements to ensure that the lamp intensity has not changed in the course of the experiment. 9. For a fluorometer with computerized data acquisition, the assay can be calibrated by subtracting the initial level of fluorescence [I(0)] from the data set, and dividing the resulting data set by the numerical difference between the fluorescence intensity of the calibration vesicles [I (•)] and the new 0% level. If the resulting value is multiplied by 100, the percentage value is obtained. Thus, the extent of fusion, F(t), as a percentage of maximal fluorescence, is given by

F (t ) = 100 x [I (t ) - I(0)]/[I (∞) - I(0)] where I(t) is the fluorescence intensity at time t. 10. The assay may be calibrated in an alternative manner also. Here, 50 µM of the Tb/DPA liposomes are considered to have the maximal, 100%, fluorescence. These liposomes represent the fusion product of all the Tb-liposomes and DPA-liposomes in the fusion assay. Of course, these calibration liposomes should have the same size distribution and should be used at the same lipid concentration as the total of the Tb- and DPAliposome populations (25 µM of each). One slight complication of the use of Tb/DPA liposomes is that after they are transferred to room temperature from the storage temperature (0˚C), it takes about 0.5 h for the fluorescence to reach a steady state.

3.2.2. The ANTS/DPX Assay

1. Three batches of large unilamellar liposomes are prepared with the ANTS, DPX, and ANTS/DPX solutions. Small unilamellar liposomes have not produced reliable results with this assay (see Note 6). 2. The liposomes are chromatographed on separate Sephadex G-75 or G-50 columns equilibrated with Buffer C. The first 4 ml, measured from the point where the liposomes are

Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells

223

allowed to partition into the gel, are discarded as the void volume. The next 3 ml are collected in the polystyrene culture tube. The tube is flushed gently with argon, capped tightly, sealed with parafilm, and stored in ice until use. 3. The phosphate concentration of each of the liposome preparations is measured using the assay described in Subheading 3.1.4. 4. The excitation monochromator of the fluorometer is set to 360 nm, and the emission monochromator to 530 nm, with relatively wide slit widths (10–20 nm). A high-pass filter (e.g. Corning 3–68) is placed in the emission channel to eliminate light scattering artifacts. 5. ANTS-liposomes (25 µM lipid) are placed in the appropriate amount of buffer (1–2 ml, depending on the fluorometer, with “flea” or “castle” stir-bars), and the gain of the fluorometer is adjusted to an arbitrary unit of 100%. The fluorescence of 50 µM ANTS/DPX-liposomes is taken as 0%. The latter liposomes represent the theoretical fusion product of all the ANTS vesicles and all the DPX vesicles (also 25 µM) used in the assay. 6. Putative fusogens, such as divalent cations, protons, or fusogenic peptides are added to the cuvette from a concentrated stock solution under constant stirring. Liposome fusion results in the decrease of fluorescence due to quenching of ANTS by the DPX. 3.3. Liposome Fusion: Intermixing of Lipids 3.3.1. The NBD/Rhodamine Assay

1. “Labeled liposomes” are prepared as described in Subheadings 3.1.1, 3.1.2, or 3.1.3, with 0.8 mole% each of NBD-PE and Rh-PE in the initial chloroform mixture in addition to other phospholipids (see Note 7). 2. “Unlabeled liposomes” of the same lipid composition are prepared without the fluorophores present. 3. “Calibration liposomes”, or “mock fused liposomes,” are made as in 1, but with 0.08 mole% of each fluorophore (see Note 8). 4. The phosphate concentration of each of the liposome preparations is measured (Subheading 3.1.4), and aliquots from the labeled and unlabeled liposomes are mixed at a ratio of 1:9 in 2 mL of Buffer C. For a total lipid concentration of 50 µM, this would be 5 µM-labeled liposomes and 45 µM-unlabeled liposomes. 5. The excitation and emission monochromators are set to 460 nm and 530 nm, respectively. The maximal (100%) fluorescence is set with 50 µM of the calibration liposomes. The residual fluorescence of the mixture of labeled and unlabeled liposomes is set to 0% fluorescence. The percentage of lipid mixing as a function of time is given by M(t) = 100 × [I(t) − I(0)]/ [I(•) − I(0)], where I(t) is the fluorescence intensity at time t,

224

Düzgünes¸, Faneca and Lima

I(0) is the residual fluorescence, and I(•) is the maximal fluorescence. 6. Fusion of the liposomes results in an increase in NBD fluorescence as the probes are diluted into the unlabeled liposomes. 7. Maximal fluorescence can also be set by lysing the labeled liposomes at the same concentration to be used in the assay, with a detergent that does not affect the quantum yield of NBD. Commercial preparations of Triton X-100 require a correction factor of 1.4–1.5, while some purified preparations of this detergent may have no inhibitory effect on the fluorescence. Alternative detergents are C12E8 or C12E9 (Calbiochem) 3.3.2. Inner Monolayer Mixing Assay

1. Liposomes containing 0.75 mole% each of N-NBD-PS and either N-Rh-PE or diI(5)C18 (20 mM total phospholipid) are prepared in Buffer C, as described in Subheadings 3.1.1 or 3.1.2 (see Note 9). 2. To a 10 mM liposome suspension on ice, dithionate is added to a final concentration of 80–100 mM, and the mixture incubated for 30–45 min. 3. The dithionate is removed by centrifuging100 µl aliquots through the spin columns. 4. The remainder of the assay is as described in Subheading 3.3.1 to measure the intermixing of the inner monolayers.

3.4. Fusion of Liposomes and Lipoplexes with Cultured Cells

1. THP-1 human monocytic leukemia cells are grown in the growth medium described in Subheading 2.4, at 37˚C and 5% CO2. 2. The cells are harvested by centrifugation at 180 g for 7 min at room temperature, and the pellet is resuspended in phenol red-free RPMI-1640 with HEPES buffer. The cells are centrifuged again. This procedure is then repeated twice. The final pellet of cells is suspended at a concentration of 20 × 106 cells/ml in a polypropylene culture tube and kept on ice until use the same day. Cell viability is determined by mixing a small sample of cells 1:1 with the Trypan blue solution, while counting the cells in a hemacytometer. 3. Large unilamellar liposomes containing the cationic lipid DOTAP alone, or mixed with PE or PC, are prepared first by hydration of a dried lipid film in 150 mM NaCl, 10 mM HEPES, pH 7.4, at a final concentration of 5 mM lipid (see Subheading 3.1.2). This is followed by vortexing under argon to form multilamellar liposomes, and by extrusion through polycarbonate membranes of 100 nm pore diameter. Rh-PE is included in the initial chloroform mixture of lipids at 5 mole%, at which concentration it is self-quenched.

Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells

225

4. Lipid mixing between the liposomes and cell membrane is monitored continuously in a magnetically stirred fluorometer cuvette maintained at 37˚C or other desired temperatures. Cells (100 µl) are added to the cuvette containing 1.9 ml of 150 mM NaCl, 10 mM HEPES, pH 7.4, and allowed to equilibrate to the set temperature. Liposomes are added to a final concentration of 20 µM. 5. The fluorescence is measured with the excitation wavelength at 568 nm and the emission wavelength at 586 nm. In a Spex fluorometer, the excitation and emission slits are set to 0.5 and 1 mm, respectively. The fluorescence scale is set with the initial fluorescence of liposomes and cells being 0%, and the maximal fluorescence (100%) is obtained by dissolving the liposomes with 0.5% (v/v) Triton X-100. 6. To prevent endocytosis, and hence to measure fusion at the plasma membrane only, the cells are pre-incubated with 1 µg/ml antimycin A, 10 mM NaF and 0.1% (w/v) NaN3 for 30 min at 37˚C. Fusion is monitored in the presence of these inhibitors. 7. To examine the fusion of cationic liposome-DNA complexes with cells, the complexes are prepared at varying lipid nitrogen/DNA phosphate (+/−) ratios right before they are added to the cells in the cuvette. The effect of the maturation of the complexes on fusion can also be investigated by preincubating the complexes for 1–2 h before addition to the cells (see Note 10). 3.5. Intracellular Delivery of Liposome Contents 3.5.1. Flow Cytometric Analysis

1. Large unilamellar liposomes encapsulating 80 mM calcein and containing 1 mole% Rh-PE in their membrane are prepared by extrusion of multilamellar liposomes (Subheading 3.1.2). 2. Monocytic human THP-1 cells are cultured in RPMI 1640 medium supplemented with 10% FBS. The cells are differentiated to macrophage-like cells by adding 160 nM phorbol 12-myristate 13-acetate in 24-well culture plates (106 cells per well). The cells are incubated for 5–6 days, and the culture medium is replaced with fresh medium. 3. The liposomes are added to the cells at a final phospholipid concentration of 100 µM, and incubated for various times at 37°C. The cells are then washed twice with phosphate-buffered saline (PBS) without calcium or magnesium ions. 4. The cells are detached from the plastic by adding 0.5 ml of dissociation buffer and mixed with 0.5 ml of PBS containing divalent cations, 2% FBS and 1 mg/ml propidium iodide. Rhodamine and calcein fluorescence is detected with a Becton Dickinson FACStar Plus flow cytometer (or equivalent),

226

Düzgünes¸, Faneca and Lima

controlled by a Hewlett-Packard computer with Lysis II software (Becton Dickinson, San Jose, CA). The samples are analyzed for rhodamine using excitation at 528 nm and emission at 575 nm. Calcein is excited at 488 nm and the fluorescence emission is detected at 520 nm with a 0.1 neutral density filter. For each sample, 10,000 events are recorded. Forward scatter and propidium iodide fluorescence signals are used to gate the cell subset of interest and eliminate debris, dead cells, and cell aggregates. 5. Mean rhodamine fluorescence values reflect the binding and uptake of liposomes by the cells. The mean calcein fluorescence indicates the intracellular dequenching of the dye. The calculated ratio of calcein to rhodamine fluorescence is taken to measure the amount of aqueous marker released intracellularly per cell-associated liposome. The initial calcein to rhodamine fluorescence ratio of liposomes bound to the cells, in the absence of endocytosis, is obtained by incubating the liposomes with the cells at 4°C (see Note 11). 3.5.2. Image Analysis

1. THP-1 cells (105/well) are differentiated for 5–6 days in Lab-Tek chambered coverglasses for tissue culture, obtained from Nunc (Naperville, IL), by adding 160 nM phorbol 12-myristate 13-acetate to the culture medium. Cells are washed with cold RPMI media without phenol red, containing 20 mM HEPES buffer, pH 7.4. 2. Liposomes are added to cells at a final phospholipid concentration of 200 µM and incubated for 1 h at 4°C in phenol red-free RPMI medium with 20 mM HEPES buffer. After this pre-binding step, the cells are washed with cold media and the initial calcein and rhodamine fluorescence images are recorded using the Photon Technology International (PTI) ratio imaging system. To evaluate the kinetics of calcein dequenching, the cold medium is removed, medium at 37°C is added, and the cells are incubated for various times. 3. The cells are washed with cold medium, and the calcein and rhodamine fluorescence images are recorded. Cells are observed in a Nikon Diaphot epi-fluorescence microscope (Melville, NY) using a 100× objective and filters for FITC/ TXRD obtained from Chroma (Rockingham, VT). Averages of 16 snapshots are taken to reduce the background. Ratio images are produced using PTI (or equivalent) software. 4. Histograms of calcein to rhodamine ratio for each cell are determined using a square of 100 × 100 pixels, representing the approximate area of one cell at the magnification used. This provides a measure of calcein dequenching per cell. Average histograms for each liposome composition and time points are then calculated in Excel.

Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells

227

5. Medians for each average histogram are calculated from cumulative curves. To calculate the growth of the area occupied by the calcein that is dequenched, the highest ratio in the control experiment (incubation at 4°C) is taken as the cut-off. The sum of the pixels with ratios higher than the cutoff ratio, are taken to estimate the dequenched area at various time points (see Note 12). 3.5.3. Fluorometric Measurements

1. THP-1 cells (5 × 106) are differentiated by incubation for 5–6 days in 5–ml of medium containing 160 nM phorbol 12-myristate 13-acetate in 6-well culture plates. The medium is removed and the cells are washed with phenol red-free RPMI medium with 20 mM HEPES buffer. 2. Cells are incubated with liposomes at a final phospholipid concentration of 150–200 µM for different times at 37°C in phenol red-free RPMI medium/20 mM HEPES buffer, and then washed. 3. Two milliliters of dissociation buffer are added, and the cells are incubated for 10 min at 37°C. The cells are detached from plastic with disposable scrapers (Costar Corporation, Cambridge, MA) and transferred into disposable fluorometer cuvettes (Hughes & Hughes Limited, Tonedale, Wellington). 4. Fluorescence measurements are performed in a SPEX Fluorolog 2 fluorometer (SPEX Industries, Inc. Edison, NJ), Perkin Elmer LS50 fluorometer or equivalent instrument. Calcein fluorescence is read at excitation and emission wavelengths of 490 and 520 nm, respectively, using 0.5 mm excitation and 1.0 mm emission slits. Rhodamine fluorescence is measured at excitation and emission wavelengths of 568 and 600 nm, respectively, using 0.5 mm excitation and 1.5 mm emission slits. 5. The sample chamber is adjusted to front-face configuration to prevent cell-scattering effects and is equipped with a magnetic stirrer. The temperature is maintained at 20°C with a thermostatic water circulator (see Note 12).

3.6. Liposome Permeability 3.6.1. Release of Carboxyfluorescein or Calcein

1. Large unilamellar liposomes are prepared using 50 mM carboxyfluorescein or calcein (Na salt) with 10 mM TES, pH 7.4. This concentration is iso-osmotic with Buffer C. 2. Small unilamellar liposomes are prepared using 100 mM carboxyfluorescein or calcein, 10 mM TES, pH 7.4. Larger concentrations of encapsulated material can be used in this case, since these liposomes are not active osmotically. 3. The unencapsulated material is separated by gel filtration on Sephadex G-75, using Buffer C as the elution buffer. 4. The phosphate concentration of the liposomes is measured as in Subheading 3.1.4.

228

Düzgünes¸, Faneca and Lima

5. The excitation wavelength is set to 490 nm and the emission wavelength to 530 nm, using 0.5 mm excitation and 1.0 mm emission slits in the Spex fluorometer, and 5 nm and 10 nm slits in the Perkin-Elmer fluorometer. 6. The maximal fluorescence (100%) of a certain concentration of liposome is set by lysing the vesicles with 0.1% (w/v) Triton X-100 or 0.8 mM C12E8. 7. The extent of release, R(t) may also be calculated from R(t) = 100 × [I(t) − I(0)]/[I(•) − I(0)], where I(t) is the fluorescence intensity at time t, I(0) is the initial residual fluorescence of the liposomes before any membrane-active agents are added, and I(•) is the fluorescence obtained when the liposomes are lysed with detergent (see Note 13). 3.6.2. Release of the Tb/DPA Complex

1. Liposomes are prepared using the Tb/DPA solution appropriate for large or small unilamellar liposomes (Subheading 3.1). 2. The lipsomes are passed through the column to eliminate unencapsulated material, using Buffer A for elution. 3. The phosphate concentration of the liposomes is measured as in Subheading 3.1.4. 4. A stock solution is prepared in Buffer B at a lipid concentration tenfold higher than that to be used in the final assay. The stock solution (0.1 ml) is diluted into 0.9 ml of Buffer B, diluting the final EDTA concentration to 0.1 mM. 5. If the liposomes are not affected by the presence of Ca2+ (i.e. they do not contain a high-mole fraction of anionic phospholipids), this cation should be included in the external buffer to facilitate the dissociation of the Tb/DPA complex. 6. The initial fluorescence (excitation at 278 nm, emission at 545 nm, with the 3–68 cutoff filter) of the liposomes is set to 100%. After placing the liposomes into buffer, it is necessary to allow some time for the fluorescence to stabilize. The minimum fluorescence, corresponding to maximal leakage, can be set by lysing the liposomes with detergent (0.1% (w/v) Triton X-100 or 0.8 mM C12E8 (final concentrations)). 7. The extent of release, R(t), may be calculated from R(t) = 100 × [I(0) − I(t)]/[I(0) − I(•)], where I (t) is the fluorescence intensity at time t, I(0) is the initial fluorescence of the Tb/DPA liposomes before any membrane-active agents are added, and I(•) is the fluorescence obtained when all the liposome contents leak. The liposomes are lysed with detergent (see Note 14).

3.6.3. The Release of ANTS/DPX

1. Large unilamellar liposomes are prepared to encapsulate the ANTS/DPX solution. This assay is not suitable for use with small unilamellar liposomes.

Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells

229

2. The liposomes are passed through a Sephadex column to eliminate unencapsulated material, using Buffer C for elution. 3. The phosphate concentration of the liposomes is measured as in Subheading 2.1.4. 4. The fluorometer is set as follows: Excitation wavelength at 360 nm, the emission wavelength at 530 nm, using relatively wide slit widths (10–20 nm), and a high-pass filter (Corning 3–68) in the emission channel to eliminate light-scattering artifacts. 5. The assay is calibrated to maximal fluorescence by lysing the vesicles with 0.1% (w/v) Triton X-100 or 0.8 mM C12E8 (final concentrations). The extent of release as a function of time can be calculated according to the formula in Subheading 3.6.1. (see Note 15).

4. Notes 1. The reverse-phase evaporation method was first reported by Szoka and Papahadjopoulos [1] and further developed by Düzgünes¸ et al. [2]. 2. Extrusion of multilamellar liposomes to obtain large unilamellar liposomes was developed by Olson et al. [3] and later by Mayer et al. [4]. 3. Small unilamellar liposomes prepared by sonication were first reported by Papahadjopoulos and Miller [5]. The method described here is adapted from Düzgünes¸ et al. [2]. 4. The method for determination of inorganic phosphate described here was first described by Bartlett [6]. The details described here were developed in the Papahadjopoulos laboratory at UCSF by T. Heath and D. Alford. 5. The interaction of Tb and DPA results in the formation of the fluorescent [Tb(DPA)3]3- chelation complex. Internal energy transfer from DPA to Tb results in an increase in the fluorescence intensity of the latter by four orders of magnitude. The reactants are encapsulated in different populations of liposomes. Membrane fusion results in an increase in fluorescence [7, 8]. Contents that are released into the external medium are prevented from interacting by the presence of EDTA in the medium. Divalent cations used to induce fusion in same cases further inhibit the interaction. The presence of citrate in the Tb-liposomes is essential to prevent the interaction of Tb3+ with anionic phospholipids in the liposome

230

Düzgünes¸, Faneca and Lima

membrane. If phosphatidic acid is used as an anionic lipid, it is necessary to use a stronger chelator of Tb3+, such as nitrilotriacetic acid. 6. This assay is based on the collisional quenching of ANTS fluorescence by DPX. These reagents are encapsulated in separate populations of liposomes [9]. Membrane fusion results in the intermixing of ANTS and DPX and the quenching of ANTS fluorescence. If the contents are released and diluted into the medium, fluorescence quenching does not occur, since high concentrations of DPX are required for quenching. 7. The lipid-mixing assay measures the dilution of a fluorescence energy donor/acceptor pair from “labeled” liposomes to “unlabeled” liposomes as a result of membrane fusion. The efficiency of resonance energy transfer is then decreased, since the rate and efficiency of the transfer depend on the inverse sixth power of the distance between the two fluorophores (as well as the overlap of the emission spectrum of the donor and the absorption spectrum of the acceptor). Low concentrations of probes, usually less than 1 mole% of total lipid, are used for the assay, minimizing the extent of membrane perturbation. We describe a resonance energy transfer pair that has been used extensively by a number of laboratories [10, 11]. The fluorophores are attached to the headgroups of phosphatidylethanolamine and thus do not cause any appreciable perturbation of bilayer packing. 8. To avoid the use of detergents that may affect NBD fluorescence, liposomes corresponding to the fusion product of the labeled and unlabeled liposomes can be used to calibrate the assay to 100% fusion [12]. 9. The interaction of ions, proteins and other fusogens with the outer monolayer of liposomes containing fluorescent probes may alter the fluorescence intensity or lateral distribution of the probes. If the fluorophores were located only in the inner monolayer of liposomes, the fluorescence would not be affected by ion or protein binding to the liposome surface. In this method, the fluorophores exposed on the outer monolayer of the liposome are reduced with dithionate, thereby eliminating the fluorescence of outer monolayer fluorophores. N-NBD-phosphatidylserine (PS) was found to be more suitable for these experiments than N-NBD-PE, since it is less prone to transbilayer movement following reduction [13]. 10. Results obtained by this method have been described by Pires et al. [14].

Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells

231

11. This method was first described by Slepushkin et al. [15] to evaluate the ability of sterically stabilized pH-sensitive liposomes to deliver their contents into cultured cells. 12. Results obtained with this method may be found in Simões et al. [16]. 13. The use of liposomal carboxyfluorescein and calcein was described by Weinstein et al. [17] and Allen et al. [18], respectively. The effect of bacterial toxins on liposome membrane permeability, using these fluorophores, was reported by Kayalar and Düzgüneş [19]. 14. The dissociation of the Tb/DPA complex was initially used to measure liposome permeability during membrane fusion [20]. 15. The ANTS/DPX release assay was first developed by Ellens et al. [21].

References 1. Szoka F Jr, Papahadjopoulos D (1978) Procedure for preparation of liposomes with large internal aqueous space and high capture by reverse-phase evaporation. Proc Natl Acad Sci USA 75:4194–4198 2. Düzgüneş N, Wilschut J, Hong K, Fraley R, Perry C, Friend DS, James TL, Papahadjopoulos D (1983) Physicochemical characterization of large unilamellar vesicles prepared by reverse-phase evaporation. Biochim Biophys Acta 732:289–299 3. Olson F, Hunt CA, Szoka FC, Vail WJ, Papahadjopoulos D (1979) Preparation of liposomes of defined size distribution by extrusion through polycarbonate membranes. Biochim Biophys Acta 557:9–23 4. Mayer LD, Hope MJ, Cullis PR (1985) Vesicles of variable sizes produced by a rapid extrusion procedure. Biochim Biophys Acta 858:161–168 5. Papahadjopoulos D, Miller N (1967) Phospholipid model membranes. I. Structural characteristics of hydrated liquid crystals. Biochim Biophys Acta 135:624–638 6. Bartlett GR (1959) Phosphorus assay in column chromatography. J Biol Chem 234:466–468 7. Wilschut J, Düzgüneş N, Fraley R, Papahadjopoulos D (1980) Studies on the mechanism of membrane fusion: Kinetics of Ca2+induced fusion of phosphatidylserine vesicles followed by a new assay for mixing of aqueous vesicle contents. Biochemistry 19:6011–6021

8. Düzgüneş N, Wilschut J (1993) Fusion assays monitoring intermixing of aqueous contents. Methods Enzymol 220:3–15 9. Ellens H, Bentz J, Szoka FC (1985) H+- and Ca2+-induced fusion and destabilization of liposomes. Biochemistry 24:3099–3106 10. Struck DK, Hoekstra D, Pagano RE (1981) Use of resonance energy transfer to monitor membrane fusion. Biochemistry 20:4093–4099 11. Hoekstra D, Düzgüneş N (1993) Lipid mixing assays to determine fusion in liposome systems. Methods Enzymol 220:15–32 12. Düzgüneş N, Allen TM, Fedor J, Papahadjopoulos D (1987) Lipid mixing during membrane aggregation and fusion. Why fusion assays disagree. Biochemistry 26:8435–8442 13. Meers P, Ali S, Erukulla R, Janoff AS (2000) Novel inner monolayer fusion assays reveal differential monolayer mixing associated with cation-dependent membrane fusion. Biochim Biophys Acta 1467:227–243 14. Pires P, Simões S, Nir S, Gaspar R, Düzgüneş N, Pedroso de Lima MC (1999) Interaction of cationic liposomes and their DNA complexes with monocytic leukemia cells. Biochim Biophys Acta 1418:71–84 15. Slepushkin VA, Simões S, Dazin P, Newman MS, Guo LS, Pedroso de Lima MC, Düzgüneş N (1997) Sterically stabilized pH-sensitive liposomes: intracellular delivery of aqueous contents and prolonged circulation in vivo. J Biol Chem 272:2382–2388

232

Düzgünes¸, Faneca and Lima

16. Simões S, Slepushkin V, Düzgüneş N, Pedroso de Lima MC (2001) On the mechanisms of internalization and intracellular delivery mediated by pH-sensitive liposomes. Biochim Biophys Acta 1515:23–37 17. Weinstein JN, Yoshikami S, Henkart P, Blumenthal R, Hagins WA (1977) Liposomecell interaction: transfer and intracellular release of a trapped fluorescent marker. Science 195:489–492 18. Allen TM, Cleland LG (1980) Serum-induced leakage of liposome contents. Biochim Biophys Acta 597:418–426

19. Kayalar C, Düzgüneş N (1986) Membrane action of colicin E1: detection by the release of carboxyfluorescein and calcein from liposomes. Biochim Biophys Acta 860:51–56 20. Bentz J, Düzgüneş N, Nir S (1983) Kinetics of divalent cation-induced fusion of phosphatidylserine vesicles: correlation between fusogenic capacities and binding affinities. Biochemistry 22:3320–3330 21. Ellens H, Bentz J, Szoka FC (1984) pHinduced destabilization of phosphatidylethanolamine-containing liposomes: role of bilayer contact. Biochemistry 23:1532–1538

Chapter 17 The Use of Isothermal Titration Calorimetry to Study Multidrug Transport Proteins in Liposomes David Miller and Paula J. Booth Abstract Biophysical measurements of multidrug transporters in vitro can often be of limited relevance to the natural in vivo behavior. In particular, the properties of transporters when removed from their native bilayer and solubilized in detergents or lipids can differ significantly from their in vivo properties, reducing the value of in vitro measurements for the design of antagonists to the transporters. This problem can be addressed by studying the transport protein in liposomes in which the properties of the liposome bilayer are altered through systematic changes in lipid composition. Isothermal titration calorimetry can be used to determine the properties of the lipid-reconstituted protein in bilayers of different lipid compositions as well as to quantify the percentage recovery of functional protein in different lipids. Both these measurements lead to an accurate analysis of substrate binding activity. The approach is illustrated here for the small multidrug transport protein, EmrE from Escherichia coli. The percentage of functional EmrE successfully reconstituted into liposome depends on lipid composition. Differences in ligand binding and subtle differences in the secondary structure also occur in different lipid compositions. Key words: Small multidrug transporters, Ligand binding activity, Dissociation constant, Isothermal titration calorimetry, Reconstitution

1. Introduction Multidrug exporters are of great pharmaceutical interest because of their role in drug resistance of bacterial pathogens and their overexpression in a large number of cancerous cells resulting in resistance to anti-cancer pharmaceuticals. The study of multidrug transporters in vitro is essential for a complete understanding of the mechanism and for the successful design of antagonists to the transporters. The small multidrug transporter EmrE is capable

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_17, © Humana Press, a part of Springer Science+Business Media, LLC 2010

233

234

Miller and Booth

of recognising and exporting a broad range of hydrophobic cationic substrates from Escherichia Coli (E. coli) (1) and has homologues in a number of bacterial species. The behavior of multidrug transporters in synthetic lipid bilayers is influenced by the bilayer composition and the quantity of functionally incorporated protein. The heterogenous nature of membrane bilayers complicates the detailed biophysical studies of integral membrane proteins. Interactions between the integral membrane proteins and the different regions of the bilayer are complex, with hydrophobic interactions occurring between the protein and the central hydrophobic region of the lipid bilayer, whilst polar and charge interactions exist with the lipid headgroups and aqueous solvent on either side of the bilayer. Furthermore, different lipid compositions, varying in headgroup and aliphatic chain composition, have significant effects on the protein function within the membrane, and there are variations in composition between different cellular compartments and different species (2). Altering the bilayer composition affects the properties of a number of integral membrane proteins, such as changes in the activity of enzymes involved in membrane lipid metabolism (Reviewed in (3, 4)), mechanosensitive channel opening properties (5–7), and the stability of the potassium channel KscA (8, 9). The consequences of lipid composition are not always considered in the structural and functional studies of membrane protein, which are often determined in detergent micelles, or in lipid environments that differ greatly from the native bilayer. The amount of functional transporter incorporated into the lipid bilayers will affect the kinetic parameters determined for transport or activity. We have used an isothermal titration calorimetry (ITC) based assay to quantify the amount of functional protein in synthetic liposomes, reconstituted from a ligand-binding competent, detergent-solubilized state in vitro. The ITC assay accurately quantifies the amount of functional protein from the ligand-binding activity and provides information on the ligandbinding properties in different lipid compositions. Specifically, the dissociation constant, Kd, is determined. In the ITC experiment, ligand is added to the transporter protein reconstituted into liposomes. The amount of heat released in the exothermic binding reaction between ligand and protein is measured for increasing ligand concentrations. This generates a binding isotherm from which Kd can be found by fitting the binding data to a particular binding model. The assay presented here is likely to be applicable for the study of other classes of transporters and receptors, provided a suitable ligand is available.

Use of Isothermal Titration Calorimetry

235

2. Materials 2.1. Protein Purification

All chemicals are purchased from Sigma unless otherwise stated. 1. Solubilization buffer: 40 mM Tris HCl, pH 8.2, 100 mM NaCl, 10 mM 2-mercaptoethanol (b-ME), and 4% (w/v) n-Dodecyl-b-D-Malopyranoside (DDM) from Anatrace. All buffers containing b-ME are freshly prepared before use. 2. Ni2+-NTA agarose (QIAGEN) 3. Ni2+ wash buffer 1: 20 mM Tris-HCl, pH 8.3, 400 mM NaCl, 15 mM imidazole, 0.1% w/v DDM, and 5 mM b-ME. 4. Ni2+ wash buffer 2: 20 mM Tris-HCl, pH 8.3, 20 mM NaCl, 15 mM imidazole, 0.1% w/v DDM, and 5 mM 2-mercaptoethanol. 5. Ni2+ elution buffer: 20 mM Tris-HCl, pH 8.3, 25 mM NaCl, 200 mM imidazole, 0.1% w/v DDM, and 5 mM 2-ME. 6. NH4DDM buffer: 15 mM Tris-HCl, pH 7.5, 190 mM NH4Cl, 0.08% (w/v) DDM, and 5 mM 2ME. 7. Denaturing wash buffer: 10 M urea and 5% w/v SDS (173 mM, CMC ~8 mM or ~0.2% w/v) in 20 mM Tris-HCl, pH 8.3. 8. Denaturing elution buffer: 10 M urea, 5% w/v SDS, 20 mM Na Acetate, pH 4.0. 9. Amicon Ultra 50,000 MWCO centrifugal concentrator (Millipore). 10. HiLoad 16/160 Superdex 200 gel filtration column (Amersham).

2.2. EmrE Reconstitution

1. 1,2-Dioleoyl-sn-Glycero-3-Phosphocholine (DOPC), 1,2-Dioleoyl-sn-Glycero-3-[Phospho-rac-(1-glycerol)](DOPG), and 1,2-Dioleoyl-sn-Glycero-3-Phosphoethanolamine (DOPE) (Avanti polar lipids, as powder) prepared as 50 mg/ml stock solutions in either 1:1 chloroform:methanol or cylclohexane. 2. n-Dodecyl-b-D-Malopyranoside (DDM) and Octyl-b-DGlucoside (OG) were purchased from Anatrace and prepared fresh as a 20% (w/v) stock in water. 3. Reconstitution buffer: 15 mM TRIS HCl, pH 7.5 or 8.5, and 190 mM NH4Cl. 4. 1 ml HisTrap columns (Amersham).

2.3. Circular Dichroism

1. Resuspension buffer: 5 mM TRIS HCl, pH 8.0, and 190 mM NH4Cl. 2. Dilution buffer: 20 mM TRIS HCl at pH 7.5 or pH 8.5.

236

Miller and Booth

3. TPP+: tetraphenyphosphonium bromide in 20 mM TRIS HCl, pH 7.5, freshly prepared from a 20 mg/ml stock in water. 4. Ethidium: Ethidium bromide in 20 mM TRIS HCl, pH 8.5, freshly prepared from a 10 mg/ml stock and stored in dark. 2.4. Protein Quantification

1. 1:1 chloroform:methanol solution prepared fresh and stored, wrapped in foil to prevent exposure to light.

3. Methods Efficient reconstitution of membrane protein is essential for ITC and circular dichroism (CD) measurements as large amounts of protein are typically required, and yields of overexpressed membrane protein are often low. Subsequent quantification of reconstituted transporter in lipid bilayers is a prerequisite for accurate activity measurements. The choice of detergent for purification and as a prelude to reconstitution is important as detergents such as DDM, which are very good at stabilising membrane proteins, are very difficult to remove from proteolipid samples, and may interfere with functional assays. Detergents such as OG, although easy to remove from proteolipid samples, do not suitably stabilize EmrE. Detergent choice is particularly important where activity assays involve a pH gradient across the bilayer as this cannot be maintained when residual DDM is present. A pH gradient is required in transport assays of EmrE. However, measurement of ligand binding is not required for the maintenance of a pH gradient, enabling the use of DDM. ITC binding assays provide a quantitative measure of ligand binding through the measurement of Kd. The stoichiometry of binding (n) can also be calculated from the binding isotherm. The binding of two ligands to EmrE is presented: the tight binding TPP+ (small Kd) and the weaker binding ethidium (larger Kd). ITC also has the advantage of determining the quantity of functional protein. Ligand binding to the small multidrug exporters occurs within the transmembrane region, and as ligand binding is a prerequisite for transport, it can be used as a quantitative measure of the amount of functional protein present. A prior knowledge of n enables the amount of functional reconstituted transporter to be determined by ITC (through non-linear curve fitting of the binding data). This is the case for EmrE, where n is known to be 0.5; one ligand binds to an EmrE dimer in DDM (10). The amount of protein in the reconstituted sample can be determined by various other methods. However, these often report on the amount of protein present and not whether it is also functional. A simple method for quantification

Use of Isothermal Titration Calorimetry

237

of the amount of protein involving solubilization of reconstituted protein in chloroform:methanol and quantification by A280 using standards prepared in the same solvent is presented. CD is also used to assess the overall secondary structure content of EmrE reconstituted into different lipid compositions. The use of a high intensity UV source with a detector geometry, suitable for collecting scattered light, is essential for accurate CD measurements of liposome containing samples. A synchrotron radiation source with correct detector geometry is the preferred method for collection of CD data of proteolipid samples (11). Scattering of light from samples and failure to collect a significant proportion of the scattered light will result in significant artefacts in data, most notably at shorter wavelengths. For an accurate comparison between different lipid compositions, concentration of inserted protein must be accurately determined (i.e., by chloroform/methanol solubilization). However, a qualitative comparison of a single lipid composition in the presence and absence of a ligand is possible without an accurate protein concentration, provided all the proteins present are functional. 3.1. Sample Preparation: EmrE Overexpression and Purification

1. EmrE-His (cloned into pT7-7) is overexpressed and purified as previously described (12, 13). Briefly, after protein expression, cultures are harvested and disrupted using a cell disrupter. Membranes are isolated by centrifugation, solubilized at 4°C for at least 2 h in solubilising buffer. Solubilized material is diluted 1:1 with water, and insoluble material removed by centrifugation at 35,000× g for 30 min at 4°C. NaCl and imidazole are added to final concentrations of 350 mM and 15 mM, respectively. 2. DDM solubilized protein is subsequently incubated with 1.5 ml of Ni-NTA agarose for 1.5 h at 4°C. Non-specifically bound protein is removed by washing with 20 columns of Ni2+ wash buffer 1, followed by 20 column volumes of Ni2+ wash buffer 2. 3. EmrE is eluted with 10 column volumes of Ni2+ elution buffer. Eluted protein is concentrated to 2 ml using an Amicon Ultra-15 Ultracel 50 k centrifugal concentrator (Millipore) and applied to a HiLoad Superdex 200 gel filtration column (Amersham), equilibrated with NH4DDM buffer. 4. A280 of eluted protein is monitored and fractions collected. Fractions which contain a symmetric protein elution peak, determined to contain EmrE, are collected; fractions outside of this range are discarded, including those which contributed to small shoulders on the EmrE elution peak. Eluted EmrE is concentrated to approximately 1–2 mM as determined by A280 (A280 of 1 mg/ml = 2.56) (10), snap frozen in liquid N2 and stored at −80°C until required.

238

Miller and Booth

3.2. Reconstitution of EmrE in Liposomes

1. Fresh stock solutions of the lipid: DOPC, DOPE, and DOPG at 50 mg/ml are prepared in cyclohexane and mixed to the desired molar ratios, evaporated to a thin film at room temperature and freeze-dried overnight. Dried lipid mixtures are rehydrated to 6.25 mg/ml (for ITC TPP+ binding studies) or 12.5 mg/ml (for ethidium binding) in reconstitution buffer at pH 7.5 or 8.5 for TPP+ or ethidium binding experiments, respectively (the higher pH being required for the weaker binding ethidium substrate). 2. Lipids are stirred for 15 min at room temperature (~22°C) and OG added from a 20% (w/v) stock solution to 0.65% or 1% (w/v) for TPP+ and ethidium binding experiments, respectively, and stirred for an additional hour. Lipid vesicles are extruded to 200 nm diameter by passaging 10 times through a polycarbonate filter (Nucleopore Track-Etch Membrane from Whatman) using a LIPEXTM extruder (Northern Lipids Inc). 3. EmrE is reconstituted to a calculated maximum of 10 µM (for TPP+ experiments) or 50 µM (for ethidium) by rapid mixing of 2.2 ml (for TPP+) or 4.4 ml (for ethidium) of lipid vesicle preparation to purified EmrE, giving a final EmrE concentration of 10 or 25 µM for TPP+ and ethidium, respectively, followed by incubation at room temperature for 30 min. Higher concentrations of EmrE are required for weaker binding ligands (Note 1) 4. Lipid vesicles are diluted 90-fold into the appropriate pH reconstitution buffer, stirred at room temperature for 1 h. 5. Proteoliposomes are recovered by centrifugation at 370,000× g for 50 min, followed by resuspension in 2.2 ml of reconstitution buffer. Reconstitution efficiency varies depending on the lipid composition. Therefore, quantification of the amount of functional protein present in the bilayer is essential for activity measurements of either transport or binding assays.

3.3. Isothermal Titration Calorimetry (ITC)

1. All ITC experiments are performed with a VP-ITC microcalorimeter (MicroCal), using supplied Origin 5.0 software (MicroCal) for data analysis. 2. ITC experiments provide a large amount of information, but are susceptible to a number of artifacts that are not normally observed in other binding assays. It is essential that buffers used in ITC experiments are identical (Notes 2–4). 3. All samples are thoroughly degassed and equilibrated to 5°C below the experimental temperature before use with a Thermovac (MicroCal). 4. Experiments are performed at 25°C, monitoring the change in power to the measurement cell compared to the reference cell.

Use of Isothermal Titration Calorimetry

239

The integral over time of the differential power is equal to the thermal energy (DH), and is dependent on the binding reaction monitored in the particular buffer conditions (Note 5). Typically, experiments are performed using 1 initial injection of 7.5 µl of ligand, followed by 17 injections of 15 µl or using 1 initial injection of 10 µl of ligand, followed by 14 injections of 20 µl for TPP+ and ethidium binding experiments, respectively. Ligand concentrations between 20 and 40 µM TPP+ and 100–200 µM ethidium are used depending on the lipid composition and the percentage reconstitution of EmrE. Higher concentrations of ligand are required in experiments with a greater percentage of reconstituted EmrE to ensure saturation of binding sites after addition of all ligands (Note 6). 5. Control experiments to determine the heats of dilution of ligand are performed using lipid vesicles in the absence of EmrE (for TPP+ and ethidium binding experiments) and lipid vesicles containing a non-functional EmrE mutant, E14C (for ethidium binding experiments). Control heats of dilution for ligand-binding experiments were subtracted from either heats of dilution at saturation of EmrE or controls performed with ligand and lipid vesicles containing no protein; these two approaches are often identical. Controls are difficult to subtract from ethidium binding data because of differences in the background ethidium binding properties of liposomes. Control heats of dilution are determined to be linear for ethidium plus liposomes with no protein, and ethidium plus lipid vesicles containing inactive E14C EmrE. Heats of dilution for ethidium binding experiments were determined by subtracting a linear background determined from binding data after saturation. 6. The first data point in ITC binding experiments is always excluded from analysis because of premixing of the ligand solution in the injection syringe tip. After subtraction of a linear correction for control heats of TPP+ and ethidium dilution, binding data are fitted to a single-site binding model using the equations (14): 2    4X t  nM t ∆HV o  Xt Xt 1 1  1+ + − 1 + + − Q =  nM t nKM t 2 nM t nKM t  nM t    

Where DH is the molar heat of ligand binding, Vo is the working volume of the calorimeter cell, Q is the total heat content of Vo, Xt is the bulk concentration of ligand, Mt is the bulk concentration of receptor, and n is the number of binding sites. The heat evolved from the ith injection can be calculated as:

240

Miller and Booth

∆Q (i) = Q (i) +

dVi  Q (i) + Q (i − 1)   − Q (i − 1) Vo  2 

here DQ(i) is the heat released from the ith injection and Vi W is the injection volume. Fitting of the experimental data to the above equations is handled by Origin 5.0 (MicroCal) to determine the values for n, K, and DH. Kd is calculated as 1/K. Representative data of TPP+ and ethidium binding to EmrE in liposomes consisting of 100% DOPC is shown in Fig. 1. 3.4. Circular Dichroism Spectroscopy

1. Circular Dichroism (CD) measurements are performed at a Synchrotron Radiation Source (those shown were collected at Daresbury, UK). 2. Preparation of samples for CD measurements is essentially the same as required for ITC. EmrE for ligand-binding experiments was reconstituted as described above, with the exception of EmrE at 20 µM and lipid at 6.25 mg/ml.

Fig. 1. ITC data of EmrE binding the ligand (a) and (b): TPP+ in DOPC lipids, (c) and (d): EmrE binding ethidium in DOPC. Raw data is shown in (a) and (c) and corrected and integrated data shown in (b) and (d). Values of nexp calculated from the non-linear fitting of the binding isotherm (and resulting values for percentage reconstitution) of functional protein are 0.26 (52%) and 0.23 (46%) for TPP+ and ethidium binding experiments shown in (b) and (d), respectively

Use of Isothermal Titration Calorimetry

241

3. After recovery of liposomes by ultra-centrifugation, samples are resuspended in 5 mM TRIS HCl, pH 8.0, and 190 mM NH4Cl. Samples are diluted in 20 mM TRIS HCl at pH 7.5 with ligand at 50 µM or buffer only. Assuming maximum recovery of sample, the final EmrE and lipid concentrations is 30 µM and 9.4 mg/ml, respectively. Any insoluble material is removed by centrifugation at 13,000× g for 10 min, prior to measurement. 4. All CD measurements are performed in a 0.1 mm cuvette (Helma) at 25°C using a 1 nm wavelength increment and slit width. Duplicate samples are prepared and the resulting scans averaged. The amount of time required for measurement at each wavelength is dependent on the UV source used. A nonsynchrotron source will require significantly longer measurement times. CD data determined for EmrE in different lipid conditions in the presence and absence of the ligand TPP+ is shown in Fig. 2.

Fig. 2. Circular dichroism of EmrE, reconstituted in different lipid compositions in the presence and absence of the ligand TPP+. Data are shown for ligand binding in (a): 1 mole fraction (100%) DOPC, (b): 1 mole fraction (100%) DOPG, (c): 0.5 mole fraction DOPE+ 0.5 mole fraction DOPC (50% DOPE, 50% DOPC), and (d): 0.5 mole fraction DOPE + 0.5 mole fraction DOPG (50% DOPE, 50% DOPG). Solid lines: EmrE in the absence of ligand, dashed lines: EmrE in the presence of saturating concentrations of TPP+. Samples + and − ligand are prepared from identical proteolipid stock solutions and are directly comparable within each lipid composition

242

Miller and Booth

3.5. Protein Quantification by Solubilization of Lipid Vesicles with Organic Solvent

1. Proteolipid samples are solubilized in 1:1 chloroform:methanol, resulting in a final mix of 43:43:14 chloroform:methanol:sample (v/v). Other ratios may be used, as long as the resulting mixture does not phase separate. 2. The absorbance spectrum is measured between 250 and 310 nm to determine the protein absorbance at 280 nM (Note 7). Although only absorbance at 280 nM is required for protein concentration determination, it is useful to check the absorbance around this region to determine the contribution of other components to the measured absorbance. 3. The absorbance of EmrE standards of known concentration in DDM is determined in the region from 250 to 310 nm. 4. A linear baseline is subtracted from the absorbance spectra between 255 and 305 nm to eliminate non-protein absorbance and the UV absorbance band integrated. The band intensity can also be used for calibration, but integration of the area will provide a more accurate concentration. 5. Non-linear regression is used to determine the sample protein concentration. An example of representative data is shown in Fig. 3.

3.6. Protein Quantification by Non-linear Curve Fitting of ITC Binding Data

1. An ITC binding isotherm of the protein of interest is obtained and a value of nexp from the non-linear curve fitting obtained. To increase the accuracy of the determined nexp, EmrE can be reconstituted to a higher concentration of 50 mM in 12.5 mg/ml, and ITC experiments performed using TPP+ at 200 mM using 5 ml injections.

Fig. 3.Protein quantification by solubilization of proteolipid samples in chloroform:methanol. Raw data are shown for standards (black lines) and an EmrE sample in DOPC (grey line). Standards are prepared from accurately determined EmrE samples in DDM solubilized in chlorofrm:methanol. Calibration curve derived from the integration of protein absorbance is shown as an inset. ITC data of TPP+ binding to the sample is shown in Fig. 1

Use of Isothermal Titration Calorimetry

243

2. Percentage reconstitution is calculated from: % reconstitution = nexp/nbinding × 100, where nexp is the experimentally determined value for n and nbinding is the known stoichiometry of the interaction (Notes 8 and 9).

4. Notes 1. The range of binding affinities that can be obtained by ITC is limited at the lower end (high affinity) by the sensitivity of the instrument, and at the upper end (low affinity) by the solubility and available concentrations of receptor. Realistically, for membrane proteins in liposomes, this limits the range from low nanomolar to low micromolar affinities. The lower limits of detection for ITC measurement require receptor at a concentration of ~ 5 µM; if ligand-binding affinity is weak then concentrations will have to be significantly higher. Although it is possible to work with low concentrations of protein to determine a dissociation constant, Kd, for weak binding, this will not produce enough data for fitting all parameters and values for the ligand:protein binding stoichiometry, n, will typically need to be defined. This is not possible if accurate protein concentrations are unknown or when reconstitution is inefficient. 2. Identical buffers for ITC experiments can be obtained either by dialysis of the protein in the required buffer, followed by the preparation of the ligand in dialysate, or ideally by gel filtration of the protein and preparation of the ligand in the gel filtration buffer. This has an advantage over dialysis of a final polishing purification of the protein and complete buffer exchange into the required buffer for ITC in less time than required by dialysis. 3. Ligand pH should be checked after solubilsation to ensure that there is no significant deviation in pH between ligand and protein solutions. Any pH differences must be adjusted, or significant buffer dilution effects will be observed. 4. Care should be taken with reducing agents in ITC as these can result in a significant deviation of the baseline during an experiment. Reducing agents such as DTT or TCEP should be avoided and concentrations of b-ME should be reduced. 5. ITC measurements may not be possible due to a small enthalpy change of the binding reaction under the observed conditions. Altering the temperature of a given reaction will result in a change in the observed enthalpy. If the binding reaction involves a protonation or deprotonation event, a buffer such as TRIS which has a large enthalpy of protonation can be used

244

Miller and Booth

to increase the signal. Care should be taken in the analysis of the resulting enthalpy of binding in such cases, as this will be a sum of all enthalpy changes in the cell including changes observed from the buffer. 6. The concentration of ligand required for binding experiments is dependent on the percentage reconstitution of functional protein. An approximate value for this can be determined from an initial binding experiment using an estimated concentration of ligand. Ideally upon completion of an ITC binding experiment, ligand should be at approximately two-fold excess to the available binding sites to ensure that a complete binding isotherm is obtained. 7. Evaluation of the protein absorbance band is necessary to ensure that the measured sample concentrations are accurate. After solubilization in organic solvent, the protein absorbance from approximately 250--300 nm should be determined and compared to control samples. 8. Accurate measurements of binding stoichiometry and quantification of functional protein are dependent on an accurate protein concentration – errors in protein concentration will be observed in any calculated stoichiometry. Such errors are typically obvious if the binding stoichiometry of fully functional protein in detergent is determined to be a non-integer value. Colorimetric assays for protein concentration may not provide an accurate value for membrane proteins, even when accounting for the presence of detergents. An accurate extinction coefficient for protein absorbance at 280 nm is ideal which can be determined by amino acid analysis. 9. Care must be taken with the assumption that the amount of ligand binding is proportional to the amount of functional protein. This may not be valid for integral membrane proteins in which ligands bind to a soluble domain that may function independently of the membrane-spanning region such as some classes of receptors. In the case of small multidrug transporters, the ligand-binding region is within the membrane and very little of the structure extends beyond the bilayer.

Acknowledgments This work was supported by the BBSRC (B19845), Leverhulme Trust (F/00182/AW) and The Royal Society. PJB holds a Royal Society Wolfson Research Merit Award. We thank Kalypso Charalambous, Paul Curnow, Mark Lorch, and other members of PJB’s research group for discussions of experimental work and Kath Moreton for excellent technical assistance.

Use of Isothermal Titration Calorimetry

245

References 1. Rotem D, Schuldiner S (2004) EmrE, a multidrug transporter from Escherichia coli, transports monovalent and divalent substrates with the same stoichiometry. J Biol Chem 279(47):48787–48793 2. Morein S, Andersson A, Rilfors L, Lindblom G (1999) Wild type Escherichia coli cells regulate the membrane lipid composition in a “window” between gel and non-lamellar structures. J Biol Chem 271:6801–6809 3. Johnson AE, van Waes MA (1999) The translocon: a dynamic gateway at the ER membrane. Annu Rev Cell Dev Biol 15:799–842 4. Cornell R, Arnold R (1996) Modulation of the activities of enzymes of membrane lipid metabolism by non-bilayer-forming lipids. Chem Phys Lipids 81(2):215–227 5. Perozo E, Kloda A, Cortes DM, Martinac B (2002) Physical principles underlying the transduction of bilayer deformation forces during mechanosensitive channel gating. Nat Struct Biol 9(9):696–703 6. Moe P, Blount P (2005) Assessment of potential stimuli for mechano-dependent gating of MscL: effects of pressure, tension, and lipid headgroups. Biochemistry 44(36): 12239–12244 7. Perozo E, Cortes DM, Sompornpisut P, Kloda A, Martinac B (2002) Open channel structure of MscL and the gating mechanism

8.

9.

10.

11.

12.

13.

14.

of mechanosensitive channels. Nature 418(6901):942–948 van den Brink-van der Laan E, Chupin V, Killian JA, de Kruijff B (2004) Stability of KcsA tetramer depends on membrane lateral pressure. Biochemistry 43(14):4240–4250 van den Brink-van der Laan E, Chupin V, Killian JA, de Kruijff B (2004) Small alcohols destabilize the KcsA tetramer via their effect on the membrane lateral pressure. Biochemistry 43(20):5937–5942 Butler PJ, Ubarretxena-Belandia I, Warne T, Tate CG (2004) The Escherichia coli multidrug transporter EmrE is a dimer in the detergentsolubilised state. J Mol Biol 340(4):797–808 Wallace BA (2000) Synchrotron radiation circular-dichroism spectroscopy as a tool for investigating protein structures. J Synchrotron Radiat 7(Pt 5):289–295 Tate CG, Kunji ER, Lebendiker M, Schuldiner S (2001) The projection structure of EmrE, a proton-linked multidrug transporter from Escherichia coli, at 7 A resolution. Embo J 20(1–2):77–81 Muth TR, Schuldiner SA (2000) membraneembedded glutamate is required for ligand binding to the multidrug transporter EmrE. Embo J 19(2):234–240 MicroCal (1998) ITC Data Analysis in Origin. MicroCal

Chapter 18 Studying Lipid Organization in Biological Membranes Using Liposomes and EPR Spin Labeling Witold K. Subczynski, Marija Raguz, and Justyna Widomska Abstract Electron paramagnetic resonance (EPR) spin-labeling methods provide a unique opportunity to determine the lateral organization of lipid bilayer membranes by discrimination of coexisting membrane domains or coexisting membrane phases. In some cases, coexisting membrane domains can be characterized without the need for their physical separation by profiles of alkyl chain order, fluidity, hydrophobicity, and oxygen diffusion-concentration product in situ. This chapter briefly explains how EPR spin-labeling methods can be used to obtain the above-mentioned profiles across lipid bilayer membranes (liposomes). These procedures will be illustrated by EPR measurements performed on multilamellar liposomes made of lipid extracts from cortical and nuclear fractions of the fiber cell plasma membrane of a cow-eye lens. To better elucidate the major factors that determine membrane properties, results for eye lens lipid membranes and simple model membranes that resemble the basic lipid composition of biological membranes will be compared. Key words: Liposomes, Lipid bilayer, Membrane domain, Cholesterol, Lens lipid, Hydrophobic barrier, Fluidity, Order, Oxygen permeation, Spin label, EPR

1. Introduction Electron paramagnetic resonance (EPR) spin-labeling methods provide information about the lateral organization of lipid membranes and also about the molecular dynamics and structure in the direction of the membrane depth. Using lipid spin labels, with EPR monitoring groups (free radical nitroxide moieties) located at different depths in the membrane, profiles of different membrane properties across the lipid bilayer can be obtained. Because of the overall similarity of the molecular structures of

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_18, © Humana Press, a part of Springer Science+Business Media, LLC 2010

247

248

Subczynski, Raguz, and Widomska

Fig. 1. Chemical structures of lipid spin labels including phospholipid analogs: n-PC, T-PC, and n-SASL and cholesterol analogs: CSL and ASL. Chemical structures of POPC and cholesterol molecules are included to illustrate approximate localizations of these molecules and nitroxide moieties of spin labels across the membrane

these spin labels with phospholipids and cholesterol (Fig. 1), they should – to a certain degree – approximate the distribution of phospholipid and cholesterol molecules between membrane domains, as well as cholesterol–phospholipid and cholesterol– cholesterol interactions in the membrane. These spin labels can be distributed between different membrane domains or membrane phases, which makes it possible not only to discriminate these domains and phases but also to characterize them without the need for their physical separation by profiles of certain membrane properties obtained in coexisting domains and phases. This is the case for the raft domain that coexists within bulk lipids (1–3) or the liquid-ordered phase that coexists with the liquid-disordered or solid-ordered phases (4, 5). Information concerning coexisting membrane domains is practically missing from membrane research. For some membrane compositions, and membrane lateral organization, the distribution of the lipid spin label is unique; it allows

Lipid Organization in Biological Membranes

249

Fig. 2. Schematic drawing of the raft domain in bulk lipids (a) and the pure cholesterol crystalline domain in the bulk phospholipid-cholesterol bilayer (b). The distribution and approximate localization of lipid spin labels in these domains are also shown. Phospholipid spin labels 5-, 10-, 16-, T-PC, and 9-SASL are distributed between the raft domain and bulk lipids (a) and are only located in the bulk phospholipid-cholesterol domain when it coexists with the pure cholesterol crystalline domain (b). Spin-labeled cholesterol analogues ASL and CSL are distributed between both domains for coexisting raft domain and bulk lipids (a) and for coexisting cholesterol crystalline and bulk phospholipid-cholesterol domains (b). The nitroxide moieties of spin labels are indicated by black dots

additional information about the structure and dynamics of coexisting membrane domains to be obtained. This is the case for membranes overloaded with cholesterol in which pure cholesterol crystalline domains are formed (6). Figure 2 is a schematic drawing illustrating the cases mentioned above. Both phospholipid-type and cholesterol analogue spin labels are distributed between raft and bulk domains (Fig. 2a), which allows these domains to be identified using the discrimination by oxygen transport (DOT) method, and profiles of the oxygen transport parameter (oxygen diffusion-concentration product) across each domain

250

Subczynski, Raguz, and Widomska

can be obtained (7). In membranes with the cholesterol crystalline domain, phospholipid-type spin labels should partition only into the bulk phospholipid-cholesterol domain (Fig. 2b). Thus, in addition to the profile of the oxygen transport parameter, profiles of the order parameter, fluidity, and hydrophobicity can be obtained using these spin labels (which should only describe properties of the bulk phospholipid-cholesterol domain without “contamination” from the cholesterol crystalline domain) (6). Cholesterol analogues should distinguish between both domains (Fig. 2b), and thus only ASL and CSL can detect and discriminate both the coexisting domains and allow information about the pure cholesterol crystalline domain to be obtained (6). The DOT method has already been described in Methods in Molecular Biology (7). This method has been successfully applied to discriminate the domains in reconstituted membranes crowded with integral membrane proteins (8), as well as in influenza virus envelope membranes, which contain cholesterol- and protein-rich raft domains (1). In model membranes, made from binary mixtures of phosphatidylcholine and cholesterol or sphingomyelin and cholesterol, liquid-ordered, liquid-disordered, and solidordered phases were distinguished and characterized in different regions of a phase diagram when they formed a single phase or when two phases coexisted (4, 5). In membranes made from a ternary raft-forming mixture, the raft domain was also distinguished from bulk lipids by the DOT method (2, 3). Similarly, the DOT method can be used to study the lipid organization in lipid bilayer membranes (liposomes), derived from the lipid extract of certain biological membranes. The main focus of this chapter will be on new applications of EPR spin-labeling methods to study membranes overloaded with cholesterol, like those of eye-lens fiber cell plasma membranes (6, 9, 10). The presence of the pure cholesterol crystalline domain in these membranes broadens the possibilities of the DOT method, which, in combination with conventional EPR spin-labeling approaches, enables rather complete information about membrane structure and its physical properties to be obtained (6). 1.1. EPR Spin-labeling Approaches for Profiles of Membrane Properties

To obtain detailed profiles across lipid bilayers, a variety of lipid spin labels are incorporated into the membrane for probing at specific depths and specific membrane domains (Fig. 1). EPR spin-labeling methods apply conventional EPR and saturationrecovery EPR techniques. Information obtained from conventional EPR spectra includes profiles of the order parameter (11) and hydrophobicity (12). Saturation-recovery EPR signals primarily provide information about collisions between paramagnetic molecules, including nitroxide–nitroxide (13), nitroxide–oxygen (14, 15), and nitroxide–paramagnetic metal ions (12). These collisions can be extracted from conventional EPR measurements

Lipid Organization in Biological Membranes

251

using the line-broadening approach (16) or continuous wave power saturation measurements (17), however, with less detail and less precision. Profiles of spin-lattice relaxation time, obtained directly from saturation-recovery EPR signals, also contain useful information about the dynamics (fluidity) of the membrane interior. All of the profiles presented here were obtained for liquid-crystalline phase membranes, except for hydrophobicity profiles. Hydropho bicity profiles were drawn based on measurements performed for a frozen suspension of membranes (liposomes) at −150 to −165°C. This is necessary to distinguish between motional and solvent effects on the EPR spectra (12). 1.2. Detailed Profiles Across Lipid Bilayers

To determine the correct profile across the lipid bilayer, knowledge of the position (depths) at which the nitroxide moiety is located in the membrane is very important. However, vertical fluctuations of the nitroxide moiety of stearic acid spin labels (n-SASL) and phospholipid spin labels (n-PC) toward the polar surface of the lipid bilayer have been reported (13, 18). It can be presumed from these studies that distribution of the vertical positions of the nitroxide moiety of n-SASL and n-PC in the membrane exists, with the mean value of each distribution shifting toward the center as the quantum n increases. Positions of carbon atoms in the alkyl chain have been determined by neutron diffraction (19), which shows that the mean positions of these carbons (or nitroxide moieties attached to these carbons) can be defined with an accuracy of ±1 Å, even in the liquid-crystalline state. Assuming a Gaussian distribution of the labeled segments in the projection on the bilayer normal, these authors reported that time-averaged positional fluctuations increase from 1.5 Å for the C4 position to 3.4 Å for the C12 position. It can be concluded that a nitroxide moiety stays at the position determined by neutron diffraction for most of the time (see Note 1).

1.3. Necessity of Measurements for Simple Membrane Models

It is strongly recommended that the results obtained for membranes made of lipids extracted from biological membranes and those obtained for simple two- or three-component membranes, made of commercially available lipids and that resemble the basic lipid composition of biological membranes, are compared. This comparison makes it possible to better elucidate the major factors (to indicate the major membrane components) that determine certain membrane properties. In the presented example of coweye lens lipid membranes, the simple models are membranes made of an equimolar binary mixture of POPC and cholesterol and of pure POPC (9, 10). This comparison allows us to conclude that the high cholesterol content in the lipid extracts from fiber cell plasma membranes is responsible for the unique membrane properties observed with EPR spin-labeling methods (see Subheading 3.5).

252

Subczynski, Raguz, and Widomska

2. Materials 1. Phospholipid spin labels (1-palmitoyl-2-(n-doxylstearoyl) phosphatidylcholine (n-PC, where n = 5, 7, 10, 12, 14, or 16), or tempocholine-1-palmitoyl-2-oleoylphosphatidic acid ester (T-PC)) can be purchased from Avanti Polar Lipids, Inc. (Alabaster, AL). n-Doxylstearic acid spin labels (n-SASL, where n = 5, 7, 9, 10, 12, or 16), cholestane spin label (CSL), and androstane spin label (ASL) can be purchased from Sigma (St. Louis, MO). Spin labels are dissolved in chloroform at 1 mM and stored in a freezer at −70°C. 2. Chloroform solutions of the total lipids extracted from a biological material (usually 5–20 mg/mL) are kept in a freezer at −70°C (see Note 2). 3. Stock solutions of commercially available lipids (phospholipids and cholesterol) from Avanti Polar Lipids, Inc. (Alabaster, AL) in chloroform (usually, 20–50 mg/mL) are kept in a freezer at −70°C. These lipids are used to form simple two- or threecomponent membrane models (see Subheading 1.3). 4. Buffers: Typically, 10 mM PIPES and 150 mM NaCl; pH 7.0 is used as a buffer. For samples with n-SASL, 0.1 M borate at pH 9.5 is used as a buffer. In this case, a rather high pH is chosen to ensure that all SASL probe carboxyl groups are ionized in lipid bilayer membranes (20, 21) (see Note 3).

3. Methods 3.1. Spin Labeling

1. Chloroform solutions of extracted lipids and an appropriate spin label are mixed to obtain a final concentration of 0.5 or 1 mol% of spin labels in the lipid bilayer (see Note 4). 2. n-PC, T-PC, and n-SASL are spin labels that allow hydrophobicity profiles and profiles of the oxygen transport parameter across the lipid bilayer to be obtained (see Fig. 1 for their structures and approximate localization in the lipid bilayer). 3. n-PC and n-SASL allow profiles of the alkyl chain order parameter across the lipid bilayer to be obtained. 4. n-PC, T-PC, and n-SASL allow discrimination of raft and bulk domains (1–3) or liquid-ordered, liquid-disordered, and solid-ordered phases (4, 5) (see Fig. 2a). 5. Only the spin-labeled cholesterol analogue ASL allows discrimination of the cholesterol crystalline domain from the bulk phospholipid-cholesterol domain (6) (see Fig. 2b and Subheading 3.5.4).

Lipid Organization in Biological Membranes

3.2. Preparation of Multilamellar Liposomes

253

Membranes used for EPR measurements are usually multilamellar dispersions of lipids (multilamellar liposomes (see Note 5)) prepared in the following way: 1. A chloroform solution of lipids and spin label (usually, 2–4 mg of total lipid per sample) (see Subheading 3.1, step 1) is placed in a glass test tube, and chloroform is evaporated with a stream of nitrogen. The lipid film on the bottom of the test tube is thoroughly dried under reduced pressure (about 0.1 mmHg) for about 12 h. 2. A buffer solution (usually, 0.5–1.0 mL) is added to the dried film at a temperature above the phase transition temperature of investigated membranes (for lipids extracted from biological materials, 40°C is sufficient) and vortexed vigorously. 3. The lipid dispersion is centrifuged briefly (15 min at 4°C with an Eppendorf bench centrifuge at 16,000× g), and the loose pellet (about 20% lipid, w/w) is used for EPR measurements.

3.3. C onventional EPR

1. For all EPR measurements, the loose pellet is transferred to a capillary made of gas-permeable methyl-pentene polymer known as TPX (see Note 6), and the end of the capillary is sealed with Baxter Miniseal wax B4425.1 (see Notes 7 and 8). This plastic is permeable to oxygen, nitrogen, and other gases, and is substantially impermeable to water. 2. The TPX capillary is fixed inside the EPR dewar insert in the resonator of the X-band EPR spectrometer with a special Teflon holder (see Note 9) and equilibrated with nitrogen gas, which is used for temperature control. 3. The sample is thoroughly deoxygenated, yielding correct EPR line shape (see Note 10). 4. To obtain profiles of the order parameter, EPR spectra are recorded for spin labels with the nitroxide moiety at different depths in the membrane (see Subheading 3.1, step 3). Only one type of spin-label molecule is present in each sample. Recording conditions include modulation amplitude of 0.5–1.0 G and an incident microwave power of about 5 mW. 5. The order parameter S is calculated using the equation (22)

S = 0.5407(A¢II – A¢⊥)/a0, where a0 = (AII¢ + 2A⊥¢)/3

(1)

The values used for the calculation of the hydrocarbon chain order parameter, AII¢ and A¢⊥, are measured directly from EPR spectra as indicated in Fig. 3. 6. To obtain hydrophobicity profiles across the membrane, the z-component of the hyperfine interaction tensor, AZ, for spin labels with the nitroxide moiety at different depths in the membrane is determined directly from EPR spectra for samples frozen at about –165°C as indicated in

254

Subczynski, Raguz, and Widomska

Fig. 3. Panel of EPR spectra of 5-, 10-, and 16-PC in membranes made of cortical and nuclear cow-lens lipids. Spectra were recorded at 25°C. Measured values for evaluating the order parameter are indicated. The positions of certain peaks were evaluated with a high level of accuracy by monitoring them at 10 times higher receiver gain and, when necessary, at higher modulation amplitude

Fig. 4. EPR spectra of 16-PC in membranes made of cortical and nuclear cow-lens lipids. Spectra were recorded at −163°C to cancel motional effects. The measured 2AZ value is indicated

Fig. 4 (see Subheading 3.1, step 2). Only one type of spin-label molecule is present in each sample. Recording conditions include modulation amplitude of 2 G and an incident microwave power of 2 mW (12). In hydrophobicity profiles, 2AZ is plotted as a function of the approximate position of the nitroxide moiety in the lipid bilayer (see Note 11). 3.4. SaturationRecovery EPR (see Note 12)

The saturation-recovery EPR method for measuring electron spin-lattice relaxation time (T1) is a technique in which the recovery of the EPR signal is measured at a low level microwave field

Lipid Organization in Biological Membranes

255

(weak observing microwave power) after the end of the saturating microwave pulse. The time scale of this recovery is characterized by the spin-lattice relaxation time, T1. 1. The sample, prepared as described in Subheading 3.2, is transferred to the TPX capillary, positioned in the loop-gap resonator of the saturation-recovery EPR spectrometer, and deoxygenated by blowing nitrogen gas around the TPX capillary (see Notes 8 and 10). Thorough deoxygenation allows correct measurement of the spin-lattice relaxation time T1. The saturation-recovery signal is recorded at the required temperature for the spin label, with the nitroxide moiety at a fixed depth in the membrane. Only one type of spin-label molecule is present in each sample. 2. To obtain values of the oxygen transport parameter, the saturation-recovery signal is also recorded for the same sample, which is equilibrated with the required partial pressure of oxygen at the required temperature (see Note 13). 3. The same procedure is repeated for other spin labels, with the nitroxide moieties at different depths in the membrane (see Subheading 3.1, step 2). 4. T1s of spin labels in the absence and presence of molecular oxygen are determined by analyzing the saturation-recovery signal of the central line obtained by short-pulse saturationrecovery EPR (see Note 14). 5. The pulse length for short-pulse experiments is in the range of 0.1–0.5 ms. Pump power is selected to maximize the amplitude of the saturation-recovery signal and is typically in the range of 2–3.5 G. Observing power is selected to be as high as possible without affecting the time constant of the recovery. The minimum time between the end of the pulse and the beginning of observation of the recovery is determined by the ring-down time of the resonator and the switching transients, and is usually longer than 0.35 ms. Typically, 105–106 decays are acquired with 2,048 data points on each decay. Sampling intervals are from 1 to 32 ns, depending on the sample, temperature, and oxygen tension. The total accumulation time is typically 2–5 min. 6. Saturation-recovery signals are fitted by single and double exponentials and compared (see Fig. 5). If there is no substantial improvement in the fitting when the number of exponentials is increased from one, recovery curves are analyzed as single exponentials (see Fig. 5a–e, and Note 15). This is often the case for samples equilibrated with nitrogen (Fig. 5a, c, e). For samples equilibrated with a different partial pressure of oxygen, the saturation-recovery signal can often be fitted successfully with the double-exponential curve (as shown in Fig. 5f), indicating the presence of two coexisting domains or two coexisting phases (see Note 16).

256

Subczynski, Raguz, and Widomska

Fig. 5. Representative saturation-recovery signals with fitted curves and the residuals (the experimental signal minus the fitted curve) for 7-PC (a, b), CSL (c, d), and ASL (e, f) in membranes made of lens lipids isolated from the nuclear fraction of fiber cells of cow eyes. Signals were recorded at 25°C for samples equilibrated with 100% nitrogen gas (a, c, e) and with a gas mixture of 50% air and 50% nitrogen (b, d, f). Saturation-recovery signals for 7-PC and CSL were satisfactorily fitted to a single-exponential function in both the absence and presence of molecular oxygen with time constants of 4.37 ± 0.01 ms (a), 2.05 ± 0.01 ms (b), 2.95 ± 0.01 ms (c), and 1.91 ± 0.01 ms (d). For ASL, the saturation-recovery signal in the presence of molecular oxygen can be fit satisfactorily only with a double-exponential curve with time constants of 1.50 ± 0.28 ms and 0.55 ± 0.04 ms (compare the upper residual for single and lower residual for double-exponential fit in f), whereas single-exponential fit with a time constant of 2.71 ± 0.01 ms was satisfactory in the absence of molecular oxygen (e). Additional criteria for the goodness of single- and double-exponential fits are explained in Note 16

7. Calculation of the oxygen transport parameter from single-exponential decays is shown in Fig. 5a–e. The oxygen transport parameter is calculated using the equation (14)

W(x) = T1–1(Air, x) – T1 –1 (N2, x) = AD (x)C (x).

(2)

T1(Air, x) and T1(N2, x) are spin-lattice relaxation times of nitroxides in samples equilibrated with atmospheric air and nitrogen, respectively. Note that W(x) is normalized to the sample equilibrated with atmospheric air. W(x) is proportional to the product of the local translational diffusion coefficient D(x) and the local concentration C(x) of oxygen at a depth x

Lipid Organization in Biological Membranes

257

in the membrane, which is in equilibrium with atmospheric air (see Note 17). 8. Calculation of the oxygen transport parameter from doubleexponential decays is shown in Fig. 5f. In membranes consisting of two lipid environments with different oxygen transport rates – the fast oxygen transport (FOT) domain and the slow oxygen transport (SLOT) domain – the saturation-recovery signals recorded for samples equilibrated with air and nitrogen are simple double-exponential curves with time constants of T1–1(Air, FOT), T1–1(Air, SLOT), and T1−1(N2, FOT), T1–1(N2, SLOT), respectively (1, 4, 5, 7, 8) (see Note 18). Thus, values of the oxygen transport parameter in each domain can be calculated using the equations:

W(FOT) = T1–1 (Air, FOT) – T1–1(N2, FOT)

(3)

W(SLOT) = T1–1(Air, SLOT) – T1–1(N2, SLOT)

(4)

Here, “x” from (2) is changed to the two-membrane domains FOT and SLOT, and the depth fixed (the same spin label is distributed between the FOT and SLOT domains). 3.5. Profiles of Membrane Properties Across Homogeneous Membranes and Coexisting Membrane Domains

The final goal in the study of lipid organization in biological membranes using liposomes and EPR spin labeling is not only to characterize membranes by single (at one depth) spectral parameters but also to obtain detailed profiles of these parameters across membranes. These detailed profiles contain unique information about membrane structure and dynamics. Additionally, these profiles can often be obtained in coexisting membrane domains without the need for their physical separation, which provides unique opportunities in studies of physical properties of domains in situ. Using various spin-labeling techniques as well as conventional and saturation-recovery EPR spectroscopy (covering a time scale of 100 ps–10 ms), membrane molecular organization and dynamics can be investigated in the ps-to-ms range. The profiles of four parameters that were obtained with EPR spin-labeling methods and that describe the different properties of biological and model membranes are presented in the following sections, together with a short explanation of the information that can be extracted from these profiles.

3.5.1. Order Parameter

In the membrane, the alkyl chain of n-PC or n-SASL with the nitroxide moiety attached at the Cn position (see Fig. 1) undergoes a rapid anisotropic motion about the long axis of the spin label and a wobbling motion of the long axis within the confines of a cone imposed by the membrane environment. The order parameter (Eq. 1) is a measure of the amplitude of the wobbling motion. Increase in the order parameter indicates that the angle

258

Subczynski, Raguz, and Widomska

of the cone, responsible for the wobbling motion of the alkyl chain, decreases. Moreover, moving from the bilayer surface to the membrane interior, deviations in the alkyl chain segment direction from the bilayer normal accumulate. Consequently, ordering of the alkyl chain induced by stearic contact with the plate-like portion of cholesterol also causes ordering of the distal fragment of the alkyl chain, even though the rate of wobbling fluctuations can be higher (9, 13) (see Subheading 3.5.2). Although the order parameter indicates the static property of the lipid bilayer, for brevity, the change in the order parameter is most often described as a change in spin-label mobility, and thus as a change in membrane fluidity. Profiles of the molecular order parameter obtained at 25°C for the bulk phospholipid-cholesterol domain of cortical and nuclear cow-lens lipid membranes are displayed in Fig. 6a. In both membranes, values of the order parameter measured at the same depths are practically the same and are close to those measured for membranes made of the equimolar POPC/cholesterol mixture (Fig. 6b). They are, however, significantly greater than those measured for the pure POPC membrane (Fig. 6b), indicating that a saturating amount of cholesterol is responsible for the rigidity of the lensmembrane. In all membranes, profiles have an inverted bell shape and alkyl chain order that gradually decreases with an increase in membrane depth.

Fig. 6. Profiles of the molecular order parameter at 25°C obtained with n-PC and n-SASL across membranes made of cortical and nuclear cow-lens lipids are presented in (a) and across membranes made of the POPC/Chol equimolar mixture and of pure POPC are presented in (b). Approximate localizations of nitroxide moieties of the spin labels are indicated by arrows. For POPC and POPC/Chol membranes, data were taken from Refs. (9) and (11)

Lipid Organization in Biological Membranes 3.5.2. Fluidity

259

There are no easily obtainable EPR spectral parameters for lipid spin labels that can describe the profiles of membrane fluidity related to dynamic membrane properties. Only in the membrane center can membrane fluidity be evaluated directly from conventional EPR spectra of 16-PC (or 16-SASL) by measuring the effective rotational correlation time of the nitroxide moiety of the lipid spin label and assuming its isotropic rotational motion (23). However, already, 14-PC shows anisotropic rotational motion. Also, in the presence of 30 mol% cholesterol, motion of 16-PC moves into the slow-tumbling regime, where no conventional parameterization has been established (see Note 19). As was indicated in the subheading 3.5.1, the order parameter, which is most often used as a measure of membrane fluidity, describes, in principle, the static membrane properties – namely, the amplitude of the wobbling motion. Fortunately, the spin-lattice relaxation time (T1) is a spectral parameter that can be obtained from saturationrecovery EPR measurements with lipid spin labels (see Subheading 3.4). This parameter depends primarily on the rate of motion of the nitroxide moiety within the lipid bilayer, and thus describes the dynamics of the membrane environment at a depth at which the nitroxide fragment is located. It should be mentioned here that both the rotational motion (24) and the Brownian translational motion (25) are mechanisms involved in the spin-lattice relaxation process of nitroxide spin labels. Thus, T1 can be used as a conventional quantitative measure of membrane fluidity that indicates the rate of motion of phospholipid alkyl chains (or nitroxide free radical moieties attached to those chains). If T1 is measured for n-PC or n-SASL spin labels, a fluidity profile across the lipid bilayer can be obtained that reflects the membrane dynamics. In principle, these fluidity profiles can be obtained in coexisting domains and coexisting phases without the need for their physical separation (see Note 18). Fluidity profiles (T1 versus depth in the membrane) for the bulk phospholipidcholesterol domain of the cortical and nuclear cow-lens lipid membranes obtained at 25°C are presented in Fig. 7a. As expected, membrane fluidity (membrane dynamics) increases toward the membrane center, and profiles in both membranes are inverted bell-shaped, and are similar to profiles obtained for membranes made of the equimolar POPC/cholesterol mixture and pure POPC (Fig. 7b). As shown by comparing the profiles for the pure POPC bilayer and the POPC/cholesterol bilayer (Fig. 7b), cholesterol decreases the membrane fluidity close to the membrane surface and increases it at the membrane center. This confirms the earlier results (13, 15, 26) (see Note 20). The order parameter (the static membrane property) cannot differentiate the effects of cholesterol at different depths (see Subheading 3.5.1), while another dynamic parameter – namely, the oxygen transport parameter – clearly shows the differences between the membrane

260

Subczynski, Raguz, and Widomska

Fig. 7. Profiles of the electron spin-lattice relaxation time T1 for n-PC spin labels at 25°C across membranes made of cortical and nuclear cow-lens lipids are presented in (a) and across membranes made of the POPC/Chol equimolar mixture and of pure POPC are presented in (b). Approximate localizations of nitroxide moieties of the spin labels are indicated by arrows. For POPC membranes, data were taken from Ref. (40)

Fig. 8. Hydrophobicity profiles (2AZ) across membranes made of cortical and nuclear cow-lens lipids are presented in (a) and across membranes made of the POPC/Chol equimolar mixture and of pure POPC are presented in (b). Upward changes indicate increases in hydrophobicity. Because T-PC contains a different nitroxide moiety than n-PC and n-SASL, its points are not connected with other points. However, the relative changes of the hydrophobicity in the polar headgroup region can be evaluated (see Note 13). Broken lines indicate hydrophobicity in the aqueous phase (see Note 13). Approximate localizations of nitroxide moieties of the spin labels are indicated by arrows. For POPC and POPC/Chol membranes, data were taken from Refs. (9) and (11)

region where the cholesterol ring structure is located and the dipper region where the isooctyl chain of cholesterol is located (see Subheading 3.5.4). 3.5.3. Hydrophobicity

Figure 8a shows hydrophobicity profiles across the bulk phospholipid-cholesterol domain of cortical and nuclear cow-lens lipid membranes. Here, 2AZ data, obtained as described in

Lipid Organization in Biological Membranes

261

Subheading 3.3, step 5, are presented as a function of the approximate position of the nitroxide moiety of the spin label within the lipid bilayer. Smaller 2AZ values (upward changes in the profiles) indicate higher hydrophobicity. In both membranes, hydrophobicity profiles show a similar rectangular shape, with an abrupt increase of hydrophobicity between the C9 and C10 positions. 2AZ values in the center of both membranes (positions 10-, 12-, 14-, and 16-PC) indicate that the hydrophobicity in this region is only slightly lower than for pure hexane (e = 2) and can be compared to that of dipropylamine (see Note 21). It can be seen from Fig. 8a that the center of the bulk phospholipid-cholesterol domain of the nuclear membrane is less hydrophobic than that of the cortical membrane. A similar difference is also observed close to the membrane surface (5- and 7-PC positions). These interesting findings are in good agreement with an earlier observation (12) that cholesterol causes a significant increase in hydrophobicity of the PC lipid bilayer center when its concentration increases up to ~30 mol%; this is followed by a moderate decrease in hydrophobicity when the cholesterol concentration increases further, up to 50 mol%. The addition of cholesterol (from 0 to 50 mol%) monotonically decreases the hydrophobicity in the region close to the membrane surface. This confirms that the cortical bulk phospholipid-cholesterol domain has not yet reached saturation with cholesterol, while the nuclear bulk phospholipidcholesterol bilayer is saturated with cholesterol. Profiles presented in Fig. 8a are similar to those measured for the membrane made of the equimolar POPC/cholesterol mixture (Fig. 8b). These profiles also show a rectangular shape and an abrupt increase in hydrophobicity between the C9 and C10 positions, but differ from the typical bell-shaped profile of the pure POPC membrane (Fig. 8b), with a gradual increase in hydrophobicity toward the bilayer center. Also, the change in hydrophobicity between the membrane surface and the center is significantly lower for the pure POPC membrane than for the cortical and nuclear membranes. It can be concluded that the rectangular shape of the hydrophobicity profiles is characteristic of membranes with high cholesterol content. 3.5.4. Oxygen Transport Parameter

The oxygen transport parameter was introduced as a conventional quantitative measure of the rate of collision between spin label and molecular oxygen (Eq. 2). Kusumi et al. (14) conclude that the oxygen transport parameter is a useful monitor of membrane fluidity that reports on translational diffusion of small molecules. The profiles of the oxygen transport parameter for the bulk phospholipid-cholesterol domain of the cortical and nuclear cow-lens lipid membranes obtained at 25°C are presented in Fig. 9a (see Note 22). All profiles have a rectangular shape with an abrupt increase in the oxygen transport parameter between the C9 and C10 positions. This abrupt increase is as large as ~3 times, and the overall

262

Subczynski, Raguz, and Widomska

Fig. 9. Profiles of the oxygen transport parameter (oxygen diffusion-concentration product) at 25°C across membranes made of cortical and nuclear cow-lens lipids are presented in (a) and across membranes made of the POPC/Chol equimolar mixture and of pure POPC are presented in (b). Broken lines indicate the oxygen transport parameter in the aqueous phase. Approximate localizations of nitroxide moieties of spin labels are indicated by arrows. For POPC and POPC/Chol membranes, data were taken from Refs. (9) and (40). (PCD: phospholipid-cholesterol domain; CCD: cholesterol crystalline domain)

change of the oxygen transport parameter across the membrane becomes as large as ~6 times. The oxygen transport parameter from the membrane surface to the depth of the ninth carbon is as low as in gel-phase PC membranes, and at locations deeper than the ninth carbon, as high as in the fluid-phase membranes (4, 9, 11, 15, 26). These profiles are practically identical if we take into account the accuracy of the measurements (evaluated as 10%). Profiles are also very similar to those for the membrane made of the equimolar mixture of POPC and cholesterol (Fig. 9b). These profiles also show a rectangular shape and an abrupt increase in the oxygen transport parameter between the C9 and C10 positions. However, profiles for cortical and nuclear membranes differ from the bell-shaped profile across the pure POPC bilayer (Fig. 9b). This additionally confirms that high cholesterol content is responsible for the unique properties of lens lipid membranes. It should also be indicated that the abrupt change in hydrophobicity (Fig. 8) and oxygen transport parameter profiles (Fig. 9) is observed between the C9 and C10 positions, which is approximately where the steroid-ring structure of cholesterol reaches into the membrane. ASL shows two exponential saturation-recovery signals in nuclear membranes equilibrated with air/nitrogen mixture (Fig. 5f), indicating the presence of two membrane environments around ASL (see (Eq. 3) and (Eq. 4), and Notes 18 and 22). Comparing values of the oxygen transport parameter obtained with ASL and obtained with phospholipid-type spin labels in the bulk phospholipid-cholesterol domain allows the conclusion that ASL molecules, which give a greater oxygen transport

Lipid Organization in Biological Membranes

263

parameter, are located in the bulk phospholipid-cholesterol domain (see profile of the oxygen transport parameter for the nuclear lens lipid membrane with added points for the cholesterol analogue spin labels in Fig. 9a). The greater oxygen transport parameter value monitored with ASL is practically the same as the value obtained with 10-PC. This confirms that in the bulk phospholipid-cholesterol bilayer the nitroxide moiety of ASL is located close to the C10 position (6). The second value of the oxygen transport parameter monitored by ASL is about 5 times smaller than the value monitored by ASL in the bulk phospholipid-cholesterol domain and can be considered as that characterizing the pure cholesterol crystalline domain. In cortical membranes, ASL in the presence and absence of oxygen shows single-exponential saturation-recovery signals, indicating a homogenous environment (Fig. 5a). The oxygen transport parameter value detected by ASL is practically the same as that detected by 10-PC. It allows the assumption that cortical membranes consist of the bulk phospholipid-cholesterol domain without a detectable cholesterol crystalline domain. CSL in both cortical and nuclear membranes detects only homogenous environments, showing single-exponential saturation-recovery signals. Thus, the oxygen transport parameter values detected in the polar headgroup region of the nuclear lens lipid membrane, with CSL located in the bulk phospholipid-cholesterol domain and in the cholesterol crystalline domain, are very similar and cannot be distinguished by the DOT method (see Note 23). These results are in agreement with results obtained with ASL and CSL in piglens lipid membranes in which the cholesterol crystalline domain was induced by adding excess cholesterol (6). Values of the oxygen transport parameter measured with CSL and ASL in the nuclear lens lipid membrane are indicated in Fig 18.9a, and the profile of the oxygen transport parameter across the cholesterol crystalline domain, which coexists with the bulk phospholipidcholesterol domain, is shown. We would like to direct readers to Ref. (41), which was recently accepted for publication in Biochimica et Biophysica Acta. This paper describes in detail our investigation of the physical properties of membranes derived from the total lipid extract from the lens cortex and nucleus of a two-year-old cow using EPR spin-labeling methods.

4. Notes 1. The thickness of the lens lipid membrane is assumed to be the same as the thickness of the POPC/Chol = 1/1 membrane. It is also assumed that the location of the alkyl chain carbon atom in the membrane changes linearly with the position of

264

Subczynski, Raguz, and Widomska

the alkyl chain. The nitroxide moieties of n-PC and n-SASL are located at the same depth as the carbon atoms of the 2-chain of phospholipid. For details, see (10, 15, 26). 2. The commonly used procedure for extracting lipids from animal tissues was described by Folch and co-authors in 1957 (27). This procedure was also used with minor modification to extract lipids from cow-eye lenses (9, 10). The choice of extraction procedure should depend on the nature of the tissue matrix as well as other factors, such as specific lipid classes. The principles and good practice of tissue handling and lipid separation has been described in many specialist articles, books, and reviews (28–30). 3. At a pH of 7.0, a mixture consisting of two forms of SASLs can be presented in the lipid bilayer (with protonated and ionized carboxyl groups), and nitroxide moieties can be positioned at two different depths in the membrane. 4. A concentration of spin labels in the lipid bilayer that is too high affects the measurement of spectral parameters, especially spinlattice relaxation time. If the composition of the lipid extract (see Note 2) is known, the concentration of spin labels can be easily calculated. Otherwise, the average molecular weight of phospholipids in the total lipid extract has to be assumed and used for the calculation of spin-label concentration (see (9)). 5. Membranes in samples made from multilamellar liposomes are tightly packed, giving much better signal-to-noise ratio than in samples made from unilamellar liposomes. 6. For measurements at X-band, sample tubes are machined from TPX with dimensions of 0.6 mm ID, 0.1 mm wall thickness, and 25 mm length. Capillaries are machined from TPX rods, which can be purchased from Midland Plastic (Madison, WI). 7. This sealant is no longer commercially available but can be found in many laboratories. Other tube sealants can be used, including Critoseal (Fisher Scientific) and X-Sealant (Bruker Biospin). 8. It is often desirable to additionally concentrate the sample inside the TPX capillary by centrifugation in order to improve the signal-to-noise ratio (31). 9. TPX capillaries, together with the Teflon holder, can be obtained from Molecular Specialties (Milwaukee, WI). 10. Molecular oxygen is paramagnetic, having a triplet ground state, and bimolecular collisions of molecular oxygen with spin labels affect the EPR spectral parameters including line width and spin-lattice relaxation time (32). 11. Hydrophobicity profiles are constructed based on EPR measurements of 2AZ for n-PC and n-SASL spin labels. For SASLs in the aqueous phase, 2AZ was calculated as described

Lipid Organization in Biological Membranes

265

in (12). In addition, T-PC is used to probe the membrane polar headgroup region. Because the chemical structure of the nitroxide moiety of T-PC differs from that for n-PC and n-SASL, 2AZ values measured with T-PC cannot be directly compared with those for n-PC and n-SASL. However, T-PC shows relative hydrophobicity changes in polar headgroup regions after the addition of certain membrane modifiers (like cholesterol). 2AZ values measured with T-PC can also be used to compare the hydrophobicity of polar headgroup regions in different membranes. 12. The state-of-the-art X- and Q-band saturation-recovery EPR spectrometers were built and are available at the National Biomedical EPR Center, Medical College of Wisconsin, Milwaukee, WI, USA. The mission of the Center is to make advanced EPR research resources available to investigators nationally, regionally, and locally (see the link for Center use: http://www.mcw.edu/display/router.asp?docid=3211). Another X-band saturation-recovery spectrometer was built and is located at the Department of Biophysics, Faculty of Biotechnology, Jagiellonian University, Krakow, Poland. Presently, Bruker produces EPR spectrometers that are capable of saturation-recovery measurements at X-band. Pulse saturation recovery is possible on an E-580 FT/EPR system equipped with DC-AFC and LCW (low power CW arm) options and combined with an AmpX CW microwave power amplifier. Saturation recovery is treated as an accessory to the E-580 and is not usually a stand-alone configuration. 13. Switching the gas around the TPX capillary from nitrogen to the air/nitrogen mixture allows one to equilibrate the sample with the required partial pressure of oxygen for oximetry measurements and for obtaining T1s in the presence of molecular oxygen. Because the same gas mixture is used for temperature control, samples are equilibrated with oxygen at the required temperature. The mixture of air and nitrogen is adjusted with flow meters (Matheson Gas Products, Montgomeryville, PA, Model 7631 H-604). 14. The short-pulse method is favorable for multiexponential decays in oximetry measurements (33, 34). For a short pulse, only populations of the irradiated transition are affected; for a long pulse, all populations are altered because of transverse relaxations. Ref. (35) addresses the long- and short-pulse saturation-recovery methods in more detail. 15. Additional criteria for a good single-exponential fit are the negligible preexponential coefficient for the second component, the large standard deviation of T1 for the second component, and the repetition of the fit for different recording conditions such as the number of points and time increment.

266

Subczynski, Raguz, and Widomska

16. Although saturation-recovery signals in the absence of molecular oxygen cannot differentiate between these two domains, in the presence of oxygen the recovery curves are very different in each domain, indicating that the collision rate of molecular oxygen, or the oxygen transport parameter, is quite different. This is a good illustration of why the method is named DOT – because different domains can be clearly discriminated and characterized only in the presence of molecular oxygen. 17. A is remarkably independent of the hydrophobicity and viscosity of the solvent and of spin-label species (36–38). 18. When located in two different membrane domains, the spin label alone most often cannot differentiate between the two, giving very similar conventional EPR spectra and similar T1 values (T1−1(N2, FOT) » T1−1(N2, SLOT)). However, even small differences in lipid packing in these domains will affect oxygen partitioning and oxygen diffusion, which can be easily detected by observing the different T1s from spin labels in these two locations in the presence of oxygen. 19. In principle, the rotational motion of the 16-PC molecule as a whole, which is anisotropic, has to be distinguished from segmental motion, which comes from gauche-trans isomerization of the alkyl chain. For carbon atoms near the terminal methyl group (16-PC position), it is assumed that the segmental motion is not restricted, and as a result, motion of the nitroxide moiety is approximately isotropic. At higher temperatures, the motion of all molecules becomes so great that segmental motion dominates. At lower temperatures, the segmental motion is diminished and the motion of 16-PC becomes highly anisotropic. 20. To the best of our knowledge, the profile of the spin-lattice relaxation time is used here for the first time as a monitor of membrane dynamics and fluidity. It should be indicated that T1 is sensitive to the molar concentration of spin labels in the lipid bilayer. Therefore, the concentration of all spin labels in the lipid bilayer should be the same, as well as in membranes that are compared. This is not an easy task, especially for lipid bilayer membranes (liposomes) derived from the lipid extract of certain biological membranes. This may be the reason that profiles for cortical and nuclear lens lipid membranes presented in Fig. 7a are shifted relative to each other. Other EPR spectral parameters (molecular order, hydrophobicity, and oxygen transport parameter) are affected much less by spin-label concentration. Profiles for POPC and POPC/Chol presented in Fig. 7b are based on data obtained in 2003 (40) and 2007 (9), respectively. The effects of cholesterol on T1 of the lipid spin labels located in the membrane center and close

Lipid Organization in Biological Membranes

267

to the membrane surface are also clearly shown in Refs. (13, 15, 26). 21. For brevity, we refer to Fig. 2 in Ref. (12) to show the relation of the local hydrophobicity as observed by 2AZ to the hydrophobicity (or e) of the bulk organic solvent. 2AZ in bulk solvent provides a convenient yardstick for describing the local hydrophobicity in the membrane, and comparison of two 2AZ values may help to develop a “feel” for local hydrophobicity. Such a comparison is only semi-quantitative and could be made operationally because small changes in 2AZ correspond to large changes in e and the mechanism by which the presence of water affects 2AZ is not well understood (12, 39). 22. Restrictions on the distribution of lipid spin labels in membranes containing the cholesterol crystalline domain indicate that only spin-labeled cholesterol analogues can discriminate this domain (6). These analogues should approximate the distribution of cholesterol molecules in the membrane because of the overall similarity of CSL, ASL, and cholesterol molecular structures. Phospholipid spin labels, which should not partition into the cholesterol crystalline domain, cannot discriminate these domains. Indeed, in both cortical and nuclear membranes, saturation-recovery signals for n-PC, 9-SASL, and T-PC are single-exponential signals and are assigned as signals characterizing the bulk phospholipidcholesterol bilayer. 23. Results presented in Figs. 5 and 9 are an excellent illustration of the advantages and limitations of the DOT method. To detect membrane domains, lipid spin labels have to be distributed between these domains (like CSL and ASL, but unlike n-PC (see Fig. 2b)). Even when located in two different membrane domains, the spin label alone most often cannot differentiate between the two, giving very similar T1 values (in the absence of oxygen, saturation-recovery signals for both CSL and ASL were single-exponential signals (Fig. 5), indicating that T1 values in both environments are very close). In membranes equilibrated with air and consisting of two lipid environments with different oxygen transport rates – the fast oxygen transport (FOT) domain and the slow oxygen transport (SLOT) domain – the saturation-recovery signal should be a double-exponential curve with time constants of T1(air, FOT) and T1(air, SLOT) (see Subheading 3.5 and (Eq. 3) and (Eq. 4)). This is the case with ASL, which shows doubleexponential saturation-recovery signals for the sample equilibrated with the air/nitrogen mixture (Fig. 5f). CSL cannot distinguish these domains (Fig. 5c, d), probably because collision rates between oxygen and the nitroxide moiety of CSL located in the polar headgroup region are similar.

268

Subczynski, Raguz, and Widomska

Acknowledgments This work was supported by grants EY015526, EB002052, and EB001980 of the National Institutes of Health. References 1. Kawasaki K, Yin J-J, Subczynski WK, Hyde JS, Kusumi A (2001) Pulse EPR detection of lipid exchange between protein-rich raft and bulk domains in the membrane: methodology development and its application to studies of influenza viral membrane. Biophys J 80:738–748 2. Wisniewska A, Subczynski WK (2006) Accumulation of macular xanthophylls in unsaturated membrane domains. Free Radic Biol Med 40:1820–1826 3. Wisniewska A, Subczynski WK (2006) Distribution of macular xanthophylls between domains in a model of photoreceptor outer segment membranes. Free Radic Biol Med 41:1257–1265 4. Subczynski WK, Wisniewska A, Hyde JS, Kusumi A (2007) Three-dimensional dynamic structure of the liquid-ordered domain as examined by a pulse-EPR oxygen probing. Biophys J 92:1573–1584 5. Wisniewska A, Subczynski WK (2008) The liquid-ordered phase in sphingomyelincholesterol membranes as detected by the discrimination by oxygen transport (DOT) method. Cell Mol Biol Lett 13:430–451 6. Raguz M, Widomska J, Dillon J, Gailard ER, Subczynski WK (2008) Characterization of lipid domains in reconstituted porcine lens membranes using EPR spin-labeling approaches. Biochim Biophys Acta 1778:1079–1090 7. Subczynski WK, Widomska J, Wisniewska A, Kusumi A (2007) Saturation-recovery electron paramagnetic resonance discrimination by oxygen transport (DOT) method for characterizing membrane domains. In: McIntosh TJ (ed) Methods in molecular biology. Lipid rafts, vol 398. Humana Press, Totowa, pp 143–157 8. Ashikawa I, Yin J-J, Subczynski WK, Kouyama T, Hyde JS, Kusumi A (1994) Molecular organization and dynamics in bacteriorhodopsin-rich reconstituted membranes: discrimination of lipid environments by the oxygen transport parameter using a pulse ESR spin-labeling technique. Biochemistry 33:4947–4952

9. Widomska J, Raguz M, Dillon J, Gaillard ER, Subczynski WK (2007) Physical properties of the lipid bilayer membrane made of calf lens lipids: EPR spin labeling studies. Biochim Biophys Acta 1768:1454–1465 10. Widomska J, Raguz M, Subczynski WK (2007) Oxygen permeability of the lipid bilayer membrane made of calf lens lipids. Biochim Biophys Acta 1768:2636–2645 11. Subczynski WK, Lewis RNAH, McElhaney RN, Hodges RS, Hyde JS, Kusumi A (1998) Molecular organization and dynamics of 1-palmitoyl-2-oleoylphosphatidylcholine bilayers containing a transmembrane a-helical peptide. Biochemistry 37:3156–3164 1 2. Subczynski WK, Wisniewska A, Yin J-J, Hyde JS, Kusumi A (1994) Hydrophobic barriers of lipid bilayer membranes formed by reduction of water penetration by alkyl chain unsaturation and cholesterol. Biochemistry 33:7670–7681 13. Yin J-J, Subczynski WK (1996) Effects of lutein and cholesterol on alkyl chain bending in lipid bilayers: a pulse electron spin resonance spin labeling study. Biophys J 71:832–839 14. Kusumi A, Subczynski WK, Hyde JS (1982) Oxygen transport parameter in membranes as deduced by saturation recovery measurements of spin-lattice relaxation times of spin labels. Proc Natl Acad Sci U S A 79: 1854–1858 15. Subczynski WK, Hyde JS, Kusumi A (1989) Oxygen permeability of phosphatidylcholinecholesterol membranes. Proc Natl Acad Sci U S A 86:4474–4478 16. Smirnov AI, Clarkson RB, Belford RL (1996) EPR linewidth (T2) method to measure oxygen permeability of phospholipids bilayer and its use to study the effect of low ethanol concentration. J Magn Reson B 111:149–157 17. Altenbach C, Froncisz W, Hyde JS, Hubbell WL (1989) Conformation of spin-labeled melittin at membrane surface investigated by pulse saturation recovery and continuous wave power saturation electron paramagnetic resonance. Biophys J 56:1183–1191

Lipid Organization in Biological Membranes 18. Merkle H, Subczynski WK, Kusumi A (1987) Dynamic fluorescence quenching studies on lipid mobilities in phosphatidylcholine-cholesterol membranes. Biochim Biophys Acta 897: 238–248 19. Zaccai G, Büldt G, Seelig A, Seelig J (1979) Neutron diffraction studies on phosphatidylcholine model membranes II. Chain conformation and segmental disorder. J Mol Biol 134:693–706 20. Egreet-Charlier M, Sanson A, Ptak M, Bouloussa O (1978) Ionization of fatty acids at lipid–water interface. FEBS Lett 89:313–316 21. Kusumi A, Subczynski WK, Hyde JS (1982) Effects of pH on ESR spectra of stearic acid spin labels in membranes: probing the membrane surface. Fed Proc 41:1394 22. Marsh D (1981) Electron spin resonance: spin labels. In: Grell E (ed) Membrane Spectroscopy. Springer-Verlag, Berlin, pp 51–142 23. Berliner LJ (1978) Spin labeling in enzymology: spin-labeled enzymes and proteins. Rotational correlation times calculation. Methods Enzymol 49:466–470 24. Atkins PW, Kivelson D (1966) ESR linewidth in solution. II. Analysis of spin-rotational relaxation data. J Chem Phys 44:169–174 25. Robinson BH, Hass DA, Mailer C (1994) Molecular dynamics in lipid spin lattice relaxation of nitroxide spin labels. Science 263:490–493 26. Subczynski WK, Hyde JS, Kusumi A (1991) Effect of alkyl chain unsaturation and cholesterol intercalation on oxygen transport in membranes: a pulse ESR spin labeling study. Biochemistry 30:8578–8590 27. Folch J, Lees M, Sloane Stanley GH (1957) A simple method for the isolation and purification of total lipids from animal tissues. J Biol Chem 226:497–509 28. Markham JE, Li J, Cahoon EB, Jaworski JG (2006) Separation and identification of major plant sphingolipid classes from leaves. J Biol Chem 281:22684–22694 29. Bodennec J, Pelled D, Futerman AH (2003) Aminopropyl solid phase extraction and 2 D TLC of neutral glycosphingolipids and neutral lysoglycosphingolipids. J Lipid Res 44:218–226 30. Christie WW (2003) Lipid analysis: isolation, separation, identification and structural analysis of lipids, 3rd edn. Oily, Bridgwater

269

31. Subczynski WK, Felix CC, Klug CS, Hyde JS (2005) Concentration by centrifugation for gas exchange EPR oximetry measurements with loop-gap resonators. J Magn Reson 176:244–248 32. Subczynski WK, Swartz HM (2005) EPR oximetry in biological and model samples. In: Eaton SS, Eaton GR, Berliner LJ (eds) Biological magnetic resonance, biomedical epr-part a: free radicals, metals, medicine, and physiology, vol 23. Kluwer/Plenum, New York, pp 229–282 33. Yin J-J, Hyde JS (1987) Spin-label saturationrecovery electron spin resonance measurements of oxygen transport in membranes. Z Phys Chem (Munich) 153:57–65 34. Hyde JS, Yin J-J, Feix JB, Hubbell WL (1990) Advances in spin label oximetry. Pure Appl Chem 62:255–260 35. Yin J-J, Hyde JS (1989) Use of high observing power in electron spin resonance saturation-recovery experiments in spin-labeled membranes. J Chem Phys 91:6029–6035 36. Hyde JS, Subczynski WK (1984) Simulation of ESR spectra of the oxygen-sensitive spin-label probe CTPO. J Magn Reson 56:125–130 37. Hyde JS, Subczynski WK (1989) Spin-label oximetry. In: Berliner LJ, Reuben J (eds) Biological magnetic resonance, vol 8. Plenum, New York, pp 399–425 38. Subczynski WK, Hyde JS (1984) Diffusion of oxygen in water and hydrocarbons using an electron spin resonance spin-label technique. Biophys J 45:743–748 39. Griffith OH, Dehlinger PJ, Van SP (1974) Shape of the hydrophobic barrier of phospholipids bilayers (evidence for water penetration into biological membranes). J Membr Biol 15:159–192 40. Subczynski WK, Pasenkiewicz-Gierula M, McElhaney RN, Hyde JS, Kusumi A (2003) Molecular dynamics of 1-palmitoyl-2oleoylphosphatidylcholine membranes containing transmembrane a-helical peptides with alternating leucine and alanine residues. Biochemistry 42:3939–3948 41. Raguz M, Widomska J, Dillon J, Gaillard ER, Subczynski WK (2009) Physical properties of the lipid bilayer membrane made of cortical and nuclear bovine lens lipids: EPR spinlabeling studies. Biochim Biophys Acta, doi:10.1016/j.bbamem.2009.09.005

Chapter 19 Membrane Translocation Assayed by Fluorescence Spectroscopy Jana Broecker and Sandro Keller Abstract Assessing the ability of biomolecules or drugs to overcome lipid membranes in a receptor-independent way is of great importance in both basic research and applications involving the use of liposomes. A combination of uptake, release, and dilution experiments performed by steady-state fluorescence spectroscopy provides a powerful, straightforward, and inexpensive way of monitoring membrane translocation of fluorescent compounds. This is particularly true for peptides and proteins carrying intrinsic tryptophan residues, which eliminates the need for attaching extrinsic labeling moieties to the compound of interest. The approach encompasses three different kinds of fluorescence titrations and some simple calculations that can be carried out in a spreadsheet program. A complete set of experiments and data analyses can typically be completed within two days. Key words: Membrane binding, Membrane permeability, Membrane permeation, Flip–flop, Transbilayer movement, Uptake, Release, Dilution, Tryptophan fluorescence, Vesicles

1. Introduction Determining the permeability of lipid membranes to peptides, proteins, small molecules, and other biologically relevant compounds is a frequent task in many fields of liposome research. Heerklotz et al. (1–3) established a combination of so-called uptake and release experiments as a widely applicable solution to this problem on the basis of isothermal titration calorimetry (ITC). This approach has since been exploited to monitor membrane translocation (also referred to as membrane permeation or flip–flop) of such diverse compounds as photoactivatable nucleotide precursors (4), small molecules developed for conditional gene expression (5) or ion channel gating (6), surfactants (7–10),

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_19, © Humana Press, a part of Springer Science+Business Media, LLC 2010

271

272

Broecker and Keller

and peptides (11). A detailed step-by-step guide to microcalorimetric uptake and release experiments can be found in the literature (12). We have recently adapted (11) this approach to assess membrane translocation by spectroscopic methods. Fluorescence spectroscopy lends itself particularly well to this purpose, because it is available in many laboratories, generates datasets that can be interpreted in a straightforward way, and sensitively monitors membrane binding over a broad range of fluorophore concentrations. If, in addition, the molecule of interest is a peptide (or a protein) whose association with lipid membranes is accompanied by changes in the emission spectrum of intrinsic fluorescent amino acid residues, there is no need for attaching extrinsic fluorescent labels, which might drastically affect interactions with lipids. For the sake of simplicity, we will henceforth refer to the fluorophore at hand as peptide; however, the approach can easily be extended to any other membrane-interacting molecule whose fluorescence properties change upon membrane binding. Even if this is not the case, dialysis may be used to adapt the assay to the needs of other spectroscopic methods such as simple absorbance readings (13). The rationale underlying the fluorescence-spectroscopic approach is simple, involving three different titrations referred to as uptake, release, and dilution experiments (Fig. 1): In the uptake experiment, lipid vesicles are titrated into a peptide solution. a

uptake

+

– or

release

b

–

+ or

c

+

dilution

– or

Fig. 1. Schematic representation of (a) uptake, (b) release, and (c) dilution experiments to assess the translocation of a peptide (black) across lipid vesicle membranes (grey). The two extreme cases of one-sided membrane binding (−) and transbilayer equilibration (+) are depicted. (Reproduced with permission from ref. 11. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.)

Membrane Translocation Assayed by Fluorescence Spectroscopy

273

Uptake of the peptide occurs into the outer membrane leaflet only or into both leaflets depending on the peptide’s ability to translocate across the membrane. In the release experiment, lipid vesicles are first homogeneously preloaded with peptide on both leaflets and then titrated into pure buffer, causing release of the peptide from the membrane into the aqueous phase. Importantly, complete desorption can be achieved only if the peptide is membrane-permeant; otherwise, the fraction trapped inside the vesicles will remain adsorbed to the inner leaflet. In the dilution experiment, peptide is first applied externally to preformed lipid vesicles, and this mixture is then injected into buffer. Irrespective of membrane translocation, the peptide should thus completely desorb from the membrane upon strong dilution. This means that the release and dilution experiments will yield identical results only if the peptide can equilibrate across lipid bilayers, whereas differences between the two setups indicate one-sided membrane binding or incomplete translocation on the experimental time scale. A quantitative comparison between the release and dilution experiments is best accomplished by calculating the fraction of membrane-bound peptide, J, according to

J =

∑ (F (l ) − F

aq

∑ (F

b

)(

(l ) F b (l ) − F aq (l )

(l ) − F aq (l )

)

2

)

(1)

Here, F(l) is the concentration-normalized fluorescence intensity at wavelength l; F aq(l) is the corresponding value of free peptide in aqueous (aq) solution, which can be measured in the absence of lipid vesicles; and F b(l) is the corresponding value of completely membrane-bound (b) peptide, which is obtained in the presence of excess lipid (11). This chapter provides a recipe for performing uptake, release, and dilution experiments with the aid of intrinsic tryptophan fluorescence, and for analyzing such data on the basis of Eq. (1). The approach is exemplified using the cationic cell-penetrating peptide (CPP), penetratin (14), which carries two intrinsic tryptophan residues and avidly interacts with negatively charged phospholipid membranes (11, 13). The mechanisms of cellular internalization of penetratin and other CPPs have been the subject of controversy for several years (15, 16), and passive transbilayer diffusion has been invoked as a possible route of entry (17, 18). Using the fluorescence-spectroscopic approach presented here, however, we found no evidence of translocation across pure lipid membranes (11), which is in agreement with the results obtained by a number of other methods (13). The protocol, as presented here, turned out to be optimal for the specific case of penetratin. Other peptides, and even more

274

Broecker and Keller

non-peptidic fluorophores, will most likely require adaptations with respect to lipid and fluorophore concentrations. We, therefore, recommend to initially perform the uptake experiment to get an idea of the range of lipid concentrations required to observe near-complete membrane binding. The release and dilution experiments should then be carried out using lipid concentrations that lead to virtually complete membrane binding in the titrant solutions while allowing for desorption of >80% of the bound peptide after the first injection into buffer. At the same time, the peptide concentration has to be chosen so as to afford a reliably measurable fluorescence signal over the entire concentration range.

2. Materials 2.1. Stock Solutions

1. Lyophilized peptide (~100 nmol) or peptide stock solution (~1 mL, 100 µM). Peptide purity should be >95% by analytical high-performance liquid chromatography (HPLC). Most lyophilized peptides can be stored at −20°C for several months. The stability of a peptide solution depends on the peptide being used and has to be checked before starting any experiment. 2. Phospholipid (~45 mg, assuming an effective molar mass of 780 g/mol for a “typical” phospholipid (12)) dissolved in organic solvent (typically, a chloroform/methanol mixture) or in powder form. Lipids can be purchased from several suppliers, for instance, Avanti Polar Lipids (Alabaster, USA), Biosynth (Staad, Switzerland), or Genzyme Pharmaceuticals (Liestal, Switzerland). A zwitterionic phospholipid frequently used to mimic uncharged outer leaflets of mammalian plasma membranes is 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), whereas negatively charged bacterial membranes or inner leaflets of mammalian membranes can be imitated by using a mixture of POPC and either of the anionic phospholipids 1-palmitoyl-2-oleoyl-sn-glycero-3[phosphorac-(1-glycerol)] (POPG) or 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoserine (POPS) at a molar ratio of 3:1. Lipid purity should be >99%. Phospholipid dissolved in organic solvent or in powder form can be stored at −20°C for several months when overlaid with argon or nitrogen to prevent oxidation and hydrolysis. Phospholipid powder is potentially harmful if ingested or inhaled; avoid contact with eyes, skin, or clothing. 3. Buffer of choice (~20 mL; see Note 1). We routinely use 10 mM phosphate (NaH2PO4 and Na2HPO4) buffer containing 154 mM NaF and adjusted to pH 7.4. Buffer should be

Membrane Translocation Assayed by Fluorescence Spectroscopy

275

sterile-filtered and stored at 4°C to prevent microbial contamination. 4. Plastic vials (~2 mL, ~20×) and pipette tips; if necessary siliconized (see Note 2). 5. Round-bottomed glass vials (~5 mL, 2×) with tightly sealed lids. 6. If lipids are purchased in powder form, chloroform (~3 mL) of highest available purity (see Note 3). Store in a cool place away from light and ignition sources. Chloroform is dangerous to health; do not ingest, inhale, or get into contact with eyes, skin, or clothing. Always work under a hood and wear laboratory coat, goggles, and appropriate gloves that do not allow fast permeation of chloroform (made from, e.g., polyvinyl alcohol (PVA) or Viton). Waste material should be handled according to your institution’s waste disposal guidelines. 7. Disposable glass Pasteur pipette. 8. Nitrogen or argon gas source. 9. Exsiccator connected to high vacuum (~10−2–10−3 mbar). 2.2. Vesicle Preparation

1. Titanium-tip ultrasonicator (e.g., Labsonic L from B. Braun Biotech, Melsungen, Germany; Sonopuls HD 2070 from Bandelin electronic, Berlin, Germany) with clamp, ring stand, and 200-mL beaker. 2. Eppendorf or similar tubes (2 mL, 5×) suitable for centrifugation.

2.3. Fluorescence Spectroscopy

1. Fluorescence spectrometer (fluorometer). Light sources such as xenon lamps emit highly intense visible and ultraviolet (UV) radiation, which can seriously damage eyes. Never look directly into the light source and always wear safety glasses. 2. Quartz cuvette (1 × 1 cm; Hellma, Mühlheim, Germany). 3. Magnetic stir bar (<10 mm in length) or, if the fluorometer has no integrated stirrer unit, 1-mL pipette (if necessary, with siliconized pipette tips, see Note 2). 4. Hellmanex (~10 mL, 2% solution; Hellma, Mühlheim, Germany) or other cleaning agent for quartz cuvettes. 5. Methanol (~15 mL) of highest available purity. Methanol is harmful if inhaled or ingested. Always work under a hood. Wear laboratory coat, goggles, and appropriate gloves that do not allow fast permeation of methanol (made from, e.g., polyisobutylene, polyvinyl chloride (PVC), or Viton). Waste material should be handled according to your institution’s waste disposal guidelines.

276

Broecker and Keller

3. Methods 3.1. Preparation of Reagents for all Uptake, Release, and Dilution Experiments

1. Prepare 1 mL of a 100 µM peptide stock solution by either dissolving lyophilized peptide in buffer or diluting an existing stock solution (see Note 4). 2. Prepare two round-bottomed glass vials (~5 mL) with tightly sealed lids. Vial 1 is for the uptake and the dilution experiments and vial 2 for the release experiment; label the vials accordingly. Determine the tara weights of the vials (with lids on) on a microbalance, taking at least five significant digits (see Notes 5–6). 3. Add the following amounts of lipid dissolved in organic solvent to the vials (see Notes 7–9): ~40 mg (e.g., ~2 mL of a 20 mg/mL solution) to vial 1 (uptake and dilution) and ~5 mg (e.g., ~240 µL of a 20 mg/mL solution) to vial 2 (release). If lipid was purchased in powder form, first dissolve it in chloroform to yield a concentration of 20 mg/mL and then proceed as above. 4. Attach a glass Pasteur pipette to a nitrogen or argon gas port and pass a gentle gas stream through the pipette over the lipidcontaining organic solutions in the two vials (see Note 10). During this procedure, gently shake the vials such that a thin lipid film is deposited on the bottom and the walls of each vial while chloroform evaporates (see Note 11). Once all organic solvent has seemingly gone, continue for 5 min to thoroughly dry the lipid films. 5. Transfer all vials into an exsiccator and remove traces of chloroform by applying high vacuum (~10–2−10−3 mbar; see Note 12) overnight. 6. Take the vials out of the exsiccator and immediately seal them tightly to prevent water sorption from the air (see Note 13). Weigh the closed vials and subtract the tara weights determined above (3.1.2) to calculate the net weights of the lipid films (see Note 14).

3.2. Vesicle Preparation 3.2.1. Lipid Vesicles for Uptake Experiment

1. Suspend the dry lipid film in vial 1 in an appropriate volume of buffer (typically, ~1 mL; prewarmed to room temperature) by shaking and vortexing for at least 5 min to yield a final lipid concentration of 50 mM (see Note 15). This procedure results in the formation of large and polydisperse multilamellar vesicles (MLVs), which give the suspension a turbid appearance (see Notes 16–18). 2. Fill the 200-mL beaker with ice and water, and mount the vial clamp on the ring stand to position vial 1 in the ice/water bath in such a way that no water can splash into the open vial.

Membrane Translocation Assayed by Fluorescence Spectroscopy

277

Lower the titanium tip of the ultrasonicator into the MLV suspension and prepare small unilamellar vesicles (SUVs) (see Notes 19–21) by ultrasonication for 20 min at a power of ~80 W (see Notes 22–24). 3. Transfer the vial content into a 2-mL centrifuge tube and spin down titanium debris at 20,000× g for 5 min in a tabletop centrifuge. Transfer the supernatant into a new glass or plastic vial, referred to as vial 3. 3.2.2. Mixed Vesicles for Release Experiment

1. Add 500 µL of the 100 µM peptide stock solution (3.1.1) to 1,500 µL buffer to yield 2 mL of a 25 µM peptide solution. 2. Suspend the dry lipid film in vial 2 in an appropriate volume of this peptide solution (typically, ~ 1.2 mL; prewarmed to room temperature) by shaking and vortexing for at least 5 min to yield a final lipid concentration of 5 mM (see Note 15). This procedure results in the formation of MLVs that are preloaded with peptide on both membrane leaflets (see Notes 16–18). 3. Transform the MLVs into peptide-preloaded SUVs by ultrasonication and get rid of the titanium debris as detailed above for pure lipid vesicles (3.2.1.2–3.2.1.3; see Notes 22–25). After centrifugation (3.2.1.3), transfer the supernatant into a new glass or plastic vial, referred to as vial 4.

3.2.3. Lipid Vesicles for Dilution Experiment

1. Remove 100 µL of the SUV dispersion from vial 3 (3.2.1.3) and combine this with 400 µL buffer to yield 500 µL containing 10 mM lipid in a new vial, referred to as vial 5.

3.3. Fluorescence Spectroscopy

1. Initialize the emission scan mode of your fluorescence spectrometer, that is, the mode in which the emission spectrum is scanned while the excitation wavelength is held constant.

3.3.1. Instrument Setup

2. Set the excitation wavelength to 280 nm, which roughly corresponds to the absorption maximum of tryptophan. 3. Set the emission wavelength to scan the range from 300 nm to 500 nm (see Note 26). 4. Choose appropriate values for the slit widths of both excitation and emission monochromators (see Note 27).

3.3.2. Uptake Experiment

1. Add 30 µL of the 100 µM peptide stock solution (3.1.1) to 2,970 µL buffer to yield 3 mL of a 1 µM peptide solution. 2. Combine 5 µL of the peptide stock solution (3.1.1) with 95 µL buffer, and add this solution to 400 µL of the SUV dispersion in vial 3 containing 50 mM lipid (3.2.1.3) to prepare 500 µL of a titrant stock solution containing 40 mM lipid and 1 µM peptide (see Note 28). This solution is referred

278

Broecker and Keller

to as titrant solution 5 for the uptake experiment (abbreviation: U5) because it is the fifth titrant solution that will be injected during the uptake experiment. 3. Prepare four different 200-µL dilutions of this titrant stock solution using the 1 µM peptide solution prepared above (3.3.2.1): First, add 100 µL U5 to 100 µL 1 µM peptide to yield 20 mM lipid and 1 µM peptide (U4). Second, add 25 µL U5 to 175 µL 1 µM peptide to yield 5 mM lipid and 1 µM peptide (U3). Third, add 10 µL U5 to 190 µL 1 µM peptide to yield 2 mM lipid and 1 µM peptide (U2). Fourth, add 5 µL U5 to 195 µL 1 µM peptide to yield 1 mM lipid and 1 µM peptide (U1). 4. Fill the quartz cuvette with 1.5 mL buffer and place it in the sample compartment of the fluorescence spectrometer, set the excitation and emission wavelengths to 280 nm and 500 nm, respectively, and zero the system. 5. Empty and clean the cuvette. Flush 10 times with distilled water and 10 times with ultrapure water, rinse 2 times with methanol and completely remove the organic solvent by drying under a stream of nitrogen. 6. Put 1.5 mL of the 1 µM peptide solution (3.3.2.1) into the cuvette and place the cuvette in the sample compartment. 7. Record a fluorescence emission spectrum using the settings specified above (3.3.1). If necessary, adjust the sensitivity of the photodetector (see Note 29) to allow for reasonable signal intensity and good signal-to-noise ratio (see Note 30), and repeat the measurement. 8. Inject a 15-µL aliquot of titrant solution U1 (1 mM lipid, 1 µM peptide) into the cuvette containing 1 µM peptide. Carefully mix the cuvette content, either by pipetting up and down at least 3 times using a 1-mL pipette or, if the fluoro meter is fitted with an integrated stirrer unit, with the help of a small magnetic stir bar (see Note 31). Record an emission spectrum, wait for 3 min, and record a second spectrum to check if the system has reached equilibrium. If this is not the case, continue recording spectra in short time intervals until equilibrium is reached. 9. Repeat the previous step 9 times with 15-µL aliquots of U1 (1 mM lipid, 1 µM peptide), 10 times with 15-µL aliquots of U2 (2 mM lipid, 1 µM peptide), 10 times with 15-µL aliquots of U3 (5 mM lipid, 1 µM peptide), 1 time with a 150-µL aliquot of U4 (20 mM lipid, 1 µM peptide), and finally 1 time with a 150-µL aliquot of U5 (40 mM lipid, 1 µM peptide). 10. Empty and clean the cuvette. Flush 2 times with 2% Hellmanex and wash exhaustively, that is, 10 times with distilled water

Membrane Translocation Assayed by Fluorescence Spectroscopy

279

and 10 times with ultrapure water, rinse 2 times with methanol and completely remove the organic solvent by drying under a stream of nitrogen. 3.3.3. Release Experiment

1. The solution in vial 4 containing 5 mM SUVs preloaded with 25 µM peptide (3.2.2.3) is titrant solution 2 for the release experiment (R2). Add 20 µL R2 to 180 µL buffer to prepare a titrant solution containing 0.5 mM lipid and 2.5 µM peptide (R1). 2. Fill the cuvette with 1.5 mL buffer, place it in the sample compartment, set the excitation and emission wavelengths to 280 nm and 500 nm, respectively, and zero the system. 3. Inject a 15-µL aliquot of titrant solution R1 (0.5 mM lipid, 2.5 µM peptide) into the cuvette containing buffer. Mix the cuvette content, record an emission spectrum, wait for 3 min, and record a second spectrum, as detailed before for the uptake experiment (3.3.2.8). 4. Repeat the previous step 9 times with 15-µL aliquots of R1 (0.5 mM lipid, 2.5 µM peptide), 10 times with 15-µL aliquots of R2 (5 mM lipid, 25 µM peptide), and finally 10 times with 75-µL aliquots of R2 (5 mM lipid, 25 µM peptide; see Note 32). 5. Empty and clean the cuvette as detailed above (3.3.2.10).

3.3.4. Dilution Experiment

1. Combine 375 µL of the original peptide stock solution (3.1.1) with 375 µL buffer and add 750 µL of the SUV dispersion of vial 5 containing 10 mM lipid (3.2.3.1) to prepare 1.5 mL of a titrant solution containing 5 mM lipid and 25 µM peptide. This is titrant solution 2 for the dilution experiment (D2). 2. Add 20 µL of D2 to 180 µL buffer to prepare 200 µL of titrant solution containing 0.5 mM lipid and 2.5 µM peptide (D1). 3. Fill the cuvette with 1,500 µL buffer, place it in the sample compartment, set the excitation and emission wavelengths to 280 nm and 500 nm, respectively, and zero the system. 4. Inject a 15-µL aliquot of titrant solution D1 (0.5 mM lipid, 2.5 µM peptide) into the cuvette containing buffer. Mix the cuvette content, record an emission spectrum, wait for 3 min, and record a second spectrum, as detailed above for the uptake experiment (3.3.2.8). 5. Follow the same titration scheme as for the release experiment (3.3.3.4): Repeat the previous step 9 times with 15-µL aliquots of D1 (0.5 mM lipid, 2.5 µM peptide), 10 times with 15-µL aliquots of D2 (5 mM lipid, 25 µM peptide), and finally 10 times with 75-µL aliquots of D2 (5 mM lipid, 25 µM peptide; see Note 32). 6. Empty and clean the cuvette as detailed above (3.3.2.10).

280

Broecker and Keller

3.4. Data Evaluation 3.4.1. Uptake Experiment

1. Calculate the lipid concentration in the cuvette after each injection of the uptake experiment. For the injection schedule given above (3.3.2.8–3.3.2.9), the lipid concentration amounts to 9.90, 19.6, 29.1, 38.5, 47.6, 56.6, 65.4, 74.1, 82.6, 90.9, 108, 125, 142, 158, 174, 190, 205, 220, 235, 250, 289, 328, 366, 403, 440, 476, 512, 547, 581, 615, 2,000, and 4,533 µM after the first, second, …, and 32nd injection, respectively. 2. Import the fluorescence data recorded before the first injection and after the last injection into a spreadsheet program (such as Microsoft Office Excel, MicroCal Origin, or Systat SigmaPlot) and arrange them as follows: column 1: wavelength, l; column 2: fluorescence intensity before the first injection, that is, in the absence of lipid, F aq(l); column 3: fluorescence intensity after the last injection, that is, in the presence of excess lipid, F b(l). Save the spreadsheet in its current form, as this will be used for all subsequent steps during the evaluation of uptake, release, and dilution experiments. 3. Import the fluorescence data recorded after the first injection (at a lipid concentration of 9.90 µM) and copy the fluorescence intensity, F(l), into column 4 of the spreadsheet. 4. Determine the fraction of membrane-bound peptide, J, after the first injection according to Eq. (1) (see Note 33). To this end, proceed as follows: In column 5, calculate the term in the numerator of Eq. (1), (F(l)–F aq(l))(F b(l)−F aq(l)), by setting column 5 = (column 4−column 2)(column 3−column 2). In column 6, calculate the term in the denominator of Eq. (1), (F b(l)–F aq(l))2, by setting column 6 = (column 3−column 2)2. In an empty cell, divide the sum of the values in column 5 by the sum of the values in column 6 to obtain J. 5. Repeat the previous two steps for each spectrum recorded during the uptake experiment to obtain J for each lipid concentration calculated above (3.4.1.1). 6. Plot J as a function of lipid concentration in an x–y-diagram. An example is given in Fig. 2b for the uptake of penetratin into SUVs composed of POPC/POPG 3:1 (11).

3.4.2. Release Experiment

1. Calculate the lipid and peptide concentrations after each injection of the release experiment. For the injection schedule given above (3.3.3.3–3.3.3.4), the lipid concentration amounts to 4.95, 9.80, 14.6, 19.2, 23.8, 28.3, 32.7, 37.0, 41.3, 45.5, 90.1, 134, 177, 219, 261, 302, 342, 381, 420, 458, 640, 808, 963, 1,107, 1,241, 1,367, 1,484, 1,594, 1,697, and 1,794 µM and the peptide concentration to 0.0248, 0.0490, 0.0728, 0.0962, 0.119, 0.142, 0.164, 0.185, 0.206, 0.227, 0.450, 0.670, 0.885, 1.10, 1.30, 1.51, 1.71, 1.91, 2.10, 2.29, 3.20, 4.04, 4.81, 5.54, 6.21, 6.83, 7.42, 7.97, 8.48, and 8.97 µM after the first, second, …, and 30th injection, respectively.

Membrane Translocation Assayed by Fluorescence Spectroscopy

320 100

360

l (nm) 400

440

480

200 300 cL (mM)

400

281

a

F (au)

80 60 40 20 0 1.0

b

0.8

J

0.6 0.4 0.2 0.0 0

100

Fig. 2. Uptake experiment with penetratin and POPC/POPG 3:1 SUVs. Lipid (1–40 mM) vesicles were injected into a penetratin (1 µM) solution. (a) Fluorescence emission intensity, F, versus wavelength, l, upon excitation at 280 nm. Lipid concentrations were cL = 0, 20, 57, 91, 158, 328, and 4,533 mM (in order of ascending F values at l = 350 nm). Thirteen out of a total of 33 spectra are shown. (b) Fraction of membrane-bound penetratin, J, versus lipid concentration, cL. Experimental data (circles; filled symbols correspond to spectra depicted in (a)) were obtained from emission spectra using Eq. (1). Shown is also a fit (solid line) based on a membrane partitioning model assuming onesided membrane binding (10, 11). (Adapted with permission from ref. 11. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.)

2. Import the fluorescence data recorded after the first injection (at lipid and peptide concentrations of 4.95 µM and 0.0248 µM, respectively) into a spreadsheet and normalize the fluorescence intensity such that it corresponds to a peptide concentration of 1 µM (as used in the uptake experiment). To this end, multiply the measured intensity by 1 µM and divide it by the actual peptide concentration calculated above (3.4.2.1; i.e., 0.0248 µM after the first injection).

282

Broecker and Keller

3. Copy the concentration-normalized fluorescence intensity, F(l), calculated in the previous step into column 4 of the spreadsheet prepared above (3.4.1.2) and determine the fraction of membrane-bound peptide, J, after the first injection according to Eq. (1) as explained above (3.4.1.4; see Note 33). 4. Repeat the previous two steps for each spectrum recorded during the release experiment to obtain J for each lipid concentration calculated above (3.4.2.1). 5. Plot J as a function of lipid concentration in an x–y-diagram. An example is given in Fig. 3b for the release of penetratin from SUVs composed of POPC/POPG 3:1 (11). 3.4.3. Dilution Experiment

1. Calculate the lipid and peptide concentrations after each injection of the dilution experiment. For the injection schedule given above (3.3.4.4–3.3.4.5), the concentrations are the same as those obtained for the release experiment (3.4.2.1). 2. Import the fluorescence data recorded after the first injection (at lipid and peptide concentrations of 4.95 µM and 0.0248 µM, respectively) into a spreadsheet and normalize the fluorescence intensity such that it corresponds to a peptide concentration of 1 µM (as used in the uptake experiment). To this end, multiply the measured intensity by 1 µM and divide it by the actual peptide concentration calculated above (3.4.2.1; i.e., 0.0248 µM after the first injection). 3. Copy the concentration-normalized fluorescence intensity, F(l), calculated in the previous step into column 4 of the spreadsheet prepared above (3.4.1.2) and determine the fraction of membrane-bound peptide, J, after the first injection according to Eq. (1) as explained above (3.4.1.4; see Note 33). 4. Repeat the previous two steps for each spectrum recorded during the dilution experiment to obtain J for each lipid concentration calculated above (3.4.2.1). 5. Plot J as a function of lipid concentration in an x-y-diagram. An example is given in Fig. 4b for the dilution of penetratin and SUVs composed of POPC/POPG 3:1 (11).

3.4.4. Interpretation

1. Compare the plots created for release (3.4.2.5) and dilution (3.4.3.5) experiments (see Note 34). If the J values obtained for both kinds of titrations superimpose, it can be concluded that the peptide could equilibrate across the vesicle membranes during the titrations. If, by contrast, the J values for the dilution experiment are lower than those for the release

Membrane Translocation Assayed by Fluorescence Spectroscopy

320 100

360

l (nm) 400

440

480

200 300 cL (mM)

400

283

a

F (au)

80 60 40 20 0 1.0

b

0.8

J

0.6 0.4 0.2 0.0 0

100

Fig. 3. Release experiment with penetratin and POPC/POPG 3:1 SUVs. Lipid (5 mM) vesicles preloaded on both leaflets with penetratin (25 µM) were injected into buffer. (a) Concentration-normalized fluorescence emission intensity, F, versus wavelength, l, upon excitation at 280 nm. Lipid concentrations were cL = 4.95, 14.6, 23.8, 41.3, 90.1, 177, 342, 640, and 1,697 µM (in order of ascending F values at l = 350 nm). Nine out of a total of 30 spectra are shown. (b) Fraction of membrane-bound penetratin, J, versus lipid concentration, cL. Experimental data (circles; filled symbols correspond to spectra depicted in (a)) were obtained from emission spectra using Eq. (1). Shown is also a fit (solid line) based on a membrane partitioning model assuming one-sided membrane binding (10, 11). (Adapted with permission from ref. 11. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.)

experiment, the peptide was not able to equilibrate across the vesicle membranes on the experimental time scale. In the example shown in Figs. 3b and 4b, it is obvious from the difference between the release and dilution experiments that penetratin cannot cross the membranes of SUVs composed of POPC/POPG 3:1 (11).

284

Broecker and Keller

320 100

360

l (nm) 400

440

480

200 300 cL (mM)

400

a

F (au)

80 60 40 20 0 1.0

b

0.8

J

0.6 0.4 0.2 0.0 0

100

Fig. 4. Dilution experiment with penetratin and POPC/POPG 3:1 SUVs. Lipid (5 mM) vesicles externally loaded with penetratin (25 µM) were injected into buffer. (a) Concentrationnormalized fluorescence emission intensity, F, versus wavelength, l, upon excitation at 280 nm. Lipid concentrations were cL = 4.95, 14.6, 23.8, 41.3, 90.1, 177, 342, 640, and 1,697 µM (in order of ascending F values at l = 350 nm). Nine out of a total of 30 spectra are shown. (b) Fraction of membrane-bound penetratin, J, versus lipid concentration, cL. Experimental data (circles; filled symbols correspond to spectra depicted in (a)) were obtained from emission spectra using Eq. (1). Shown is also a fit (solid line) based on a membrane partitioning model assuming one-sided membrane binding (10, 11). (Adapted with permission from ref. 11. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.)

4. Notes 1. Ultrapure water (Milli-Q Element System, Millipore, Molsheim, France) with a resistivity ≥18 MW cm (at 25°C) and a total organic content of <5 ppb should be used for preparing the buffer. 2. Some peptides strongly adsorb to plastic and/or glass surfaces. If this is the case, try using siliconized tubes and pipette tips (19, 20) for all steps involved in peptide handling.

Membrane Translocation Assayed by Fluorescence Spectroscopy

285

3. Chloroform deteriorates rather quickly and therefore should be redistilled before use. After distillation, add 0.7–1.0% (vol/vol) ethanol for stabilization. 4. In case the peptide is not readily soluble in the buffer used, try dissolving it in a defined volume of tridistilled water before adding the same volume of double-concentrated buffer. Repulsive electrostatic forces, which may help keeping the peptide in solution, become weaker with increasing ionic strength. 5. Organic solvents such as chloroform are difficult to handle in a volumetrically precise and reproducible way. Therefore, the amounts of lipid transferred to the vials have to be determined gravimetrically. 6. Take care not to change the tara weights later by putting tape or writing on the vials. 7. When lipid mixtures are to be used, the different lipid species should be dried and weighed one by one in the same vial to allow for accurate quantification of membrane composition. 8. Lipids may be dissolved in other organic solvents such as tertiary butanol or cyclohexane, if chloroform is not suitable. 9. Usually, phospholipids are dissolved at concentrations of 10–20 mg/mL of chloroform; however, higher concentrations are often possible, though rarely necessary. 10. Avoid using too high pressure as that bears the risk of jerking lipid-containing droplets out of the vial. 11. Alternatively, and especially for volumes >5 mL, chloroform or other organic solvents can be removed in a rotary evaporator. However, be careful not to boil organic solvent traces left in the lipid film. 12. For small volumes (<1 mL), desiccating lipids by passing nitrogen or argon over the organic solution and then over the lipid film may be enough. 13. Vials without lid can be closed with Parafilm as well. 14. It may be helpful to note the net weight of the lipid both on the vial and in a laboratory book. 15. Always maintain temperatures above the lipid’s gel-to-liquid– crystalline phase transition temperature during hydration and vesicle formation. Some lipid species may require prolonged vigorous shaking or stirring (~1 h) for complete hydration. Some charged lipids tend to form viscous gels when hydrated at low ionic strength; adding additional salt may help circumvent this problem. 16. MLV suspensions can be stored under nitrogen or argon at −20°C for months.

286

Broecker and Keller

17. Apart from sonication, extrusion is a popular method for downsizing MLVs (21–23). Extrusion yields large unilamellar vesicles (LUVs), with diameters ranging from ~50 nm to several hundred nanometres, depending on the pore size used. Before extrusion, MLVs should be subjected to 5–10 freeze–thaw cycles using liquid nitrogen and a water bath. This is particularly important when preparing vesicles for the release experiment where the peptide needs to distribute uniformly over both membrane leaflets. For extrusion, connect two gas-tight 0.5-mL or 1-mL glass syringes (Hamilton, Reno, USA) to a mini extruder (Avestin, Ottawa, Canada) and force the MLV suspension through two stacked polycarbonate membranes (Nucleopore, Pleasanton, USA) having a defined pore size. 18. Light scattering can cause serious problems when performing fluorescence-spectroscopic experiments (24). Owing to their smaller diameters and ensuing reduced scattering effects, SUVs are preferable over LUVs. Scattering effects can also be taken into account by running suitable control experiments (24). 19. Freshly prepared SUVs have to be stored at room temperature. Because of their pronounced membrane curvature, SUVs are mechanically metastable and will irreversibly agglomerate when incubated at low temperature. Therefore, SUV dispersions should be used for no longer than 2–3 days after preparation and must always be stored under nitrogen or argon. 20. Sonication can be applied to many lipid mixtures, including lipid extracts from natural sources. In the latter case, it is often helpful to rupture MLVs prior to sonication by 5–10 freeze– thaw cycles using liquid nitrogen and a water bath. 21. It has been reported (25) that letting MLV suspensions age overnight facilitates downsizing and sharpens the vesicle size distribution. 22. Sonication is accompanied by a change in appearance of the lipid suspension from very hazy to translucent or even clear, depending on the lipid. 23. Bath sonication is gentler than tip sonication and offers a better temperature control, but can be applied to sample volumes <1 mL only. 24. It is highly recommended to check the size and polydispersity (size distribution) of the vesicles after sonication (or extrusion) by dynamic light scattering (DLS). 25. For the release experiment, it should also be confirmed that the presence of the peptide during sonication (or extrusion) has no influence on the size and polydispersity of the vesicles.

Membrane Translocation Assayed by Fluorescence Spectroscopy

287

26. As a rule of thumb, the lowest wavelength included in the emission spectrum should equal the sum of the excitation wavelength and both monochromator slit widths. For instance, excitation at 280 nm with slit widths of 5 nm each requires a start value of ≥290 nm. 27. Slit widths primarily determine the spectral resolution and also affect the radiation intensities transmitted from the light source to the sample and from the sample to the photomultiplier. Narrowing the slit widths thus reduces the signal intensity but enhances the spectral resolution, although the latter is of limited relevance to the present purpose. More importantly, photobleaching effects can be strongly reduced by keeping the excitation slit width as narrow as possible, whereas the emission slit width may be increased in order to allow for reasonably strong signal intensity. At a fixed excitation slit width, a twofold increase in the emission slit width will evoke a fourfold increase in the fluorescence emission intensity. Photobleaching effects can be further minimized by closing the shutters whenever no data are being collected and by any other means that reduces the sample’s exposure time to the light source. 28. Peptide is included in the titrant solution of the uptake experiment to avoid fluorophore dilution upon addition of lipid vesicles to the peptide solution. 29. As the release and dilution experiments span a very wide range of fluorophore (peptide) concentrations, it may be necessary to adjust the sensitivity of the photomultiplier during these titrations. In such a case, the raw emission data have to be normalized later not only with respect to the peptide concentration (3.4.2.2 and 3.4.3.2) but also with respect to the instrument sensitivity. To this end, record several emission scans of the same sample at different sensitivities, determine the intensity ratio at a given sensitivity as compared with the sensitivity used for the uptake experiment, and divide each spectrum by the corresponding intensity ratio. 30. The signal-to-noise ratio can also be improved by increasing the number of accumulations recorded by the spectrometer. Note that the signal-to-noise ratio is proportional to the square root of the number of accumulations; that is, a twofold improvement in signal-to-noise ratio will require a fourfold increase in the number of accumulations. 31. As fluorescence spectroscopy is very sensitive, do not forget to thoroughly clean the magnetic stir bar before and between the experiments. Wash carefully as described for cuvettes (3.3.2.10). 32. To assess slow membrane partitioning or translocation, you may want to perform a control experiment at the strongest

288

Broecker and Keller

dilution tested and incubate the sample for several hours before checking if the emission spectrum changes on the experimental time scale. 33. Rather than following the deconvolution procedure described in the text, you may also rely on a single wavelength, l, to calculate the fraction of membrane-bound peptide, J. Equation (1) then simplifies to J = (F(l)−Faq(l))/(Fb(l)−Faq(l)). This approach is more straightforward than employing Eq. (1); however, it does not make use of all the information contained in the emission spectra and therefore usually results in greater scatter when plotting J as a function of lipid concentration. If you nevertheless decide to use the simplified equation, make sure to choose a wavelength close to the emission peak (~350 nm for tryptophan). 34. On top of the qualitative comparison of release and dilution experiments presented here, uptake, release, and dilution experiments can be fitted globally using a simple membrane partition equilibrium model accounting for electrostatic effects by means of Gouy–Chapman theory (10, 11). For the example of penetratin, a satisfactory fit can be achieved only under the assumption of one-sided membrane binding; the fitted titration data are depicted as solid lines in Figs. 2 – 4b.

Acknowledgments We thank Heike Nikolenko (FMP) and Matthias Böthe (Robert Koch Institute, Berlin, Germany) for excellent technical assistance and Sebastian Fiedler (FMP) for helpful comments on the manuscript. We are indebted to Dr. Michael Beyermann, Dagmar Krause, and Bernhard Schmikale for synthesis and purification and to Drs. Eberhard Krause and Michael Schümann (all FMP) for mass-spectrometric characterization of penetratin peptide. This work was supported by the European Commission with grant No. QLK3-CT-2002-01989 to S.K. References 1. Heerklotz HH, Binder H, Epand RM (1999) A “release” protocol for isothermal titration calorimetry. Biophys J 76:2606–2613 2. Heerklotz H, Seelig J (2000) Titration calorimetry of surfactant–membrane partitioning and membrane solubilization. Biochim Biophys Acta 1508:69–85

3. Heerklotz H (2004) The microcalorimetry of lipid membranes. J Phys Condens Matter 16:R441–R467 4. Hagen V, Dekowski B, Nache V, Schmidt R, Geissler D, Lorenz D, Eichhorst J, Keller S, Kaneko H, Benndorf K, Wiesner B (2005) Coumarinylmethyl esters for ultrafast release

Membrane Translocation Assayed by Fluorescence Spectroscopy

5.

6.

7.

8.

9.

10.

11.

12.

13.

14. 15.

of high concentrations of cyclic nucleotides upon one- and two-photon photolysis. Angew Chem Int Ed 44:7887–7891 Cambridge SB, Geissler D, Keller S, Cürten B (2006) A caged doxycycline analogue for photoactivated gene expression. Angew Chem Int Ed 45:2229–2231 Gilbert D, Funk K, Dekowski B, Lechler R, Keller S, Möhrlen F, Frings S, Hagen V (2007) Caged capsaicins: new tools for the examination of TRPV1 channels in somatosensory neurons. ChemBioChem 8:89–97 Heerklotz H, Szadkowska H, Anderson T, Seelig J (2003) The sensitivity of lipid domains to small perturbations demonstrated by the effect of Triton. J Mol Biol 329:793–799 Tsamaloukas A, Szadkowska H, Slotte PJ, Heerklotz H (2005) Interactions of cholesterol with lipid membranes and cyclodextrin characterized by calorimetry. Biophys J 89:1109–1119 Tsamaloukas A, Szadkowska H, Heerklotz H (2006) Thermodynamic comparison of the interactions of cholesterol with unsaturated phospholipid and sphingomyelins. Biophys J 90:4479–4487 Keller S, Heerklotz H, Blume A (2006) Monitoring lipid membrane translocation of sodium dodecyl sulfate by isothermal titration calorimetry. J Am Chem Soc 128: 1279–1286 Keller S, Böthe M, Bienert M, Dathe M, Blume A (2007) A simple fluorescence-spectroscopic membrane translocation assay. ChemBioChem 8:546–552 Tsamaloukas AD, Keller S, Heerklotz H (2007) Uptake and release protocol for assessing membrane binding and permeation by way of isothermal titration calorimetry. Nat Protoc 2:695–704 Bárány-Wallje E, Keller S, Serowy S, Geibel S, Pohl P, Bienert M, Dathe M (2005) A critical reassessment of penetratin translocation across lipid membranes. Biophys J 89:2513–2521 Derossi D, Chassaing G, Prochiantz A (1998) Trojan peptides: the penetratin system for intracellular delivery. Trends Cell Biol 8:84–87 Drin G, Déméné H, Temsamani J, Brasseur R (2001) Translocation of the pAntp peptide

289

and its amphipathic analogue AP-2AL. Biochemistry 40:1824–1834 16. Persson D, Thorén PEG, Esbjörner EK, Goksör M, Lincoln P, Nordén B (2004) Vesicle size-dependent translocation of penetratin analogs across lipid membranes. Biochim Biophys Acta 1665:142–155 17. Thorén PEG, Persson D, Karlsson M, Nordén B (2000) The Antennapedia peptide penetratin translocates across lipid bilayers – the first direct observation. FEBS Lett 482:265–268 18. Terrone D, Sang SLW, Roudaia L, Silvius JR (2003) Penetratin and related cell-penetrating cationic peptides can translocate across lipid bilayers in the presence of a transbilayer potential. Biochemistry 42:13787–13799 19. Chico DE, Given RL, Miller BT (2003) Binding of cationic cell-permeable peptides to plastic and glass. Peptides 24:3–9 20. Persson D, Thorén PEG, Herner M, Lincoln P, Nordén B (2003) Application of a novel analysis to measure the binding of the membrane-translocating peptide penetratin to negatively charged liposomes. Biochemistry 42:421–429 21. Hope MJ, Bally MB, Webb G, Cullis PR (1985) Production of large unilamellar vesicles by a rapid extrusion procedure. Characterization of size distribution, trapped volume and ability to maintain a membrane potential. Biochim Biophys Acta 812:55–65 22. Mayer LD, Hope MJ, Cullis PR (1986) Vesicles of variable sizes produced by a rapid extrusion procedure. Biochim Biophys Acta 858:161–168 23. MacDonald RC, MacDonald RI, Menco BPM, Takeshita K, Subbarao NK, Hu LR (1991) Small-volume extrusion apparatus for preparation of large, unilamellar vesicles. Biochim Biophys Acta 1061:297–303 24. Ladokhin AS, Jayasinghe S, White SH (2000) How to measure and analyze tryptophan fluorescence in membranes properly, and why bother? Anal Biochem 285:235–245 25. Chatterjee S, Banerjee DK (2002) Preparation, isolation, and characterization of liposomes containing natural and synthetic lipids. Methods Mol Biol 199:3–16

Chapter 20 Interaction of Lipids and Ligands with Nicotinic Acetylcholine Receptor Vesicles Assessed by Electron Paramagnetic Resonance Spectroscopy Hugo Rubén Arias Abstract Electron paramagnetic resonance (EPR) spectroscopy is a powerful technique that permits the study of membrane-embedded proteins in its lipid environment by assessing the interaction of spin labels with the protein in its natural environment (i.e., native membranes) or in reconstituted systems prepared with exogenous lipid species. Nicotinic acetylcholine receptors (AChRs) contain a large surface in intimate contact with the lipid membrane. AChRs, members of the Cys-loop receptor superfamily, have essential functional roles in the nervous system and its malfunctioning has been considered as the origin of several neurological diseases including Alzheimer’s disease, drug addiction, depression, and schizophrenia. In this regard, these receptors have been extensively studied as therapeutic targets for the action of several drugs. The majority of the marketed medications bind to the neurotransmitter sites, the so-called agonists. However, several drugs, some of them still in clinical trials, interact with non-competitive antagonist (NCA) binding sites. A potential location for these binding sites is the proper ion channel, blocking ion flux and thus, inhibiting membrane depolarization. However, several NCAs also bind to the lipid–protein interface, modulating the AChR functional properties. The best known examples of these NCAs are local and general anesthetics. Several endogenous molecules such as free fatty acids and neurosteroids also bind to the lipid–protein interface, probably mediating important physiological functions. Phospholipids, natural components of lipid membranes interacting with the AChR, are also essential to maintain the structural and functional properties of the AChR. EPR studies showed that local anesthetics bind to the lipid–protein interface by essentially the same dynamic mechanisms found in lipids, and that local and general anesthetics preferably decrease the phospholipid but not the fatty acid interactions with the AChR. This is consistent with the existence of annular and non-annular lipid domains on the AChR. Key words: Nicotinic acetylcholine receptors, Lipid–protein interface, Native and reconstituted vesicles, Electron paramagnetic resonance spectroscopy, Spin labels, Local anesthetics, General anesthetics, Fatty acids, Phospholipids, Steroids, Gangliosides

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_20, © Humana Press, a part of Springer Science+Business Media, LLC 2010

291

292

Arias

1. Introduction Pentameric ligand-gated ion channels (LGICs) play key roles in chemical synapses throughout the nervous system. They include receptors activated by neurotransmitters such as acetylcholine (ACh), g-aminobutyric acid, glycine, and serotonin (1). These receptors are known as the Cys-loop receptor superfamily because all subunits contain a pair of disulfide-bonded cysteines, separated by 13 residues in their extracellular amino-terminal domains. The analysis of the evolutionary relationships within the superfamily suggests that the ancestor receptor was probably homooligomeric and appeared 2,500 million years ago (2). Nicotinic acetylcholine receptors (AChRs) are involved in important physiological functions such as neuromuscular and motor autonomous transmission, learning and memory, and in several pathophysiological processes (3, 4). In addition, different AChR types are expressed in nonneuronal cells, modulating important physiological functions (5). Cys-loop receptors are composed of either five identical subunits (i.e., homopentamers) or different subunits (i.e., heteropentamers) arranged around an axis perpendicular to the plane of the membrane. Based on the presence of two adjacent Cys residues corresponding to positions 192 and 193 of the Torpedo a1 subunit, which participates in agonist binding, the AChR subunit classes are classified as a (if they contain both Cys residues) and non-a (if they do not contain the Cys residues). To date, several nicotinic subunits have been identified including a1−a10, b1−b4, d, and g or e (depending on the muscle development stage) subunits (reviewed in refs. 1, 3, 4, 6). There is evidence supporting the existence of AChRs containing one, two, three, or even four different subunits (4). The muscle-type AChR is a heteropentameric protein formed by four distinct subunits. The adult muscletype AChR is composed of a12b1ed, whereas the combination a12b1gd is present in embryonic or denervated muscle as well as in electric organs from Torpedo and Electrophorus species. The existence of a high number of structurally distinct receptor entities, each with different ligand sensitivities, suggests that each AChR class may have a distinct physiological function (4). 1.1. Overall Structure of the AChR

All Cys-loop receptor subunits can be divided structurally and functionally into three domains: the extracellular, transmembrane, and cytoplasmic domains. Figure 1 depicts the Torpedo AChR structure derived from cryo-electron microscopy at ~4 Å resolution, where these three domains can be identified (7, 8). Extracellular domain: the N-terminal hydrophilic extracellular portion bears the neurotransmitter binding sites, several glycosylation sites, and the 15-residue Cys-loop (1, 3, 4, 6, 9, 10).

Interaction of Lipids and Ligands with AChRs Vesicles

293

E

TM

C

Fig. 1. Molecular model of the Torpedo AChR at 4 Å resolution (modified from ref. 8). The coordinates were taken from the Protein Data Bank with accession code 2BG9B (8). Three main domains can be identified: the extracellular domain (E) bearing the agonist/ competitive antagonist binding sites, the transmembrane domain (TM) forming the ion channel and the lipid–protein interface, and the amphiphatic helix of the cytoplasmic domain (C)

Details of the neurotransmitter binding site for the AChR have been revealed at near atomic resolution using a molluscan ACh binding protein (AChBP) (11). This protein has, therefore, become a functional and structural surrogate of the extracellular domain of Cys-loop receptors (4, 10). Results from affinity labeling and site-directed mutagenesis studies, which were later supported by the structural model of AChBP, revealed that the binding site is formed at the interfaces between subunits (1, 9, 10). One subunit, called the principal subunit, contributes three loops that span b-strands and harbor key aromatic residues. The adjacent subunit, called the complementary subunit, contributes three b-strands that harbor aromatic and hydrophobic residues, and a fourth loop that harbors negatively charged residues. Cytoplasmic domain: this hydrophilic portion is located between segments M3 and M4. Part of this domain, particularly the amphipathic helix, was resolved in Unwin´s molecular model (see Fig. 1). It carries several phosphorylation sites that modulate receptor function (12). This domain is also involved in anchoring these receptors to cytoplasmic proteins (4). Protein-protein contacts may affect the receptor assembly, trafficking, clustering, targeting, modulation, and turnover, as well as gating kinetics.

294

Arias

Transmembrane domain: each subunit is formed by four highly hydrophobic segments designated as M1, M2, M3, and M4. Each one of these membrane-spanning segments has a dimension of 30–35 Å (~40 Å including the phospholipid headgroup portion). The five M2 transmembrane a-helices form the wall of the ion channel. The M1, M3, and in particular the M4 transmembrane segment, are in contact with the lipid membrane. Using cryo-electron microscopy, a model of the closed pore from the Torpedo AChR has been reported (7, 8). This model shows that the pore is shaped by an inner ring of five a-helices (M2 segments). In addition, an outer ring of 15 a-helices (M1, M3, and M4 segments), which coil around each other, shields the inner ring from the lipids. The pore is maximally constricted in the middle of the membrane due to side-to-side interactions between hydrophobic residues of neighboring helices. This tight hydrophobic girdle, which may correspond to the gate, creates an energetic barrier to ions across the membrane (7). These authors suggested that during gating, twisting of the M2 helices might disrupt the girdle and, in turn, allow ion flux. There are two categories of domains within muscle-type AChR channels (1, 9, 10). There is an uncharged domain formed by a series of different rings, vectorially disposed from the extracellular to the intracellular channel portion in the following order: valine ring (position 13¢), leucine ring (position 9¢), serine ring (position 6¢), and threonine ring (position 2¢). In turn, this uncharged portion is framed by two negatively charged regions: an anionic ring located at the extracellular portion of the channel [i.e., the outer ring (position 20¢)) and two more anionic rings [i.e., the intermediate ring (position 2¢) and the inner ring (position 5¢)] located near the cytoplasmic portion of the channel. The structure of the AChR at the lipid interface has been characterized by the use of the hydrophobic probe 3-trifluoromethyl3-(m-[125I]-iodophenyl)diazirine ([125I]TID) (13, 14). [125I]TID is incorporated into M1, M3, and M4 segments in a non-specific, agonist-insensitive, manner. The pattern of labeling of these transmembrane domains by [125I]TID revealed that the M4 domain is the most external one and that it has the most extensive contact with membrane lipids. Additional photoaffinity labeling studies allowed the identification of the amino acids that interact with the surrounding lipids (13, 15). For instance, [3H]azicholesterol, a photosensitive cholesterol analog, labeled acidic residues (e.g., a1-Asp407) located at either end of each transmembrane helix (15). Experimental evidence has also shown that these lipidexposed segments contribute to channel gating kinetics (1, 10). By using fluorescence spectroscopy techniques, Jones and McNamee (16) distinguished the presence of annular and

Interaction of Lipids and Ligands with AChRs Vesicles

295

non-annular lipid domains at the AChR that show different lipid selectivity. The perimeter of the AChR is surrounded by lipids, i.e. the annular lipid domain. However, the exact location of the non-annular lipid domain on the AChR is unknown. Indirect determinations have suggested that this domain may be located either between the five subunits of the AChR and/ or between the crevices existing within the four transmembrane segments (M1–M4) from each subunit (16) (1, 10). In a more recent work, Brannigan et al. (17) studied the contribution of cholesterol to the structural stability of the AChR membrane domain by using molecular modeling. They hypothesized the existence of three layers of organized cholesterol molecules around the central pore. Some of the cholesterol molecules were additionally found to be important for the functional properties of the AChR as they stabilize the anchors between the b1-b2 loop and M2-M3 loop. The proposed locations for annular and non-annular lipid domains on the AChR are shown in Fig. 2. In addition to the structural importance of lipid molecules surrounding the AChR protein, some lipids such as fatty acids and steroids are pharmacologically active, and might have crucial physiological roles in the endogenous modulation of these receptors (18). Finally, other clinically valuable compounds such as local and general anesthetics interact with the lipid–protein interface of the AChR, modulating its properties (9, 19, 20).

Fig. 2. Location of cholesterol molecules at the AChR (modified from ref. 17). (a) Transmembrane domain of the Torpedo AChR showing potential cholesterol sites located in the groove behind M4 and in direct contact with the phospholipid (in yellow), at the interface between subunits, bordered by M1 and M2 of one subunit and M2 and M3 of the adjacent subunit (in orange), and in the subunit center, bordered by M1, M2, M3, and yellow sites (in red). (b) Docking of cholesterol molecules to proposed binding sites. M1 is purple, M2 is green, M3 is blue, and M4 is cyan

296

Arias

2. Materials 2.1. Preparation of Torpedo AChR Membranes

1. Torpedo marmorata electric fish (Bay of Arcachon, France) were maintained in acquaria until use. 2. Torpedo electric organs were freshly dissected and stored at −80°C. Alternatively, frozen Torpedo californica electric organs can be obtained from Aquaric Research Consultants (CA, USA). 3. Homogenization buffer (21): 10 mM sodium phosphate buffer, pH 7.4, containing 0.4 M NaCl, 5 mM EDTA, 5 mM EGTA, and the following proteinase inhibitors: 3 mM phenylmethanesulfonyl fluoride, 5 mM iodoacetamide, 5 mg/mL leupeptin, 5 mg/mL chymostatin, and 5 mg/mL antipain. Alternatively, other proteinase inhibitors can be used (22). The proteinase inhibitors leupeptin, chymostatin, and antipain (Sigma-Aldrich, St. Louis, MO, USA) are stored at −15°C, whereas the other drugs can be stored at room temperature (RT). 4. Sucrose (Sigma-Aldrich, St. Louis, MO, USA) was dissolved at different concentrations (i.e., 50%, 39%, and 35% (w/v)) using 10 mM sodium phosphate buffer, pH 7.4, containing 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 0.02% NaN3 (Sigma-Aldrich, St. Louis, MO, USA). 5. “Stored buffer”: 10 mM sodium phosphate buffer, pH 7.4, containing 0.25 M sucrose, 1 mM EDTA, and 0.02% NaN3.

2.2. Purification of AChRs

1. Vesicle dialysis (VD) buffer: 10 mM MOPS, 100 mM NaCl, 0.1 mM EDTA, 0.02% NaN3, pH 7.4 (Sigma-Aldrich, St. Louis, MO, USA). 2. Sodium cholate (USB Corp., Cleveland, OH, USA) was dissolved in VD buffer at 1% (w/v) final concentration. 3. Affi-Gel 10 (Bio-Rad Laboratories, Richmond, CA, USA). 4. Bromoacetylcholine bromide (1.5 g) (Research Biochemical International, Natick, MA, USA) was dissolved in 50 mL of 10 mM sodium phosphate, 100 mM NaCl, 0.1 mM EDTA, pH 7.0. 5. Cystamine (Sigma-Aldrich, St. Louis, MO, USA) solution (50 mg/mL) was prepared in 0.1 M HEPES, pH 7.5. 6. 200 mM Tris-HCl, pH 8.0, was prepared in the presence of 20 mM Dithiothreitol (Sigma-Aldrich, St. Louis, MO, USA) (see Note 1). 7. Asolectin (a crude soybean lipid extract) (0.2 mg/mL) (Sigma-Aldrich, St. Louis, MO, USA) was dissolved in 1% cholate in VD buffer.

Interaction of Lipids and Ligands with AChRs Vesicles

297

8. Carbamylcholine chloride (CCh) (10 mM) (Sigma-Aldrich, St. Louis, MO, USA) was dissolved in 1% cholate in VD buffer. 9. Assay solution to detect sulfhydryls: 1 mL 10 mM dithiobisnitrobenzoic acid (prepared in ethanol) and 30 mL of 0.1 M Tris-HCl, pH 8.0 (Sigma-Aldrich, St. Louis, MO, USA). 10. Diisopropylfluorophosphate Louis, MO, USA).

(DFP)

(Sigma-Aldrich,

St.

11. Filter paper Whatman No 1 (Whatman Inc., Clifton, NJ, USA). 12. Washing buffer: 10 mM sodium phosphate, 100 mM NaCl, 0.1 mM EDTA, pH 7.0. 2.3. Total Protein Determination

1. Bovine serum albumin (BSA) (Sigma-Aldrich, St. Louis, MO, USA) was dissolved at a concentration of 2 mg/mL with bidistilled water and stored at 4°C. 2. Bicinchoninic acid (BCA) protein assay kit (Thermo Fisher Scientific, Rockford, IL, USA).

2.4. AChR Membrane Specific Activity

1. N-[3H]Propionyl-a-bungarotoxin (specific activity 107 Ci/ mmol) (General Electric Healthcare Life Sciences, Buckinghamshire, UK) was stored in an aqueous solution containing 0.02% BSA at −15°C. 2. DEAE-cellulose disks (DE81) (Whatman Inc., Clifton, NJ, USA). 3. Dansyltrimethylamine ([1-(dimethylamino) napthalene-5sulfonamido] ethyltrimethylammonium perchlorate) (Thermo Fisher Scientific, Rockford, IL, USA) was prepared in methanol at a final concentration of 3.8 mM. 4. Phencyclidine hydrochloride (Sigma-Aldrich, St. Louis, MO, USA) was prepared in VD buffer at a final concentration of 10 mM. 5. Suberyldicholine dichloride (Sigma-Aldrich, St. Louis, MO, USA) was prepared in VD buffer at a final concentration of 22 mM.

2.5. EPR Experiments

1. Spin-labeled stearic acid isomers including 6-, 9-, 12-, and 14-SASL (Molecular Probes, Eugene, OR, USA) were dissolved in absolute ethanol. 2. Spin-labeled phosphatidylcholine isomers (14-PCSL and 16-PCSL), spin-labeled phosphatidic acid (14-PASL), spinlabeled phosphatidylethanolamine (14-PESL), spin-labeled phosphatidylglycerol (14-PGSL), spin-labeled phosphatidylserine (14-PSSL), 1,2-dioleoyl-sn-glycero-3-phosphotempocholine (HGSL), and dimyristoylphosphatidylcholine (Avanti Polar Lipids, Alabaster, AL, USA) were dissolved in absolute ethanol.

298

Arias

3. Spin-labeled cholestane (4¢,4¢-dimethylspiro[5-a-cholestane3,2¢-oxazolidin]-3’-xyloxyl) (Syva Research Chemicals, Palo Alto, CA, USA) was dissolved in absolute ethanol. 4. Egg yolk phosphatidylcholine (PC) and phosphatidic acid (PA) (Lipid Products, South Nutfield, U.K.). 5. Yeast phosphatidylinositol (PI) (~50 mg) (Lipid Products, South Nutfield, U.K.) was thoroughly dried by repeated evaporation from solution in dry benzene, and then dissolved in dry chloroform (~5 mL) to which acetic anhydride (1 mmol) and dimethylaminopyridine (500 mmol) (Sigma-Aldrich, St. Louis, MO, USA) were added. 6. Ethyl chloroformate (2.2 g) was dissolved in 50 mL ethyl acetate (E. Merck, Darmstadt, Germany). 7. Dimyristoylphosphatidylcholine (Avanti Polar Lipids, Inc., Alabaster, AL, USA). 8. Triethylamine (Sigma-Aldrich, St. Louis, MO, USA). 9. 2-Diethylaminoethanethiol hydrochloride (Sigma-Aldrich, St. Louis, MO, USA). 10. Methyl ethyl ketone (Sigma-Aldrich, St. Louis, MO, USA). 11. Acetone (Sigma-Aldrich, St. Louis, MO, USA). 12. Diazomethane (Sigma-Aldrich, St. Louis, MO, USA) was freshly dissolved in ether. 13. Hydrazine hydrate (Sigma-Aldrich, St. Louis, MO, USA). 14. 4-Pyrrolidinopyridine (Sigma-Aldrich, St. Louis, MO, USA). 15. 2,4,8-Triisopropylbenzenesulfonyl chloride (Sigma-Aldrich, St. Louis, MO, USA) (~47 mg) was dissolved in pyridine. 16. Benzocaine, procaine hydrochloride, and tetracaine hydrochloride (Sigma-Aldrich, St. Louis, MO, USA) were dissolved in ethanol at a final concentration of 1 mg/mL. 17. Procaine hydrochloride (2.4 g) (Sigma-Aldrich, St. Louis, MO, USA) was dissolved in N,N-dimethylformamide (DMF). 18. 1-Hexanol (Sigma-Aldrich, St. Louis, MO, USA). 19. Cholesterol (Sigma-Aldrich, St. Louis, MO, USA). 20. K2CO3 (Sigma-Aldrich, St. Louis, MO, USA). 21. Na2SO4 (Sigma-Aldrich, St. Louis, MO, USA). 22. Sodium acetate buffer, pH 8.0, was prepared in the presence of 0.1 M CaCl2 (Sigma-Aldrich, St. Louis, MO, USA). 23. Sodium bicarbonate (Sigma-Aldrich, St. Louis, MO, USA). 24. Acetic acid (E. Merck, Darmstadt, Germany). 25. Acetic anhydride (E. Merck, Darmstadt, Germany). 26. Ninhydrin (E. Merck, Darmstadt, Germany).

Interaction of Lipids and Ligands with AChRs Vesicles

299

27. Diisopropylethylamine (E. Merck, Darmstadt, Germany). 28. 2-Diethylaminoethylamine (E. Merck, Darmstadt, Germany). 29. N-hydroxysuccinimide (E. Merck, Darmstadt, Germany). 30. N,N¢-dicyclohexylcarbodiimide Germany).

(E.

Merck,

Darmstadt,

31. Thionyl chloride (E. Merck, Darmstadt, Germany). 32. Silica gel 60 F254 TLC plates (E. Merck, Darmstadt, Germany). 33. Porcine pancreatic phospholipase A2 (Boehringer, Mannheim, Germany). 34. 9-Fluorenylmethyl chloroformate (Fmoc Cl) (7.8 mg) (Fluka, Buchs, Switzerland) was dissolved in 1 mL n-hexane. 35. Isoflurane (Anaquest, Anaheim, CA, USA). 36. Solvent system A: chloroform:methanol:water (127:63:10, by vol.) (Sigma-Aldrich, St. Louis, MO, USA). 37. Solvent system B: chloroform:methanol:water (3:48:47, by vol.). 38. Solvent system C: ether:methanol (95:5, v/v) (Sigma-Aldrich, St. Louis, MO, USA). 39. 0.8 M KOH was dissolved in dry methanol (15 mL) saturated with argon (Sigma-Aldrich, St. Louis, MO, USA). 40. Solvent system D: chloroform:methanol:2.5 M ammonia (60:35:8, by vol.) (Sigma-Aldrich, St. Louis, MO, USA). 41. Solvent system E: 2-propanol:n-hexane:water (55:40:5, by vol.) (E. Merck, Darmstadt, Germany). 42. Solvent system F: chloroform:methanol:2.5 M ammonia (65:25:4, by vol.). 43. 1-Oxyl-2,2,5,5-tetramethylpyrroline-3-carboxylic acid (3.6 g) (E. Merck, Darmstadt, Germany) was dissolved in dry pyridine (5 mL) and benzene (30 mL). 44. Anydride ester (2.6 g) was dissolved in 80 mL DMF:toluene (1:1, v/v) (E. Merck, Darmstadt, Germany). 45. Solvent system G: chloroform:0.1 N HCl (1:1, v/v). 46. Solvent system H: dichloromethane:methanol:6 N NH4OH (63:34:3, by vol.) (Sigma-Aldrich, St. Louis, MO, USA).

3. Methods 3.1. Preparation of Torpedo AChR Native Vesicles

1. T. marmorata electric fish were killed by pithing, and the electric organs dissected and stored at −80°C or in liquid N2 until use (see Note 2).

300

Arias

2. Electric organ tissue was homogenized using a Virtis model 60 homogenizer at a speed of 10,000 rpm for 1 min (15 s intervals), with 1 volume of homogenization buffer (see Note 3). 3. The homogenate was pelleted at 9,000× g for 10 min at 4°C, and the supernantant was filtered through cheesecloth, and subsequently pelleted by centrifugation at 32,000× g for 1 h at 4°C. 4. The pellet was suspended in 30% sucrose (w/v) prepared in 10 mM sodium phosphate buffer, and liquots (~7 mL) of the suspension were layered on a discontinuous gradient of sucrose (i.e., 50%, 39%, and 35%; w/v). 5. The gradient was overloaded with ~2 mL buffer and centrifuged overnight at 90,000× g at 4°C. 6. Bands at the interface of 50–39% and 39–35% were collected separately, diluted with 1 volume of bidistilled water, homogenized, and pelleted by centrifugation at 34,000× g for 1 h at 4°C. 7. Pellets were resuspended in “stored buffer” and kept at −80°C until use (see Note 4). 3.2. Preparation of the AffinityChromatography Column

1. 25 mL Affi-Gel 10 (in isopropanol) is first washed with bidistilled water by quick filtration using a Buchner funnel with filter paper Whatman No 1 (see Note 5). 2. The gel (~25 mL) is treated overnight with ~100 mL of cystamine solution (50 mg/mL) at 4°C (see Note 6). 3. The treated gel is packed in a chromatographic column by gravity and washed with several column volumes of the washing buffer. 4. The gel is resuspended by gentle rotation (~30 min) in 50 mL of Tris-HCl buffer, packed, and washed with at least 10 column volumes of the washing buffer. 5. To check for free sulfhydryls, the following protocol was used: remove ~50 mL-aliquot of column eluent and add to 1 mL of assay solution. Bright yellow color is positive. After wash it is completely clear; take a sample of the gel matrix, remove liquid, and then add 1 mL of the assay solution to gel beds. The solution must turn bright yellow with an absorbance of A412 > 3. 6. Add bromoacetylcholine bromide solution, and gently rotate or shake the column overnight at 4°C (see Note 7). 7. The derivatized Affi-Gel 10 is packed, washed, and then equilibrated with ~15 column volumes of the asolectin-detergent solution for about 15 h at 4°C. Use a peristaltic pump set at 0.2 mL/min.

Interaction of Lipids and Ligands with AChRs Vesicles

3.3. Purification of AChRs by Affinity Chromatography

301

1. 1 mg/mL Torpedo AChR native membranes were solubilized in 1% sodium cholate in VD buffer. Suspension was centrifugated (70,000× g for 1 h at 4°C) to remove the insoluble material (see Note 8). 2. Solubilized AChRs were treated with DFP and then slowly applied (~24 h, at 4°C) to the affinity column prepared previously. Use peristaltic pump set at 0.3 mL/min. 3. The column was washed extensively for about 15 h with ~15 column volumes of the asolectin-detergent solution (see Notes 9 and 10). 4. AChRs were eluted from the column using the asolectindetergent solution (0.2 mg/mL lipid) containing 10 mM CCh. 5. Peak protein fractions (A280 × 0.6) were pooled and the lipid– protein content adjusted to the desired molar ratio.

3.4. Lipid Reconstitution of Purified AChRs

1. Pooled fractions were dialyzed against 2 L of VD buffer during four days, with buffer change once every day. During this procedure, AChRs are reconstituted into membranes containing asolectin-lipid, and CCh is removed (see Note 10). 2. Reconstituted AChR membranes were aliquoted (0.25 mg per tube) and stored at −80ºC (see Notes 4 and 11).

3.5. Protein Determination of AChR Membranes

1. Total protein concentration of native membrane preparations was determined by the method of Lowry et al. (23), with BSA as standard. 2. Alternatively, the BCA protein assay can be used with BSA as standard (e.g., see ref. 24).

3.6. Determination of AChR Membrane Specific Activity 3.6.1.The Specific Activity of AChR Membranes was Determined by the Method of Schmidt and Raftery (25)

3.6.2. Alternatively, a fluorescence titration method can also be used (e.g., (26))(this is not a heading)

1. A fixed concentration of AChR membranes (0.3 mg/mL) was incubated with increasing concentrations of [3H]-abungarotoxin (i.e., 1–10 nM) in 10 mM sodium phosphate, pH 7.4, for 1 h at room temperature (RT) (see Note 12). 2. AChR-bound [3H]-a-bungarotoxin was separated from the free radioligand by ion exchange using DE81 filter discs (see Note 13). 3. Disc radioactivity corresponding to AChR-bound [3H]-abungarotoxin was determined by scintillation counting using a Beckman counter (see Note 14). 1. The AChR membrane suspension (0.3 mg/mL) is incubated with the fluorescent agonist dansyltrimethylamine (6.6 µM, prepared in methanol) in the presence of 100 µM phencyclidine (a non-competitive antagonist that blocks the unspecific interaction of dansyltrimethylamine with the AChR ion channel).

302

Arias

2. Suberyldicholine (a more potent agonist compared with dansyltrimethylamine) is then titrated to the suspension until dansyltrimethylamine fluorescence (excitation wavelength = 295 nm; emission wavelength = 546 nm, a 530-nm cutoff filter is used to reduce the stray-light effects) is minimal and remains constant at higher concentrations of suberyldicholine. 3. The concentration of AChR binding sites is determined from the point of inflection at constant fluorescence and by considering the protein concentration of the membrane preparation (see Note 14). 3.7. Synthesis of Spin-labeled Molecules 3.7.1. Spin-labeled Bisphosphatidylglycerol (Spin-labeled Cardiolipin), 14-bisPGSL, was Synthetized According to the Method Described in Keana et al. (27)

1. 14-PGSL (~50 mg) was first dissolved in 40 mL of solvent system A, and the solution cooled to 0°C in an ice bath. 2. Cold 0.1 N HCl (4 mL) was added and the mixture vortexed for 2 min and then centrifuged for 5 min. The acidic upper layer was discarded and the lower layer was washed twice with 4 mL of cold solvent system B. 3. The organic solution was stirred with ethereal diazomethane for 30 min at 0°C. The excess of diazomethane was removed with a stream of N2, and the solvent rotary evaporated. 4. The crude methyl phosphoester derivative was purified by preparative TLC using solvent system C, previously deoxygenated with dry N2 gas. 5. The purified compound and PA (~33 mg) were dissolved in 1 mL of dry pyridine, and 0.5 mL of the phosphorylation promoting compound 2,4,8-triisopropylbenzenesulfonyl chloride was added, the mixture was stirred for 8 h at 25°C, and then cooled to 0°C. 6. The excess of 2,4,8-triisopropylbenzenesulfonyl chloride was hydrolyzed with 1 mL water, the solvent was evaporated under reduced pressure, the product was dissolved in 80 mL of system A, and then washed with 20 mL of cold 0.1 N HCl in system B. 7. The organic layer containing the methylated derivative was concentrated and then treated with an excess of ethereal diazomethane for 30 min at 0°C. 8. The dimethyl ester product was purified by preparative TLC using solvent system C, dissolved in 1 mL of dry methyl ethyl ketone, and subsequently didemethylated with NaI (~6 mg) by reflux for 30 min, and then cooled to 25°C. 9. The mixture was cooled to 0°C, centrifuged for 5 min, the supernatant was removed, and the precipitate washed three times with 0.3 mL of cold methyl ethyl ketone to give 14-bisPGSL.

Interaction of Lipids and Ligands with AChRs Vesicles 3.7.2. Spin-labeled Gangliosides Including 14-GD1bSL, 14-GM1SL, 14-GM2SL, and 14-GM3SL, were Prepared According to Schwarzmann and Sandhoff (28)

303

1. Each ganglioside (50 mmol) is suspended in freshly prepared 0.8 M KOH solution and stirred at 100°C for 20 h. 2. The solution is cooled to 20°C, neutralized by addition of acetic acid, and the excess of methanol removed by a stream of N2. 3. The residue is suspended in water (3 mL), dialyzed at 20°C for 6 h against 10 L of distilled water (two changes), and then purified by silica gel column chromatography using the solvent system D. All ninhydrin-positive fractions (i.e., deacylated ganglioside) were pooled and lyophilized from water following removal of the solvents. 4. The deacylated ganglioside (30 mmol) in 0.1 M sodium bicarbonate (3 mL) is mixed with diethyl ether (3 mL) and cooled in a freezer until the aqueous phase turns solid. 5. Then, 9-fluorenylmethyl chloroformate (N-Fmoc Cl) solution is added, stirred vigorously at ~7°C (cold room) for 24 h, and the mixture treated with acetic anhydride (added in drops of 10 mL) at 20°C for 3 h, until pH 5.5–6 is attained. 6. Following evaporation of ether and most of the water, the salts were removed by dialysis as described in step 3, the product is freeze-dried, and finally purified by silica gel column chromatography using solvent system E. 7. The product, N-Fmoc-lysoganglioside (20 mmol), is treated with ammonia, frozen under liquid N2, kept at 20°C for ~3 h, frozen again, and finally ammonia is evaporated at RT. 8. The residue is purified by silica gel column chromatography using solvent system F. All ninhydrin-positive fractions (i.e., lysoganglioside) were pooled and freeze-dried after solvent evaporation. 9. A solution of lysoganglioside (5 mmol) in DMF (0.4 mL) and diisopropylethylamine (20 mL) is mixed with N-succinimidyl ester of 14-SAL (10 mmol) (see step 10. for its preparation), the mixture kept in the dark for 2 days under argon at 30°C, and finally purified by silica gel column chromatography using solvent system E. Spin-labeled gangliosides can be stored as powder at −20°C until use. 10. 14-SASL (50 mmol) is dissolved in dry ethyl acetate (1 mL) and N-hydroxysuccinimide (50 mmol) is added. After dissolution, N,N¢-dicyclohexylcarbodiimide (50 mmol) is added and the mixture stirred overnight under argon at 25°C. After centrifugation, the supernatant is dried in a stream of N2 and the residue (N-succinimidyl ester of 14-SASL) is further used in step 9. N-Succinimidyl ester of 14-SASL can be stored at −20°C under argon until use.

304

Arias

3.7.3. Spin-labeled Local Anesthetics (LASLs) were Synthesized According to Hideg et al. (29)

1. Thionyl chloride was added dropwise to 1-oxyl-2,2,5,5tetramethylpyrroline-3-carboxylic acid solution at 10°C, and stirred for 30 min at RT 2. Benzocaine (3.3 g) was added to the mixture, stirred for 1 h at RT, and then poured into 5% HCl (100 mL) at 0°C. 3. The organic layer was separated, washed with 10% K2CO3 (20 mL), dried over anhydrous Na2SO4, and evaporated, giving ethyl 4-(1-oxyl-2,2,5,5-tetramethylpyrroline-3-carbonyl) aminobenzoate (i.e., spin-labeled benzocaine). 4. Spin-labeled benzocaine (6.6 g) in ethanol (50 mL) was added to 1 N NaOH (40 mL), incubated overnight at RT, acidified to pH 4 with diluted HCl at 0°C, and the precipitate recrystallized from chloroform. 5. The product (6.1 g) and triethylamine (2 g) were dissolved in dry ethyl acetetate (50 mL), and ethyl chloroformate solution was added dropwise at 0°C. The mixture was stirred at RT for 3 h, then filtered, and the filtrate evaporated to dryness and finally crystallized from ether. 6. This product (3.8 g) was dissolved in DMF solution (40 mL), and incubated overnight with 2-diethylaminoethanethiol hydrochloride (1.7 g). After adding ether, the hydrochloride salt of spin-labeled thioprocaine was crystallized. 7. In addition, the DMF solution prepared in step 6 was heated for 1 h with 2-diethylaminoethylamine, treated as in step 8, finally producing spin-labeled procainamide after crystallization with chloroform/hexane. 8. The anhydride ester solution was mixed with the procaine solution, heated at 80°C for 1 h, and evaporated to dryness. The residue dissolved in 50 mL chloroform and treated as in step 3 gives 2-diethylaminoethyl 4-(1-oxyl-2,2,5,5-tetramethylpirroline-3-carbonyl)aminobenzoate (i.e., spin-labeled procaine).

3.7.4. Spin-labeled Phosphatidylinositol, 14-PISL, was Synthetized According to the Method Described in Mantipragada et al. (30)

1. Yeast PI suspension was stirred at RT for 24 h with continuous bubbling of N2 gas, washed with system G, and further purified by precipitation from cold (−20°C) acetone. 2. The pentaacetyl-PI so obtained was dispersed in sodium acetate buffer, and porcine pancreatic phospholipase A2 was added in 1 mg aliquots every 30 min over a period of ~4 h, until the desired pentacetyllyso-PI was obtained. 3. The lysolipid was extracted into dichloromethane:methanol (2:1, v/v) and further purified by silica gel column chromatography using solvent system H as eluent. 4. The pentacetyllyso-PI was acylated with the anhydride of 14-SASL, using 4-pirrolydinopyridine as catalyst, and the

Interaction of Lipids and Ligands with AChRs Vesicles

305

product was purified by preparative TLC using solvent system H. 5. The 14-pentaacetyl-PISL was deacetylated by hydrazinolysis, where 10 mL of hydrazine hydrate was added to the spinlabeled lipid solution in 0.5 mL of methanol:water (3:1, v/v) and incubated at 60°C for 6 h with stirring and then cooled on ice. 6. The precipitate formed was dissolved in dichloromethane, washed, and further purified by chromatography on silica gel column using solvent system H. 7. TLC results indicate that 14-PISL migrated as a single spot slightly ahead of authentic PI as is common for spin-labeled phospholipids, and quenched the fluorescence of the TLCembedded fluorescent indicator, indicating the presence of the spin-label group. 3.8. Electron Paramagnetic Resonance (EPR) Spectroscopy Experiments

1. AChR vesicles were centrifuged for 15 min at 20,000× g rpm, and the pellet (~0.6 mg of protein) was resuspended in 1 mL final volume of 10 mM sodium phosphate buffer, pH 7.4 (LASL experiments; see ref. 31), or alternatively resuspended in 10 mM HEPES-HCl, pH 8.0 (spin-labeled lipid experiments; see ref. 30), giving a final concentration of ~0.5 mM AChRs. 2. The AChR membrane suspension was labeled with 10 mg CSL or 20 mg of each spin label (1% and 2% of the total volume, respectively), and allowed to interact at RT for 30–45 min (see Note 15). 3. The labeled vesicles were centrifuged at 90,000× g for 45 min. When necessary the membranes were washed again in buffer to remove the unincorporated spin label. 4. The final pellet was transferred into EPR capillaries (i.d. = 1 mm) and concentrated in a bench-top centrifuge. Even if the spin label incorporated completely, the spin label/lipid+cholesterol ratio did not exceed 3 mol%. 5. To minimize the dielectric loss (i.e., aqueous signal), samples were trimmed to a height of 10 mm by carefully removing the excess supernatant. 6. For the competition experiments with local anesthetics, the resuspended membrane pellet (0.35 mg of protein in 1 mL) was preincubated for 30 min with benzocaine, procaine, or tetracaine, before adding the spin-labeled molecules. Membranes were then collected by centrifugation, without washing. The estimated LASL/AChR ratio in the membrane was ~15 mol/mol, by assuming a partition coefficient of 10 for the LASLs.

306

Arias

3.9. EPR Recordings

1. EPR spectra were recorded at 9 GHz on a Varian E-12 Century Line spectrometer equipped with a nitrogen gasflow temperature regulation system (30, 31). Temperatures were measured with a thermocouple placed just above the cavity adjacent to a sample capillary. 2. EPR spectra were collected on an IBM personal computer with a Labmaster interface (12 bit A/D resolution) using a software written by Dr. M. D. King (Max-Planck-Institut für Biophysikalische Chemie, Götingen, Germany). 3. Spectral subtractions were performed as described previously (32–34) by using extracted lipids from Torpedo AChR membranes and sonicated dimyristoylphosphatidylcholine vesicles in the gel phase for reference spectra (see Note 16).

3.10. Assignment of Spectral Components

1. The association of the ligand or lipid molecule to the lipid– protein interface can be modeled as an exchange equilibrium between spin-labeled hydrophobic molecules (SL) (e.g., spinlabeled lipids or LASLs) and native lipids (L), competing for sites at the hydrophobic surface of the receptor protein (R) (9, 32–34):

R – L + SL « R – SL +L

(1)

Since the motion of the SL molecules becomes slower when they interact with the protein, the signal provided by the SL molecules at the lipid–protein interface will be distinguishable from the signal in the bulk lipid membrane. In practice, both signals can be separated by spectral subtraction. This consists in subtracting the signal provided by the SL molecules in a membrane suspension of previously extracted lipids (the membrane mobile component) of the tissue or cell under study from the total signal (both protein perturbed and membrane mobile components) of the same spin-label in either the native or reconstituted membrane. 3.11. Spectral Subtraction

1. As an example of spectral subtraction, Fig. 3a shows a typical EPR spectrum of the spin-labeled analog of procaine thioster (compound III) (see its molecular structure in the upper part of Fig. 4) in native membranes from T. marmorata (31). The spectrum is dominated by the extremely narrow peaks from the proportion of the spin-labeled molecules remaining in the aqueous phase. The contribution of this aqueous spectral component to the integrated intensity is small: from quantitative double integration, a fraction of 8% was obtained for compound III (Fig. 3a), and the fraction varied between 8% and 32% for the other LASLs. 2. By comparison of the EPR spectra obtained in the native membrane and in the aqueous dispersions of the extracted

Interaction of Lipids and Ligands with AChRs Vesicles

307

Fig. 3. Spectral subtractions with a typical EPR spectrum of spin-labeled procaine thioester(modified from ref. 31). (a) Torpedo AChR membranes; (b) lipid-associated component; (c) protein-associated component. Spectra (c) and (d) were obtained by subtracting the aqueous signal and resolving the two membrane-associated components by inter-subtractions between these difference spectra. Spectrum (d) is shown together with a simulated slow-motion line shape (dotted line). Total scan width = 100 G; T = 22°C

lipids, the presence of a third spectral component was evident in the outer wings of the membrane spectra. The two membrane-associated spectral components could be resolved by subsequent pairwise inter-subtractions with membrane spectra containing different proportions of these two components, as illustrated in Fig. 3. The sharp spikes in spectra C

308

Arias

Fig. 4. Location of 14-PISL (PtdIns) (30) and of procaine-SL [chemical structure in upper panel; (31)] at the AChR–lipid interface is indicated schematically (modified from ref. 34). Electron microscopy profile of the Torpedo AChR viewed from within the membrane plane (42). Vertical arrows indicate the motionally restricted, protein-interacting spin-label component in each spectrum. T = 22°C. Total scan width = 10 mT

and D arise from slight mismatches in the subtraction of the aqueous component and do not contribute appreciably to the total integrated spectral intensity. The two spectral components, which were observed in both native membranes and dispersions of the extracted lipids, were assigned to LASLs, partitioning between the aqueous and fluid lipid phases (spectra B and C in Fig. 3). 3. The third component was observed only in the native membranes and therefore, was assigned to anesthetic molecules undergoing restricted motion at the lipid–protein interface (spectrum D in Fig. 3). This assignment is consistent with previous EPR results obtained with spin-labeled fatty acids, phospholipids, and steroids, which have also been interpreted in terms of populations in the fluid bilayer and at the interface of intrinsic membrane proteins (30) (9, 32–34). 4. The protein-associated component was analyzed quantitatively by a second spectral subtraction, and as expected, the quality of these EPR difference spectra depended critically on

Interaction of Lipids and Ligands with AChRs Vesicles

309

the purity of the difference spectrum from the first subtraction used to remove the aqueous component. 5. The protein-associated component displayed a spectral anisotropy and line shape typical for slow motion (2 Amax = 63.7 G at 22°C) and could be simulated with the modified Bloch equations (dotted line in Fig. 3d). Considering the interaction of spin-labeled molecules with the lipid–protein interface of the AChR, an appreciable f value for spin-labeled molecules including phospholipids (Table 1), fatty acids (Table 2), steroids (Table 3), and local anesthetics

Table 1 Association of spin-labeled phospholipids to the hydrophobic surface of the AChR Spin-labeled phospholipid

DDG SL,d kJ/mol

AChR Membrane

Temperature °C fa

Kr/Kr

4-PCSL

Native

4

0.32 ± 0.01

1.0

0

(48)

14-PCSL

Native Native Reconstituted Reconstituted Reconstituted

22 0 0 0 0

0.12 0.33 0.32 ± 0.01 0.50 0.55–0.56

1.0 1.0 1.0 (1.1)c 1.0

0 0 0 0 0

(30) (49) (39) (39) (41)

16-PCSL

Native Reconstituted

−4 0

< 0.10 0.34

1.0 (1.0)c

0 0

(50) (38)

14-PASL

Reconstituted

0

0.35 ± 0.01

1.1

16-PASL

Reconstituted

0

0

(2.7)

14-PESL 16-PESL

Native Reconstituted

22 0

0.06 0.

0.5 (1.1)c

+1.8 –

(30) (38)

14-PSSL

Native Reconstituted

22 0

0.26 0.51 ± 0.01

2.7 2.2

−2.4 −1.8

(30) (39)

16-PSSL

Reconstituted

0

0.

(0.7)c

+0.8

(38)

14-PGSL

Native

22

0.18

1.7

−1.3

(30)

14-PISL

Native

22

0.38

4.7

−3.8

(30)

14-bisPGSL

Native

22

0.40

5.1

−4.0

(30)

PCSL b

−0.3

Reference

(39) (38)

c

The fraction of protein-associated component was quantitated by spectral subtraction as described in Subheading 3.11 The association lipid constants relative to spin-labeled phosphatidylcholine were calculated according to Eq. (4). For this purpose, we used the corresponding f and fPCSL (the protein-associated component for spin-labeled phosphatidylcholine) values for each particular experimental condition (e.g., temperature and AChR membrane preparation) c Association constants obtained by Brotherus-type analysis d The differential free energy change of association values were calculated according to Eq. (5) a

b

310

Arias

Table 2 Association of spin-labeled fatty acids to the hydrophobic surface of the AChR DDG SL,d kJ/mol

Reference

−0.3

[48]

Spin-labeled fatty acid AChR membrane

Temperature °C f a

4-SASL

Native

10

0.35 ± 0.10 1.14

6-SASL

Reconstituted

25

0.45

[51]

9-SASL

Reconstituted

20

0.43

[51]

12-SASL

Reconstituted

15

0.38

[51]

14-SASL

Reconstituted

−10 −5 0 5

0.49 0.48 0.45 0.43

[51]

Native

0

0.44

1.60

Native

22

0.39

4.90

Reconstituted

0

0.54

1.17 (3.0)

Reconstituted

0

Native Reconstituted

16-SASL

Kr/K

PCSL b r

−1.1

[49]

−3.9

[30]

−0.4 (−2.5)

[41]

0.49 ± 0.02 2.14

−1.7

[39]

−4

0.35

4.85

−3.5

[50]

0

0.60

2.91 (4.1)c

−2.4 (−3.2)e

[38]

c

e

The fraction of protein-associated component was quantitated by spectral subtraction as described in Subheading 3.11 The association lipid constant relative to spin-labeled phosphatidylcholine were calculated according to Eq. (4). For this purpose, we used the corresponding f and fPCSL (the protein-associated component for spin-labeled phosphatidylcholine) values for each particular experimental condition (e.g., temperature and AChR membrane preparation) shown in Table 1 c Association constants obtained by Brotherus-type analysis d The differential free energy change of association values were calculated according to Eq. (5) e Values in parentheses were obtained taking into account the association constants relative to DOPC a

b

(Table 4), was observed. Figure 4 shows the interaction of 14-PISL (30) and procaine-SL (31) at the lipid–protein interface of the Torpedo AChR and their respective EPR spectra. Interestingly, the f values for a series of spin-labeled gangliosides, including 14-GD1bSL, 14-GM1SL, 14-GM2SL, and 14-GM3SL for 14-PESL (30) and CSL (35) were similar to that for 14-PCSL. These results indicate that there is no selectivity for mono- and di-sialogangliosides, for certain phospholipids such as PE and PC, and for cholesterol, for the AChR-lipid interface. Depending on the used AChR membrane preparation, the temperature and pH at which the EPR measurement was performed, or the lipid/protein molar ratio of the reconstituted system, a broad range of f values is obtained. For instance, the f values for the methyl-doxylamine intracaine derivative (C6SL)

Interaction of Lipids and Ligands with AChRs Vesicles

311

Table 3 Association of spin-labeled steroids to the hydrophobic surface of the AChR Spinlabeled steroid

AChR membrane

Temperature °C f a

ASL

Native Native Reconstituted Reconstituted

22 15 0 0

CSL

Native

22

0.43 ± 0.03 0.38 ± 0.01 0.59 0.58 ~0.16

Kr/Kr

PCSL b

4.27 1.30 2.79 (4.3)c 1.38 (3.5)c ~1.08

DDG SL,d kJ/mol

Reference

−3.6 −0.6 −2.3 (−3.3)e −0.7 (−2.8)e

(52) (48) (38) (41)

~ −0.2

(35)

The fraction of protein-associated component was quantitated by spectral subtraction as described in Subheading 3.11 The association lipid constants relative to spin-labeled phosphatidylcholine were calculated according to Eq. (4). For this purpose, we used the corresponding f and fPCSL (the protein-associated component for spin-labeled phosphatidylcholine) values for each particular experimental condition (e.g., temperature and nAChR membrane preparation) shown in Table 1 c Association constants obtained by Brotherus-type analysis d The differential free energy change of association values were calculated according to Eq. (5). e Values in parentheses were obtained taking into account the association constants relative to DOPC a

b

depended on the lipid:protein ratio (36). In fact, the f value changed from ~0.77 to ~0.14 when DOPC:AChR molar ratios of about 150–560 were, respectively, used (see Table 4). The same basic result was obtained by using the quaternary derivative C6SL-MeI (37). In this case the f values decreased from 0.67 to 0.23 when the DOPC:AChR molar ratio augmented from 97 to 284 (see Table 4). Local anesthetics including procaine, tetracaine, and benzocaine inhibit the interaction of 14-PISL to the lipid–protein interface of the AChR (see Table 5). A similar effect was observed for several general anesthetics including 1-hexanol, diethyl ether, and urethane, on the interaction of spin-labeled phospholipids with the AChR-lipid interface (see Table 5). Interestingly, the inhibitory action of these anesthetics on the interaction of spinlabeled fatty acids (i.e., 14-SASL and HGSL, with the nitroxy group located at position 14 or at the headgroup, respectively) and of spin-labeled cholestane (i.e., CSL), a cholesterol analog, is not appreciated (see Table 5). These results indicate that there are two different binding domains for lipids, one for phospholipids that are susceptible to the action of anesthetics and another for fatty acids and cholesterol that are not susceptible to the action of anesthetics. This result supports a previous model where the AChR contains annular (for phospholipids) and nonannular (for fatty acids and cholesterol; see Fig. 2) lipid domains (16, 17).

312

Arias

Table 4 Association of spin-labeled local anesthetics to the lipid–protein interface of the AChR Affinity High

AChR membrane

Native C6SL Reconstituted C6SL-MeI Native IX Benzocaine-SL (I) Thioprocaine-SL (III)

Intermediate Native

Low

Spin-labeled local anesthetica f b ~0.60 0.52–0.67 0.36 0.35 0.34

Kr/Kr

DDG SL,d kJ/mol

− − 2.4 2.3 2.2

− − −2.1 −2.0 −1.9

PCSL c

V VI+/Me X Procainamide-SL (IV) VIII VI

0.31 0.31 0.31 0.29

1.9 1.9 1.9 1.7

−1.6 −1.6 −1.6 −1.4

0.28 0.27

1.7 1.6

−1.2 −1.1

Native Procaine-SL (II) Reconstituted C6SL-MeI

0.23 0.23

1.3 −

−0.6 −

Reference (36) (37) (31)

(37)

The molecular structure of LASLs are shown in Horváth et al. (31) and in Arias (19). The protein perturbed component for each LASL was obtained by means of spectral subtractions as described in Subheading 3.11 c The relative association constant of LASL with respect to 14-PCSL (fPCSL = 0.19; (31)) was obtained according to Eq. (4). d The differential association free energy change of each LASL was calculated according to Eq. (5) a

b

3.12. Determination of the Equilibrium Association Constant

1. Considering the exchange equilibrium described in Eq. (1), the equilibrium constant for association (Kr) of SL at firstshell sites, relative to that for lipids in the membrane, is given by ( (32, 33)): Kr = [R – SL] / [L] [R – L][SL]

(2)

where [R–SL] and [R–L] are the moles of labeled and unlabeled lipids (or local anesthetics) that are associated at a particular site on the transmembrane surface of the protein, and [SL] and [L] are similarly defined for the free (non-associated) lipid populations. The average relative association constant, Kr that is measured by EPR at low label concentration is given by (30) (32–34):

Kr = (1/Nb)Sini Ki; (i = 1to m)

(3)

where ni is the number of association sites of type i and Nb is the total number of lipid association sites at the transmembrane surface of the protein.

Interaction of Lipids and Ligands with AChRs Vesicles

313

Table 5 Effect of anesthetics on the association of spin-labeled lipids to the hydrophobic surface of the AChR Spin-labeled lipid

Anesthetic

AChR membrane

fa

Reference

14-PISL

None Tetracaine Procaine Benzocaine

Native

0.38 0.22 0.29 0.30

(30)

14-PCSL

None 1-Hexanol None 1-Hexanol Diethyl ether Urethane Ethanol

Reconstituted

0.56 0.56–0.59 0.33 0.20 0.19 0.24 0.30

(41)

14-PGSL

none 1-Hexanol

Native

0.42 0.33

(49)

14-SASL

None 1-Hexanol None Tetracaine Procaine Benzocaine None Ethanol Isoflurane 1-Hexanol

Native

0.44 0.40 0.39 0.31 0.39 0.37 0.31 0.34–0.40 0.30 0.34–0.36

(49)

HGSL

CSL

None Tetracaine, procaine, and benzocaine

Native

Native

Reconstituted

Native Native

~0.16 ~0.16

(49)

(30)

(53)

(35)

The fraction of protein-associated component was quantitated by spectral subtraction as described in Subheading 3.11 a

3.13. Determination of the Relative Association Constant

1. Since the Ki and thus, the Kr constants in Eq. (20.3) are difficult to calculate experimentally, the association constant for each SL relative to the association constant of PCSL (KrSL/KrPCSL) was instead calculated according to the following equation [30) (9, 32–34): KrSL 1KrPCSL = [(1 – fPCSL)fPCSL]/[(1 – fSL)/fSL]

(4)

In this regard, we compared the fSL value of each spin-labeled lipid or LASL with the value obtained for PCSL (fPCSL) measured in the same experimental conditions. Assuming that the specificity arises from differences in SL affinity, rather than from the stoichiometry of the SL-protein interaction, the

314

Arias

relative association constant can be used as a measure of the SL affinity. The dependence of the f values with the specific conditions of each experiment (see Tables 1–4) can be eliminated using this approach. The relative association constants were summarized in Tables 1–4. 3.14. Determination of the Differential Free Energy of Association

1. The energetics of the ligand selectivity of the lipid–protein interaction can be obtained relative to PCSL. For this purpose, the differential free energy of association of each spin-labeled lipid or LASL with respect to PCSL (DG SL − DG PCSL = DDG SL) was calculated by using the expression [30) (9, 32, 33): DDG SL = – RT ln(KrSLKrPCSL)

(5)

where T is the absolute temperature used in the EPR experiment and R is the gas constant (8.314 J/mol/K). The calculated values were also included in Tables 1–4. Considering that the DDG SL term shown in Tables 1–4 is a direct indication of the strength of the interaction, all SLs, except n-PESL and n-PCSL (Table 1), and CSL (Table 3), present a higher affinity for the hydrophobic surface of the AChR than for the lipid membrane (i.e., negative DDG SL values). Interestingly, the selectivity for fatty acids is pH dependent (38, 39), whereas the selectivity for other negatively charged lipids such as 14-PSSL and 14-PASL, and probably for 14-PISL as well, are not, indicating that selectivity is not entirely due to charge (39). Regarding the selectivity for local anesthetics, the relative constants summarized in Table 4 allows us to discriminate among local anesthetics interacting with high (between −2.1 and −1.9 kJ/mol), intermediate (between −1.6 and −1.1 kJ/ mol), and low (approximately −0.6 kJ/mol) specificity with the lipid–protein interface of the Torpedo AChR (31). 3.15. Lipid–Protein Stoichiometry

Since PC is the majority lipid in the Torpedo AChR membrane (40), it may be used to estimate the number of lipid sites associated with the protein [see Nb in Eq. (3)) (30) (32, 33). n-PCSL is associated to a level of 10–15% with the integral membrane protein (see Table 1). Therefore, the total number of association sites on the protein corresponds to 12% of the total lipids, which is consistent with 42 ± 7 sites (27 phospholipid and 15 cholesterol molecules) per AChR molecule. The obtained value is compatible with those determined previously [40 ± 7 (39); 56 ± 2 (41)) by using the Brotherus analysis. Interestingly, this stoichiometric ratio corresponds very well with the number of lipids that can be accommodated around the perimeter of the protein determined at ~4 Å resolution (see Fig. 1; ref. 42),

Interaction of Lipids and Ligands with AChRs Vesicles

315

as well as with the minimum number of lipids that are required to support the functional properties of the AChR (43). 3.16. Conclusions

EPR spectroscopy is a powerful method that can be used to determine the interaction of spin labels with the lipid–protein interface of the AChR. The conclusions that different laboratories arrived by using this technique is that the lipid–protein interface of the AChR is a dynamic domain where specific interactions with lipids and anesthetics occur. In addition, an important correlation between the structure and function of the first lipid shell surrounding the AChR protein can be envisioned.

4. Notes 1. Add DTT immediately prior to its use. 2. Electric organs can be stored at −80°C or in liquid N2 for years, without decreasing its AChR content. 3. The homogenization buffer can be prepared the day before the homogenization protocol and kept at 4°C until its use, whereas PMSF must be prepared fresh every time, dissolving it in ~200 mL absolute ethanol before it is added to the buffer. The whole homogenization process must be performed on ice. 4. Torpedo AChR membranes stored at −80°C can be maintained for long time with only ~1% degradation per year. 5. Do not let the Affi-gel 10 to dry during the filtration process. 6. Add cystamine immediately prior to its use and gently rotate to maintain the gel resuspended. 7. Affinity-gel can be maintained in buffer at 4°C for approximately one month. 8. Sodium cholate, but not other detergents such as Triton X-100, octyl glucoside, or Tween 20, stabilizes the AChR in the resting state (44). 9. This extensive wash ensures complete exchange of the endogenous lipids for the asolectin-lipid mixture (15, 45). 10. AChRs reconstituted in lipid vesicles composed only of DOPC are in the desensitized state, whereas liposomes formed by DOPC, dioleoylphosphatidic acid, and cholesterol retain the ion gating activity of AChRs (46). 11. The purity of purified AChRs can be assessed by SDS-PAGE (e.g., see ref. 45). The purified AChR usually comprises more than 90% of protein.

316

Arias

12. [3H]a-Bungarotoxin can be sealed under vacuum or an inert gas (e.g., N2) to maintain activity for longer time. 13. DE81 disks can be washed with 10 mM sodium phosphate buffer, pH 7.4, before its use, to obtain the best performance. 14. Specific activities of 1.2–2.4 nmol binding sites/mg protein were obtained for native AChR membranes. These values correspond to 0.6–1.2 AChRs/mg protein, respectively, considering that the Torpedo AChR has two binding sites. 15. In order to avoid spin–spin broadening of the spectra permitting a more reliable quantitation, it is essential to use rather low spin-label concentrations (30, 38). 16. For the preparation of liposomes, total lipids can be first extracted from Torpedo AChR membranes by the method of Folch et al. (47), the organic solvent evaporated under N2, and the lipids resuspended in buffer by sonication.

Acknowledgments This research was supported by grants from the Science Foundation Arizona and Stardust Foundation and from the College of Pharmacy, Midwestern University. The author thanks the comments by Dr. Blanton (Texas Tech University Health Sciences Center, Lubbok, TX, USA) on AChR purification procedures. References 1. Arias HR (2006) Ligand-gated ion channel receptor superfamilies. In: Arias HR (ed) Biological and biophysical aspects of ligandgated ion channel receptor superfamilies. Research Signpost, Kerala, India, pp 1–25 2. Ortells MO, Lunt GG (1995) Evolutionary history of the ligand-gated ion-channel superfamily of receptors. Trends Neurosci 18:121–127 3. Changeux J-P, Taly A (2008) Nicotinic receptors, allosteric proteins and medicine. Trends Mol Med 14:93–102 4. Albuquerque EX, Pereira EF, Alkondon M, Rogers SW (2009) Mammalian nicotinic acetylcholine receptors: from structure to function. Physiol Rev 89:73–120 5. Arias HR, Richards V, Ng D, Ghafoori ME, Le V, Mousa S (2009) Role of non-neuronal nicotinic acetylcholine receptors in angiogenesis. Int J Biochem Cell Biol 41:1441–1451 6. Arias HR (2001) Thermodynamics of nicotinic receptor interactions. In: Raffa RB (ed)

7. 8. 9.

10.

11.

Drug-receptor thermodynamics: introduction and applications. Wiley, USA, pp 293–358 Miyazawa A, Fujiyoshi Y, Unwin N (2003) Structure and gating mechanism of the acetylcholine receptor pore. Nature 423:949–955 Unwin N (2005) Refined structure of the nicotinic acetylcholine receptor at 4 Å resolution. J Mol Biol 346:967–989 Arias HR, Bouzat CB (2006) Modulation of nicotinic acetylcholine receptors by noncompetitive antagonists. In: Arias HR (ed) Biological and biophysical aspects of ligandgated ion channel receptor superfamilies. Research Signpost, Kerala, India, pp 61–107 Arias HR, Bhumireddy P, Bouzat C (2006) Molecular mechanisms and binding site locations for noncompetitive antagonists of nicotinic acetylcholine receptors. Int J Biochem Cell Biol 38:1254–1276 Brejc K, van Dijk WJ, Klaassen RV, Schuurmans M, van der Oost J, Smit AB, Sixma TS (2001)

Interaction of Lipids and Ligands with AChRs Vesicles

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

Crystal structure of an ACh-binding protein reveals the ligand-binding domain of nicotinic receptors. Nature 411:269–276 Swope SL, Moss SJ, Blackstone CD, Huganir RL (1992) Phosphorylation of ligand-gated ion channels: a possible mode of synaptic plasticity. FASEB J 6:2514–2523 Blanton MP, Cohen JB (1994) Identifying the lipid–protein interface of the Torpedo nicotinic acetylcholine receptor: secondary structure implications. Biochemistry 33:2859–2872 Hamouda AK, Sanghvi M, Chiara DC, Cohen JB, Blanton MP (2007) Identifying the lipid– protein interface of the a4b2 neuronal nicotinic acetylcholine receptor: hydrophobic photolabeling studies with 3-(trifluoromethyl)3-(m-[125I]iodophenyl)diazirine. Biochemistry 46:13837–13846 Hamouda AK, Chiara DC, Sauls D, Cohen JB, Blanton MP (2006) Cholesterol interacts with transmembrane ahelices M1, M3, and M4 of the Torpedo nicotinic acetylcholine receptor: Photolabeling studies using [3H] azidocholesterol. Biochemistry 45:976–986 Jones OT, McNamee MG (1988) Annular and nonannular binding sites for cholesterol associated with the nicotinic acetylcholine receptor. Biochemistry 27:2364–2374 Brannigan G, Hénin J, Law R, Eckenhoff R, Klein ML (2007) Embedded cholesterol in the nicotinic acetylcholine receptor. Proc Natl Acad Sci USA 105:14418–14423 Arias HR (1998) Binding sites for exogenous and endogenous non-competitive inhibitors of the nicotinic acetylcholine receptor. Biochim Biophys Acta 1376:173–220 Arias HR (1999) Role of local anesthetics on both cholinergic and serotonergic ionotropic receptors. Neurosci Biobehav Rev 23: 817–843 Arias HR, Blanton MP (2002) Molecular and physicochemical aspects of local anesthetics acting on nicotinic acetylcholine receptorcontaining membranes. Mini Rev Med Chem 2:385–410 Arias HR, Alonso-Romanowski S, Disalvo EA, Barrantes FJ (1994) Interaction of merocyanine 540 with nicotinic acetylcholine receptor membranes from Discopyge tschudii electric organ. Biochim Biophys Acta 1190:393–401 Pedersen SE, Dreyer EB, Cohen JB (1986) Location of ligand-binding sites on the nicotinic acetylcholine receptor alpha-subunit. J Biol Chem 261:13735–13743 Lowry OH, Rosebrough NJ, Farr L, Randall RJ (1951) Protein measurement with the

24.

25. 26.

27.

28.

29. 30.

31.

32. 33. 34.

35.

317

Folin phenol reagent. J Biol Chem 193: 265–275 Arias HR, Bhumireddy P, Spitzmaul G, Trudell JR, Bouzat C (2006) Molecular mechanisms and binding site location for the noncompetitive antagonist crystal violet on nicotinic acetylcholine receptors. Biochemistry 45:2014–2026 Schmidt J, Raftery MA (1973) A simple assay for the study of solubilized acetylcholine receptors. Anal Biochem 52:349–354 Sanghvi M, Hamouda AK, Jozwiak K, Trudell JR, Blanton MP, Arias HR (2008) Identifying the binding site(s) for antidepressants on the Torpedo nicotinic acetylcholine receptor: [3H]2-Azidoimipramine photolabeling and molecular dynamics studies. Biochem Biophys Acta 1778:2690–2699 Keana JFW, Shimizu M, Jernstedt KK (1986) A short, flexible route to symmetrically and unsymetrically substituted diphosphatidylglycerol (cardiolipins). J Org Chem 51:2297–2299 Schwarzmann G, Sandhoff K (1987) Lysogangliosides: synthesis and use in preparing labeled gangliosides. Methods Enzymol 138:319–341 Hideg K, Lex L, Hankovszky HO, Tigyi J (1979) Synthesis of spin-labelled procaine and its derivatives. Synth Commun 9:781–788 Mantipragada SB, Horváth LI, Arias HR, Schwarzmann G, Sandhoff K, Barrantes FJ, Marsh D (2003) Lipid–protein interactions and the effect of local anesthetics in acetylcholine receptor-rich membranes from Torpedo marmorata electric organ. Biochemistry 42:9167–9175 Horváth LI, Arias HR, Hankovszky HO, Hideg K, Barrantes FJ, Marsh D (1990) Association of spin-labeled anesthetics at the hydrophobic surface of acetylcholine receptor in native membranes from Torpedo marmorata. Biochemistry 29:8707–8713 Marsh D (2008) Electron spin resonance in membrane research: protein-lipid interactions. Methods 46:83–96 Marsh D (2008) Protein modulation of lipids, and vice-versa, in membranes. Biochim Biophys Acta 1778:1545–1575 Marsh D, Páli T (2004) The protein-lipid interface: perspectives from magnetic resonance and crystal structures. Biochim Biophys Acta 1666:118–141 Arias HR, Sankaram MB, Marsh D, Barrantes FJ (1990) Effect of local anaesthetics on steroidnicotinic acetylcholine receptor interactions in native membranes of Torpedo marmorata electric organ. Biochim Biophys Acta 1027:287–294

318

Arias

36. Earnest JP, Wang HH, McNamee MG (1984) Multiple binding sites for local anesthetics on reconstituted acetylcholine receptor membranes. Biochem Biophys Res Commun 123:862–866 37. Earnest JP, Limbacher HP, McNamee MG, Wang HH (1986) Binding of local anesthetics to reconstituted acetylcholine receptors: the effect of protein surface potential. Biochemistry 25:5809–5818 38. Ellena JF, Blazing MA, McNamee MG (1983) Lipid–protein interactions in reconstituted membranes containing acetylcholine receptor. Biochemistry 22:5523–5535 39. Raines DE, Miller KW (1993) The role of charge in lipid selectivity for the nicotinic acetylcholine receptor. Biophys J 64:632–641 40. Rotstein NP, Arias HR, Barrantes FJ, Aveldaño MI (1987) Composition of lipids in elasmobranch electric organ and acetylcholine receptor membranes. J Neurochem 49:1333–1340 41. Abadji VC, Raines DE, Dalton LA, Miller KW (1994) Lipid–protein interactions and protein dynamics in vesicles containing the nicotinic acetylcholine receptor: a study with ethanol. Biochim Biophys Acta 1194:25–34 42. Unwin N (1995) Acetylcholine receptor channel imaged in the open state. Nature 373:37–43 43. Jones OT, Eubanks JH, Earnest JP, McNamee MG (1988) A minimum number of of lipids are required to support the functional properties of the nicotinic acetylcholine receptor. Biochemistry 27:3733–3742 44. McCarthy MP, Moore MA (1992) Effects of lipids and detergents on the conformation of the nicotinic acetylcholine receptor from Torpedo californica. J Biol Chem 267: 7655–7663 45. Hamouda AK, Chiara DC, Blanton MP, Cohen JB (2008) Probing the structure of

46.

47.

48.

49.

50.

51.

52.

53.

the affinity-purified and lipid-reconstituted Torpedo nicotinic acetylcholine receptor. Biochemistry 47:12787–12794 Fong TM, McNamee MG (1986) Correlation between acetylcholine receptor function and structural properties of membranes. Biochemistry 25:830–840 Folch J, Lees M, Sloane Stanley GH (1957) A simple method for the isolation and purification of total lipids from animal tissues. J Biol Chem 226:497–509 Dreger M, Krauss M, Herrmann A, Hucho F (1997) Interactions of the nicotinic acetylcholine receptor transmembrane segments with the lipid bilayer in native receptor-rich membranes. Biochemistry 36:839–847 Fraser DM, Louro SRW, Horváth LI, Miller KW, Watts A (1990) A study of the effect of general anesthetics on lipid–protein interactions in acetylcholine receptor enriched membranes from Torpedo nobiliana using nitroxide spin-labels. Biochemistry 29:2664–2669 Rousselet A, Devaux PF, Wirtz KW (1979) Free fatty acids and esters can be immobilized by receptor rich membranes from Torpedo marmorata but not phospholipids acyl chains. Biochem Biophys Res Commun 90:871–877 Raines DE, Wu G, Dalton LA, Miller KW (1995) Elestron spin resonance studies of acyl chain motion in reconstituted nictonic acetylcholine receptor membranes. Biophys J 69:498–505 Marsh D, Watts A, Barrantes FJ (1981) Phospholipid chain immobilization and steroid rotational immobilization in acetylcholine receptor-rich membranes from Torpedo marmorata. Biochim Biophys Acta 645:97–101 Seto T, Firestone LL (2000) Effects of normal alcohols and isoflurane on lipid headgroup dynamics in nicotinic acetylcholine receptor-rich lipid vesicles. Biochim Biophys Acta 1509:111–122

Chapter 21 Environmental Scanning Electron Microscope Imaging of Vesicle Systems Yvonne Perrie, Habib Ali, Daniel J. Kirby, Afzal U.R. Mohammed, Sarah E. McNeil, and Anil Vangala Abstract The structural characteristics of liposomes have been widely investigated and there is certainly a strong understanding of their morphological characteristics. Imaging of these systems, using techniques such as freeze-fracturing methods, transmission electron microscopy, and cryo-electron imaging, has allowed us to appreciate their bilayer structures and factors that influence this. However, there are a few methods that study these systems in their natural hydrated state; commonly, the liposomes are visualized after drying, staining and/or fixation of the vesicles. Environmental scanning electron microscopy (ESEM) offers the ability to image a liposome in its hydrated state without the need for prior sample preparation. We were the first to use ESEM to study the liposomes and niosomes, and have been able to dynamically follow the hydration of lipid films and changes in liposome suspensions as water condenses onto, or evaporates from, the sample in real-time. This provides an insight into the resistance of liposomes to coalescence during dehydration, thereby providing an alternative assay for liposome formulation and stability. Key words: Liposomes, Surfactant vesicles, Niosomes, Non-ionic surfactant vesicles, Lipoplexes, ESEM analysis

1. Introduction Since the early images of liposomes e.g.,(1), which demonstrated the concentric bilayers of multilamellar vesicles, there has been a rangeofmethods used to visualize the different structures liposomes may form depending on their preparation, composition, and loaded drug from the “spaghetti and meatballs” (2) freezefracture micrographs of cationic liposome/DNA complexes, the cryo-electron pictures of cationic liposome/DNA complexes (3),

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_21, © Humana Press, a part of Springer Science+Business Media, LLC 2010

319

320

Perrie et al.

and liposomal adjuvant systems (4, 5). These have allowed us to visualize the systems at any instant in time. However, liposomes are dynamic systems, sensitive to many external factors. Yet morphological techniques can also be used for more than visualization of vesicles; the phase behavior and structural changes of these systems under varying conditions can be monitored (6, 7). In addition, data from micrographs have supported other physicochemical studies of pharmaceutical systems with, for example, micrographs of niosomes demonstrating that the enhanced flow and dye retention of cholesterol-rich niosomes (hexadecyl diglycerol ether:Chol:polyoxyethylene 24 cholesteryl ether; 49:49:2) was a result of their spherical nature compared to the poor flow characteristics of C16G2:SolulanC24 (91:9) niosomes which formed polyhedral structures (6). Unfortunately, many of these imaging techniques require manipulation of fixation of the samples prior to analysis, and this can influence the structural attributes of liposomes. For example, liposomes visualized by the scanning electron microscopy (SEM) are stained and the water fixed before imaging, with the disadvantage that images can be poorly representative (8). Alternatively, freeze-fracture can also influence the liposome morphology due to the mechanical stresses encountered during specimen preparation (9). Therefore, the ability to image liposomes in their natural hydrated state without prior preparation has many advantages. Atomic force microscopy (AFM) does allow sample analysis in the wet mode, by the analysis and interpretation of interactions between a sub-microscopic probe and the sample surface in both contact as well as non-contact modes. Unfortunately, the contact mode involves the probe touching the sample surface and can thereby cause physical damage to the sample (or damage to the probe in cases of hard sample surfaces), and running AFM with the non-contact mode is less sensitive. A second method that offers the ability to image wet samples without prior preparation is environmental scanning electron microscopy (ESEM); this method is a particularly useful method for morphological investigations in biological applications (10–13). 1.1. Principles of ESEM

Following the initial observations in the late 1970s (14–16) and the subsequent development of the gaseous detection device (16, 17), ESEM became commercially available towards the end of the 1980s. The main distinguishing feature of ESEM as compared to SEM is the presence of vapor (usually water) in the sample chamber, made possible by a system of differential pumping zones, which maintains the required high vacuum for the electron gun (10−6 torr) whilst allowing for partial vacuum in the sample chamber (10 torr) (18, 19).

ESEM Imaging of Vesicle Systems

321

B & W IN PRINT

Primary electron beam

Detector

Gaseous atom Positive ion Secondary electron

Sample

Fig. 1. Schematic representation of the gaseous amplification process used in ESEM. Primary electron beam generates secondary electrons (black arrows) from the sample, which in turn generate secondary electrons from vapor molecules present in sample chamber, leading to a cascade amplification of the signal before reaching the detector

The presence of water vapor molecules within the sample chamber plays a vital role in ESEM imaging (Fig. 1). The primary electron beam generates secondary electrons from the sample surface, which then encounter the vapor molecules, which in turn become ionized and generate further secondary electrons, which in turn encounter adjacent vapor molecules, and so on, leading to a “cascade” amplification of the signal before reaching the detector (Fig. 1.) (11, 19, 20). Further, the positive ions resulting from the ionization of the water vapor molecules are attracted to the negatively charged surface of the sample, compensating for the negative charge build up, hence precluding the need for conductive coating of the sample (13, 19). Consequently, ESEM has previously been used to investigate several particulate formulations, including liposomes (7, 20), niosomes (21), microspheres (22), thermo-responsive microspheres (23), polymeric surfactant micelles (24), dendrimers (24), and colloidal latex dispersions (25). In addition to these attributes, ESEM also allows for the variation of parameters within the sample chamber, such as temperature, pressure, and gas composition, permitting realtime studies into the effect of environmental changes on the sample (20).

322

Perrie et al.

2. Materials 2.1. Lipids

The lipids used throughout these experiments were dissolved in a chloroform:methanol (9:1 v/v) solution at the desired concentration. These should be stored at −20°C in sealed (preferably brown) glass containers. 1. Phosphatidylcholines: Egg phosphatidylcholine (PC) (grade I) (Lipid Products, Epsom, Surrey, U.K), Dioleoyl-sn-Glycero3-Phosphocholine (DOPC) (Avanti lipids, Alabaster, AL, USA). These are prepared at concentrations of 100 mg/ml. 2. Phosphatidylethanolamines: 1,2-dioleoyl-sn-glycero-3-phosphatidylethanolamine (DOPE) (Avanti lipids, Alabaster, AL, USA). These are generally prepared at concentrations of 10 mg/ml. 3. Cationic lipids: 3b-(N-(N¢,N¢-Dimethylaminoethane)carbamoyl) Cholesterol (DC-Chol) (Sigma-Aldrich, Poole, Dorset, UK). Dimethyldioctadecylammonium (DDA) (both from Avanti Polar lipids, Alabaster, USA). These are generally prepared at concentrations between 10 and 100 mg/ml. 4. Non-ionic lipids: 1-Monopalmitoyl glycerol (C16:0) (MP) (Sigma-Aldrich Company Ltd, Poole, UK). Non-ionic lipids are used to replace phospholipids to produce non-ionic surfactants (niosomes) (see Note 1). 5. Cholesterol (Chol) (Sigma-Aldrich Company Ltd, Poole, UK). This can be prepared at a concentration of 100 mg/ml. 6. Immunomodulators: a,a¢-trehalose 6,6¢-dibehenate (TDB) (Avanti Lipids, Alabaster, AL, USA). We prepare this at 10 mg/ml.

2.2. Drugs/Antigens For Inclusion with the Liposome Systems

A wide range of drugs and antigens can be incorporated within liposomes. In this study, we have considered three types of drugs. 1. Low-solubility drugs: Ibuprofen and propofol (SigmaAldrich Company Ltd, Poole, UK) can be incorporated with the bilayer of liposomes. These are dissolved in a chloroform:methanol (9:1 v/v) solution at 100 mg/ml. 2. Plasmid DNA: Plasmid DNA encoding for the Hepatitis B surface antigen (pRc/CMV HBS) (Aldevron, Fargo, USA) can be used to form lipoplexes with the cationic systems. 3. Antigens: In our studies, the TB antigen (Ag85B-ESAT-6) used was supplied by Peter Andersen et al. Department of Infectious Disease Immunology, Adjuvant Research, Statens Serum Institut, DK-2300 Copenhagen, Denmark. However, there are a range of model antigens, such as OVA that are commonly used.

ESEM Imaging of Vesicle Systems

2.3. Environmental Scanning Electron Microscope

323

In our studies, ESEM analysis was performed using a Philips XL30 ESEM-FEG (Philips Electron Optics (FEI), Eindhoven).

3. Methods 3.1. Preparation of Liposomes with Bilayer Loaded Drugs

MLV are easily prepared by the well-established film technique (i.e. the assembly of phospholipids in closed lipid bilayers with excess water), first observed by Bangham in the 1960s (25). A schematic of this process is shown in Fig. 2. 1. The required lipid solutions, at the desired concentrations, are placed in a 50 ml round-bottom spherical Quick-fit flask. 2. At this stage, the low-solubility drugs (ibuprofen or propofol; 1.25 mg) dissolved in the chloroform/methanol mixture are also added at the required concentrations if we wish to load the drugs within the bilayers. 3. The solvent is removed by rotary evaporation at about 37°C. This yields a thin lipid film on the walls of the flask, which is then flushed with oxygen-free nitrogen in order to ensure complete removal of all solvent traces (see Note 2).

B & W IN PRINT

4. The dry lipid film is hydrated by addition of 2 ml of double distilled water (ddH2O) and agitated vigorously until the thin lipid film is completely dissolved and transformed into a milky suspension. If the liposomes are to be prepared in a buffered solution, replace the 2 ml of ddH20 with the appropriate buffer, such as phosphate buffered saline (PBS; pH 7.4).

Solvent solutions of lipids and drug mixed at the desired concentrations

Aqueous medium

Removal of solvent

Hydrate above Tc

Allow to cool

Fig. 2. Schematic representation of the lipid hydration method where drugs are loaded within the bilayer rather than the aqueous phase

324

Perrie et al.

5. The hydration of the lipid film should be maintained above the gel-liquid crystal transition temperature (Tc) of the phospholipid (>Tc). PC has a Tc of 0°C and DMPC has a Tc of 23°C; however, DSPC has a Tc above 55°C. Therefore, water added to hydrate the lipid film should be pre-warmed to a temperature above 55°C and the liposome solution should be maintained at this temperature during liposome formation (see Note 3). 6. Drug-free liposomes should be prepared as controls by omitting step 2 to allow for comparisons. 3.2. Preparation of DehydrationRehydration Niosomes with Entrapped Antigen

The dehydration-rehydration procedure (26) can be used for the entrapment of antigens within liposomes, and this method was used for the incorporation of antigen (Ag85B-ESAT-6) into niosomes (21) (Fig. 3). 1. 2 ml of multilamellar niosomes are prepared using 16 mmol of MP, 8 mmol of cholesterol, and 4 mmol of DDA in the presence or absence of TDB (0.5 mmol) as described in Item 1 of subheading 2.1 without the addition of low-solubility drug (Step 2). 2. These multilamellar niosomes are disrupted using sonic energy to fracture the large vesicles into smaller structures (<~100 nm). In this instance, a probe sonicator (Soniprep 150) was used to produce SUV. The tip of the sonication probe (diameter ~4 mm) is placed on the surface of the 2 ml MLV mixture for the amount of time required to produce vesicles ~100 nm in size. This time varies depending on the lipid composition (see Note 4). The milky MLV suspension transforms into clear SUV suspension with a slight blue tint. 3. The SUV suspension is then centrifuged at 3,500 g for 10 min to remove any titanium debris released from the probe during sonication. 4. These SUV are mixed with 20 mg of Ag85B-ESAT-6, frozen at −70°C, and freeze-dried overnight.

Mix

Antigen

Freeze-dry

Rehydrate

SUV

Fig. 3. Schematic of the dehydration-rehydration process used to entrap antigens within liposomes or niosomes

ESEM Imaging of Vesicle Systems

325

5. Controlled rehydration (26) of the dried powder leads to the formation of antigen containing multilamellar dehydrationrehydration vesicles (DRV). 3.3. Cationic Lipid-DNA Complexes

To investigate the lipoplexes by ESEM, cationic SUV are prepared and mixed with plasmid DNA as follows. 1. Cationic MLV are prepared from a combination of a cationic lipid (DC-Chol) and either cholesterol or DOPE combined at a 1:1 lipid ratio as outlined in Item 1 of subheading 2.1. 2. These cationic MLV are then sonicated, as outlined in item 2 of subheading 2.1 (see Note 4). 3. Cationic SUV are mixed with plasmid DNA at a charge ratio of 1.5 to produce lipoplexes.

3.4. Environmental SEM Analysis

Both liposomes and dried lipid films can be analyzed using a Philips Electron Optics ESEM.

3.4.1. ESEM Imaging of Lipid Film Hydration

1. Dry lipid films are prepared using the same method used to prepare MLV in which the required lipid solutions, at the desired concentrations (PC:Chol; 4:1 molar ratio), are mixed. However, rather than placing this in a round-bottom flask, the solvent/lipid mixture is placed on the top of a glass cover side and the solvent allowed to evaporate, to produce the required dry lipid film. 2. The ESEM sample holder is loaded with the dry lipid film. 3. Dynamic formation of liposomes is monitored by controlled hydration of dried lipid film chamber under an operating pressure maintained at 1.9 torr and a working temperature of 5°C. An example of the images from ESEM, showing the formation of liposomes via hydration of a lipid film, is shown in Fig. 4.

Fig. 4. Environmental scanning electron micrographs of a dry lipid film containing a mixture of PC and cholesterol (4:1 molar ratio) subjected to controlled hydration in the ESEM sample chamber under an operating pressure maintained at 1.9 torr. The formation of spherical structures is clearly observed from the film in both (a) and (b). Reproduced from [7] with permission from Elsevier Science

326

Perrie et al.

3.4.2. ESEM Imaging of Liposomes

1. The ESEM sample holder is loaded with liposome formulation prepared previously (as described in the previous sections), and examined under saturated water vapor conditions. 2. Gradual reduction of pressure from 4.00 torr to ~1 torr in the sample chamber resulted in controlled dehydration of the sample environment (see Note 5). Examples of ESEM images of liposomes are shown in Figs. 5–9. 3. The effect of the hydration medium on the stability of drug-free liposomes under controlled dehydration conditions can be investigated using PBS and distilled water, respectively. 4. The influence of bilayer drug (ibuprofen) loading on liposome stability can also be analyzed using controlled dehydration of samples to define a coalescence pressure. This is shown in Fig. 5. Examples of the images from ESEM, showing the stability of POPC liposomes during dehydration, are shown in Fig. 6. 5. A similar study using antigen-containing niosomes is shown in Fig. 7 and images of lipoplexes studied by ESEM is shown in Fig. 8 (see Note 6).

Fig. 5. ESEM micrographs of pre-formed drug-free MLV (PC:Chol; 4:1 molar ratio) (Fig. 2a, b, c) and ibuprofen-loaded MLV (PC:Chol; 4:1 molar ratio) (Fig. 2d, e, f) suspended in 0.01M PBS (pH 7.4). Vesicles were subjected to controlled dehydration in the ESEM sample chamber and ibuprofen-loaded liposomes were shown to be resistant to pressure as low as 1.4 Torr, compared to drug-free MLV which were observed to coalesce at pressure around 2.9 Torr. Magnification of samples held at an operating pressure of 1.9 Torr reveals small grains or specks on the liposome surface (Fig. 2e). Reproduced from [7] with permission from Elsevier Science

ESEM Imaging of Vesicle Systems

327

Fig. 6. ESEM micrographs of propofol-loaded liposomes formed using POPC and 11 total molar% cholesterol suspended in double distilled water. Vesicles were subjected to controlled dehydration in the ESEM sample chamber. At an operating pressure of 4.0Torr, vesicles appeared as spherical structures. (a) Gradual decrease of the operating pressure to 2.4 and 1.4 Torr showed intact, stable liposome structures that evidently resisted any significant morphological changes (b and c, respectively)

Fig. 7. ESEM micrographs of niosome-based vesicles (MP:Chol:DDA:TDB; 16:16:4:0.5 µmol) suspended in an aqueous buffer (pH. 7.4). The clarity of the image can be seen to improve as the moisture content is reduced (Fig. 5 a, b, and c), with a zone of evaporating water surrounding the vesicles being observed (Fig. 5d). Salt crystals can be seen to be forming on (Fig. 5d) and around (Fig. 5e) the niosomes. Images reproduced from (21) with permission from Pharm Press

328

Perrie et al.

Fig. 8. ESEM of liposome-DNA complexes composed of DOPE:DC-Chol (1:1 molar ratio, at a charge ratio of 1.5:1). Various structures from small spherical vesicles (a) to larger aggregates (b) and long tubules (c) can be seen. Images reproduced from [20] with permission from Informa Health

4. Notes 1. Non-ionic surfactant vesicles or niosomes are prepared using non-ionic surfactants rather than phospholipids, and may offer advantages in terms of chemical stability. Like liposomes, they can be prepared in a range of sizes and can incorporate drugs in a similar fashion to liposomes. 2. The temperature used at this stage need to only support evaporation and does not need to support the transition temperature of the lipids, as is sometimes reported. 3. Generally, incubation of the liposomes for 30 min is recommended. However, this will depend on the amount of lipid to be evaporated. At higher temperatures, care should be taken to avoid evaporation of the hydration media and the volume at the end of the process should be measured. 4. The time of sonication will vary greatly depending on the lipids used, the volume, and the concentration. Therefore, it is advisable to perform a preliminary study where the effect of sonication time against liposome size is measured. 5. Using buffers during ESEM can result in the formation of crystals from the buffers present (e.g. Fig. 9). The buffer salts should be removed prior to analysis by dialysis or washing via centrifugation. Alternatively, controls of buffer solution will allow for the identification of crystal structures formed due to the presence of such salts. 6. ESEM, whilst offering good visualization of these systems does not give information on the internal structure of these systems. Some indications of the bilayer nature of the liposomes are shown in Fig. 6, however this is limited. For detailed examination of internal compartments, freeze-fracture or cryo-SEM is more advantageous (Fig. 10).

ESEM Imaging of Vesicle Systems

329

Fig. 9. The presence of geometric salt crystals can be seen at reduced pressures in some instances (Fig. 10a) and the growth of these crystals can be precipitated by focusing the ESEM beam for enhanced durations at single sites on the grid (Fig. 10b). Images reproduced from (20) with permission from Informa Health

Fig. 10. Multilamellar vesicles imaged using various EM systems. Multiple bilayers and vesicle aggregates can be visualized using TEM (a and b), freeze-fracture (Fig. 6c; image kindly supplied by Brigette Sternberg-Papahadjopoulos) and cryo-EM (Fig. 6d; image courtesy of Peter Frederik). Images reproduced from (20) with permission from Informa Health

330

Perrie et al.

Acknowledgments The authors gratefully acknowledge the financial support rendered: during the time of this work, Afzal Mohammed’s research was supported by Pfizer Global Research and The School of Life and Health Sciences, Aston University; Anil Vangala was awarded an Aston Scholarship; Daniel Kirby was funded by the European Commission (contract no. LSHP-CT-2003-503367), Habib Ali and Sarah McNeil were both funded through EPSRC Case awards, with additional support from Lipoxen Technologies Ltd for Sarah McNeil. References 1. Segal AW, Willis EJ, Richmond JE, Slavin G, Black CD, Gregoriadis G (1974) Morpho logicalobservations on the cellular and subcellular destination of intravenously administered liposomes. Br J Exp Pathol 55:320–7 2. Sternberg B, Sorgi FL, Huang L (1994) New structures in complex formation between DNA and cationic liposomes visualized by freeze—fracture electron microscopy. FEBS Lett 356:361–6 3. Xu Y, Hui S-W, Frederik P, Szoka FC Jr (1999) Pysicochemical characterisation and purification of cationic lipoplexes. Biophysical J 77: 341–53 4. Perrie Y, Frederik PM, Gregoriadis G (2001) Liposome-mediated immunisation: the effect of vesicle composition. Vaccine 19:3301–10 5. Davidsen J, Rosenkrands I, Christensen D, Vangala A, Kirby D, Perrie Y, Agger EM, Andersen P (2005) Characterization of cationic liposomes based on dimethyldioctadecylammonium and synthetic cord factor from M. tuberculosis (trehalose 6, 6¢-dibehenate) – a novel adjuvant inducing both strong CMI and antibody responses. Biochim Biophys Acta 1718(1–2):22–31 6. Arunothayanun P, Turton JA, Uchegbu IF, Florence AT (1999) Preparation and in vitro/ in vivo evaluation of luteinizing hormone releasing hormone (LHRH)-loaded polyhedral and spherical/tubular niosomes. J Pharm Sci 88:34–8 7. Mohammed AR, Weston N, Coombes AGA, Fitzgerald M, Perrie Y (2004) Liposome formulation of poorly water soluble drugs: optimisation of drug loading and ESEM analysis of stability. Int J Pharm 285:23–34 8. Donald AM (1998) Taking SEMs into a new environment. Mater World: 399–401

9. Egelhaaf RM, Epand RF, Maekawa S (2003) The arrangement of cholesterol in membranes and binding of NAP-22. Chem Phys Lipids 122:33–9 10. McKinlay KJ, Allison FJ, Scotchford CA, Grant DM, Oliver JM, King JR, Wood JV, Brown PD (2004) Comparison of environmental scanning electron microscopy with high vacuum scanning electron microscopy as applied to the assessment of cell morphology. J Biomed Mater Res 69A:359–66 11. Muscariello L, Rosso F, Marino G, Giordano A, Barbarisi M, Cafiero G, Barbarisi A (2005) A critical overview of ESEM applications in the biological field. J Cell Physiol 205:328–34 12. Robinson VN (1975) A wet stage modification to a scanning electron microscope. J Microsc 103:71–7 13. Moncrieff DA, Robinson VNE, Harris LB (1978) Charge neutralisation of insulating surfaces in the SEM by gas ionisation. J Phys D: Appl Phys 11:2315–25 14. Moncrieff DA, Barker PR, Robinson VNE (1979) Electron scattering by gas in the scanning electron microscope. J Phys D: Appl Phys 12:481–8 15. Danilatos GD (1993) Introduction to the ESEM instrument. Microsc Res Tech 25:354–61 16. Danilatos GD (1990) Theory of the gaseous detector device in the ESEM. Adv Elect Electron Phy 78:1–102 17. Danilatos GD (1988) Foundations of environmental scanning electron microscope. Adv Elect Electron Phy 71:109–250 18. Donald AM, He C, Royall CP, Sferrazza M, Stelmashenko NA, Thiel BL (2000) Applications of environmental scanning electron microscopy to colloidal aggregation and film formation. Colloid Surface Physicochem Eng Aspect 174:37–53

ESEM Imaging of Vesicle Systems 19. Stokes DJ (2003) Recent advances in electron imaging, image interpretation and applications: environmental scanning electron microscopy. Philos Transact A Math Phys Eng Sci 361:2771–87 20. Perrie Y, Mohammed AR, Vangala A, McNeil SE (2007) Environmental scanning electron microscopy studies of liposomes and niosomes. J Liposome Res 17(1):27–37 21. Vangala AK, Kirby D, Rosenkrands I, Agger E-M, Andersen P, Perrie Y (2006) A comparative study of cationic liposomes and niosomebased adjuvant systems for protein subunit vaccines Characterisation, Environmental Scan ning Electron Microscopy analysis and immunisation studies. J Pharm Pharmacol 58:787–99 22. Elvira C, Fanovich A, Fernandez M, Fraile J, San RJ, Domingo C (2004) Evaluation of drug delivery characteristics of microspheres

331

of PMMA-PCL-cholesterol obtained by supercritical-CO2 impregnation and by dissolution-evaporation techniques. J Control Release 99:231–40 23. D’Emanuele A, Dinarvand R (1995) Preparation, characterisation, and drug release from thermoresponsive microspheres. Int J Pharm 118:237–42 24. Cao Y, Li H (2002) Interfacial activity of a novel family of polymeric surfactants. Eur Polymer J 38:1457–63 25. Bangham AD, Standish MM, Watkins JC (1965) Diffusion of univalent ions across the lamellae of swollen phospholipids. J Mol Biol 13:328–52 26. Kirby C, Gregoriadis G (1984) Dehydrationrehydration vesicles: a simple method for high yield drug entrapment in liposomes. Biotechnology 2:979–84

Chapter 22 Freeze-Fracture Electron Microscopy on Domains in Lipid Mono- and Bilayer on Nano-Resolution Scale Brigitte Papahadjopoulos-Sternberg Abstract Freeze-fracture electron microscopy (FFEM) as a cryo-fixation, replica, and transmission electron microscopy technique is unique in membrane bilayer and lipid monolayer research because it enables us, to excess and visualize pattern such as domains in the hydrophobic center of lipid bilayer as well as the lipid/gas interface of the lipid monolayer. Since one of the preparatory steps of this technique includes fracturing the frozen sample and, since during this fracturing process the fracture plane follows the area of weakest forces, these areas are exposed allowing us to explore the pattern built up by lipids and/or intrinsic proteins and which are also initiated by peptides, drugs, and toxins reaching into these normally hard to access areas. Furthermore, FFEM as a replica technique is applicable to objects of a large size range and combines detailed imaging of fine structures down to nano-resolution scale within images of larger biological or artificial objects up to several ten’s of micrometers in size. Biological membranes consist of a multitude of components which can self-organize into rafts or domains within the fluid bilayer characterized by lateral inhomogeneities in chemical composition and/ or physical properties. These domains seem to play important roles in signal transduction and membrane traffic. Furthermore, lipid domains are important in health and disease and make an interesting target for pharmacological approaches in cure and prevention of diseases such as Alzheimer, Parkinson, cardiovascular and prion diseases, systemic lupus erythematosus and HIV. As a cryofixation technique FFEM is a very powerful tool to capture such domains in a probe-free mode and explore their dynamics on a nano-resolution scale. Key words: Freeze-fracture electron microscopy, Domain exploration, Liposomal bilayer, Lipid monolayer, Lipid-stabilized gas bubbles, Membranes

1. Introduction Originally developed for material science (1) freeze-fracture electron microscopy (FFEM) was adapted to imaging of virus particles (2), and extended to biological objects such as cell membranes (3). Over the past 45 years FFEM has developed into a standard V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_22, © Humana Press, a part of Springer Science+Business Media, LLC 2010

333

334

Papahadjopoulos-Sternberg

tool providing a wealth of information about cellular structures and leading to fundamental advances in our understanding of cell structure and function (4–7). Furthermore FFEM is widely used to characterize artificial superstructure organizations (8–10) such as lipid bilayer- and non-bilayer arrangements (11, 12) as well as monolayer-based structures (13). Key requirement for any electron microscopy technique is avoiding potential artifact formation and providing sample stability in the electron beam of the microscope. In the early 1960s the concern about artifacts led to cryofixation, a radical new approach in maintaining exactly the original sample morphology under real-life conditions without using chemical fixatives and/or cryoprotectant. Living cells, artificial model systems, and even tissues are rapidly frozen at temperature of liquid nitrogen and the shockfrozen sample is broken open and fractured under vacuum. The freshly broken fracture planes are then shadowed and replicated with a thin layer of platinum-carbon. The thoroughly cleaned replica is then ready for examination in a transmission electron microscope where, different from the original sample, they show excellent stability in the electron beam over a period of time. Furthermore, they reveal extraordinary three-dimensional clarity of cellular, sub-cellular, and artificial cell structures and allow visualization of even intrinsic proteins (14–16) as well as biocolloids (17–19) at near-molecular resolution. A selection of electron micrographs obtained by applying freezefracture technique will be presented depicting lipid, drug, toxin, as well as protein domains formed in liposomal bilayer. Furthermore, more ordered domains (D) will be shown recently detected in lipid monolayer, stabilizing hydrophobic gas bubbles (20).

2. Materials 2.1. Rapid Freezing

1. Schleicher & Schuell Filter papers # 4 (Whatman®), qualitative, circles, 55 mm Ø. 2. Copper specimen carriers, roughened, for fluid samples (flat, BU 012056-T) and for fluid and tissue samples (with depression, BU 012054-T) from BAL-TEC Electron Microscopy Preparation Equipment, Accessories & Consumables. 3. Ultrasound water bath and Branson OR formulated cleaning concentrate for cleaning the copper specimen carriers. 4. Syringes (Hamilton®) or capillaries (Drummond Microcaps®). 5. Tweezers with fine tips and insulated by paint or rubber coating at the holding end. 6. Liquid nitrogen from 160 L storage tank.

Freeze-Fracture Electron Microscopy

335

7. Propane from a camping bottle. 8. Home-made propane liquidizing equipment in Dewar cylinder. 9. Home-made storage containers containing small metal tracks where the sandwiches can be tagged into small slots which keep them neatly together, and stored under liquid nitrogen in specimen baskets until the fracturing process will be carried out. 2.2. Fracturing the Cryo-fixed Samples and Replicating the Fracture Faces

1. JED-9000 Freeze-etching equipment (Jeol) adapted to the sandwich technique. 2. Hinged double replica devices with two sandwich slots each device, built for JED-9000 at UC Berkeley workshop. 3. Home-made loading station to transfer shock-frozen sandwiches from the storage container into slots of hinged double replica device and further on, by using the Jeol transfer rod, into the preparation chamber of the JED-9000. This entire step is carried out at liquid nitrogen temperature. 4. Braycote 803 Vacuum Grease from Ted Pella, Inc. 5. Microscope slides (glass, 75 × 38 × 1 mm). 6. Platinum Carbon electrodes (Platinum Carbon Chip, Item SMP# 315082 from Böckeler Instruments, Inc. Tucson). 7. Carbon electrodes (Carbon Chip, Item SMP# 315081 from Böckeler Instruments, Tucson). 8. Tungsten wire for filaments (99.90% Tungsten, size 0.02″ diameter) from Electron Microscopy Sciences (www.emsdiasum.com). 9. Homemade holder for opened replica device to take out sandwich halves, carrying replicas of sample fracture planes, safely without destroying replicas.

2.3. Cleaning the Replica and Examination in a Transmission Electron Microscope

1. Porcelain spot dish, 6 or 12 wells (Ted Pella, Inc.). 2. Plastic storage boxes with tightly fitting cover (11¾″ × 8″ × 2¾″). 3. Pasteur capillary pipettes (glass, 9″ long) and rubber bulbs for glass pipettes. 4. De-ionized Water (ACS 18 MW-cm resistivity, ACME Analytical Solutions, Inc.). 5. Loops made from Platinum wire for cleaning replicas with fuming nitric acid. 6. Titanium Tweezers with very sharp pointed tips (Dumont Biology, Ted Pella, Inc.). 7. Nitric acid, fuming (“Baker Analyzed” Reagent, 90.0% min) and Acetone (“Photrex”, 99.5% min) from J. T. Backer, Inc. and Chloroform (Optima, 99.0% min).

336

Papahadjopoulos-Sternberg

8. Baking soda. 9. Au-grids (400mesh, 50/vial, Gilder) and Grid storage box (100-grid-capacity, Gilder) from EMS, Inc. 10. Negatives (Kodak EM films 4489, 3½ × 4″, 250/pk) and photo paper (multigrade IV RC Deluxe, 8 × 10″, 250/pk). 11. Developer for negatives (Kodak Professional D-19), developer for paper (Kodak Professional Dektol), fixer for plates and paper (Kodak Professional Fixer), and Kodak hypo clearing agent for film and paper. 12. Darkroom equipment.

3. Methods Three practical operations are needed for freeze-fracturing of sample dispersions or even powders in preparation for electron microscopic examination: Rapid freezing of the sample (described in Subheading 3.1.), fracturing the sample under vacuum and producing a Pt/C replica (described in Subheading 3.2.), and cleaning and then examining the replica in a transition electron microscope (described in Subheading 3.3.) (8). 3.1. Rapid Freezing

Ultrarapid freezing is used, ideally, to preserve the morphology of the structures and to maintain all components at real life conditions from which the sample was frozen without the need for chemical fixative or cryoprotectans. During the cryofixation procedure the momentary distribution of all components in the systems is retained. In non-ideal solutions such as cell suspensions, liposome dispersions, micellar solution, or molecular aggregates, the association and orientations of interacting particles should be instantaneously fixed. This means that high spatial resolution for the morphology of the membrane components and fast resolution for dynamic processes can be achieved. Furthermore, ice crystal formation has to be avoided at shock freezing of biological samples known for their high water content (in most cases). Ice crystals easily destroy cells, especially plant cells because of their large vacuoles. Therefore the cryofixation procedure should be so fast that vitrified ice is obtained embedding all sample components like a matrix. For ultra-rapid quenching, the sandwich technique in which the sample is mounted between two copper foils in a thin layer has been found to be highly efficient (21). 1. Placing at least 20 specimen carriers on a filter paper, taped to the top of the lab bench, for each sample (see Note 1). 2. Mounting sample onto specimen carriers in a thin layer with the help of a syringe or glass capillary for liquid suspensions

Freeze-Fracture Electron Microscopy

337

or with a glass rod or spatula for creams or powders (see Note 2). 3. Forming a sandwich with the help of tweezers, precisely aligned, from two specimen carriers enclosing the sample inside the sandwich. 4. Blotting-off excess sample with small filter paper pieces. 5. Liquidizing propane from a camping bottle by leading it through a copper spiral cooled with liquid nitrogen, and storing it in a small metal reservoir, all cooled by liquid nitrogen in a Dewar cylinder. 6. Grasp sample-containing sandwich with tweezers and push it into liquid nitrogen cooled propane (83 K) and twirl it for 2 s to get a quenching rate of about 104 K/s (see Note 3). This is the same order of magnitude as obtained by jet freezing, which produces a slightly higher cooling rate (22) than sandwich freezing. This is still two magnitudes of order faster than required to prevent molecular rearrangement of the lipids in the bilayer on cooling through the main phase transition of lipids which is 2 s (23). 7. Transfer the sandwich with the shock-frozen sample as quickly as possible into liquid nitrogen and store it in a small metal container (one for each sample) marked with a number and covered by a lid. These homemade containers contain small metal tracks where the sandwiches can be tagged into small slots which keep them neatly together, and stored under liquid nitrogen (77 K) in specimen baskets until the fracturing process will be carried out (the usual storage time is not exciding 2 h). As an example of temperature-induced rearrangements of membrane lipids, smooth fracture planes are observed for liposomes made from the synthetic lipid 1,2-dipalmitoyl-sn-glycero-3 phosphocholine (DPPC) quenched at temperatures (Tq>Tc) above the phospholipids bilayer gel to liquid-crystalline transition temperature (Tc~41°C) at a cooling rate faster than 104 K/s (see Fig. 2A in (8)). Jumbled ripples are noticeable on fracture planes of DPPC liposomes quenched at Tq>Tc, but with a slower cooling rate than 104 K/s (Fig. 2B in (8)). At a similar cooling rate but quenched at temperature below Tc , the liposomal fracture planes display two kinds of ridges, termed l/2 (zigzag) and l (wave-like) ridges (Fig. 2C in (8)). The l ridges indicate molecular ordering characteristics of the Pb’ phase an intermediate bilayer phase between gel (Lb’) and liquid-crystalline state (La). The l/2 ridges are characteristic for the Lb’ gel phase (24). In similar proteoliposomes, made from 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) and the intrinsic protein bacteriorhodopsin, the protein particles decorate the lipid ridges or are localized in structural defects of these lipid ridges (Fig. 4a–f in (14)).

338

Papahadjopoulos-Sternberg

3.2. Fracturing and Shadowing of the Fracture Faces

1. Setting up freeze-fracture device Jeol JED-9000 Freeze-EtchingEquipment by servicing the gun chamber with fresh Pt/C and C electrodes and fresh filaments, if needed, and positioning the electrodes neatly in the upper center of the filament. 2. Add liquid nitrogen to the diffusion pump Dewar when a total pressure of <2 × 10−2 PA is attained. Cool down the cold stage and shroud by filling the Dewar of the specimen chamber with liquid nitrogen. When the temperature indicated is below −150°C and the vacuum has reached 2 × 10−5 Torr samples may be loaded. 3. Load the shock-frozen sandwiches from storage container into the slots of the hinged double replica device, pre-cooled by liquid nitrogen, by using the homemade loading station. 4. Transfer the sandwich holder at low temperature and high vacuum onto the table of the preparation chamber of the freeze-fracture machine by using the Jeol transfer rod. 5. Carrying/Carry-out the fracturing procedure by opening the hinged double replica device with a hook (see Note 4). During the fracturing procedure the fracture plane follows the area of weakest forces exposing the protoplasmic fracture face (PF, convex fracture face, shadow behind the structure, Fig. 1a) and the exoplasmic fracture face (EF, concave, shadow in front of the structure, Fig. 1b) in the case of bilayer structures such as cell membranes (25) or liposomes (8). Figure 2a. depicts medium-size liposomes displaying concave as well as convex fracture faces (8, 26). In the case of lipid monolayer

Fig. 1. Bilayer splitting of cells/liposomes by freeze-fracturing (a and b) reveal protoplasmic (PF in a) and exoplasmic (EF in b) fracture faces. The true outer (OS, in c) and inner (IS, in d) surfaces of the cell/liposomal bilayers are exposed by freeze-etching (c and d), i.e., lowering of the ice (I) level by sublimation

Freeze-Fracture Electron Microscopy

339

Fig. 2. Additional structural information from fracture behavior: (a) Liposomes display concave (shadow in front of the structures, some marked by **) and convex (shadow behind the structures) fracture faces. (b) Lipid-stabilized gas bubbles display exclusively concave (shadow in front of the structure, some marked by **) fracture planes. (c) Protein particles as hard-core particles show shadow exclusively behind the structures. Bars on all freeze-fracture electron micrographs represent 100 nm and the shadow direction runs from bottom to top

stabilizing gas bubbles, the fracture plane follows the gas/lipid interface where the fatty acid tails of the lipid molecules are reaching into the hydrophobic gas bubble. As noticed in Fig. 2b. concave (shadow in front) fracture planes are revealed this way (13). Furthermore, hard-core particles such as quantum dots (27), protein particles (14–16), or spherical micelles show always shadows behind their structures (18, 28) as seen in Fig. 2c. Observing their fracture behavior additional information about the nature of the samples is obtained as demonstrated in Figs. 2a–c. The fracture behavior typical for bilayer

340

Papahadjopoulos-Sternberg

structures has been used to identify the spaghetti-type structure as a bilayer tube coating super-coiled plasmid DNA (29). 6. If etching is desired adjust the thermal controller to the desired etching temperature, typically at −100°C, and sublimate some of the ice for some seconds (typically 30 s and for deep etching 60–90 s) in order to lower the ice level and to expose the true outer and inner surfaces of the sample (Fig. 1c and d, respectively). 7. Quickly tilt the stage to the desired shadowing angle (typically 45 degrees). 8. Replicating the freshly exposed fracture plane by evaporating Platinum for 25 s in the chosen shadowing angle and, after resetting the specimen angle to “0” degrees, with carbon for 35 s (2 kV/ 60–70 mA). Since the replica preparation is quite a difficult process and since the quality of the electron micrographs is crucially dependant on the particle size of the evaporation layer it is recommended to produce, at least, three but better about 10 replica sets from each sample. Although this procedure results in a strong carbon backing to the replica, it can still fragment. Replica of quantum dots, micelles, or other small objects have a stronger tendency to fragment than replicas of larger objects such as large multilamellar vesicles (MLV) or large extended bilayer or non-bilayer structures. 9. If rotary shadowing is required, release the stage lock switch and apply the desired rotational speed for the entire platinum as well as carbon evaporation process. Rotary shadowing (30) is useful for direct measurement of the shape and size of structural entities such as intrinsic membrane proteins or lipidic particles (31). Rotary shadowing is an adequate solution to minimize the “cohort” effect observed on unidirectional shadowed replicas, wherein the shadow of a particle influences the image of a neighboring particle. This way it causes problems in authentic identification and size measurements. However, rotary shadowing is not universally applicable since particle heights are more difficult to measure in rotary-shadowed replicas. 3.3. Cleaning and Examination of the Replica in a Transmission Electron Microscope

1. Transfer of the hinged double replica device containing the sandwich holder with the platinum/carbon replica still attached to the fracture faces of the frozen sample out of the preparation chamber of the Jeol JED-9000 onto a homemade holder. 2. Take out sandwich halves from the replica device with fine pointed tweezers, warm them up to room temperature, and let the replica slide from the copper foil onto the surface of distilled water in one of the wells of the first row of a Porcelain spot dish.

Freeze-Fracture Electron Microscopy

341

3. Clean the replica totally free from the still adhered sample by transferring it into fuming nitric acid placed in the next well of the second row of a 6 or 12 wells spot dish. 4. Immerse the replica for two days in the fuming nitric acid at room temperature while the Porcelain spot plates are stored in the plastic containers covered tightly with the plastic lid. That way none of the fumes can get out of the box but rather clean the replica aggressively. 5. After, at least, two days of agitation time transfer the cleaned replica out of the fuming nitric acid with a platinum wire loop into distilled water placed into the next well in the third and final row. Normally one box houses two 12 wells plates or four six wells plates (24 spots totally) allowing cleaning of the replicas of eight preparations simultaneously by using three spots for each preparation (see Note 5). 6. Neutralization of residual fuming nitric acid by adding baking soda. 7. If it turns out that further cleaning is needed especially on samples showing high lipid concentrations, repeated agitation of the replica (non-mounted as well as mounted onto grids) with chloroform or chloroform/methanol (1:1 by vol.) has been proven to be successful (see Note 6). 8. Fish the replicas cleaned this way onto 400 mesh square gold-coated (G-400-Au) grids, dry them on filter paper, store the dried grids with the replica of the sample mounted in grid storage boxes, and examine them at a transmission electron microscope (JEOL 100 CX). If the transmission electron microscope has no digital camera attached negatives have to be exposed (usually 12 for each sample), developed, and selected for the final prints (usually 8½ × 11″in size). This way high quality of the freeze-fracture electron micrographs is obtained. The resolution observed in the final micrographs using this technique is limited by the size of the emitted Pt/C particles forming the evaporation layer (replica) and not by the resolution of the transmission electron microscope. A typical resolution of 20 Å for periodical structures is one order of magnitude less than the resolution obtained by the negative staining technique. It can be improved by using heavy metals other than Platinum (such as Tungsten), applying ultrahigh vacuum (<10−9 torr), and lower temperature (23 K) (32). The replicas produced this way represent the heart piece of FFEM. They are very stable in the electron beam of the microscope and also in time. Furthermore, they enable us to study objects of a large size range ranging from 2 nm (resolution limit for periodical structures) to light microscopy size range (~50 mm).

342

Papahadjopoulos-Sternberg

Therefore, FFEM combines detailed imaging of fine structures at high resolution within larger biological objects such as the interaction of the spaghetti-type structure with cultured skin cells (Fig. 3a) or artificial molecular arrangements such as polymerstabilized gas bubbles (Fig. 3b) given as examples. Figure 3a shows intact crossing of a spaghetti-type structure through the cytoplasmic membrane of a HaCaT cell. The part of the spaghetti-type structure, still outside of the cell and connected to a “meatball”-complex, is marked by ↑↑. The part of

Fig. 3. Examples of large size range detectable by FFEM: (a) Cell interaction of Spaghetti-meatball complex (▲Spaghetti part inside of cell, ↑↑Spaghetti part outside of cell, *meatball complex, CM=cytoplasm) Reproduced with permission from (10). (b) PVA-stabilized gas bubble (GB), diameter ~3 mm, shell thickness ~300 nm, some of the PVA chain/aggregates are marked by ↑. Bars on all freeze-fracture electron micrographs represent 100 nm and the shadow direction runs from bottom to top

Freeze-Fracture Electron Microscopy

343

this spaghetti-type structure, that has already penetrated the cell membrane and localized inside the cytoplasm of the cell, is marked by ◄. (For more details see (9, 10, 29)). This is an example for the power of FFEM in following the fate of a pretty small structure (~9 nm measured by FFEM for the spaghetti diameter) during its interaction with a much larger structure such as a cultured skin cell (~1 mm in diameter). The polymer-stabilized gas bubble in Fig. 3b. represents another example for FFEM imaging of tiny structures in a much larger, this time artificial, molecular arrangement. Polyvinyl alcohol chains/chain bundles, marked by ↑ are forming a ~300 nm shell and stabilizing a ~3 mm in diameter gas bubble. (Sample provided by Dr. G. Paradossi, Universita di Roma Tor Vergata, for more details see (33, 34)). 3.4. Domain Formation in Lipid Mono and Bilayer

Microdomains/lipid rafts within biological membranes as well as artificial mono and bilayers represent lateral inhomogeneities in chemical composition and/or physical properties. They are small (~10–500 nm), highly dynamic and made of a variety of substance classes. Among all techniques used to detect and characterize domains FFEM is a forgotten one. Therefore examples will be given demonstrating the unique advantages of FFEM in domain exploration in a probe-free mode and on nano-resolution scale. Freeze-fracture electron micrographs are shown of domains formed in lipid monolayer (Fig. 4) as well as in liposomal bilayers (Fig. 5a–c) and made of lipid (Fig. 4, 5a and d), drug (Fig. 5b), toxin (Fig. 5c), and protein (Fig.5e). Figure 4 displays domains (some of them marked by D) surrounded by matrix (marked by M), localized in a lipid monolayer, stabilizing a gas bubble. More than 10 domains with diameters ranging from 100 to 250 nm and a step height of about 20 nm are detectable in the lipid monolayer, coating the gas bubble shown. The diameter of the lipid-monolayer, stabilized gas bubble, represents 2.6 mm and similar to Fig. 2b concave fracture behavior is observed (Sample provided by Dr. Mark A. Borden/ Dr. Katherine W. Ferrara, Dept. Biomedical Engineering, UC Davis). For further details see (20). Fig. 5a depicts lipid-domain formation in liposomal bilayers made of egg lecithin:bola-lipid (50:50 mol%). The bola-lipid has been extracted from the cytoplasmic membranes of Thermoplasma acidophilum cells. Since these thermoacidophilic archaebacteria live at high temperature (39–59°C) and low pH ~1 to 2 unique bipolar and membrane-spanning lipids were found to protect the cell from the fatal environment. The main glycophospholipid, representing approx. 50% of the total membrane lipid, has been characterized as a diisopranol-2,3-glycerotetraether, modulated by the attachment of different headgroups (phosphoryl- and monoglycosyl) at both 1-positions of the two glycerols. Due to its

344

Papahadjopoulos-Sternberg

Fig. 4. Domain formation in lipid-monolayer stabilized gas bubble. Some of the domains are marked by D, matrix-areas marked by M. Bar represent 100 nm and the shadow direction runs from bottom to top

bipolarity and membrane-spanning character of the saturated methyl-branched C40 hydrocarbon (diphytanyl) chains this bolalipid forms a “monolayer” of about the dimensions of a common phospholipid bilayer when suspended in water. Such bola-lipid membranes do not fracture within their apolar plane like common membrane lipids and show cross fractions exclusively (35). Liposomes prepared from mixtures of egg lecithin/bola-lipid display bola-lipid domains (marked by BL in the larger, 500 nm in diameter liposome in Fig. 5a). With increasing bola-lipid content, however, such liposomes show more and more cross fractions similar to the smaller, ~100 nm in diameter liposomes in Fig. 5a. (marked by CF). Figure 5b. displays drug domain formation in a liposomal Amphotericin B preparation. Among drug delivery systems lipidbased formulations such as liposomes have been successfully used to deliver anti-fungal agents such as Amphotericin B. This polyene antibiotic produced by Streptomycetes nodosus is highly hydrophobic (insoluble in water at pH 6 to 7). Therefore in lipid

Freeze-Fracture Electron Microscopy

345

Fig. 5. Domain formation initiated by a variety of substance classes. (a) Bola-lipid domains in egg-lecithin liposomes (some of the domains are marked by BL) and cross fractures of bola-lipid rich liposomes (some of them marked by CF), (b) Liposomal Amphotericin B (some of the numerous drug domains marked by D), (c) Liposomal reconstituted LtxA toxin (some of the LtxA domains are marked by D, tubule formation marked by T), (c’) Inset shows LtxA-domain fine structure at high magnification), (d and e) Domain formation in proteoliposomes composed of BR/DMPC, (d) at a protein to lipid molar ration of 1:187 (some of the BR-free DMPC domains are marked by D) (Reproduced with permission from (14)), and (e) made of BR/DMPC/PMLipid at molar ratios of 1:14:9 (Reproduced with permission from (15)), liposome is covered all over with well-organized 2D-BR arrays. Bars on all freeze-fracture electron micrographs represent 100 nm and the shadow direction is running from bottom to top

bilayers it is accumulated in the hydrophobic region formed by the lipid fatty acid chains and away from the hydrophilic lipid head groups and the lipid bilayer/water interface. Here at higher drug concentrations it forms small drug domains showing diameters up to 120 nm (some of them are marked by D). In Fig. 5c LtxA toxin domains are visible in a giant liposome prepared by repeating freeze-thawing cycles. Aggregatibacter (Actinobacillus) actinomycetemcomitans, a gram-negative oral pathogen, produces leukotoxin LtxA. Such oral microbiota are responsible for dental caries and periodontitis and may represent a predisposing factor in systemic conditions such as coronary heart disease, diabetes and low birth weight (36). Reconstitution of LtxA into the liposomal bilayer causes domain formation (some domains are marked by D in Fig. 5c and 5c’, a blow-up of a highly structured domain) and the formation of tubular structures (marked by T in Fig. 5c). The overall circular domains have diameters ranging from 200 to 500 nm and appear to be made up of many individual LtxA molecules that are oriented perpendicular to the

346

Papahadjopoulos-Sternberg

liposome fracture plane. At higher magnification (Fig. 5c’) some domains display very regular pattern of particle rows reminiscent of gap-junctions, which are a sophisticated type of a pore. (Sample provided by DMD Edward T. Lally, School of Dental Medicine, University of Pennsylvania). The purple membrane (PM) of Halobacterium halobium cells belongs to the few naturally occurring 2D-crystals of membranespanning proteins and represents a very special case of domain formation in biological membranes. The formation of well ordered arrays is absolutely essential for structural elucidation of membrane proteins by electron microscopy in high resolution. Learning from the PM as a model we reconstituted bacteriorhodopsin (BR) in liposomal bilayers of BR/lipid complexes in order to get some insights into the minimal conditions for obtaining well-ordered 2D-BR arrays. These are (1) high content of BR (less than 100 lipid molecules per one BR molecule), (2) one of the essential lipid species of the PM containing twofold negatively charged head groups such as 2,3-di-O-phytanyl-sn-glycero-1-phosphoryl-3¢-snglycerol 1¢-phosphate and 2,3-di-O-phytanyl-sn-glycero-1phosphoryl-3¢-sn-glycerol 1¢-sulphate but also commercially available cardiolipin, and (3) high salt concentrations (4 M NaCl). Under no circumstances we found 2D-arrays of BR in complexes with the synthetic lipid 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) as shown in Fig. 5d. Here DMPC lipid domains with diameters ranging from 150 to 250 nm (some of them are marked by D) are well detected surrounded by high concentration of statistically distributed BR molecules (14). Different from this figure Fig. 5e displays well ordered 2D-BR arrays mediated by the PM lipid mixture which contains some proportions of the highly negatively charged head-group lipids (15, 16).

4. Notes 1. Forming each sandwich from one flat specimen carrier (BU 0120056-T) and one specimen carrier with a depression (BU 012054-T) gives sufficient fast cooling rates and a large enough sample volume for statistical sample evaluations. Furthermore, 10 sandwiches/replica preparations from each sample instead of three only increases the success rate in obtaining excellent replicas. 2. Using sandwich technique allows us to apply FFEM on a big variety of materials such as biological as well as artificial suspensions at all concentrations, creams, powders, gels, and even solids (whatever is suitable to be mounted between the two copper sandwiches in a thin layer).

Freeze-Fracture Electron Microscopy

347

3. In order to obtain an optimal cooling rate attention should be paid to keep the propane always liquid. Avoid its transformation into liquid nitrogen/propane sludge by adding fresh liquidized propane to the freezing chamber. 4. Open the hinged double replica devise with caution to avoid the flying out of especially the upper sandwich halves. For even more protection little metal “flags” were added to the replica device shielding the sandwiches from falling out during the opening process. 5. Fuming nitric acid (69.0-70.0 %) is a poison, very corrosive, and can cause serious damage to all body tissues. Handling even small amounts demands extreme caution, wearing glasses, gloves, and a face mask even when working under a well-aired hood. Residual fuming nitric acid, not used up during the sandwich cleaning process, is neutralized by adding baking soda. 6. Further cleaning of the replica with chloroform or chloroform/ methanol can be carried out even if the replicas are already mounted onto the grids. Self-made wire-mesh supports are placed into the spots of the Porcelain spot plates, the grids with the replicas mounted, placed at these supports and cleaned from still adhering hydrophobic sample by adding the solvent with a pipette.

Acknowledgments The author would like to thank Mr. Stephen Kuzmic, S & J Services, Santa Clara for all his technical support especially in building all the home-made devices, Mr. John Ayou, Microanalytical Laboratories, Inc., Emeryville for excess to the JEOL 100CX, Mr. Alexander Veynberg, UC Berkeley for his excellent workshopwork, and Dr. Jack Ackrell for all the helpful discussions and his technical help especially on preparation days. References 1. Hall CBE (1950) A low temperature replica method for electron microscopy. J Appl Phys 21:61–62 2. Steere RL (1957) Electron microscopy of structural detail in frozen biological specimens. J Biophys Biochem Cytol 3:45–60 3. Moor H, Mühlethaler K (1963) Fine structure in frozen-etched yeast cells. J Cell Biol 17:609–628 4. Pinto da Silva P, Branton D (1970) Membrane splitting in freeze-etching: covalently Bound

Ferritin as a Membrane Marker. J Cell Biol 45:598–605 5. Branton D (1971) Freeze-etching studies of membrane structures. Phil Trans Roy Soc Lond B 261:133–138 6. Singer SJ, Nicolson GL (1972) The fluid mosaic model of the structure of cell membranes. Science 175:720–731 7. Bullivant S (1974) Freeze-etching technique applied to biological membranes. Phil Trans R Soc London 268:5–14

348

Papahadjopoulos-Sternberg

8. Sternberg B (1992) Freeze-fracture electron microscopy of liposomes. In: Gregoriadis G (ed) Liposome Technology, 2nd edn. CRC Press Vol. I, Boca Raton, Ann Arbor, London, Tokyo, pp 363–383 9. Sternberg B (1996) Liposomes as a model for membrane structures and structural transformations: a liposome album. In: Lasic DD, Barenholz Y (eds) Handbook of nonmedical applications of liposomes. From gene delivery and diagnostics to ecology. CRC Press, Boca Raton, New York, London, Tokyo, pp 271–297 10. Sternberg B (1998) Ultrastructural morphology of cationic liposome-DNA complexes for gene therapy. In: Lasic DD, Papahadjopoulos DP (eds) Medical applications of liposomes. Elsevier Amsterdam. Lausanne, New York, Oxford, Shannon, Singapore, Tokyo, pp 395–427 11. Angelova A, Angelov B, PapahadjopoulosSternberg B, Bourgaux C, Couvreur P (2005) Protein driven pattering of selfassembled cubosomic nanostructures: Long oriented nanoridges. J Phys Chem B 109(8):3089–3093 12. Angelov B, Angelova A, PapahadjopoulosSternberg B, Lesieur S, Sadoc J-F, Ollivon M, Couvreur P (2006) Detailed structure of dimond-type lipid cubic nanoparticles. J Am Chem Soc 128(17):5813–5817 13. Brancewicz C, Rasmussen DH, Papahadjopoulos-Sternberg B (2006) Hydrophobic gas bubble formation in Definity®: A freeze-fracture electron microscopy study. J Disp Sci and Techn 27:761–765 14. Sternberg B, Gale P, Watts A (1989) The effect of temperature and protein content on the dispersive properties of bR from H. halobium in reconstituted DMPC complexes free of endogenous purple membrane lipids: a freeze-fracture electron microscopy study. Biochim Biophys Acta 980:117–126 15. Sternberg B, Hostis CL, Whiteway CA, Watts A (1992) The essential role of specific Halobacterium halobium polar lipids in 2D-array formation of bacteriorhodopsin. Biochim Biophys Acta 1108:21–30 16. Sternberg B, Watts A, Cejka Z (1993) Lipid induced modulation of the protein packing in two-dimensional crystals of Bacteriorhodopsin. J Structural Biology 110:196–204 17. Lee Kan P, Papahadjopoulos-Sternberg B, Wong D, Waigh RD, Watson DG, Gray AI, McCarthy D, McAllister M, Schätzlein AG, Uchegbu IF (2004) Highly hydrophilic fused aggregates (microsponges) from a C12

Spermine Bolaamphiphile. J Phys Chem B 108:8129–8135 18. Qu X, Khutoryanskiy VV, Stewart A, Rahman S, Papahadjopoulos-Sternberg B, Dufes Ch, McCarthy D, Wilson CG, Lyons R, Carter KC, Schätzlein A, Uchegbu IF (2006) Carbohydrate-based micelle clusters which enhance hydrophobic drug bioavailability by up to 1 order of magnitude. Biomacromolecules 7(12):3452–3459 19. Bell PC, Hurley CA, Nicol A, Guenin E, Wong JB, Pilkington-Miksa MA, Sarkar S, Writer MJ, Barker SE, PapahadjopoulosSternberg B, Ayazi Shamlou P, Hailes HC, Hart SL, Zicha D, Tabor AB (2007) Biophysical characterization of an integrintargeted lipopolyplex gene delivery vector. Biochemistry 46:12930–12944 20. Borden MA, Martinez GV, Ricker J, Tsvetkova N, Longo M, Gillies RJ, Dayton PA, Ferrara KW (2006) Lateral phase separation in lipid-coated microbubbles. Langmuir 22(9):4291–4297 21. Costello MJ (1980) Ultra-rapid freezing of thin biological samples. Scan Electron Microsc Pt 2:361–370 22. Costello MJ, Fetter R, Höchli M (1982) Simple procedures for evaluating the cryofixation of biological samples. J Microsc 125:125–136 23. Tenchov BG, Lis LJ, Quinn PJ (1987) Mechanism and kinetics of the subtransition in hydrated L-dipalmitoyl-phosphatidylcholine. Biochim Biophys Acta 897:143–151 24. Copeland BR, McConnell HM (1980) The rippled structure in bilayer membranes of phosphatidylcholine and binary mixtures of phosaphtidylcholine and cholesterol. Biochim Biophys Acta 599:95–109 25. Branton D (1966) Fracture faces of frozen membranes. Proc Natl Acad Sci 55:1048–1056 26. Torchilin VP, Levchenko TS, Rammohan R, Volodina N, Papahadjopoulos-Sternberg B, D’Souza GG (2003) Cell transfection in vitro and in vivo with nontoxic TAT peptide-liposome-DNA complexes. Proc Natl Acad Sci USA 100(4):1972–1977 27. Weng KC, Noble CO, PapahadjopoulosSternberg B, Chen FF, Drummond DC, Kirpotin DB, Wang D, Hom YK, Hann B, Park JW (2008) Targeted tumor cell internalization and imaging of multifunctional quantum dot-conjugated immunoliposomes in vitro and in vivo. Nano Lett. Published on Web 08/20/2008 28. Torchilin VP, Lukyanov AN, Gao Z, Papahadjopoulos-Sternberg B (2003) Immunomicelles: targeted pharmaceutical

Freeze-Fracture Electron Microscopy

29.

30. 31. 32.

carriers for poorly soluble drugs. Proc Natl Acad Sci USA 100(10):6039–6044 Sternberg B, Sorgi FL, Huang L (1994) New structures in complex formation between DNA and cationic liposomes visualized by freeze-fracture electron microscopy. FEBS Lett 356:361–366 Margaritis LH, Elgsaeter A, Branton D (1977) Rotary replication for freeze-etching. J Cell Biol 72:47–56 Ververgaert PHJTh, Verkley AJ (1978) A view on intramembraneous particles. Experrientia 34:454–455 Gross H (1987) High resolution metal replication of freeze-dried specimens. In: Steinbrecht RA, Zierold K (eds) Cryotechniques in Biological Electron Microscopy. SpringerVerlag, Berlin, pp 205–228

349

33. Paradossi G, Cavalieri F, Chiessi E, Ponassi V, Martorana V (2002) Tailoring of physical and chemical properties of macro- and microhydrogels based on telechelic PVA. Biomacromolec 3(6):1255–1262 34. Cavalieri F, El Hamassi A, Chiessi E, Paradossi G (2006) Tethering functional ligands onto shell of ultrasound active polymeric microbubbles. Biomacromolec 7(2):604–611 35. Sternberg B, Rudolph P (1992) Unusual fracture behaviour of membranes made of bipolar lipids of Thermoplasma acidophilum. Electron Microscopy 3, EUREM 92, Granada, Spain, 85–86. 36. Henderson B, Wilson M, Sharp L, Ward JM (2002) Actinobacillus actinomycetemcomitans. J Med Microbiol 51:1013–1020

Chapter 23 Atomic Force Microscopy for the Characterization of Proteoliposomes Johannes Sitterberg, Maria Manuela Gaspar, Carsten Ehrhardt, and Udo Bakowsky Abstract Atomic Force Microscopy (AFM) is a useful tool for the visualization of soft biological samples in a nanoscale resolution. In the study presented here, the surface morphology ofP-selectin and Transferrin modified proteoliposomes were investigated in air and under water. The proteins were visualized without prefunctionalization or staining. Key words: AFM, Proteoliposomes, P-Selectin, Transferrin, Surface Modification

1. Introduction Visualization of surface morphologies and structures can be used to deepen the understanding of physical, chemical and biological phenomena. Atomic Force Microscopy (AFM) allows the investigation of biological specimens in their natural environment with high lateral (below 1 nm) and vertical (approx. 0.1 nm) resolutions. Specimens are not compromised during sample preparation, because no fixation or staining is needed. In addition, with this microscopical approach, it is possible to measure physical properties, including friction, softness and viscoelasticity, and charge density at a nanometer scale, in addition to the topography of a sample. Thus, AFM is an extremely useful tool that extends to studies of large objects, such as whole cells down to smaller structures, such as membranes or model membranes (1), proteins (2), chromosomes (3) and nucleic acids (4).

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_23, © Humana Press, a part of Springer Science+Business Media, LLC 2010

351

352

Sitterberg et al.

In recent years, AFM was intensively used to study proteins within membranes, the topography of biomembranes (1, 5) and single molecules, at resolutions down to ~1 nm (6). Such high resolutions are only possible in densely packed amorphous or 2D crystal arrangements of proteins. These arrangements can be located in or on supported planar lipid bilayers (SPBs) or directly immobilized onto solid supports like mica (7). Proteins associated with SPBs do not allow studying protein functions such as transport across a membrane. However, this can be achieved when proteins are incorporated into liposomal membranes, forming the so called “proteoliposomes”. Proteoliposomes are spherical vesicles of diameters in the range of 100 nm to several µm. They consist of a lipid bilayer shell surrounding an aqueous core. The conditions on either side of the liposomal membrane can be closely controlled in order to investigate proteinrelated transport processes across membranes, e.g., the water transport mechanism of aquaporins (8) or the ATP-dependent transport of the chemotherapeutic, doxorubicin, against a concentration gradient (9). Moreover, proteoliposomes can be used in drug carrier research in pharmacological and biopharmaceutical sciences (10–12). Proteoliposomes also allow studies of the formation of ligand-receptor interactions of membrane-associated proteins. For proteoliposomes modified with antibodies, it could be demonstrated that the lateral structure depends on the coupling strategy and the density of the antibody (13). By means of AFM, the specific adhesion of unilamellar vesicles to functionalized surfaces was investigated and the changes in size and shape during the adhesion process could be measured (14). Due to the complexity of such studies, the use of AFM in this field, however, has only been poorly explored. Nevertheless, AFM is a technique with unique capabilities in native-state protein research. Here, we present some fundamental techniques of proteoliposome preparation and the characterization of two different types of proteoliposomes using AFM.

2. Materials 2.1. Preparation of P-Selectin-modified Proteoliposomes

1. P-selectin (155 kDa including natural glycosilation and transmembrane domain) was a gift from Dr. U. Rothe und Dr. H. Sann (Martin-Luther-Universitaet, Halle Wittenberg, Germany) who extracted the protein according to (15). 2. Tris buffered saline (TBS): Dissolve 0.24 g Tris and 0.87 g NaCl in 100 ml water. Store at 4°C (stable for at least 6 weeks).

AFM for the Characterization of Proteoliposomes

353

3. Octyl-beta-D-gluco-pyranoside is added to TBS to achieve a final concentration of 25 mM. Store at 4°C (stable for at least 6 weeks). 4. Lipid films: Dissolve 10 mg phosphatidylcholine (PC, purity >99%, Avanti Polar Lipids, Cologne, Germany) in 10 ml chloroform:methanol (2:1 v/v) (see Note 1) and subsequently complete evaporation of the solvents in a glass flask (see Note 2). 5. Dialysis membrane (Carl Roth, Karlsruhe, Germany, cutoff size: 14 kDa). 2.2. Liposome Preparation of Transferrin-modified Liposomes

1. 1,2-Distearoyl-sn-glycero-3-phosphocholine (DSPC), 1,2distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy (polyethylene glycol)-2000] (ammonium salt) (DSPE-PEG) and 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N[carboxy (polyethylene glycol) 2000] (ammonium salt) (DSPE-COOH-PEG) (Avanti Polar Lipids, Alabaster, AL). All lipids had a purity of >99% and were stored at −20°C. 2. Cholesterol (Chol) (also stored at −20°C), human holotransferrin (Tf) (stored at 4°C) and HEPES (stored at room temperature) (Sigma-Aldrich,Dublin, Ireland). 3. N-(3-Dimethylaminopropyl)-N¢-ethylcarbodiimide hydrochloride (EDC) (Fluka, Dublin, Irleand); the EDC working solution (2 mg EDC per 1 µmol of lipid in PBS needs to be prepared freshly before use. 4. N-Hydroxysulfosuccinimide (Sulfo-NHS) (Pierce, Rockford, IL). 5. BCA protein assay kit (Pierce, Rockford, IL).

2.3. Atomic Force Microscopy

1. As substrate for AFM-sample preparation silicon wafers with a natural silicon oxide surface layer (thickness 3.8 nm) and a surface roughness of 0.3 nm were used (Wacker Chemie AG, Munich, Germany). The wafers were split into small pieces of about 1 × 1 cm. The pieces were cleaned in a bath sonicator for 20 min in chloroform:methanol (2:1 v/v), then dried in a dry air stream. For better handling and to avoid artifacts (e.g., fingerprints), one side of the wafers was labeled as “bottom” surface. Store in a dust free atmosphere. 2. For AFM imaging in ambient conditions, NSC 16/Cr–Au intermittent contact cantilevers from Anfatec Instruments AG (Oelsnitz, Germany) with a nominal force constant of 45 N/m, a resonance frequency of 170 kHz and a length of 230 µm were used. The tip of the radius was smaller than 10 nm (manufacturer’s datasheet). 3. For AFM imaging under water, CSC 21 AlBS cantilevers from Anfatec Instruments AG were used. Cantilevers have a nominal force constant of 2 N/m and a resonance frequency

354

Sitterberg et al.

of 105 kHz (in air). The tip of the radius was smaller than 10 nm (manufacturer’s datasheet).

3. Methods 3.1. Preparation of Proteoliposomes

To study membrane proteins in a proteoliposomal bilayer, several factors have to be taken into consideration optimizing the protein reconstitution: the homogeneity of the protein, the molar lipid-toprotein ratio, the final orientation of the protein and the size and permeability of the liposomal membrane. For most proteins, the direction of the protein within the lipid bilayer is of great importance and has to be taken into account when optimizing the reconstitution method. There are several methods for the insertion of membrane proteins into artificial membranes using mechanical treatment, freeze-thawing, organic solvents, or detergents (16). Since, due to their hydrophobic character, solubilization of membrane proteins generally requires the use of detergents, it is very convenient to use an insertion method which also involves the use of detergents to prepare proteoliposomes. P-selectin has been chosen as an example for this chapter, but other transmembrane proteins can be incorporated into liposomal bilayers following the same procedures. Other types of proteoliposomes have the protein attached to linker lipids that are part of the phospholipids mixture comprising the membrane. The proteins can be conjugated either directly to the lipid, i.e., residing at the membrane, or be coupled to the end of spacers (e.g., polyethylene glycol chains). Proteoliposomes of the latter two types are commonly used in drug targeting and to investigate ligand-receptor interactions (17). We have chosen a modification with the serum protein, transferrin, as an example for this chapter.

3.2. Preparation of P-selectin-Modified Proteoliposomes

1. Disperse the purified P-selectin in TBS containing octyl-betaD-gluco-pyranoside to achieve a final protein concentration of 50 µg/ml (see Fig. 1). 2. Add 1 ml of this micellar dispersion to the flask containing the lipid film. 3. Ultrasonicate the flask for 5 min at 25°C. 4. Incubate this dispersion inside a dialysis membrane against 5 L distilled water to remove the octylglycoside. Change the water every 12 h for a total duration of 2 days.

3.3. Preparation of Transferrin-Modified Proteoliposomes

1. Liposomes are prepared by the thin film hydration method. The lipid composition is DSPC, Chol, DSPE-PEG and DSPE-PEG-COOH at the molar ratio of 1.85:1:0.132:0.018 (see Fig. 1).

AFM for the Characterization of Proteoliposomes

355

Fig. 1. Scheme of the investigated proteoliposomes

2. Prior to weighing, lipids should be brought to room temperature and weighed using an analytical balance, e.g., 9.7 mg DSPC, 2.6 mg Chol. 2.5 mg DSPE-PEG and 0.34 mg DSPE-PEG-COOH for preparing to 2 ml of liposomes at a phospholipid concentration of 10 µmol/ml. 3. All lipids are then dissolved in an organic solvent, e.g., chloroform. Once the lipid has been fully dissolved, the solvent must be removed completely creating a lipid film using a rotary evaporator (see Note 2). 4. The dried lipid film is then hydrated with the desired buffer to form multilamellar vesicles. The hydration of the lipid film has to be performed at a temperature at approximately 5°C above the temperature of the highest melting lipid component at least for 30 min. In this case hydration was performed at 60°C. 5. After the hydration step, and in order to produce a homogenous liposomal suspension, liposomes are submitted to extrusion using a Lipex extruder (Northern Lipids Vancouver, BC) at moderate pressures of 200–500 psi, by sequentially filtering the suspension through polycarbonate membranes until an average vesicle size of 0.1 µm was achieved. This step has to be also performed at 60°C. 6. The size of the liposomes in suspension can be determined by dynamic light scattering, e.g., in a ZetaSizer, Nano Series (Malvern Instruments, Malvern, UK). As a measure of particle size distribution of the dispersion, the equipment reports the polydispersity index ranging from 0 for an entirely monodisperse sample up to 1.0 for a polydisperse suspension. 7. The coupling of transferrin to liposomes is performed through an amide bound between the carboxylic groups of

356

Sitterberg et al.

the linker lipid, DSPE-PEG-COOH, and free amino groups of the transferrin in the presence of a water-soluble carbodiimide, EDC. The reaction of carboxyl group in the presence of EDC forms an amine-intermediate, which after reacting with Sulfo-NHS will form a semi-stable amine reactive NHS ester that afterwards will react with amino groups of Tf leading to a stable amide bond (17). 8. After the phospholipid quantification of the liposomal suspension, 1 ml of the buffer selected for liposomes preparation is added to 1 ml of liposomes at a lipid concentration of 10 µmol/ ml, 180 µl of Sulpho-NHS (0.25 M) and 180 µl of EDC (0.25 M) freshly prepared using the selected buffer (see Note 3). This mixture is allowed to incubate for 10 min at room temperature. After this incubation period, 125 µg of Tf/µmol of lipid are added and gently agitated overnight (see Note 4). 9. The unbound protein is separated from the liposomes after dilution of the suspension in the work buffer, followed by an ultracentrifugation step at 250,000×g for 3 h. The pelleted liposomes are re-suspended and their physicochemical properties are characterized in terms of mean size, surface charge, phospholipid concentration and Tf binding efficiency. 3.4. Atomic Force Microscopy

The methods are described for a Digital Nanoscope IV Bioscope (Veeco Instruments, Santa Barbara, CA) but can easily be adapted to other makes and models. The atomic force microscope was vibration- and acoustically damped (for detailed description see Oberle et al. (18). All measurements were performed in tapping™ mode (see Note 5). The applied force to the sample surface was adjusted to a minimum to avoid damage to the sample (see Note 6). The specimen was investigated and scanned under a constant force. The scan speed was proportional to the scan size and the scan frequency was between 0.5 and 1.5 Hz. Images were obtained by displaying the amplitude signal of the cantilever in the trace direction, and the height signal in the retrace direction, both signals being simultaneously recorded. The results were visualized either in height (the measured height of the sample in a resolution of 0.3 nm) or in amplitude mode (the damping of the cantilever oscillation due to tip-sample-interactions) (see Note 7).

3.4.1. AFM of P-SelectinModified Proteoliposomes

Various methods are available for the sample preparation (see Note 8). Proteoliposomes containing P-selectin are very fragile and have a tendency to spread across interfaces (sample support/water interface). Furthermore, these proteoliposomes show a loss of stability in diluted media. 1. The liposomal dispersion is directly transferred to a glass chamber with a piece of silicon wafer as a sample support immobilized at the bottom. During 1 h of incubation, liposomes are

AFM for the Characterization of Proteoliposomes

357

Fig. 2. Visualisation of P-selectin modified liposomes. The AFM measurements were performed under water. (a) Height mode overview. (b) higher resolution. Visualisation of the surface of a single P-selectin modified liposome. The small P-selectin molecules are visible as bright dots. The protein raises the lipid bilayer about 0.6 nm. The diameter of the protein is about 6.5 nm. (c) surface of a plain liposome. The surface is smooth (d) Section analysis of the surface in b

allowed to adsorb to the support. AFM imaging is conducted under water in the liposomal dispersion (see Fig. 2). 2. The cantilever is brought into contact with the sample surface in contact mode. 3. The resonance frequency of the cantilever is determined in the liposomal dispersion near the silicon support (see Note 9). The drive frequency of the piezo crystal is set slightly lower than the resonance frequency (see Note 10). 4. The AFM cantilever is approached to the sample surface until it contacts the surface. 5. Scanning of the sample is performed. The scanning speed is set proportional to the scan size to achieve constant speed of the AFM tip across the surface. The ideal speed for scanning, highly depends on the height of the sample and found to be 1 µm/s or slower. 6. The “speed” of the feedback loop (the so called integral gain or I-gain), which keeps the cantilever sensor system at a constant height relative to the sample, is adjusted. It should be as high as possible without visible “noise” in the AFM picture. This is monitored best using a topography oscilloscope visualizing each line that is scanned.

358

Sitterberg et al.

7. When the ideal settings for the sample are found, the imaging is restarted to avoid artifacts of sub-ideal settings. 8. After an overview of the sample is retrieved by scanning of at least three independent spots on the sample of an area of 10 × 10 µm², single proteoliposomes are investigated at higher resolution of 1.5 × 1.5 µm² up to 100 × 100 nm to visualize the liposome surface (Perform steps 3–6.). 3.4.2. AFM of TransferrinModified Liposomes

A very convenient procedure for the preparation of liposomes is the self-assembly technique (19). Small pieces of silicon wafers (about 1 × 1cm) as sample support material are immersed in the liposomal dispersion for 20 min at room temperature. During this time, the liposomes adsorb on the support surface under equilibrium conditions. After 20 min, the silicon supports are removed from the dispersions. The samples are dried in a dry air flow at room temperature and investigated within 2 h. 1. The resonance frequency of the AFM cantilever is determined by the manufacturer’s software. The driving frequency is set slightly smaller than resonance frequency (see Note 9). As preliminary settings, the standard settings of the AFM software can be used (see Fig. 3).

Fig. 3. Visualisation of Tf modified liposomes. The AFM measurements were performed in air (60% r.h.). (a) Height mode overview. The liposomes are visible as small bright colored objects. (b), higher resolution, but no surface features are visible in this height mode image. (c) 3-D image of a single liposome, some Tf molecule could be visualized at the surface (marked with arrows, height image), (d, e) Amplitude mode images. The Tf molecules are clearly visible. (f) Section analysis of the surface. The diameter of the Tf molecules could be measured and is between 3.0 and 3.5 nm

AFM for the Characterization of Proteoliposomes

359

2. The AFM cantilever is approached to the sample surface until it contacts the surface. 3. Then, adjusting and scanning is performed as in steps 3–8 in Subheading 3.4.1. 4. For imaging of Transferrin - liposomes, a scan speed of 1.5 µm/s is found to be ideal.

4. Notes 1. High performance liquid chromatography (HPLC) grade solvents should be used to avoid contamination by impurities. 2. Special care should be taken during evaporation of the solvent in order to prevent certain lipid components from crystallizing, and thus leading to a less homogenous lipid mixture. 3. The pH and composition of the re-suspending buffer is of utmost importance. When preparing Tf-modified proteoliposomes, PBS (pH 5.9 and 7.4) and 10 mM sodium citrate containing 140 mM NaCl (pH 5.9) showed the best results at coupling efficiencies of 90–98%. For comparison, 10 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) containing 140 mM NaCl (pH 7.4) resulted in a coupling efficiency of only ~56%. 4. Two different incubation temperatures are generally used for the Tf coupling reaction, room temperature and 4°C. In a side-by-side comparison, we could find slightly better results when performing this reaction at 4°C. 5. Tapping mode™ or intermittent contact mode is a key advance in AFM of soft, adhesive or fragile samples, i.e., biological materials. In this mode, the cantilever is oscillating near its resonance frequency and touches the sample only at its lower amplitude. During sample scanning, this contact of the tip with the sample causes a reduction in the oscillation amplitude. This change in oscillation amplitude is used to identify and measure surface features. Unlike in contact mode, where the cantilever is scanning across the surface and is in touch with the surface all the time, tapping mode minimizes the lateral forces applied to the sample. This allows imaging of soft material with high resolution without damage or alteration of the sample. This technique can be used for imaging in air and also in liquid environments. 6. Although lateral forces are minimized in the tapping mode, a force perpendicular to the sample surface is still applied to the sample. This is because at each of the lower amplitude

360

Sitterberg et al.

of cantilever oscillation, the tip contacts the sample. As the dampening of the lower amplitude due to sample tip interaction is used to determine sample height, the value of this dampening has to be high enough to clearly identify tip sample interactions. On the other hand, a higher dampening is proportional to a higher force applied onto the sample, which may lead to structural alterations or destruction of the sample. Therefore, the force should be adjusted to a point where both the force is high enough for high resolution scanning and the force is still as low as possible to avoid sample destruction. 7. AFM allows visualizing of a sample in different ways: The one which is easiest to understand is the “height mode” where the color scale of the AFM image gives information on the actual height of the sample at the specific spot, like a landscape map. Another mode, which allows visualizing of smaller surface features, is the “amplitude mode”, where the damping of the cantilever oscillation is indicated. This gives information about the steepness of the sample at the specific spot. Pictures visualizing amplitude mode always look as if illuminated from either left or right and show surface details more prominently. 8. Sample preparation is crucial for reliable and reproducible AFM images. Especially when working with labile structures like liposomes, several factors have to be considered. One of them is the tendency of lipids to form self assembled lipid bilayers at interfaces. This spreading may be avoided by adjustment of the surface chemistry of the support (e.g., chemical modification via silanization or by working in a liquid environment. As rehydration of samples may cause artifacts due to structural changes of the sample, keeping the sample in the liquid all time until imaging in fluids, is highly recommended. Special fluid chambers are commercially available for most atomic force microscopes. 9. The oscillation of the cantilever is not only dependent on the cantilever properties (its dimensions and material), but also on the medium it is surrounded by. In water, a damping of the oscillations due to the viscosity of water causes a shift of the resonance to lower frequencies. This is further affected by a higher damping near a surface. To determine the driving frequency which may be used for scanning under water, it is best to determine it while in the same medium and distance (e.g., 300 nm) from the surface as in scanning. 10. The oscillation of the cantilever is damped, when it is scanning across the sample. This causes the resonance frequency to decrease slightly. This phenomenon is taken into account by setting the driving frequency of the piezo crystal to a frequency lower than the free resonance frequency of the cantilever.

AFM for the Characterization of Proteoliposomes

361

Acknowledgments This work was supported in parts by DFG Forschergruppe 627 Nanohale and JPK Instruments Berlin (Germany) (UB and JS). This work was supported in part by grants from Enterprise Ireland under the National Development Plan co-funded by EU Structural Funds and Science Foundation Ireland (CE and MMG). References 1. Lukas K, Zhifeng S (1998) The application of AFM to biomembranes Biomembrane Structures 2. Schneider S, Lärmer J, Henderson R, Oberleithner H (1998) Molecular weights of individual proteins correlate with molecular volumes measured by atomic force microscopy. Pflügers Arch 435:362–7 3. Rasch P, Wiedemann U, Wienberg J, Heckl W (1993) Analysis of banded human chromosomes and in situ hybridization patterns by scanning force microscopy. Proc Natl Acad Sci USA 90:2509 4. Hansma HG, Kasuya K, Oroudjev E (2004) Atomic force microscopy imaging and pulling of nucleic acids. Curr Opin Struct Biol 14:380–5 5. Kamruzzahan A, Kienberger F, Stroh C et al (2004) Imaging morphological details and pathological differences of red blood cells using tapping-mode. AFM Biol Chem 385:955–60 6. Engel A, Müller D (2000) Observing single biomolecules at work with the atomic force microscope. Nat Struct Biol 7:715–8 7. Müller DJ, Engel A (1999) Voltage and pHinduced channel closure of porin OmpF visualized by atomic force microscopy. J Mol Biol 285:1347–51 8. Zeidel M, Nielsen S, Smith B, Ambudkar S, Maunsbach A, Agre P (1994) Ultrastructure, pharmacologic inhibition, and transport selectivity of aquaporin channel-forming integral protein in proteoliposomes. Biochemistry 33:1606–15 9. Awasthi S, Singhal S, Pikula S et al (1998) ATP-Dependent human erythrocyte glutathioneconjugate transporter. II. Functional recons titution of transport activity. Biochemistry 37:5239–48 10. Dass C (2008) Drug delivery in cancer using liposomes. Methods Mol Biol 437:177–82

11. Lian T, Ho R (2001) Trends and developments in liposome drug delivery systems Journal of Pharmaceutical. Sciences 90:667–80 12. Opinion E (2008) Antibody-targeted liposomes in cancer therapy and imaging. Expert Opin Drug Deliv 5:189–204 13. Bendas G, Krause A, Bakowsky U, Vogel J, Rothe U (1999) Targetability of novel immunoliposomes prepared by a new antibody conjugation technique. Int J Pharm 181:79–93 14. Pignataro B, Steinem C, Galla H, Fuchs H, Janshoff A (2000) Specific adhesion of vesicles monitored by scanning force microscopy and quartz crystal microbalance. Biophys J 78:487–98 15. Moore K (1991) GMP-140 binds to a glycoprotein receptor on human neutrophils: evidence for a lectin-like interaction. J Cell Biol 112:491–9 16. Rigaud J (2002) Membrane proteins: functional and structural studies using reconstituted proteoliposomes and 2-D crystals Brazilian. J Med Biol Res 35:753–66 17. Anabousi S, Laue M, Lehr C, Bakowsky U, Ehrhardt C (2005) Assessing transferrin modification of liposomes by atomic force microscopy and transmission electron microscopy. Eur J Pharm Biopharm 60:295–303 18. Oberle V, Bakowsky U, Zuhorn I, Hoekstra D (2000) Lipoplex formation under equilibrium conditions reveals a three-step mechanism. Biophys J 79:1447–54 19. Kneuer C, Ehrhardt C, Bakowsky H et al (2006) The influence of physicochemical parameters on the efficacy of non-viral DNA transfection complexes: a comparative study. Journal of Nanoscience and Nanotechnology 6:2776–82

Chapter 24 Method of Simultaneous Analysis of Liposome Components Using HPTLC/FID Sophia Hatziantoniou and Costas Demetzos Abstract Liposomes are composed of different kind of lipids or lipophilic substances and are used as carriers of bioactive molecules. The characterization of the prepared liposomes consists of the calculation of the drug to lipid molar ratio by measuring the lipids and the encapsulated molecule. The present work describes an analytical methodology on simultaneous determination of all the lipid ingredients of the liposome formulation, using Thin Layer Chromatography coupled with a Flame Ionization Detector (TLC/FID), using the least possible sample quantity. The method consists of a chromatographic separation of the liposomal ingredients on silica gel scintillated on quartz rods and subsequent detection of the ingredients by scanning the rods by a hydrogen flame. The produced ions are detected by a Flame Ionization Detector and the signal is converted to a chromatogram. This method may be applied on every step of the liposome preparation for examining the quality of the raw materials, tracking possible errors of the preparation procedure and finally analyzing the content of the final liposomal composition. Key words: Liposome, Lipid analysis, Drug/lipid ratio, Bioactive molecule, HPTLC/FID

1. Introduction Liposome technology is widely applied to both pharmaceutical and cosmetic formulations. Liposomes are used as a suitable vehicle for bioactive molecules in order to overcome their poor water solubility or their possible non desired side effects on normal cells (1, 2). The main component of the lipid bilayers of the liposomes are acyl-phosphatidylcholines of natural or synthetic origin. Other lipid substances such as cholesterol or charged

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_24, © Humana Press, a part of Springer Science+Business Media, LLC 2010

363

364

Hatziantoniou and Demetzos

lipids may be used in order to give desired performance to the liposomal formulation (3). The characterization of liposomes requires the calculation of the drug to lipid ratio, measuring the amount of the encapsulated active ingredient, as well as all the used components (4). In the present work we describe a method of simultaneous determination of all the liposomal ingredients using Thin Layer Chromatography coupled with a Flame Ionization Detector (TLC/FID) (2, 5, 6). The sample ingredients are separated on re-usable silica coated quartz rods using the classical Thin Layer Chromatography techniques and subsequently analyzed by passing through a Flame Ionization Detector. The flame of the burner is generated by an external Hydrogen supply and atmospheric oxygen supplied by the air pump that is incorporated into the instrument. The procedure requires only one measurement per sample and it can be applied even in very small or much diluted samples. The speed of analysis and the ability to assess many samples at the same time makes this method suitable for routine assessment. This method may find application on the assessment of liposomal formulations, on the quality control of the raw materials and preparation procedure.

2. Materials 2.1. Liposomal Sample Lyophilization

1. Screwed cup glass vials of 5 ml capacity

2.2. Sample Preparation

1. Pasteur pipettes

2.3. Sample Spotting

1. Chromarods-III

2. Parafilm

2. Cotton wool

2. Glass syringe of 1ml capacity 2.4. Sample Development

1. Development tank 2. The following solvent mixture of analytical grade have been used for liposome component separation: Chloroform/methanol/d-Water 45:25:5 (v/v) (7, 8) (see Note 1). Iatroscan newMK-5 (Iatron Laboratories, INC. Tokyo, Japan). 1. Lipids and bioactive components of analytical grade for preparation of standard solutions at concentrations similar to that of the samples.

Analysis of Liposome Components by HPTLC/FID

365

3. Methods 3.1. Liposomal Sample Lyophilization

1. Prepare 500 ml aliquots of each liposome sample in screwed cup glass vials of 5 ml capacity. 2. Cover the vials with parafilm and pierce it with a needle, to create thin holes. 3. Freeze the samples (see Note 2) and place them in the freezedrier overnight (9). 4. Place the cups on the vials and store at 4°C until use.

3.2. Sample Preparation

1. Dilute the freeze-dried residue in chloroform or other suitable solvent mixture. 2. Filter the samples through cotton filters in order to remove the sugars used as cryoprotectants, or the salts of the buffers used (1) (see Note 3). 3. Wash the residue on the cotton filter with 1 ml chloroform twice and add the filtrates. 4. Remove the chloroform under nitrogen stream 5. Weigh the residue and add a proper volume of chloroform (see Note 4) to a final concentration of about 20 mg/ml.

3.3. Sample Spotting

1. Run a blank scan to ensure that the Chromarods-III are clean. 2. Using a glass syringe of 1 ml capacity spot 1 ml of the sample solutions on the zero point of the Chromarods-III (see Note 5). Apply the sample on two to three Chromarods-III to calculate the mean area of three measurements. 3. Use the last Chromarod-III for corresponding standards at concentrations similar to the samples.

3.4. Sample Development

1. Line the rear side of the development tank with a piece of filter paper. 2. Pour the mobile phase (60–70 ml of solvent mixture) in the development tank and cover it with its glass lid (see Note 1). 3. Gently move the tank to wet the filter paper lining and allow the tank to saturate with solvent vapor (see Note 6). 4. Place the rod holder in the development tank and leave them until the solvent mixture front reaches the desired height (see Note 7). 5. Remove the rod holder from the development tank and allow the excessive solvent to drain. 6. Hold the rod holder under a stream of hot air for 1 min to completely remove the solvent residue (see Note 8).

366

Hatziantoniou and Demetzos

3.5. Sample Scanning

1. Place the rod holder in the scanning frame 2. Scan the Chromarods-III at the following conditions: scanning speed 30 s/scan, airflow 2000 L/h, and hydrogen flow 160 ml/min.

3.6. Calibration Curve

1. Prepare standard solutions for each component of the sample having twice the concentration of the expected in the sample. 2. Spot gradually increasing volumes of the standard solution from 0.2–1 ml using two rods per volume. 3. Place the rod holder into the development chamber and allow the development with the same conditions to the sample (see subheading 3.4). 4. Scan the rods and obtain the chromatogram. 5. Calculate the calibration curve plotting the area under the peak against ingredient’s quantity.

3.7. Qualitative and Quantitative Analysis

1. After obtaining the chromatogram of the sample, identify each peak, comparing the retention time to that of the corresponding standard. An example of the chromatogram is shown in Fig. 1. 2. Calculate the content of each component using the peak area (Area Under Curve, AUC) of the unknown and the corresponding calibration curve.

Fig. 1. Chromatogram of a liposomal formulation containing phosphatidylcholine [1] and bioactive molecule [2]. The separation of the liposomal ingredients was achieved using the multiple development technique. The two subsequent mobile phases were: (a) CHCl3/CH3OH/d-H2O 45:25:5 (v/v) up to 5 cm (50% of the Chromarod-III), (b) Hexane/Diethyl ether 40:60 (v/v) up to 10 cm (100% of the Chromarod-III). After the separation of the ingredients and the elimination of the solvent vapor the Chromarods-III were scanned at the following conditions: scanning speed 30 s/scan, airflow 2000 L/h, and hydrogen flow 160 ml/min

Analysis of Liposome Components by HPTLC/FID

367

4. Notes 1. For enhancement of the component separation several solvent mixtures should be tried. The best technique suggested is the multiple development method in which two successive mobile phases are used: At first the samples are allowed to develop until 5 cm from zero point (50% of the rods) in Chloroform/ Methanol/d-Water 45:25:5 (v/v) in order to separate the phospholipids (2, 5, 6). Subsequently, the solvent is removed under hot air stream and the samples are placed in the second development tank containing either Hexane/Diethyl ether/ glacial acetic acid 80:20:2 (v/v) or Hexane / Diethyl ether 40:60 (v/v) until 10 cm (100% of the rods) for the separation of non-polar components from the phospholipids. 2. Place the samples in deep freezer for 20 min or immerse them in dry ice/iso-butanol bath until frozen. 3. Place cotton wool in Pasteur pipettes creating cotton filters of 1 cm height. Wash them with 1 ml chloroform twice. 4. The appropriate solvent selection is based on the sufficient solubility of the sample components. The solvent with lowest boiling point and polarity is preferred in order to produce narrow spots. 5. Hold the syringe in such a way to avoid scratching of the silica surface, allowing the sample drop to touch the surface of the Chromarod-III. The sample volume should be placed in small aliquots to produce a narrow spot. The sample spot should be as narrow as possible (less than 3 mm) to ensure good separation performance. 6. Wet the filter paper with the solvent immediately before starting the development in order to ensure complete solvent vapor saturation. 7. Do not allow the solvent front to exceed 100 cm from zero point because some separated components may be out of the scanning area. 8. If the solvent is not completely removed noise peaks will appear on the baseline signal and the results will not be reproducible.

References 1. Hatziantoniou S, Dimas K, Georgopoulos A, Sotiriadou N, Demetzos C (2006) Cytotoxic and antitumor activity of liposome-incorporated sclareol against cancer cell lines and human

colon cancer xenografts. Pharmacol Res 53(1): 80–7 2. Goniotaki M, Hatziantoniou S, Dimas K, Wagner M, Demetzos C (2004) Encapsulation

368

Hatziantoniou and Demetzos

of naturally occurring flavonoids into liposomes: physicochemical properties and biological activity against human cancer cell lines. J Pharm Pharmacol Oct 56(10):1217–24 3. Gabizon A, Papahadjopoulos D (1988) Liposomes formulations with prolonged circulation time in blood and enhanced uptake by tumors. Proc Natl Acad Sci USA 85:6949–6953 4. New RRC (1992) Characterization of liposomes. In: New RRC (ed) Liposomes: a practical approach. Oxford University Press, Oxford, pp 105–108 5. Hatziantoniou S, Demetzos C (2006) Qualitative and quantitative one step analysis of lipids and encapsulated bioactive molecules in liposome preparations by HPTLC/FID (Iatroscan). Journal of Liposome Research 16(4):321–330

6. Kaourma E, Hatziantoniou S, Georgopoulos A, Kolocouris A, Demetzos C (2005) Development of simple thiol-reactive liposome formulations, one-step analysis and physicochemical characterization. J Pharm Pharmacol 57(4):527–31 7. Henderson JR, Tocher DR (1992) Thin-layer chromatography. In: Hamilton RJ, Hamilton S (eds) Lipid Analysis: a practical approach. Oxford University Press, Oxford, pp 100–108 8. De Schriijver R, Vermeleulen D (1991) Separation and quantitation of phospholipids in animal tissues Iatroscan TLC/FID. Lipids 26(1):74–76 9. Madden TD, Boman N (1999) Lyophilization of liposomes. In: Janoff S (ed) Liposomes Rational Design. Marcel Dekker, Inc, N. York, pp 261–282

Chapter 25 Viscometric Analysis of DNA-Lipid Complexes Sadao Hirota and Nejat Düzgünes¸ Abstract DNA-cationic lipid complexes, “lipoplexes”, are used as gene carriers for molecular biology and gene therapy applications. Colloidal properties of lipoplexes can be determined by viscometric analysis. (1) The shape parameter of lipoplexes can be one of the factors affecting transfection efficiency; (2) the volume fraction of free liposomes remaining after lipoplex formation can be used as an index of purity of the lipoplex product; (3) the shear dependence of the viscosity of a diluted lipoplex suspension can be used as a macroscopic shape factor: (4) the attraction force parameter between particles can be a colloidal stability factor. These properties should be characterized and specified for process control of lipoplex production and quality control of lipoplex products. We describe an automated mini-capillary viscometer for a sample volume of 0.5 ml, and its application to the characterizations of lipoplexes. We show a procedure for viscosity measurements and provide a calculation using complexes of plant DNA-distearyldimethylammonium chloride (DDAC) at a charged ratio of 1:4 (−/+), in which the amount of DNA is less than 250 µg. The prolate/ellipsoidal axial ratio, a/b, was found to be 70. Determination of the shape parameter with a/b is found to be better than that with other shape parameters, e.g., a of the Sakurada equation, because fractionation of the particle size is not necessary. By the proposed method, colloidal parameters of lipoplexes and bioactive polymer complexes are characterized quantitatively. Key words: Shape parameter, Ellipsoid, Liposomes, Lipoplexes, Viscosity, Capillary viscometer, Quality control in large scale production

Symbols a Longer semi-diameter of ellipsoid (cm) b Shorter semi-diameter of ellipsoid (cm) c Concentration in molality (mol/kg) L Length of capillary (cm) k¢ Huggins coefficient k Attracting force parameter between particles k0k at zero shear P Pressure difference between both ends of capillary (dyn/cm2) r Distance from axis of capillary (cm) R Radius of capillary (cm) V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_25, © Humana Press, a part of Springer Science+Business Media, LLC 2010

369

370

Hirota and Düzgünes¸

Re Reynolds number S Shear stress (dyn/cm2) SR Shear stress at inner wall of capillary (dyn/cm2) S0 Yield stress (dyn/cm2) t Time (s) u Linear velocity (cm/s) at distance r from axis of capillary (cm/s) U Volume velocity, V/t, (ml/s) a Shape parameter of the Sakurada equation j Volume fraction, volume of particles / volume of suspension jav Average volume fraction ji Volume fraction of inner aqueous phase jnet Net volume fraction without inner aqueous phase jc Volume fraction of cationic liposomes mred Reduced viscosity, mred = hsp/c (L/g or L/mol) [m] Intrinsic viscosity, reduced viscosity at infinite dilution (L/g or L/mol) h Viscosity of sample liquid (poise) h0 Viscosity of solvent or suspending medium (poise) hrel Relative viscosity, hrel =h/ho hsp Specific viscosity, hsp =(hrel–1) hred Non-dimensional reduced viscosity, hred = hsp/j [h] Non-dimensional intrinsic viscosity, [h] = lim(j→0)hred

1. Introduction A capillary viscometer can provide an accurate measure of viscosity, and could be an excellent tool for quality control of liposomes and lipoplexes during large-scale production for clinical or other uses. However, it has seldom been used for this purpose. The reasons are that it requires a large amount of sample and that it necessitates long strenuous attention of a researcher at the flow time measurements. To address this problem, we have developed a mini-capillary viscometer for sample volumes of 0.5 ml with automated measurements (1). 1.1. Shape Parameter

Lipoplexes have different shapes (2) that contribute to the efficiency of gene delivery (3, 4). It is therefore important to characterize the shape of particular lipoplexes before preclinical or clinical studies are undertaken. The shape of lipoplex particles has been studied by electron microscopy, but this technique does not provide an average value of all the particles in the suspension. Moreover, sample preparation for electron microscopy can alter the shape of the particles. A viscometric analysis, however, gives an average value of the shape parameter of lipoplexes.

Viscometric Analysis of DNA-Lipid Complexes

371

Einstein (5) derived an equation for the viscosity of a dilute suspension of spherical, non-attracting particles:

h = h0(1 + 2.5j)

(1)

where h is the viscosity of the suspension, h0 is the viscosity of the suspending medium, and j the volume fraction of the suspended particles. Note that Eq. 1 does not include the particle size, and holds irrespective of the particle size or size distribution. Simha (6) has extended the Einstein equation to ellipsoidal, non-atracting particles as

h = h0 (1 + nj)

(2)

Where n is called the Simha factor, or shape parameter, which is 2.5 for spherical particles in the Einstein Eq. 1, and is invariably larger than 2.5 for ellipsoidal particles. Note that Eq. 2 also does not include particle size. It holds independent of the particle size or size distribution. Non-attracting conditions are attained when j approaches zero. From Eq. 2, n = l im(j →0)(hrel -1)/j = lim (j →1)hsp / j = lim(j →1)hred = [h] (2.1)

where hrel = h/h0: relative viscosity, hsp = hrel−1: specific viscosity, hred = hsp/j: non-dimensional reduced viscosity, [h]: non-dimensional intrinsic viscosity Here we see that the Simha factor or shape parameter, n, is equivalent to the non-dimensional intrinsic viscosity. Hirota (7) related n to the ellipsoidal axial ratio, a/b, in the range 1 < a/b < 100, for prolate (rod shaped) ellipsoids as,

n = 0.057(a/b)2 + 0.61a/b + 1.83

(3)

and for oblate (disc shaped) ellipsoids as

n = 0.001(a/b)2 + 0.59a/b + 1.90

(4)

The non-dimensional intrinsic viscosity as a function of the ellipsoidal axial ratio a/b is illustrated in Fig. 1. 1.2. Free Cationic Liposomes Remaining After Complexation

Oberle et al. (2) reported that “part of the initially formed lipoplexes are/remain unstable and eventually aggregate and/or merge further into larger complexes upon prolonged incubation.” They suggested that “this fraction may actually originate from vesicles that in the early phase of preparation still display surface-bound plasmid, or at least some uncovered part of DNA and that this fraction will be excluded from involvement of cellular transfection, as such a size will preclude cellular uptake. Thus, the shape parameter determination described above should

372

Hirota and Düzgünes¸

Nondimensional intrinsic viscosity

300

Prolate 200

100 80 60 40

Oblate

20 10

20

30

40

50

60

70

Ellipsoidal axial ratio, a/b Fig. 1. Non-dimensional intrinsic viscosity as a function of the ellipsoidal axial ratio, a/b

ideally be performed on properly formed lipoplexes and not extremely large complexes. It is also important to eliminate the uncomplexed cationic liposomes or to assess their volume fraction. During lipoplex production, a process test can be performed to confirm the absence of plain cationic liposomes in the final product. With ordinary cationic lipid materials, the density of lipoplexes is in the range 1.1–1.2, whereas the density of cationic liposomes is less than 1.0. Thus, it should be possible to separate the free liposomes from lipoplexes by differential centrifugation or density gradient centrifugation. As cationic liposomes are spherical, the volume fraction of free liposomes, jc, can be determined using Eq. 1 by measuring the relative viscosity, hrel = h/h0, of the supernatant after the centrifugation (see Note 1).

jc = (hrel – 1)/2.5

(1.1)

With an ideal product, jc should be zero. 1.3. Shear Dependence of the Viscosity of a Diluted Lipoplex Suspension

In a dilute uniform lipoplex suspension at about j <0.01, the viscosity sometimes decreases as the flow velocity increases. This results from the orientation of long rod-shaped particles in the direction of flow (8). As a dilute suspension of spherical particles

Viscometric Analysis of DNA-Lipid Complexes

373

does not show such a shear-dependence of viscosity, the shear dependence of viscosity is an indication of the shape of the particles. In capillary flow, the linear velocity, u, at a distance r from the axis of the capillary is

(

)

u = P R 2 − r 2 / 4Lh

(5)

where P is the pressure difference between both ends of the capillary, R is the radius, and L is the length of the capillary. Integrating u from r = 0 to r = R, the volume velocity, U, is given as U = pPR4/8hL

(6)

Equation (6) is known as Poiseuille’s law The shear rate at the inner capillary wall is

(du/dr)R = 4U/pR3

(7)

and shear stress at the inner capillary wall is SR = PR/2L

(8)

A flow profile in a capillary is shown in Fig. 2. Simha derived Eq. 2 assuming that particles are not oriented but uniformly directed as a result of rotational Brownian motion. This condition is attained when the shear rate approaches zero. Therefore, h/h0 should be determined at several shear rates, plotted against the shear rate and extrapolated to zero shear rate. The intercept at the ordinate gives the zero shear viscosity, which is introduced into Eq. 2.1 and nis determined. The h/h0 versus du/dr plot is not always a straight line. Then, it may be preferable

u = 0 at the wall Linear velocity of flow, u

(du/dr)R at the wall

Shear rate, du/dr

du/dr = 0 at the axis SR

Shear stress, S –R

0

R

S = 0 at the axis

Distance from capillary axis, r

Fig. 2. The flow profile in a capillary

374

Hirota and Düzgünes¸

Relative viscosity

3.0

2.0

1.0 100 200 Shear rate (1/sec)

300

Fig. 3. The shear dependence of viscosity

The head is reduced

Fig. 4. Tilting of the viscometer

Rotational Brownian motion of particles without flow

Orientation of particles in a flow

Fig. 5. The cause of the dependence of viscosity on shear

to adopt theh/h0 value at the lowest value of du/dr (Fig. 3) (see Note 2). For changing du/dr, the capillary viscometer is tilted and the head is reduced (Fig. 4). The cause of shear dependence of viscosity on shear is illustrated in Fig. 5. 1.4. Attractive Force Between Lipoplex Particles

Huggins (9) showed that reduced viscosity, mred = msp/c, increases with increasing concentration, c, as m red =

[m ]

+ k¢ [m ] c 2

(9)

Viscometric Analysis of DNA-Lipid Complexes

375

Non-dimensional reduced viscosity, hred

400 du/dr = 20 ~ 25 (1/sec)

du/dr = 50 ~ 70 (1/sec)

Attraction force parameter, k

350

Non-dimensional intrinsic viscosity, [η]

300

0

5 10 Volume fraction of lipoplex particles, φ x 104

15

Fig. 6. Dependence of the non-dimensional reduced viscosity on volume fraction

where [m] is the intrinsic viscosity and k’ is the Huggins constant, which indicates attraction between polymer molecules. In a nondimensional expression, we have

hred = [h] + k [h]2 j

(10)

where hred is the non-dimensional reduced viscosity, and [h] is the non-dimensional intrinsic viscosity. When hred is plotted against the volume fraction, j a straight line is obtained (Fig. 6). We see that Huggins’ equation can be extended to a lipoplex suspension. The slope, k, indicates attraction between particles, and is dependent on the shear rate. If the k versus (du/dr)R plot shows a straight line within a low range of the shear rate, k at zero shear is obtained by extrapolation. k0 = Lim (du/dx→0) k is a parameter of the attractive force between the particles, and may be specified as a colloidal stability factor of lipoplex suspensions (see Note 3).

2. Materials 1. Buffer: 10 mM Tris-HCl, pH 8.0 2. Distearyldimethylammonium chloride (DDAC) (Nikko Chemicals, Tokyo, Japan). 3. Plant DNA (from cucumber, 10–50 Kbp)

376

Hirota and Düzgünes¸

3. Methods 3.1. Preparing DNA Sample Used As Example for Shown Viscosity Data

1. Sonicate plant DNA, 10–50 Kbp (cucumber ) in 10 mM Tris-HCl buffer (pH 8.0) in a bath sonicator at 20 KHz, 200 W, at 10–40°C for 20 min. This procedure is expected to reduce the size of DNA to 0.5~2 Kbp, 2. Determine size of DNA by gel permeation chromatography, 3. Measure optical density at 260 and 280 nm 4. Calculate concentration of DNA, which in the example shown was found to be 0.57 mN base units. 5. Calculate the OD ratio of 260/280, which in the example shown was 1.75, indicating that the DNA purity is more than 95%.

3.2. Lipoplex Preparation

An excess of cationic lipid, 2.28 mN DDAC, which is converted to liposomes by Bangham method, in 10 mM Tris-HCl buffer (pH 8.0) and an equal volume of 0.57 mN DNA solution in 10 mM Tris-HCl buffer (pH 8.0) at a charge ratio of 1:4 (−/+) are mixed using a PLEXER reported previously (10): 1. Fill, a syringe with a DNA solution and another syringe with the lipid suspension. 2. Inject both liquids into the PLEXER at the same volume velocity of 1 ml/s. The two reaction liquids flow at the same velocity and impinge in a T-tube, connecting the two syringes at an angle of 180 degrees. A rapid stream creates turbulence in a chamber after the T-tube and is mixed uniformly. The sample is released from the instrument into a reservoir. When the volume velocity, U at each of the above syringes is 1 ml/s, the maximum shear rate (10) in the arms of the T-tubes, g¢ = 4U/pR3, is 4,000 (1/s) and the Reynolds number in the chamber with diameter of D, Re = DUr/hpr2, is 5,000. If U is controlled in the range between 0.5~2.0 ml/s, satisfactory mixing and homogenizing is attained without any danger of DNA strand breakage (12). The transit time of each liquid through the PLEXER is 0.8 s at U = 0.5 ml/s and 0.2 s at U = 2 ml/s.

3.3. Viscosity Measurements

The flow times for the sample suspension and the suspending medium to pass through the two reservoirs of the Auto Mini Capillary Viscometer (AMCV) are shown in Figs. 4 and 7. h = t1/t10 = t2/t20 = t3/t30 (Newtonian flow) hrel = t1 /t10 < t2/t20 (Plastic flow or orientation of particles)

(11) (12)

where t1 and t10 are the times for the sample suspension and the suspending medium, respectively, to pass through the first reservoir

Viscometric Analysis of DNA-Lipid Complexes

377

(a →b), and t2, t20 to pass through the the second reservoir (b→c). If the liquid is Newtonian, the second and the third terms of Eq. 11 give the same value (within a relative error of 0.3%). If two terms are not equal, the liquid is non-Newtonian and the flow is expressed by Eq. 12. In case of Newtonian flow, specific viscosity, hsp is

hsp = h rel – 1

(13)

When the difference between t1/t10 and t2/t20 is within 1%, the values are averaged for the calculation of hsp by Eq. 13. Such a slight shear dependence can be neglected and the sample can be regarded as a Newtonian fluid. When the difference between the values of t1/t10 and t2/t20 is more than 5%, the flow time measurements are repeated by tilting the capillary and reducing the head difference until the difference becomes within 5% (Fig. 4). Shear dependence of the viscosity is thus reduced. Without any effects of orientation, plasticity or dilatancy, both values become the same in a dilute suspension. When the difference between the values of [h] for a→b and b→c is within 5%, the value for b→c, where the shear effect is smaller, is adopted or the three are averaged (Simha calculated the axial ratio assuming that the rotational Brownian motion makes the direction of particle axis isotropic. This condition is attained when the shear effect diminishes in a low flow velocity). Flow-time measurements in the capillary viscometer are generally very tedious. This process is now automated as shown in Fig. 7. Passage of the meniscus through an orifice is detected by a laser beam, and the signal starts and stops a timer. The signal also starts an air pump to push up the sample liquid above the top reservoir and the measurements are repeated. The flow time data are sent to a laptop computer to determine the average flow time, t1/t10, t2/t20, their averages and coefficients of variation, (du/dr)1, (du/dr)2, (hred)1, (hred)2, etc. The coefficient of variation of the flow time (CV) seldom exceeds 0.1 %, and the data are very stable. When the CV exceeds 0.3%, the data are deleted. This fluctuation is often caused by the formation of a water-repellant layer and subsequent formation of tiny bubbles on the inner surface of the capillary. The viscometer is normally rinsed with water and heated at 500°C for 20 min in an oven before the measurements (see Note 4). 3.4. Determination of the Zero Shear Viscosity

The dependence of viscosity on shear rate is not always linear, because both an orientation effect and a plasticity effect influence the viscosity at the same time. If the viscosity data at different times fluctuate, the data are averaged and the average value is assumed as the zero shear viscosity. If the viscosity data decrease with increasing shear rate, but not in a linear manner, the viscosity at the lowest shear rate is adopted as the zero shear viscosity.

378

Hirota and Düzgünes¸

Laser source

Optical fiber

Optical fiber Orifice a

Laser sensor

Reservoir 1 Laser source

Laser source

Optical fiber

Optical fiber Orifice b

Optical fiber

Laser sensor

Reservoir 2 Optical fiber Orifice c

Laser sensor

η rel = t / t0 Electromagnetic valve for switching air pump on / off

relative viscosity

η sp = η rel – 1

Signals

Timer

specific viscosity

η red = η sp / φ

ηred

Extrapolation of ηred vs φ plot to zero φ

reduced viscosity

Samples with different φ

[η]

Display of a/b

Conversion to a/b

Fig. 7. Automation of flow-time measurements

3.5. Determination of the Non-Dimensional Intrinsic Viscosity

Non-dimensional intrinsic viscosity [h] is determined using Eq. 2.1. As there is very little inner aqueous phase in lipoplex, j can be regarded to be the net volume fraction,j net. Then, reduced viscosity, hred = hsp/jnet, is plotted against jnet. The jnet is calculated from the amount of DNA and lipids in the sample. Although rof the lipoplex is unknown, r = 1 may be used. According to the Huggins equation, the plot is on a straight line in the low hred region. When the obtained plots is not on a straight line, a least square line to the plots is drawn and extrapolated to jnet = 0. The intercept on the ordinate gives non-dimensional intrinsic viscosity (Fig. 6) (see Note 5).

3.6. Conversion of the Non-Dimensional Intrinsic Viscosity Into the Shape Parameter, a/b

The non-dimensional intrinsic viscosity is converted into the shape parameter, a/b (Fig. 1). An oblate ellipsoidal lipoplex, e.g., toroidal lipoplex, has not been reported. A prolate ellipsoid shape is usually assumed.

4. Discussion According to the Sakurada (11),

ln [h] = ln2.5 + a lnM

(14)

Viscometric Analysis of DNA-Lipid Complexes

379

When the shape parameter, a, equals 0 the particles are rigid spheres, while a = 2 means that the particles are straight filaments. Although Eq. 14 gives information on the shape of the particles, determination of a requires many measurements of the dependence of [h] on M. For this purpose, the sample needs to be fractionated and j should be determined for each fraction. The shape of the DNA-lipid complexes is also characterized by the axial ratio of ellipsoids using Eqs. 3 and 4. Since Eqs. 3 and 4 hold irrespective of the particle size, time consuming fractionations and determinations of j, for each fraction is not necessary. Viscosity data for a suspension of DNA-DDAC lipoplexes are shown in Table 1. From the obtained non-dimensional intrinsic viscosity value of 337, axial ratios of 70 as prolate ellipsoids, and 350 as oblate ellipsoids are determined. The latter value indicates very thin discs. Disc shaped lipoplexes have seldom been reported. A prolate ellipsoid with an axial ratio of 70 means, that the lipoplexes are fibrillar in shape, which is more probable than a thin disc. The formation of a lipoplex is thought to proceed via the following three steps (2):

(a) An electrostatic interaction between DNA and cationic lipid vesicles within 10 s.

(b) Coating of the initial complex with a lipid bilayer in hours.

(c) A formation of a globular complex wrapped with a lipid bilayer for days, at room temperature. The time for the solution to pass through the flow impinger is less than 1 s. Although this is enough for formation of complex A, it is too short for completing the formation of complex B. If the aggregates are dispersed in the solution with excess cationic lipid vesicles in some flow condition for a few hours, all the particles would become the final complex C. The shape of the lipoplex is thought to depend on the flow condition during the maturation. However, the relation between the flow condition and rheological shape parameter of the final lipoplex is not clarified at this time. A quantitative study to follow the time course of lipoplex maturation may be possible viscometrically using Eq. 3 and Eq. 4. In Table 1, the observed shear stress, Sobs, was much smaller than the calculated shear stress, Scal. This may be caused by the capillary force exerted on the meniscus of the sample liquid. Sobs should be used for obtaining the flow curve of du/dr versus S.

4.1. Concluding Remarks

A simple technique for the determination of shape of liposomes and DNA complexes is proposed using an automatic mini-capilllary viscometer. Non-dimensional intrinsic viscosity [h] is related to the axial ratio, a/b, of an ellipsoid which is applicable to low axial ratios (in a range of 1 < a/b < 100) by

Shear rate: du/dr = 4U/pR3 (1/s), Shear stress: Sobs = h/(du/dr)
Volume velocity: U = V/t (cc/s),

Volume of reservoir: Va→ b = 0.173 (cc), Va→c = 0.274, Vb→c = 0.101;

Capillary radius: R = 0.025 cm, Capillary length, L = 45 cm,

Temperature: 40.00 +/−0.01°C.

AV average, SD standard deviation, CV coefficient of variation %

hrel = t1/t10, t2/t20 and t3/t30 through the same reservoir, e.g. a→b, hsp = hrel −1, hred = hsp/j (non-dimensional).

t1, t2 and t3: flow time of the lipoplex suspension, from a to b (a→b), a→c and b→c;

Data: t10, t20 and t30: flow time (s) of 10 mM Tris-sHCl buffer;

Volume fraction of lipoplexes in the stock, j = 0.00142. The stock is diluted to 0.149, 0.297, 0.446 mN

DNA concentration in the stock solution is 0.57 mN in base units.

The complex has a FW of 2490 and a net density of 1.01.

(DDAC, FW 575.5), 2.28 mN in the stock suspension, at molar ratio of 1: 4, dissolved in 10 mM Tris-HCl buffer (pH 8.0).

Sample: plant DNA, sonicated to 0.5 to 2 kbp, is complexed with distearyldimethylammonium chloride, as liposomes

Table 1 Viscosity of a suspension of the DNA-DDAC complex determined by the auto-mini-capillary viscometer

380 Hirota and Düzgünes¸

a→b a→c b→c

a→b a→c b→c

7.4

11.1

2

3

4(stock) 14.2

a→b a→c b→c

a→b a→c b→c

3.7

1

a→b a→c b→c

0

Buffer

Sample j×104

248.88 665.05 416.17

226.37 604.76 378.39

202.49 539.73 337.24

179.32 478.24 298.92

249.12 664.7 415.58

226.89 605.88 378.99

202.29 539.36 337.07

179.41 479.11 299.7

159.08 423.63 264.55

2

1

158.88 423.55 264.67

time,

flow

249.02 664.22 415.2

227.39 606.94 379.55

202.36 539.73 337.37

179.14 478.57 299.43

159.14 423.65 264.51

3

t (s)

249.01 664.66 415.65

226.88 605.86 378.98

202.38 539.61 337.23

179.29 478.64 299.35

159.03 423.61 264.58

AV.

0.121 0.417 0.489

0.51 1.09 0.58

0.101 0.214 0.15

0.137 0.439 0.396

0.136 0.053 0.083

SD

0.048 0.063 0.118

0.225 0.18 0.153

0.05 0.04 0.045

0.077 0.092 0.132

0.086 0.012 0.031

CV

55.5 19.6

2.4

21.2

2.6 6.8

61.2

24.5

3 7.5

68.5

26.9

3.3 8.4

71.6

du/dr

9.5

U × 104

1.9

2.5

1.9

2.5

1.9

2.5

1.9

2.5

Scal

0.31

0.87

0.31

0.87

0.31

0.87

0.31

0.87

Sobs

1.5658 1.569 1.571

1.4266 1.4302 1.4324

1.2726 1.2738 1.2746

1.1274 1.1299 1.1314

hrel

0.5658 0.569 0.571

0.4266 0.4302 0.4324

0.2726 0.2738 0.2746

0.1274 0.1299 0.1314

hsp

398 401 402

384 387 389

368 370 371

344 351 355

hred

Viscometric Analysis of DNA-Lipid Complexes 381

382

Hirota and Düzgünes¸

n =[h] = 0.057(a/b)2 + 0.61a/b + 1.83 (for a prolate ellipsoid)

n =[h] = 0.001(a/b)2 + 0.59a/b + 1.90 (for an oblate ellipsoid)

An automation of the viscosity measurement with min-capillary viscometer is presented and its application to the characterization of DNA-lipid complexes is described. The viscometer unit is available from Hirota’s laboratory for researchers who wish to perform viscometric analysis. For DNA-lipid complexes, the Einstein viscosity equation is extended to a suspension of ellipsoids by introducing the non-dimensional intrinsic viscosity, which is related to the axial ratio, a/b. A mini-capillary viscometer which uses less than 0.5 ml of a sample suspension, at a volume fraction of less than 0.1%, is automated using laser sensors connected to optical fibers. The procedure for the viscosity measurements and calculation is shown by an example with plant DNA-DDAC complex at a charge ratio of 1:4 in which the amount of DNA is less than 250 microgram. The prolate ellipsoidal axial ratio, a/b, is found to be 70. Determination of the shape parameter with a/b is found to be better than that with a from the Sakurada equation, because fractionation of the particle size is not necessary. By the proposed method, the shape of a liposome-bioactive polymer complex is characterized quantitatively. The axial ratio of the DNA-lipid complex can be determined with less than 0.5 mg of DNA sample and corresponding lipid materials by an automatic mini-capillary viscometer. The proposed method is useful for characterizing liposomes and bioactive polymer complexes in research, in process control and in quality control during large-scale production.

5. Notes 1. A sample of the lipoplex suspension three days after DNAliposome complexation is centrifuged for 20 min at 3000 rpm and hrel is determined to be 1.0. Free cationic liposomes remaining are found to be zero, from jc calculated by Eq. 1.1. 2. In Fig. 6, the shear dependence of viscosity is found to be small. 3. From Fig. 6, k0 is found to be 42,000. 4. Viscosity is highly sensitive to temperature. The thermostat should be controlled within ±0.01°C. The capillary viscometer is immersed in the thermostat. The optical fibers connecting the orifices of the viscometer and sensors circuit beside the thermostat rustle in the water stream. These cause error signals of passages of meniscus at the orifices. In order to avoid the error signals, the optical fibers should be fixed in the water.

Viscometric Analysis of DNA-Lipid Complexes

383

Room light variations also cause an error. Avoiding a direct exposure of the orifices to the room light is necessary. 5. The flow times and viscosity data are shown in Table 1 with a lipoplex suspension without any fractionation. hred is plotted against j (Fig. 6). By extrapolating j→0, the non-dimensional intrinsic viscosity is determined to be 337. Corresponding axial ratios to 337 are obtained as 75 for a prolate ellipsoid and 350 for an oblate ellipsoid. The relative viscosity from the second reservoir, t2/t20, b→c, is always larger than that from the first reservoir, t1/t10, a→b. This is most likely caused by the orientation of particles in the direction of flow. The non-dimensional viscosity, 337, at du/ dr = 20~25 is adopted, because the orientation effect is thought to be smaller than the value at du/dr = 50~70.

Acknowledgments We wish to thank Dr. Y. Sun at the Harbin Institute of Technology (China), for preparation of the lipoplexes and the flow time measurements of viscosity, and Mr. Y. Takaoka at Tokyo Denki University for automation of the mini-capillary viscometer.

References 1. Sun Y, Li X, Düzgünes¸ N, Takaoka Y, Ohi S, Hirota S (2003) The shape parameter of liposomes and DNA-lipid complexes determined by viscometry utilizing small sample volumes. Biophys J. 85(2):1223–1232 2. Oberle V, Bakowsky U, Zuhorn IS, Hoekstra D (2000) Lipoplex formation under equilibrium conditions reveals a three-step mechanism. Biophys J 79(3):1447–1454 3. Felgner PL, Gadek TR, Holm M, Roman R, Chan HW, Wenz M, Northrop JP, Ringold GM, Danielsen M (1987) Lipofection: a highly efficient, lipid-mediated DNAtransfection procedure. Proc Natl Acad Sci USA 84(21):7413–7417 4. Sternberg B. (1998) Ultrastructural morphology of cationic liposome-DNA complexes for gene therapy. In: Medical Applications of Liposomes. Lasic DD, Papahadjopoulos D (Eds), 395. Elsevier, Amsterdam 5. Einstein A (1906) Ann Phys 1928:9

6. Simha R (1940) Viscosity and the shape of protein molecules. Science 92:132–133 7. Hirota S (2004) Viscometric determination of axial ratio of ellipsoidal DNA-lipid complex. Methods Enzymol. 375: 177–199 8. Tanford C (1961) Physical Chemistry of Macro molecules, Chapter 6 Transport properties, viscosity. John Wiley & Sons, New York 331–393 9. Huggins ML (1942) The viscosity of dilute solutions of long-chain molecules. IV. Dependence on concentration. J Am Chem Soc 64:2716–2718 10. Hirota S, de Ilarduya CT, Barron LG, Szoka FC Jr (1999) Simple mixing device to reproducibly prepare cationic lipid-DNA complexes (lipoplexes). Biotechniques 27(2):286–90 11. Sakurada I (1924) Kogyokagakuzasshi 38: 383–398 (in Japanese) 12. Lasic D, Strey H, Stuart MCA, Podgornik R, Frederik PM (1997) The structure of DNAliposome complexes. J Am Chem Soc 119(4):832–833

Chapter 26 Fluorometric Analysis of Individual Cationic Lipid-DNA Complexes Edwin Pozharski Abstract Lipoplex preparations are heterogeneous mixtures of lipoplex particles of different structures. As these structures determine the efficiency of the delivery of genetic material, it is important to characterize the distribution of particles of different types in lipoplex preparations with good statistics. We describe the application of flow fluorometry which allows producing such distributions (in terms of lipoplex particle size and composition) within minutes using basic flow cytometer (Anal Biochem 341:230–240, 2005). Key words: Flow cytometry, Lipoplex, Cationic lipid, Heterogeneity, Single particle detection, Size-stoichiometry distribution

1. Introduction Cationic lipids were first introduced in 1990s as a viable alternative to viral vectors for gene delivery (2). It was quickly established that their efficiency is relatively poor in terms of the probability of a plasmid DNA molecule entering a cell when in complex with cationic lipid (3). The possibility that transfection is associated with a small subpopulation of highly effective lipoplex particles makes it important to study the distribution of various types of particles in an intrinsically heterogeneous lipoplex preparation (4, 5). Flow cytometers appear to be perfectly suited for the task. Statistically valid amounts of data can be easily collected (up to 105 particles per minute) and individual lipoplexes characterized by size and composition. The difficulty is that lipid vesicles and lipoplexes do not scatter enough light to be detected by traditional

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_26, © Humana Press, a part of Springer Science+Business Media, LLC 2010

385

386

Pozharski

methods of flow cytometry. The method described here uses fluorescent labels for detection and characterization of particles.

2. Materials 2.1. DNA Labeling

1. If you study plasmid DNA, store it as 0.1 mg/ml solution at −70°C. Plasmid DNA can be thawed at room temperature. Storing in aliquotes sufficient for a single experiment is recommended in order to avoid multiple freeze-thaw cycles. DNA concentration can be determined spectrophotometrically by diluting the stock solution (for double-stranded DNA, e(260 nm) = 6600/M/cm in terms of phosphate charges). 2. Linear DNA is produced by shearing genomic DNA from various organisms. Herring, salmon and calf thymus DNA is available from Invitrogen (Carlsbad, CA). According to the manufacturer, most of DNA fragments are smaller than 2000 bp. 3. Ethidium homodimer-2 (EthD-2) is available from Invitrogen as 1 mM stock in DMSO (see Note 1).

2.2. Lipid Labeling

1. Cationic lipid stock solution is prepared in chloroform or methanol and stored at −70°C (see Note 5). 2. b-BODIPY FL C12-HPC (referred to as BODIPY-PC) is available from Invitrogen. Dissolve in chloroform or methanol and store at −70°C wrapped in aluminum foil. 3. Nucleopore filters, 200 nm pore size. Available from Sterlitect Corporation (Kent, WA). 4. Zwittergent 3–14 is available from Calbiochem (San Diego, CA).

3. Methods 3.1. Preparation of Labeled DNA

1. Prepare 5 mg/ml stock of EthD-2 in the buffer of your choice. Concentration can be verified spectrophotometrically (e(535 nm) = 8000/M/cm). Optical density of the stock solution is too low (~0.03) to be reliably measured on most spectrophotometers. Prepare ~50 mg/ml solution (based on concentration reported by manufacturer), measure absorbance at 535 nm and dilute appropriately. 2. Mix with DNA to achieve the labeling ratio of 60 bp per label molecule (see Note 2). (EthD-2 MW = 1293 g/mol).

Fluorometric Analysis of Individual Cationic Lipid-DNA Complexes

387

Assuming 0.1 mg/ml DNA and 5 mg/ml EthD-2 stock solutions, the DNA:label mixing volume ratio is ~1.53. Add enough buffer solution to achieve the desired final DNA concentration. 3. Incubate for 1 h at room temperature prior to data acquisition. If necessary, binding of EthD-2 to DNA can be monitored by fluorescence (label emission increases upon DNA binding) (see Notes 3 and 4). 3.2. Preparation of Labeled Lipid Vesicles

1. Mix cationic lipid and BODIPY-PC in such proportions that fluorophore constitutes 3% of the total lipid weight. Always use glass vials to prevent contamination due to dissolved plastic (see Notes 5 and 6). 2. Remove bulk solvent by gentle stream of argon. Use glass Pasteur pipettes inserted into tubing attached to the argon cylinder to direct the flow. Vial will be cold to the touch until all the solvent has evaporated. Continue for 1–2 min after no visible liquid remains. Place the vial into desiccator and subject to high vacuum for at least 2 h. 3. Rehydrate the lipid film with buffer of your choice. Briefly vortex until no more visible film is left on the wall.

3.3. Preparation of Extruded Lipid Vesicles

1. Prepare labeled lipid vesicles as described in Subheading 3.2. Set aside sufficient quantity for fluorescence emission measurements (see below). 2. Prepare extruded lipid vesicles using extrusion apparatus described in (6) (see Note 7). 3. To adjust the lipid concentration, measure BODIPY fluorescence emission intensity (lex = 507 nm, lem = 515 nm) for extruded sample treated with Zwittergent 3–14 and compare to the same measurement done with the vortexed preparation. Final lipid concentration suitable for these measurements is about 5 mg/ml (see Notes 8–10).

3.4. D ata Acquisition

1. These instructions are written for single-laser FACSCalibur from Becton Dickinson. First (FL1) and third (FL3) data channels are used to detect and quantify BODIPY and EthD-2 emission. 2. Prepare at least three lipoplex samples with excess lipid (e.g., 2:1, 4:1, and 8:1 charge ratios). These and the pure lipid vesicle suspension will be used for calibration. 3. Run blank buffer to determine the appropriate gain for the FL1 channel. Set primary threshold to FL1 and set it to zero. Use logscale for amplification, set gain to 1. Adjust voltage for the FL1 channel until less than 60 events are detected within 1 min of data acquisition (value of 400–500 is suitable for FACSCalibur).

388

Pozharski

4. Just prior to data acquisition, every sample should be diluted to produce 102–103 events per second (higher values may lead to dual events) (see Note 11). 5. Acquire data for the lipid vesicle suspension (without DNA). Some intensity will be detected in FL3 channel, since BODIPY has some residual emission above 650 nm. Adjust the voltage of FL3 channel and repeat if necessary (value of 400–500 is suitable for FACSCalibur) (see Note 12). 6. Acquire data for the rest of the samples including ones for calibration (see step 2). 3.5. Data Analysis

1. For calibration, calculate the total amount of fluorescence emission detected in FL1 (F1) and FL3 (F3) channels for the free lipid and three lipoplex samples made with excess lipid (see step 2 of data acquisition). Plot F3/F1 ratio versus DNA:lipid mixing ratio. Parameters obtained from linear fit to the following equation D:L = A+ B

F3 F1

can be used to convert f3/f1 ratios of individual lipoplex particles to DNA:lipid ratios. 2. The intensity emitted into FL1 channel by each detected particle is proportional to the amount of lipid it contains. To construct a “size distribution” for a lipoplex sample, use the following equation:

p1 ( x ) =

∑

f1

log (f1 / x ) <δ

∑ f1

This distribution is not based on the geometrical size of particles, but rather the amount of lipid they contain. This distribution is calculated on logarithmic scale. 3. The f3/f1 ratio for every particle is related to its stoichiometry and can be calculated using the conversion factors determined in step 1. The “stoichiometry” distribution of a lipoplex sample can be constructed using the following equation ∑ f1

p2 ( x ) =

f3 ⁄ f1 − x < δ

∑f

1

This distribution is on linear scale and the height of it is proportional to the amount of lipid included with particles of a particular type.

Fluorometric Analysis of Individual Cationic Lipid-DNA Complexes

389

Fig. 1. Example of the size-stoichiometry distribution of lipoplexes. Panels, left to right: pure vortexed O-Ethyldioleoylphosphatidylcholine (EDOPC) (7); equimolar complex with plasmid DNA; double lipid excess complex. Two lipoplex populations (DNA-coated vesicles and lamellar lipolexes) are observed in equimolar complex, whereas for excess lipid conditions only the latter remains

4. Similarly, the two-dimensional “size-stoichiometry” distribution can be constructed using the following equation:

∑

p2 D (x, y) =

log ( f1 / x ) < δ x ( f3 / f1 ) − y < δ y

f1

∑f

1

See Fig. 1 for an example of such distribution (see Note 13).

4. Notes 1. If DNA fluorescent label is dissolved in volatile organic solvent, it can be easily dried by a gentle stream of argon, followed by rehydration in a water-based buffer. EthD-2 is available as 1 mM solution in DMSO and is directly injected into the buffer. The amount of DMSO in the final preparation is small (~100 nM for 20 mM DNA preparation) and is unlikely to have any effect on lipoplex formation. 2. Do not exceed the labeling ration of one EthD-2 molecule per 60 bp DNA. This amount was determined to be sufficient to detect DNA presence in individual lipoplexes and compensates for approximately 3% of the negative charge carried by DNA. Further increase of the amount of label may affect the structure of lipoplexes. 3. Long-term storage of EthD-2-labeled DNA is not recommended. Labeled DNA should be prepared fresh few hours before data acquisition.

390

Pozharski

4. Using EthD-2 to label oligonucleotides was not tested and appears to be problematic. It is possible that for short oligonucleotides charge compensation will significantly affect lipid binding. Covalent labeling may be considered as an alternative. 5. Lipids are usually stored at −70°C in chloroform or methanol. Use glass vial and cap with teflon liner for long-term storage. Fill the vial with argon to protect lipid from oxidation. Before vesicle preparation, warm up the vial at room temperature to minimize the condensation. Always perform all operations involving organic solvents such as chloroform and methanol under a fume hood. 6. The number of particles detected per mol of lipid is constant if the labeling ratio by BODIPY is kept above 1%. The decrease of the number of detected particles for 0.5% labeling indicates that some vesicles will have fluorescence emission below threshold. For labeling ratios above 3% the overall fluorescence intensity starts to decrease due to self-quenching. 7. To reduce the loss of lipid during extrusion (recovery may be as low as 20%), reuse the same filter. For instance, if 1 mL of lipid vesicle suspension is to be extruded in four 250 ml steps, the lipid recovery improves from ~25% for the first aliquote to ~85% for the last. Additional lipid can be recovered by washing filters (after extruded material is collected, input syringe is filled with buffer and extrusion procedure repeated). 8. We found that the fluorescence intensity increases for samples treated with detergent to the degree which depends only on the labeling ratio by BODIPY-PC. Choice of detergent for this treatment should be considered carefully – popular Triton X-100, for instance, is a fluorescence quencher and should be avoided. 9. Labeled lipid vesicles may be stored at any temperature above freezing. Protect from light. Stability of vortexed preparations was not assessed, and freshly prepared samples (1–2 days old) were always used (primarily because they are easy to make). Extruded vesicles stored for more than a few days show the presence of larger particles, however, the size distribution returns to original after brief vortexing. This indicates that vesicles aggregate but do not fuse. 10. Lipoplexes prepared with excess lipid often precipitate upon centrifugation. 11. For lipid vesicle suspension (without DNA) the final concentration suitable for data acquisition is approximately 10 nM of lipid. Adjusting the flow rate of the instrument provides additional flexibility. Highly fused multilamellar lipoplexes

Fluorometric Analysis of Individual Cationic Lipid-DNA Complexes

391

may incorporate tens of individual vesicles and the concentration (or duration of data acquisition) should be adjusted accordingly. 12. Voltage for the FL3 channel should be small enough to prevent intensity overloads (some particles will register the maximum intensity and therefore the amount of DNA they contain will be impossible to determine). The exact setting is impossible to predict until all the samples are analyzed. This presents a problem since adjusting voltage of the FL3 channel would require that calibration runs are repeated. We use the “emission leak” from BODIPY to the FL3 channel to calibrate it properly. We found that overload problems can be avoided for most samples if the average intensity observed in the FL3 channel from lipid vesicle suspension (no DNA) is about 10% of the intensity observed in FL1 channel. 13. Standard software provided with a flow cytometer is not designed for the analysis of lipoplex particles. We have developed the set of programs in MATLAB computational environment for the data analysis. These scripts are available upon request.

References 1. Pozharski EV, MacDonald RC (2005) Analysis of the structure and composition of individual lipoplex particles by flow fluorometry. Anal Biochem 341:230–240 2. Felgner PL, Gadek TR, Holm M, Roman R, Chan HW, Wenz M, Northrop JP, Ringold GM, Danielsen M (1987) Lipofection: a highly efficient, lipid-mediated DNA-transfection procedure. Proc Natl Acad Sci U S A 84:7413–7417 3. Zabner J, Fasbender AJ, Moninger T, Poellinger KA, Welsh MJ (1995) Cellular and molecular barriers to gene transfer by a cationic lipid. J Biol Chem 270:18997–19007 4. Koynova R, Tarahovsky YS, Wang L, MacDonald RC (2007) Lipoplex formulation of superior efficacy exhibits high surface

activity and fusogenicity, and readily releases DNA. Biochim Biophys Acta 1768:375–386 5. Pozharski EV, MacDonald RC (2007) Single lipoplex study of cationic Lipoid-DNA, selfassembled complexes. Mol Pharm 4:962–974 6. MacDonald RC, MacDonald RI, Menco BPM, Takeshita K, Subbarao NK, Hu L-R (1991) Small-volume extrusion apparatus for preparation of large, unilamellar vesicles. Biochim Biophys Acta 1061:297–303 7. MacDonald RC, Ashley GW, Shida MM, Rakhmanova VA, Tarahovsky YS, Pantazatos DP, Kennedy MT, Pozharski EV, Baker KA, Jones RD, Rosenzweig HS, Choi KL, Qiu RZ, McIntosh TJ (1999) Physical and biological properties of cationic triesters of phosphatidylcholine. Biophys J 77:2612–2629

Chapter 27 Fluorescence Resonance Energy Transfer-Based Analysis of Lipoplexes Edwin Pozharski Abstract We describe three applications of the FRET technique to analysis of structural and thermodynamic properties of cationic lipids and their complexes with DNA. (1) Lipid mixing assay to determine the degree to which individual vesicles undergo fusion upon complex formation. (2) DNA binding assay to obtain cationic lipid-DNA binding curves for thermodynamic analysis or binding stoichiometry characterization. (3) DNA spacing assay to determine changes in DNA packing in the multilamellar lipoplex particle. Key words: FRET, Lipoplex, Cationic lipid, Lipid mixing, Vesicle fusion, Binding assay, DNA spacing

1. Introduction FRET is a powerful technique which allows monitoring structural changes and binding events for biological macromolecules in solution. Details of the underlying physical phenomena can be found elsewhere (1). Briefly, the fluorescence emission of a fluorophore (donor) is reduced when another fluorophore (acceptor) with suitable excitation wavelength is found in the vicinity. Events resulting in changes of average distances between fluorophores can be detected. In most cases, biological macromolecules themselves are not fluorescent and therefore fluorophore must be introduced by conjugation. Lipid bilayers and DNA molecules are suitable for FRETbased methods because they can be easily labeled. Indeed, labeling lipid bilayer just requires mixing in a small amount of lipid with

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_27, © Humana Press, a part of Springer Science+Business Media, LLC 2010

393

394

Pozharski

fluorescent marker (many are commercially available). There are many fluorescent molecules (e.g., ethidium bromide) which bind to DNA non-covalently and produce bright fluorescence. Thus, DNA is also easy to label (variety of conjugation strategies are available). Depending on the location of the donor and acceptor fluorophores, three different characteristics of lipoplexes can be assayed by FRET. 1. If both donor and acceptor are embedded in the lipid membrane, changes of the average surface area per lipid molecule in the bilayer can be determined. This parameter usually does not change upon lamellar lipoplex formation and is therefore of little interest. However, if the acceptor fraction in the bilayer changes (e.g., when labeled and unlabeled vesicles fuse) then the degree to which vesicles undergo fusion upon complex formation can be determined (2). 2. If lipid and DNA are both labeled with different fluorophores then cationic lipid-DNA binding can be detected (3). In this case DNA must be labeled covalently to ensure that fluorophore is not released upon lipid binding. 3. If DNA molecules are labeled with two fluorophores capable of FRET, average DNA-DNA distance (spacing) can be estimated upon formation of lipoplex. In this case, only acceptor molecule has to be bound covalently, because the degree of dissociation can be evaluated from control experiment in the absence of acceptor fluorophore (donor must have negligible fluorescence in free form). This approach can be used to monitor structural changes of lipoplex particles (4).

2. Materials 1. N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)-1,2-dihexadecanoylsn-glycero-3-phosphoethanolamine (NBD-PE) (Invitrogen, Carlsbad, CA). 2. Lissamine™ rhodamine B 1,2-dihexadecanoyl-sn-glycero-3phosphoethanolamine (Rh-PE) (Invitrogen, Carlsbad, CA). 3. YOYO-1 (Invitrogen, Carlsbad, CA). 4. Fluorescein/Cy3 DNA labeling LabelIT™ kits (Mirus Bio Corporation, Madison, WI). Follow manufacturer instructions for labeling procedure, including the adjustments necessary to reduce the labeling efficiency to about 300 bp/dye.

FRET-Based Analysis of Lipoplexes

395

3. Methods 3.1. Lipid Mixing Assay

1. Mix chloroform stocks of cationic lipid, Rh-PE and NBD-PE in 198:1:1 molar ratio in a glass vial (full fluorophore sample) (see Notes 2 and 3). 2. Remove bulk solvent with a gentle stream of argon. Desiccate the remaining material under high vacuum for 2 h. Rehydrate the lipid film in buffer solution and briefly vortex. Sample can be sonicated or extruded at this point depending on your experiment. See Note 1 regarding concentration adjustment if samples are extruded. 3. Repeat steps 1 and 2 for pure cationic lipid (no fluorophores) and diluted lipid (398:1:1 molar ratio of cationic lipid:RhPE:NBD-PE). 4. Mix equal volumes of pure lipid and full fluorophore sample, this is referred to as undiluted lipid. 5. For every lipoplex preparation to be studied for lipid mixing, mix DNA with diluted and undiluted lipid samples (see Note 4). 6. Measure fluorescence emission of NBD-PE (lex = 464 nm, lem = 531 nm). Following measurements should be obtained for each lipoplex preparation: F (undiluted lipid and DNA), F100% (diluted lipid without DNA), F0% (undiluted lipid without DNA). 7. Calculate the degree of lipid mixing (a) using the equation:

3.2. DNA Binding Assay

a=

F − F0% F − F100%

1. Mix an aliquot of cationic lipid stock solution with the appropriate amount of Rh-PE (both in chloroform) to give 3% w/w label incorporation. Remove bulk chloroform with a gentle stream of argon and place the mixture under high vacuum for 2 h. Rehydrate the resulting film with buffer solution and briefly vortex. Sample can be sonicated or extruded at this point depending on your experiment. Prepare the lipid vesicle suspension without fluorophore following the same protocol. 2. Prepare fluorescein labeled DNA according to manufacturer instructions (see Materials). Briefly, reactive label and DNA are incubated for 1 h at 37°C and unreacted/nonreactant dye removed by ethanol precipitation. Labeling efficiency can be adjusted by reducing the reactive dye concentration or incubation time. 3. Prepare lipoplexes with (a) labeled DNA and labeled lipid; (b) labeled DNA and unlabeled lipid, (c) unlabeled DNA and unlabeled lipid. Measure fluorescence emission of fluorescein

396

Pozharski

(lex = 494 nm, lem = 521 nm) for all three samples (Fa, Fb and Fc, respectively). The upper estimate of the fraction of unbound DNA (wupper) is calculated as follows (see Note 5) 3.3. DNA Spacing Measurements

w upper =

Fa − Fc Fb − Fc

1. Prepare Cy3 labeled DNA according to manufacturers’ instructions (see step 2 in Subheading 3.2). 2. To label DNA with YOYO-1, simply mix the fluorophore with DNA at the desired labeling ratio (see Note 6). Label both DNA and Cy3 labeled DNA (see Note 7). These samples are further referred to as donor-only and donor-acceptor DNA. 3. To determine the labeling efficiency by Cy3, measure absorbance of the Cy3-labeled DNA at 260 and 550 nm. Given the extinction coefficients of Cy3 (e(550 nm) = 150,000/M/cm) and DNA (e(260 nm) = 13,200/M/cm, per base pair), the labeling efficiency, Le, is given by (see Note 8)

L e = 11.4

OD 260 bp/dye. OD550

4. Measure YOYO-1 emission (lex = 491 nm, lem = 509 nm) for both donor-only (f0) and donor-acceptor DNA (f1). Förster radius, R0, can be estimated using the following equation and the labeling efficiency determined in step 2 (see Note 9)

R0 = 1.75L e log ( f 1 /f0 ) Å. 5. Prepare separate lipoplex samples with donor-only and donor-acceptor DNA. Measure YOYO-1 emission (F0 and F1, respectively). DNA spacing can be estimated by the following equation (see Note 10)

dD =

pR0 log ( f 1 /f0 ) 2 log (F1 /F0 )

Å.

4. Notes 1. Up to 75% of lipid can be lost upon extrusion due to adhesion of cationic lipids to Nucleopore filters. Adjust lipid concentration by measuring the fluorescence emission of detergent-treated sample (popular Triton X-100 should be avoided because of fluorescence quenching, use Zwittergent 3–14 instead). For unlabeled lipid, use phosphate assay (5).

FRET-Based Analysis of Lipoplexes

397

2. Alternative protocol for lipid mixing assay is based on preparing two lipid samples, containing only donor or only acceptor fluorophore. Upon lipid membrane fusion, FRET conditions are established and fluorescence emission of the donor fluorophore reduced. This approach, however, cannot distinguish between true fusion of lipid bilayers and vesicle aggregation (6). 3. Donor-acceptor pairs other than NBD-rhodamine can be used. There is no need for protocol modifications other than using different wavelengths for fluorescence measurements. 4. Lipoplexes prepared with excess lipid have a tendency to precipitate. While this is usually not a problem at the concentrations used for fluorescence measurements, the effect is easy to account for by measuring the direct fluorescence emission of the acceptor fluorophore (for Rh-PE, use lex = 557 nm and lem = 571 nm). 5. There is always residual donor emission in the DNA binding assay. With 3% of Rh-PE embedded in the lipid membrane, the average lateral distance between fluorophores is about 2 nm. DNA molecule binds to the surface of the bilayer and hence the acceptor fluorophore is at the average elevation of about 1 nm. Förster radius for fluorescein-rhodamine pair is ~5.5 nm, and thus the FRET efficiency must be almost 100%. 6. YOYO-1 carries four positive charges. It is therefore recommended to keep the YOYO-1/DNA labeling ratio as low as practical. With 60 bp per dye molecule, about 3% of DNA charge is compensated without much effect on lipoplex formation. 7. The recommended labeling ratio for the Cy3-labeled DNA used in DNA spacing measurements is 300 bp/dye molecule. Since about 35 bp will be within Förster radius (~6 nm), this leads to approximately 10% FRET efficiency for donor-acceptor DNA. When DNA is packed into a grid between lipid bilayers in lamellar lipoplex, the number of base pairs within a circle with Förster radius increases about threefold (assuming that individual DNA double strands are separated by 3 nm), thus leading to approximately 30% FRET efficiency. Significantly increasing the labeling efficiency could make the method insensitive to changes in DNA spacing. 8. The equations used to calculate the labeling efficiency and DNA spacing are derived based on the assumption that the FRET efficiency equals the probability to find an acceptor fluorophore within Förster radius from the donor. More sophisticated calculation which takes into account the distance dependence of the energy transfer cannot be represented as a simple formula, but gives remarkably similar results. We neglect the energy transfer across the bilayer in lamellar lipoplex because lamellar repeat in these structures is longer than the Förster radius (~6 nm for the YOYO-1/Cy3 pair).

398

Pozharski

9. Förster radius can be calculated from spectral properties of donor and acceptor fluorophore. Given the uncertainties associated with other parameters required for the calculation (geometric factor, donor quantum yield and local refractive index), the result should be treated with caution. 10. The DNA spacing measurement protocol is applicable only to the conditions when no free DNA is present. Presence of free DNA increases the apparent FRET efficiency, leading to artificial increase in DNA spacing. If DNA spacing versus DNA:lipid ratio diagram is attempted, the minimum will be observed where lipoplex is saturated with DNA. To obtain a complete curve, independent measurements of the fraction of free DNA (e.g., by ethidium bromide displacement) is required to apply the appropriate correction. References 1. Lakowicz JR (2004) Energy transfer. Principles of fluorescence spectroscopy, 3rd edn. Springer, New York, pp 443–476 2. Kennedy MT, Pozharski EV, Rakhmanova VA, MacDonald RC (2000) Factors governing the assembly of cationic phospholipidDNA complexes. Biophys J 78:1620–1633 3. Pozharski E, MacDonald RC (2003) Lipoplex thermodynamics: determination of DNAcationic lipoid interaction energies. Biophys J 85:3969–3978

4. Madeira C, Loura LMS, Prieto M, Fedorov A, Aires-Barros MR (2008) Effect of ionic strength and presence of serum on lipoplexes structure monitorized by FRET. BMC Biotechnol 8:20 5. Bartlett GR (1959) Phosphorus assay in column chromatography. J Biol Chem 234:466–468 6. Duzgunes N, Allen TM, Fedor J, Papahad jopoulos D (1987) Lipid mixing during membrane aggregation and fusion: why fusion assays disagree. Biochemistry 26:8435–8442

Chapter 28 Analysis of Lipoplex Structure and Lipid Phase Changes Rumiana Koynova Abstract Efficient delivery of genetic material to cells is needed for tasks of utmost importance in the laboratory and clinic, such as gene transfection and gene silencing. Synthetic cationic lipids can be used as delivery vehicles for nucleic acids and are now considered the most promising nonviral gene carriers. They form complexes (lipoplexes) with the polyanionic nucleic acids. A critical obstacle for clinical application of the lipid-mediated DNA delivery (lipofection) is its unsatisfactory efficiency for many cell types. Understanding the mechanism of lipid-mediated DNA delivery is essential for their successful application, as well as for a rational design and synthesis of novel cationic lipoid compounds for enhanced gene delivery. A viewpoint now emerging is that the critical factor in lipid-mediated transfection is the structural evolution of lipoplexes within the cell, upon interacting and mixing with cellular lipids. In particular, recent studies showed that the phase evolution of lipoplex lipids upon interaction and mixing with membrane lipids appears to be decisive for transfection success: specifically, lamellar lipoplex formulations, which were readily susceptible to undergoing lamellar-nonlamellar phase transition upon mixing with cellular lipids and were found rather consistently associated with superior transfection potency, presumably as a result of facilitated DNA release. Thus, understanding the lipoplex structure and the phase changes upon interacting with membrane lipids is important for the successful application of the cationic lipids as gene carriers. Key words: Cationic lipid, DNA, Lipoplex, Phase transition, Membrane lipid, Lipofection, Gene delivery

1. Introduction Important therapeutic procedures, such as gene transfection and gene silencing, require efficient delivery of genetic material to cells. Synthetic cationic lipoids, which form complexes (lipoplexes) with polyanionic DNA, are presently the most widely used non viral gene carriers (1). A critical obstacle for clinical application

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_28, © Humana Press, a part of Springer Science+Business Media, LLC 2010

399

400

Koynova

of lipid-mediated DNA delivery (lipofection) is its unsatisfactory efficiency. Progress in enhancing lipofection efficacy has been impeded because its mechanism is still largely unknown (2–8). Understanding the mechanism of lipid-mediated DNA delivery is essential for its successful application, as well as for rational design and synthesis of novel cationic lipoid compounds for enhanced gene delivery. Polar lipids are known for their ability to form an impressive variety of polymorphic and mesomorphic phases – lamellar and nonlamellar – when dispersed in aqueous media (9, 10). Phase state has been recognized as potentially important for the transfection activity of lipoplexes (11–14). Typically, lipoplexes are arranged as multilayer structures in which DNA is intercalated between the lipid bilayers (15–17). Some earlier studies suggested that the inverted hexagonal phase leads to more efficient transfection efficiency than does the lamellar phase (11, 13). Recent experiments dispute this suggestion, however, and there is considerable evidence against a direct general correlation between lipoplex structure and transfection efficiency (18–24). Furthermore, an emerging viewpoint is that the critical factor in lipid-mediated gene delivery is the structural evolution of lipoplexes upon interacting and mixing with cellular lipids (25–27). The unbinding of DNA from cationic lipid carrier when lipoplex gets inside the cell has been identified as one of the key steps in lipofection. According to the current understanding, the unbinding is a result of charge neutralization by cellular anionic lipids; indeed, experiments have revealed that addition of negatively charged liposomes to lipoplexes results in dissociation of DNA from the lipid (25, 28–31). A set of significant recent findings suggest that the structure of cationic lipid carriers changes dramatically upon interaction with cellular lipids, and furthermore that such changes may critically affect the delivery efficiency. Recent studies showed that the phase evolution of lipoplex lipids upon interaction with membrane lipids appears to be decisive for transfection success: thus, lamellar lipoplex formulations, which were readily susceptible to undergoing lamellar-nonlamellar phase transitions upon mixing with cellular lipids, were found rather consistently associated with superior transfection potency, presumably as a result of facilitated DNA release (24, 27, 32). Noteworthy is that such a concept can, in principle, also account for the considerable differences in the transfection potency of lipoplexes with different cells. Thus, protocols for studying the structure of lipoplexes and the lipid phase changes upon interacting with membrane lipids, in search for correlations with the lipoplex transfection efficiency, are summarized herein. Small-angle X-ray diffraction is used as a primary method.

Lipoplex Structure and Phase Changes

401

2. Materials 2.1. Lipids

At present, a growing list of synthetic lipid transfection reagents is commercially available. Each one has its own success on specific cell line. The present chapter is mainly focused on a particularly attractive cationic lipid class, O-substituted phosphatidylcholine derivatives, in which the phosphate oxygen of phosphocholine is substituted (generating a triester), thus converting the head group zwitterion into a cation (33, 34). These lipids are especially attractive because they are slowly metabolized and have remarkably low toxicities (28, 34). Representatives of this cationic phospholipid class have been found to exhibit high transfection activity, both in vitro and in vivo, in antitumor and anti-cystic fibrosis pharmaceuticals (35–40).

2.1.1. Cationic Phospholipids

Triflate salts were synthesized as described (34), or were purchased as the chloride salt from Avanti Polar Lipids (Alabaster, AL): 1. Ethyl-phosphatidylcholines: 1,2-dioleoyl-sn-glycero-3-ethylphosphocholine (EDOPC), 1,2-dilauroyl-sn-glycero-3ethylphosphocholine (EDLPC), 1,2-dimyristoyl-sn-glycero-3ethylphosphocholine (EDMPC), 1,2-dipalmitoyl-sn-glycero3-ethylphosphocholine (EDPPC), 2-distearoyl-sn-glycero-3ethylphosphocholine (EDSPC), 1,2-diphytanoyl-sn-glycero3-ethylphosphocholine (EDPhyPC), 1-oleoyl-2-decanoyl-snglycero-3-ethylphosphocholine (C18:1/C10-EPC), 1-stearoyl2-decanoyl-sn-glycero-3-ethylphosphocholine (C18:0/ C10-EPC), and 1,2-dierucoyl-sn-glycero-3-ethylphosphocholine (DiC22:1-EPC) – this class of cationic phospholipids has been extensively characterized and found to form lamellar lipoplexes, which exhibit from high to superior transfection activity; 2. 1,2-dioleoyl-sn-glycero-3-stearylphosphocholine(C18-DOPC), 1,2-diphytanoyl-sn-glycero-3-palmitylphosphocholine (C16-DiPhyPC) – these effectively triple-chain lipids exhibit strong hexagonal-phase preferences (24); 3. 1,2-dioleoyl-sn-glycero-3-hexylphosphocholine (C6-DOPC), 1 , 2 - d i d e c a n o y l - sn- g l y c e r o - 3 - o c t y l p h o s p h o c h o l i n e (C8-DiC10PC), 1,2-didecanoyl-sn-glycero-3-myristylphosphocholine (C14-DiC10PC) form bilayer cubic phases at physiological temperatures (24); 4. DiC22:1-EPC and C14-DiC10PC exhibit large temperature hysteresis and could form either lamellar or cubic phases at room or physiological temperature, depending on the thermal prehistory (24). These compounds thus provide a remarkable opportunity to, with minimal ambiguity, examine the effect of the initial lipid phase on lipoplex properties.

402

Koynova

5. The recent approach of fine tuning lipoplex properties by using mixtures of cationic lipids (24, 41, 42) provided the opportunity to highly improve the transfection activity by properly formulating mixtures of cationic lipids. This procedure has been found to provide up to 30× increase in transfection as a consequence of some specific properties of mixtures (see below). 2.1.2. Other Cationic Lipoids

Ditetradecyldimethylammonium-bromide (diC14DAB), dioctadecyldimethylammonium-bromide (diC18DAB) (Sigma, St. Louis, MO), dimyristoyltrimethylammonium-propane (DMTAP) (Avanti Polar Lipids, Alabaster, AL).

2.1.3. Membrane Lipids

Model membrane lipid systems: Interactions of lipoplexes with cellular membranes can be approached starting from simple lipid systems (for initial understanding of the basics of the lipoplex/ membrane lipid interactions), with progression to complex mixtures modeling membrane lipid compositions and native membrane extracts, as follows: 1. Anionic, synthetic: dioleoyl-derivatives of phosphatidylglycerol (DOPG), phosphatidylserine (DOPS), phosphatidic acid (DOPA); tetraoleoylcardiolipin (TOCL); from natural sources: brain phosphatidylserine (PS), bovine liver phosphatidylinositol (PI), heart cardiolipin (CL). 2. Zwitterionic, synthetic: dioleoylphosphatidylethanolamine (DOPE), cholesterol, lysophosphatidylcholine; recent data suggest that negatively charged lipids are not a sine qua non for DNA release from lipoplexes (43); in a sense, this is consistent with the concept about the paramount importance of phase properties/ intrinsic curvature/bilayer frustration – thus the potency of zwitterionic lipids for releasing DNA could be tested. 3. Mixtures of anionic/zwitterionic lipid (phosphatidylcholine): anionic lipids listed above can be mixed with dioleoylphosphatidylcholine (DOPC) at, e.g., 8:2 ratio – the approximate common zwitterionic/anionic lipid ratio of plasma and nuclear membranes; endosomes are reported to have composition similar to that of the plasma membrane (44, 45). 4. Cholesterol is known as strong modulator of membrane biophysical properties; 20% cholesterol could be added to the above mixtures to mimic its content in nuclear and plasma membranes. 5. Phosphatidylethanolamine is a major nonlamellar forming membrane lipid: ~20% addition to mimic the plasma/endosome membrane content could be tested. 6. Membrane mimicking composition (MM = DOPC:DOPE:DOPS: Chol, 45:20:20:15, w/w) was found to be a very efficient DNA releaser (46).

Lipoplex Structure and Phase Changes

403

7. All listed lipids are from Avanti Polar Lipids (Alabaster, AL), used without further purification. 8. Other membrane constituents could be tested for their effect on the interaction with lipoplex lipids: sphingomyelin (SM) (15–20% in plasma membrane and lysosomes, respectively). Free fatty acids, mono- and diglycerides are minor membrane constituents but are able to strongly modulate membrane properties, including intrinsic curvature and phase preferences, so their effect at ~5% could also be tested. Formation of membrane rafts (47, 48) can be generated in model membranes with compositions of the proper proportions of sphingolipids and sterol. This could enable testing for a possible role of the cell membrane liquid-ordered raft structures in lipoplex internalization and DNA delivery. 9. Acyl chain chemical structure (length, saturation) of these membrane-mimicking compositions could be also varied, to correspond to natural membranes: palmitoyl (16:0), stearoyl (18:0), linoleoyl (18:2), arachidonoyl (20:4), docosahexaenoyl (22:6) derivatives could be included, and chain asymmetry could also be considered. Acyl chain length and unsaturation are known as strong modulators of membrane properties (49); often overlooked, acyl chains were recently found to be potent modulator of transfection outcome too (24). 10. Natural lipid extracts: bovine liver, brain and heart total/polar lipid extracts (Avanti Polar Lipids, used without further purification) have been found very effective in releasing DNA from lipoplexes, and prone to form nonlamellar phases as well (46). 2.1.4. Lipid Storage

2.1.5. Purity Tests

Lipids were stored as chloroform solutions in glass containers with teflon-lined closures at −20°C, under argon. Efforts were made to use lipid chloroform solutions within 1 year of production. Saturated-chain lipids were also stored as lyophilized powders, as supplied, at the same conditions: in a glass container with a teflon closure at −20°C, under argon (see Note 1). Lipid concentrations of stock solutions were determined using a phosphate assay (50), as described in Subheading 3.10, and verified by analytical ultra-microbalance weighing. 1. Cationic lipids synthesized at the MacDonald laboratory (Northwestern University, USA) were tested by mass spectrometry and proton NMR. 2. All lipids were regularly checked by thin-layer chromatography and migrated as a single spot. 3. Microcalorimetric scans were also used as purity tests: the diluted dispersions of the saturated species showed highly cooperative chain-melting phase transitions at temperatures in agreement with published values.

404

Koynova

2.2. DNA and Other Reagents

1. Herring sperm DNA (sheared; shorter than 2 kbp according to the manufacturer) from Invitrogen (Carlsbad, CA), 10 mg/ ml solution in water, was used for preparation of lipoplexes to be used in the structural and thermodynamic experiments (X-ray diffraction, calorimetry), during the initial exploratory research stage (see Note 2). 2. CMV-b-galactosidase plasmid purchased from Clontech Laboratories Inc. (Palo Alto, CA) and propagated and purified by Bayou Biolabs (Harahan, LA) was used for cell transfection and for final verification of structural results. b-gal plasmid was used also in those techniques requiring small amount of material (FRET, Dynamic Light Scattering (DLS)), for conclusions verification. 3. DNA concentration was checked by absorbance at 260 nm, assuming that e = 6,600 M/cm in terms of phosphate charges. 4. Phosphate-buffered saline (PBS): 50 mM phosphate buffer, 100 mM NaCl, pH 7.2.

2.3. Phosphate Assay

1. Sulfuric acid 2. Perchloric acid 3. Sodium phosphate, monobasic 4. Ammonium molybdate 5. ANSA reagent (solution of 0.625 g aminonaphthalenesulfonic acid, 37.5 g sodium metabisulfite, and 2.5 g sodium sulfite in 1.5 ml distilled water)

3. Methods 3.1. Liposome and Lipoplex Preparation

1. Aliquots of cationic lipids or their mixtures were placed in borosilicate glass tubes, freed of bulk chloroform with a gentle stream of argon, and kept under high vacuum for 12–24 h (typically overnight) to remove any solvent residues. Complete removal of the solvent residues was verified by analytical ultramicrobalance weighing. 2. Subsequently, samples were hydrated with phosphate-buffered saline, PBS (50 mM phosphate buffer, 100 mM NaCl, pH 7.2) overnight at room temperature and vortex-mixed for several minutes at a temperature above their gel–liquid crystalline phase transition. 3. Thereafter, several cycles of freezing-thawing between dry ice and room temperature, accompanied by vortexing during the thawing steps, were applied for homogenization.

Lipoplex Structure and Phase Changes

405

4. DNA/lipid dispersions for X-ray diffraction experiments were prepared by adding an aqueous DNA solution to the dry lipid film at the desired ratio, and immediately vortexing (51) (see Note 3). Assuming an average nucleotide molecular weight 330, the lipid/DNA weight ratios corresponding to isoelectric lipoplexes are 2.9:1 for the most commonly used cationic phospholipid, EDOPC. 5. Samples were equilibrated for 1–2 days before measurements. 6. X-ray diffraction samples: The lipid concentration of the dispersions was 10–30 wt% (see Note 4). Samples were filled into glass capillaries (d = 1.0 or 1.5 mm) (Charles Super Co., Natick, MA), using an IEC International Clinical Centrifuge (International Equipment Co., Needham Hts., MA), and flame-sealed (see Note 5). 3.2. X-ray Diffraction

X-ray diffraction is the primary method that we use for the lipoplex structure analysis: for identification of phases formed in lipid/lipoplex dispersions, for following the phase changes, as well as for determination of the structural parameters. 1. Small-angle X-ray diffraction (SAXD) measurements were performed at stations 5IDD (DND-CAT) and 18D (BioCAT), Advanced Photon Source, Argonne National Laboratory, using 12 keV X-rays. 2. Exposure times were typically ~0.5–1 s. 3. Data were collected using 2D 2048 × 2048 MAR CCD detector (79 × 79 µm pixel size) at a sample-to-detector distance of ~200 cm. 4. Spacings were determined from axially integrated 2D images using the FIT2D program (52), and silver behenate (AgBeh) as calibration standard. 5. Temperature protocols were executed directly on samples mounted on the beam line. 6. The repeat period of the lamellar phase (d-spacing) was determined from the position of the first-order reflection in the diffraction pattern; the lattice parameter of the hexagonal phase a = 2d/Ö3 was calculated from the position (d) of the (10) diffraction peak; cubic phases are identified according to the ratio of the reciprocal spacings of their diffraction maxima (53). 7. A temperature-controlled (Linkam thermal stage) capillary sample holder was used. All measurements were started at 20°C. The sample holder was mounted on a motorized stage, and, by moving it with respect to the incident beam, it was possible to ensure that the patterns recorded were representative of the whole sample volumes.

406

Koynova

8. Readings of a thermocouple placed adjacent to the capillary in the sample holder were recorded simultaneously with the diffraction patterns. 9. Some samples with longer exposure time were checked by thin layer chromatography after the experiments. Products of lipid degradation were not detected in these samples (see Note 6), and radiation damage of the lipids was not evident neither from their X-ray patterns, nor from subsequent microcalorimetric scans. 3.3. Differential Scanning Calorimetry

Differential scanning calorimetry (DSC) is the most widely used method for studying lipid phase transitions. It is applied here for constructing phase diagrams of cationic lipid mixtures. 1. High-sensitivity microcalorimetric measurements were performed using a VP-DSC Microcalorimeter (MicroCal Inc., Northampton, MA) (54). 2. Heating and cooling scans were performed at scan rates of 0.5°C/min (4 s filtering). 3. Thermograms were analyzed using MicroCal Origin software. The onset and the completion temperatures of the phase transitions needed for the construction of phase diagrams were determined by the intersections of the peak slopes with the baseline on the thermograms. 4. The maxima of the heat capacity curves were taken as the phase transition temperatures of the pure components. For asymmetric or split line shapes of the composite aggregates, the transition temperature was defined as that temperature at which the area of the endotherm was divided into equal halves.

3.4. Fluorescence Resonance Energy Transfer

In order for DNA to be released from lipoplexes and enter the cell nucleus where it is transcribed, the cationic lipid electrostatic charge must be neutralized. DNA unbinding from cationic lipoplexes is considered a result of charge neutralization by cellular anionic lipids. Thus, intermixing of membrane lipids with lipoplex lipids is a necessary step in transfection. To quantify this mixing process (see Note 7), we used a fluorescence resonance energy transfer (FRET) assay involving a pair of fluorescent lipid dyes. 1. Cationic liposomes were prepared containing 0.5% NBD-PE and 0.5% rhodamine-PE (Rh-PE) (Molecular Probes, Eugene, OR) (25). 2. The lipid concentration of the dispersions was 0.1 mM. 3. Negatively charged liposomes were prepared at the same lipid concentration, without fluorescent labels. 4. Labeled cationic liposomes were placed in an AlphaScan fluorometer (Photon Technology International, Princeton, NJ), and treated with unlabeled negatively charged lipids at 37°C.

Lipoplex Structure and Phase Changes

407

5. Fluorescence intensity was recorded as a function of time with Ex = 320 nm, Em = 535 nm, at 37°C. 6. Fluorescence was also measured of samples in which negatively charged lipids were mixed directly with cationic lipids in chloroform and then liposomes were prepared as above; this intensity was then used for normalization of measurements (see Note 8). 3.5. Dynamic Light Scattering

Lipoplex size is important for transfection efficiency, with the bigger (within limits) lipoplexes being considered generally more efficient (55–57). Even though the size of lipid aggregates can be modified by the preparation protocol (e.g., extrusion, sonication), it is also obviously dependent on the lipid physicochemical properties. In particular, lipid phase state is well known as a modulator of the colloidal properties of lipid dispersions, including the aggregate size. The size of lipid aggregates was monitored by DLS. 1. Cationic lipid dispersions in PBS were prepared at 50 µg/ml. 2. DNA (1 mg/ml in PBS) was added to generate lipoplex samples at the desired lipid/DNA ratio. 3. Measurements were performed with a Brookhaven Instruments BI-200SM goniometer and BI-9000 digital correlator (Brookhaven, NY). 4. Measurements were carried out at 37°C. 5. Borosilicate glass, 250 µl, 3 × 30 mm, flat bottom tubes were used. 6. Delay times of 10 µs–1 s were examined. 7. The correlation data were fitted with quadratic cumulants, using the algorithm provided with the instrument. 8. At least 10 correlation curves with collection times long enough to provide good statistics were collected and averaged, thus giving an estimate of the experimental uncertainty.

3.6. Lipoplex Structure Analysis 3.6.1. Lamellar Lipoplexes, Lipid and DNA Spacings

SAXD patterns revealed that the lipoplexes formed by the majority of the cationic phospholipids are arranged in lamellar arrays, as shown by the sets of sharp reflections in the diffraction patterns (Fig. 1). It is important to note that even cationic lipids forming nonlamellar structures per se, often form lamellar lipoplexes (19). The lamellar repeat period of the lipoplexes is typically ~1.5 nm higher than that of the pure lipid bilayers, as a result of the DNA intercalation between the lipid layers. The presence of the DNA strands between the phospholipid bilayers has been verified by the electron density profiles of the lipoplexes (28). In addition to the sharp lamellar reflections, a low-intensity diffuse peak was also present in the lipoplex diffraction patterns. Such a peak has been interpreted as reflecting the in-plane

408

Koynova

Fig. 1. Small-angle X-ray diffraction profiles of: (a) EDOPC and EDLPC/EDOPC (6:4) lipoplexes, at 4:1 lipid/DNA weight ratio (arrow points to the peaks originating from the DNA–DNA in-plane correlation); (b) EDLPC/EDOPC lipoplexes at different lipid composition and 6:1 lipid/DNA weight ratio; in order to magnify the DNA diffraction peaks, the sharp lipid lamellar reflections are truncated. Diffraction data were collected for 1 s at 37°C. (Reproduced with permission from (42); copyright (2007), Elsevier)

packing of the DNA strands intercalated between the lipid lamellae (15, 16, 58). Its position is dependent on the lipid-DNA stoichiometry. 3.6.2. Columnar DNA Superlattice in Gel Phase Lipoplexes

DNA arranges into rectangular columnar superlattice between lipid bilayers in the low-temperature gel phase of the lipoplexes of the saturated cationic lipids (51, 59). This is evidenced by two or three diffuse reflections in addition to the sharp lamellar reflections; these are attributed to the DNA ordering both within the smectic layers between the lipid bilayers, as well as across bilayers, from one DNA layer to another (Fig. 2) (51). The positions of these reflections in the lipoplexes of the cationic O-ethylphosphatidylcholines index to a centered rectangular columnar lattice, Shk = √[(h/a)2 + (k/b)2], with lattice constants a and b; the constant b is determined by the lamellar spacing d (b = 2d). The DNA scattering peaks index as (1,1) and (1,3), respectively (and in some cases (1,5)).

3.6.3. Gel–Liquid Crystalline Lipid Phase Transition in Lipoplexes

Aqueous dispersions of the saturated-chain representatives of the O-ethylphosphatidylcholines exhibit gel–liquid crystalline phase transitions at temperatures close to those of their parent phosphatidylcholines; addition of DNA does not change significantly the lipid phase transition parameters (17, 28). Thus, upon heating, the lipid layers of the lipoplexes undergo a phase transition from

Lipoplex Structure and Phase Changes

409

Fig. 2. Small-angle X-ray diffraction profiles of the gel-phase ethyl-PC/DNA samples at different stoichiometries. Data for lipids with three different chain lengths are shown. In all samples, the diffuse DNA reflections index as a centered rectangular columnar lattice, depicted in the drawing in the inset. The X-ray patterns are registered at ~15°C below the transition temperature of the lipids. (Reproduced with permission from (51); copyright (2004) American Chemical Society)

the interdigitated gel phase to a liquid crystalline lamellar Lca phase, at temperatures close to those of the pure lipid dispersions: 23°C, 41°C, and 52°C, respectively, for the EDMPC, EDPPC, and EDSPC complexes. The transition temperatures are independent of the amount of DNA. This transition is associated with an expansion of the lamellar repeat distance d by 4–6 Å (Fig. 3). The lipid phase transformation was immediately followed by a rearrangement of the DNA lattice, specifically, a decrease of the DNA in-plane distance by ~1 Å and a loss of the interlamellar correla-

410

Koynova

Fig. 3. (a) Small-angle X-ray diffraction patterns recorded from EDSPC/DNA (2:1) upon heating from 30°C to 58°C at 1°C/ min. The phase transition from the interdigitated gel phase (with a short repeat distance) to the noninterdigitated liquid crystalline phase takes place at ~52–53°C. An intermediate state with broader lamellar reflections at a larger repeat period can be seen. Suggested representations of the molecular arrangements are shown on the right side; (b) Spacings of the lipoplex lamellar repeat period d and the DNA arrangement dDNA in the EDSPC/DNA (2:1) system upon heating. The lipid phase transition precedes the DNA rearrangement, and includes an intermediate, disordered state of large lamellar repeat distance; the DNA interstrand distance remains typical for the gel phase in that intermediate state (due to the scaling, the DNA diffraction peaks cannot be seen in panel A). (Reproduced with permission from (51); copyright (2004) American Chemical Society)

tion, the latter revealed by the disappearance of the multiple DNA reflections from the SAXD patterns. Thus, at temperatures above the lipid phase transition, a single broad DNA reflection, attributed to the in-plane, strand-to-strand positional correlation of the DNA strands, was seen.

Lipoplex Structure and Phase Changes 3.6.4. Hexagonal Phase Lipoplexes

411

Certain cationic phospholipids such as C18-DOPC, C16DiPhyPC, C6-DOPC (effectively triple-chained) form lipoplexes substantially different from the lamellar lipid/DNA sandwiches described above. Their structure consists of DNA coated by cationic lipid monolayers and arranged on a two-dimensional hexa gonal lattice. This arrangement is identified by their small-angle X-day diffraction pattern, with diffraction maxima of reciprocal spacings fitting the ratio 1:√3:√4 (Fig. 4). The lower intensity of the diffraction peaks from the 11 and 20 planes in the presence of DNA relative to the patterns without DNA (Fig. 4a) is a likely result of the higher electron density of DNA relative to water (60).

Fig. 4. SAXD patterns of (a) C18-DOPC cationic lipid dispersion (black) and lipoplexes (red) at 37°C; lipid/DNA 4:1 w/w (Reproduced with permission from (24); copyright (2006) Biophysical Society); (b) C6-DOPC cationic lipid dispersion (black) and lipoplexes (red) at 37°C; lipid/DNA 4:1 w/w (R. Koynova, unpublished data); top: cartoon representation of the aggregate morphology

412

Koynova

It is thus an indication of the presence of DNA in the core of the hexagonal phase cylinders, i.e., for the formation of columnar inverted hexagonal phase lipoplexes (as shown in the cartoon on Fig. 4). Remarkably, hexagonal phase lipoplexes formed by cationic phospholipids consistently exhibited lower transfection activity than the lamellar ones (24). 3.6.5. Lipoplexes with Phase Coexistence

In certain lipoplexes made of cationic lipid mixtures, coexistence of lamellar and hexagonal phase is observed. Such are some C18DOPC/EDOPC and C16-DiPhyPC/EDOPC mixtures (Fig. 5). Noteworthy, these are compositions consistently corresponding to a minimum in the transfection efficiency vs. composition curves (24). Although the diffraction patterns of these mixtures clearly show the presence of two phases, structural ambiguities remain since it is hard to specify the topological modes of phase coexistence. For example, La and HII phases might coexist either in separate aggregates or in the same aggregate, and the distinction might be critical for DNA unbinding and release. Supplemental information

Fig. 5. Coexisting lamellar and inverted hexagonal phases in lipoplexes of EDOPC/C16DiPhyPC 60:40 (a), and EDOPC/C18-DOPC 60:40 (b) at 37°C as revealed by their SAXD patterns; (c) cartoon representation of the suggested aggregate morphology. (Reproduced with permission from (24); copyright (2006) Biophysical Society)

Lipoplex Structure and Phase Changes

413

on the lipoplex organization could be hence sought through size measurement; it is specific for the HII phases to be highly aggregated and hence their dispersions tend to have large particles. Really, the pure C18-DOPC lipoplexes, which are HII structures (see above), were significantly larger than those of the lamellar phase EDOPC lipoplexes; in contrast, the C18-DOPC/EDOPC lipoplexes with intermediate lipid compositions were similar in size to those of EDOPC (24). These observations, as well as the fact that the coexisting La and HII phases are virtually epitaxially matched (aBhexB = 2dhex /Ö3 » dBlam; Fig. 5) has been interpreted as suggesting that the mixed-phase lipoplexes may have a topology like that depicted in Fig. 5. In addition to the topological relationship between the two phases, another question that arises with respect to samples having two coexisting phases is whether both phases contain DNA. Analysis of the structural parameters shows that the presence of DNA in both phases seems likely: the expanded d-spacing of the lamellar phase clearly indicates the presence of DNA; the structural characteristics of the hexagonal arrays in Fig. 5 suggest that they also contain DNA: the lattice parameter of the hexagonal-phase arrays a = 2d/Ö3 = 6.5 nm is identical to that of the C18-DOPC lipoplexes, which, as indicated by the relative intensities of the X-ray diffraction peaks, contain DNA (see above). 3.7. Lipoplex Lipidsmembrane Lipids Mixing – Kinetics, Equilibrium

These experiments provide critical background information on the structural basis of lipoplex-cellular membrane interactions. X-ray diffraction is the primary method that we use for the identification of phases formed in lipid/lipoplex dispersions, as well as for determination of their structural parameters (25, 27, 32, 61, 62).

3.7.1. Kinetics of Phase Changes upon Interaction of Lipoplexes with Cellular Lipids

This is followed by time-resolved X-ray diffraction. The high brilliance of the X-rays from a synchrotron source markedly increases sensitivity and provides the possibility to follow fast processes in real time at high resolution. This approach has, for example, been extremely useful in characterizing the mechanism and kinetics of lipid phase transitions and we have recently applied it to characterize the reorganization of lipoplexes during such transitions (51). According to our experiments, the time scale for DNA release from lipoplexes is of the order of minutes (25). Since SAXD patterns can be recorded on the millisecond to second time scale when using a synchrotron source, it is possible to monitor the kinetics of phase changes after admixing of anionic lipid dispersions to lipoplex dispersions.

3.7.2. Phase Preferences of the Mixtures of Cationic Lipids with Cellular Lipids

The phase preferences are determined by X-ray diffraction (static measurements). These measurements are a useful complement to, and a reference point for, the kinetic measurements described above.

414

Koynova

Since lipid aggregates and lipoplexes are large, their interdiffusion is slow, and the complexes can undergo time-dependent internal changes for a long time subsequent to the initial encounter, hence static measurements will provide information on the limiting state available to these samples. The phase preferences of the mixtures of cationic with anionic lipids were, in fact, found to closely correlate with their potency in releasing DNA from lipoplexes. Lipid mixtures are prepared in two ways: (1) standard, by mixing lipids as chloroform solutions (this protocol is believed to produce equilibrium mixtures), as well as (2) by mixing pre-formed lipid aqueous dispersions. This procedure is closer to the in vivo conditions and it also provides better control for the lipoplex-cellular lipid interaction experiments, which are carried out in aqueous media. 3.8. Structural Evolution of the Carrier Lipid/DNA Complexes upon Interaction with Cellular Lipids

Correlation between delivery efficiency of the cationic lipid DNA carriers and the mesomorphic phases they form when interacting with negatively charged membrane lipids was established. Specifically, formulations that are particularly effective DNA carriers, form phases of highest negative interfacial curvature when mixed with anionic lipids, whereas less effective formulations form phases of lower curvature. Structural evolution of the carrier lipid/DNA complexes upon interaction with cellular lipids was hence suggested as a controlling factor in lipid-mediated DNA delivery.

3.8.1. Synergy in Lipofection by Cationic Lipid Mixtures: EDLPC/ EDOPC Mixture

For quasi-equilibrium experiments, mixed EDLPC/EDOPC cationic lipid lipoplexes and anionic lipid (TOCL, DOPG or DOPS) dispersions were prepared separately, and the latter were added to the lipoplexes. The dispersions were then equilibrated for 1–2 days, filled into glass capillaries and flame-sealed. For kinetic measurements, lipoplexes and anionic lipid dispersions were mixed in a capillary and immediately mounted on the synchrotron X-ray beamline for measurement. Diffraction patterns were recorded once per minute. Both the cationic lipid dispersions and the lipid/DNA complexes of the EDLPC/EDOPC mixture have been found to arrange into lamellar arrays at all compositions; the same is valid for the anionic lipid dispersions. When anionic lipids interact with the lipoplexes however, nonlamellar phases were formed: thus, EDLPC/EDOPC 60:40 (the composition that exhibited particularly high transfection efficiency (41)) formed the inverted micellar cubic phase, Fd3m, while both components and mixtures of other compositions all formed phases of lower curvature (Fig. 6). The micellar cubic phase exhibits even higher interfacial curvature than the inverted hexagonal phase (10). In the case of the anionic cardiolipin, it was identified by 11 diffraction maxima with reciprocal spacings fitting the ratio: √8:√11:√12:√16:

Lipoplex Structure and Phase Changes

415

Fig. 6. Correlation between the transfection efficiency (red bars) of EDLPC/EDOPC lipoplexes (41) (quantified by expression of b-galactosidase in human umbilical artery endothelial cells) and the phase structure of their mixtures with the anionic lipid cardiolipin (CL). Cartoons of the phases are arranged vertically to give a qualitative notion about the increase of the interfacial curvature. Top row: diffraction patterns of mixtures of EDLPC/EDOPC lipoplex dispersions of different compositions with CL dispersions. The charge ratio of the cationic/anionic lipids in the samples is 1:1. Similar patterns were recorded also for mixtures of EDLPC/EDOPC with the anionic phosphatidylglycerol. (Reproduced with permission from (27); copyright (2005) American Chemical Society)

√24:√32:√35:√36:√48:√51:√56, characteristic for the cubic aspect #15 (53). The anionic lipid DOPG similarly formed micellar cubic phase when mixed with the most efficient EDLPC/EDOPC 6:4 formulation. Another anionic membrane lipid, DOPS, was found to form the micellar cubic phase transiently when mixed extempore with the EDLPC/EDOPC 6:4 formulation, as revealed by our kinetic synchrotron X-ray experiments (Fig. 7).

416

Koynova

Fig. 7. Kinetics of phase changes in the EDLPC/EDOPC lipoplexes of the highest transfection activity (6:4 wt/wt (41)), upon addition of the anionic lipid DOPS, in the time interval 10–30 min after the addition (as indicated on the right side of the diffraction patterns): bilayer cubic Pn3m phase forms initially in parallel to the lamellar phase; later, micellar cubic Fd3m could be distinguished. The charge ratio of the cationic/anionic lipids in the sample is 1:1. (Reproduced with permission from (27); copyright (2005) American Chemical Society)

3.8.2. C18/C10-EPCs

Another persuasive example of a correlation of the phase properties of the carrier lipid/cellular lipid mixtures and transfection success is provided by the C18:1/C10-EPC and C18:0/C10-EPC cationic phospholipids, the former one being 50× more effective as a DNA transfection agent (HUAEC cells) than the latter, despite their similar chemical structure and virtually identical lipoplex organization (63). A likely reason for the superior effectiveness of C18:1/C10-EPC relative to C18:0/C10-EPC was suggested by the phases that evolved when these lipoids were mixed with negatively-charged membrane lipid formulations. The saturated C18:0/C10-EPC remained lamellar in mixtures with a biomembrane-mimicking lipid formulation DOPC/DOPE/DOPS/Chol 45:20:20:15, w/w; in contrast, the unsaturated C18:1/C10-EPC exhibited a lamellar-nonlamellar phase transition in such mixtures, which took place at physiological temperatures, ~37°C (Fig. 8).

3.9. Synergy in Lipofection by Cationic Lipid Mixtures: Superior Activity at the Gel–Liquid Crystalline Phase Transition

Mixtures of two cationic lipids were found to be able to deliver therapeutic DNA considerably more efficiently than do the separate molecules. In an effort to rationalize this widespread “mixture synergism,” the phase behavior of the cationic lipid mixtures exhibiting solid–liquid crystalline phase transition has been examined and their binary phase diagrams constructed. Among a group of more than 50 mixture formulations, the compositions with maximum delivery activity resided unambiguously in the solid– liquid crystalline two-phase region at physiological temperature.

Lipoplex Structure and Phase Changes

417

Fig. 8. SAXD patterns of mixture of C18:1/C10-EPC with MM (MM = DOPC:DOPE: DOPS:Chol, 45:20:20:15, w/w), recorded upon temperature heating-cooling scan at 1°C/min. (Reproduced from (32); copyright (2007) National Academy of Sciences, USA)

Samples of cationic lipids mixtures for calorimetric experiments were hydrated in PBS overnight at room temperature and vortexmixed at a temperature above their gel–liquid crystalline phase transition. Samples were subsequently equilibrated overnight at room temperature before DSC measurement. Phase diagrams were constructed from the onset and the completion temperatures of the phase transitions, determined by the intersections of the peak slopes with the baseline on the DSC thermograms (Fig. 9). The range of compositions exhibiting phase coexistence at physiological temperature was identified by the intersection of the 37°C temperature isotherm with the solidus and liquidus lines of the phase diagram. 3.9.1. DiC14DAB/ diC18DAB Mixture

Thus, according to the temperature-composition phase diagram of the diC14DAB/diC18DAB binary mixture, compositions with £45 mol% diC18DAB are in the liquid crystalline phase whereas mixtures with >90 mol% diC18DAB are in the solid (gel) phase at physiological temperature (37°C). Mixtures with 45–90 mol% diC18DAB reside within the solid–liquid crystalline phase coexistence region, between the solidus and liquidus lines of the phase diagram; thus, solid and liquid crystalline domains are expected to coexist in the lipid bilayers of these samples at 37°C. At these compositions, lipoplexes were found to exhibit maximum transfection efficiency (63).

3.9.2. Other Cationic Lipid Mixtures

Similarly, superior transfection activity was established for sets of composition residing within the solid–liquid crystalline two-phase region also for the mixtures of: EDMPC/EDPPC, EDOPC/ diC14DAB, EDOPC/diC18DAB, EDOPC/DMTAP (63).

418

Koynova

Fig. 9. (a) Heating thermograms of dispersions of diC14DAB/diC18DAB mixtures at different ratios. (b) Relationship of transfection activity to phase diagram. Left Y-axis and lines and symbols apply to the phase diagram, which was constructed from the calorimetric data. Squares denote transition onset and circles transition end. The right Y-axis and bars depict the transfection efficacy TE of each composition compared to the average efficacy of all studied compositions: [(TE–)/] × 100; the shaded region indicates compositions that reside within the solid–liquid crystalline two-phase region at 37°C. Inset: Transfection of HUAECs by diC14DA/diC18DAB mixtures. (Reproduced with permission from (63); copyright (2007) American Chemical Society)

Lipoplex Structure and Phase Changes

3.10. Phosphate Assay

419

1. For digestion of unknowns to obtain inorganic phosphorus: Dry unknowns in glass tubes; add 3 drops of concentrated H2SO4: 70% HClO4 (2:1, vol/vol) from a Pasteur pipette and heat above 250°C in high-temperature block for 3 h (sample should be colorless when digestion is complete). 2. Preparation of standards: Add 0, 5, 10, 20, 30, 40 µl of 10 mM NaH2PO4 (50–400 nmol PO4) to test tubes; add 3 drops of concentrated H2SO4: 70% HClO4 (2:1, vol/vol) from a Pasteur pipette. 3. Color reaction: To unknowns and standards add 1.8 ml 0.5% ammonium molybdate and mix. Next, Add 10 µl of ANSA reagent and mix. Let samples stand for 20 min at room temperature. Read at 660 nm.

4. Notes 1. When transferring a portion of lipid from a container that has been stored in a freezer, it is important to allow it to reach room temperature before opening the vial in order to avoid water condensation. Glass syringes need to be used for chloroform solutions. 2. Since plasmid DNA is usually expensive, and structural methods such as X-ray diffraction require substantial amount of material, our experience showed that it is reasonable to use the less expensive, sheared herring sperm DNA during the initial exploratory experiments, and to use the plasmid only at the final stage, for verification purposes. The differences with respect to lipid structures and phase behavior were only minor to none. Another way to verify results is with parallel experiments by using small-amount of material requiring techniques such as FRET and DLS. 3. Various protocols for DNA/lipid mixing for the formation of lipoplexes have been reported in the literature. Details from these protocols such as order of addition, vortexing, etc., have been found important for the characteristics of the lipoplexes (56, 64). With respect to structural experiments, our experience showed that addition of DNA solution to the dry lipid film followed by immediate vortexing and subsequent equilibration consistently resulted in very well ordered lipoplexes. At the same time, this way prepared lipoplexes exhibited the same transfection activity as those prepared in the “conventional” way, by adding DNA solution to a preformed liposome suspension. 4. The lipid concentrations used in the SAXD experiments (~50 mM) are certainly considerably higher than bulk concentrations in

420

Koynova

the cell, however, they may be similar to the local concentrations in the cell (i.e., at membrane–membrane contacts). In any case, their relevance to physiological concentrations with respect to phase data obtained has been repeatedly checked by control experiments at low concentrations (see, e.g., (15)). 5. During temperature scans to high temperatures, it is possible to have water evaporation through some micropores formed upon sealing the capillaries. To control for such possible leaks, checking the weight of the capillaries after preparation and before and after the experiments by using an analytical ultramicrobalance was found useful. 6. Using synchrotron radiation requires careful precautions for avoiding radiation damage to samples. We used strongly attenuated X-ray beam by at least 16 (usually 32–48) 20 µm aluminum foils, and applied protocols rationalized for minimization of sample exposure to radiation. At the end of temperature scans, requiring longer cumulative exposures, sample was usually moved with respect to the beam and the diffraction pattern was recorded from a nonirradiated portion, in order to verify the lack of radiation damage. 7. Although usually referred to as a “fusion assay,” strictly, lipid mixing is what is measured by the FRET experiments and, indeed, lipid mixing is what is required for neutralization of the lipoplex lipid. In principle, monomer lipid exchange between aggregates could produce generally similar FRET results; it is usually a slow process, but since charged lipids exhibit higher solubility in water, it still could be important. Still, we emphasize that the fusion experiments described here involve oppositely charged lipid aggregates (as do the lipoplexmembrane interactions), in which fusion is activated by electrostatic attraction. Fusion of such aggregates has been clearly and repeatedly visualized (43, 65–67). 8. For normalization, the fluorescence of the lipid dispersion in the presence of detergent, such as Triton X-100, has been routinely used (68). We consider, however, the system in which the two kinds of lipids have been mixed in chloroform and then made into liposomes as a more adequate control (as 100% fusion) for this fusion/mixing assay.

Acknowledgment The author highly appreciates the expert advice of Professor Robert C. MacDonald. The present work was supported by NIH grant CA119341 (Center for Cancer Nanotechnology Excellence at Northwestern University).

Lipoplex Structure and Phase Changes

421

References 1. Felgner PL, Ringold GM (1989) Cationic liposome-mediated transfection. Nature 337:387–388 2. Lobo BA, Vetro JA, Suich DM, Zuckermann RN, Middaugh CR (2003) Structure/function analysis of peptoid/lipitoid: DNA complexes. J Pharm Sci 92:1905–1918 3. Gaucheron J, Wong T, Wong EF, Maurer N, Cullis PR (2002) Synthesis and properties of novel tetraalkyl cationic lipids. Bioconjug Chem 13:671–675 4. Niculescu-Duvaz D, Heyes J, Springer CJ (2003) Structure-activity relationship in cationic lipid mediated gene transfection. Curr Medicinal Chem 10:1233–1261 5. Subramanian M, Holopainen JM, Paukku T, Eriksson O, Huhtaniemi I, Kinnunen PKJ (2000) Characterisation of three novel cationic lipids as liposomal complexes with DNA. Biochim Biophys Acta Biomembr 1466:289–305 6. Jaaskelainen I, Sternberg B, Monkkonen J, Urtti A (1998) Physicochemical and morphological properties of complexes made of cationic liposomes and oligonucleotides. Int J Pharm 167:191–203 7. Song YK, Liu F, Chu SY, Liu DX (1997) Characterization of cationic liposome-mediated gene transfer in vivo by intravenous administration. Human Gene Ther 8:1585–1594 8. Ghosh YK, Visweswariah SS, Bhattacharya S (2002) Advantage of the ether linkage between the positive charge and the cholesteryl skeleton in cholesterol-based amphiphiles as vectors for gene delivery. Bioconjug Chem 13:378–384 9. Luzzati V, Tardieu A (1974) Lipid phases – structure and structural transitions. Annu Rev Phys Chem 25:79–94 10. Seddon JM, Templer RH (1995) Polymor phism of lipid-water systems. In: Lipowsky R, Sackmann E (ed) Handbook of biological physics. Elsevier Science, Amsterdam, pp 97–160 11. Koltover I, Salditt T, Radler JO, Safinya CR (1998) An inverted hexagonal phase of cationic liposome-DNA complexes related to DNA release and delivery. Science 281:78–81 12. Zuhorn IS, Hoekstra D (2002) On the mechanism of cationic amphiphile-mediated transfection. To fuse or not to fuse: Is that the question? J Membr Biol 189:167–179 13. Smisterova J, Wagenaar A, Stuart MCA, Polushkin E, ten Brinke G, Hulst R, Engberts JBFN, Hoekstra D (2001) Molecular shape of

the cationic lipid controls the structure of cationic lipid/dioleylphosphatidylethanolamineDNA complexes and the efficiency of gene delivery. J Biol Chem 276:47615–47622 14. Simberg D, Danino D, Talmon Y, Minsky A, Ferrari ME, Wheeler CJ, Barenholz Y (2001) Phase behavior, DNA ordering, and size instability of cationic lipoplexes – relevance to optimal transfection activity. J Biol Chem 276:47453–47459 15. Radler JO, Koltover I, Salditt T, Safinya CR (1997) Structure of DNA-cationic liposome complexes: DNA intercalation in multilamellar membranes in distinct interhelical packing regimes. Science 275:810–814 16. Lasic DD, Strey H, Stuart MCA, Podgornik R, Frederik PM (1997) The structure of DNA-liposome complexes. J Am Chem Soc 119:832–833 17. Koynova R, MacDonald RC (2003) Cationic O-ethylphosphatidylcholines and their lipoplexes: Phase behavior aspects, structural organization and morphology. Biochim Biophys Acta Biomembr 1613:39–48 18. Simberg D, Danino D, Talmon Y, Minsky A, Ferrari ME, Wheeler CJ, Barenholz Y (2003) Phase behavior, DNA ordering and size instability of cationic lipoplexes: Relevance to optimal transfection activity. J Liposome Res 13:86–87 19. Rakhmanova VA, McIntosh TJ, MacDonald RC (2000) Effects of dioleoylphosphatidylethanolamine on the activity and structure of O-alkyl phosphatidylcholine-DNA transfection complexes. Cell Mol Biol Lett 5:51–65 20. Congiu A, Pozzi D, Esposito C, Castellano C, Mossa G (2004) Correlation between structure and transfection efficiency: A study of DC-Chol-DOPE/DNA complexes. Colloids Surf B Biointerfaces 36:43–48 21. Caracciolo G, Pozzi D, Caminiti R, Castellano AC (2003) Structural characterization of a new lipid/DNA complex showing a selective transfection efficiency in ovarian cancer cells. Eur Phys J E 10:331–336 22. Caracciolo G, Caminiti R (2005) Do DC-Chol/DOPE-DNA complexes really form an inverted hexagonal phase? Chem Phys Lett 411:327–332 23. Ross PC, Hensen ML, Supabphol R, Hui SW (1998) Multilamellar cationic liposomes are efficient vectors for in vitro gene transfer in serum. J Liposome Res 8:499–520 24. Wang L, Koynova R, Parikh H, MacDonald RC (2006) Transfection activity of binary mixtures

422

Koynova

of cationic O-substituted phosphatidylcholine derivatives: The hydrophobic core strongly modulates their physical properties and DNA delivery efficacy. Biophys J 91: 3692–3706 25. Tarahovsky YS, Koynova R, MacDonald RC (2004) DNA release from lipoplexes by anionic lipids: Correlation with lipid mesomorphism, interfacial curvature, and membrane fusion. Biophys J 87:1054–1064 26. Zuhorn IS, Bakowsky U, Polushkin E, Visser WH, Stuart MCA, Engberts JBFN, Hoekstra D (2005) Nonbilayer phase of lipoplex-membrane mixture determines endosomal escape of genetic cargo and transfection efficiency. Mol Ther 11:801–810 27. Koynova R, Wang L, Tarahovsky Y, MacDonald RC (2005) Lipid phase control of DNA delivery. Bioconjug Chem 16:1335–1339 28. MacDonald RC, Ashley GW, Shida MM, Rakhmanova VA, Tarahovsky YS, Pantazatos DP, Kennedy MT, Pozharski EV, Baker KA, Jones RD, Rosenzweig HS, Choi KL, Qiu RZ, McIntosh TJ (1999) Physical and biological properties of cationic triesters of phosphatidylcholine. Biophys J 77:2612–2629 29. Xu YH, Szoka FC (1996) Mechanism of DNA release from cationic liposome/DNA complexes used in cell transfection. Biochemistry 35:5616–5623 30. Zelphati O, Szoka FC (1996) Mechanism of oligonucleotide release from cationic liposomes. Proc Natl Acad Sci USA 93:11493–11498 31. Ashley GW, Shida MM, Qiu R, Lahiri MK, Levisay PC, Jones RD, Baker KA, MacDonald RC (1996) Phosphatidylcholinium compounds: A new class of cationic phospholipids with transfection activin and unusual physical properties (abstract). Biophys J 70:88A 32. Koynova R, Wang L, MacDonald RC (2006) An intracellular lamellar – nonlamellar phase transition rationalizes the superior performance of some cationic lipid transfection agents. Proc Natl Acad Sci USA 103:14373–14378 33. Solodin I, Brown CS, Heath TD (1996) Synthesis of phosphotriester cationic phospholipids. Cationic lipids 2. Syn Lett 5: 457–458 34. MacDonald RC, Rakhmanova VA, Choi KL, Rosenzweig HS, Lahiri MK (1999) O-Ethylphosphatidylcholine: A metabolizable cationic phospholipid which is a serumcompatible DNA transfection agent. J Pharm Sci 88:896–904 35. Noone PG, Hohneker KW, Zhou ZQ, Johnson LG, Foy C, Gipson C, Jones K, Noah TL, Leigh MW, Schwartzbach C, Efthimiou J,

Pearlman R, Boucher RC, Knowles MR (2000) Safety and biological efficacy of a lipidCFTR complex for gene transfer in the Nasal epithelium of adult patients with cystic fibrosis. Mol Ther 1:105–114 36. Das A, Niven R (2001) Use of perfluorocarbon (fluorinert) to enhance reporter gene expression following intratracheal instillation into the lungs of Balb/c mice: Implications for nebulized delivery of plasmids. J Pharm Sci 90:1336–1344 37. McDonald RJ, Liggitt HD, Roche L, Nguyen HT, Pearlman R, Raabe OG, Bussey LB, Gorman CM (1998) Aerosol delivery of lipid: DNA complexes to lungs of rhesus monkeys. Pharm Res 15:671–679 38. Gorman CM, Aikawa M, Fox B, Fox E, Lapuz C, Michaud B, Nguyen H, Roche E, Sawa T, WienerKronish JP (1997) Efficient in vivo delivery of DNA to pulmonary cells using the novel lipid EDMPC. Gene Ther 4:983–992 39. Faneca H, Simoes S, de Lima MCP (2004) Association of albumin or protamine to lipoplexes: Enhancement of transfection and resistance to serum. J Gene Med 6:681–692 40. Faneca H, Cabrita AS, Simoes S, de Lima MCP (2007) Evaluation of the antitumoral effect mediated by IL-12 and HSV-Tk genes when delivered by a novel lipid-based system. Biochim Biophys Acta Biomembr 1768:1093–1102 41. Wang L, MacDonald RC (2004) New strategy for transfection: Mixtures of medium-chain and long-chain cationic lipids synergistically enhance transfection. Gene Ther 11:1358–1362 42. Koynova R, Tarahovsky Y, Wang L, MacDonald RC (2007) Lipoplex formulation of superior efficacy exhibits high surface activity and fusogenicity, and readily releases DNA. Biochim Biophys Acta Biomembr 1768:375–386 43. Gordon SP, Berezhna S, Scherfeld D, Kahya N, Schwille P (2005) Characterization of interaction between cationic lipid-oligonucleotide complexes and cellular membrane lipids using confocal imaging and fluorescence correlation spectroscopy. Biophys J 88:305–316 44. White DA (1973) The phospholipid composition of mammalian tissues. In: Ansell GB, Hawthorne JN, Dawson RMC (ed) Form and function of phospholipids, 2nd edn. Elsevier Scientific, New York, pp 441–482 45. Gennis RB (1989) Biomembranes. Molecular structure and function. Springer, New York 46. Koynova R, MacDonald RC (2007) Natural lipid extracts and biomembrane-mimicking lipid compositions are disposed to form nonlamellar phases, and they release DNA from

Lipoplex Structure and Phase Changes lipoplexes most efficiently. Biochim Biophys Acta Biomembr 1768:2373–2382 47. Simons K, Ikonen E (1997) Functional rafts in cell membranes. Nature 387:569–572 48. Brown DA, London E (1998) Structure and origin of ordered lipid domains in biological membranes. J Membr Biol 164:103–114 49. Rajamoorthi K, Petrache HI, McIntosh TJ, Brown MF (2005) Packing and viscoelasticity of polyunsaturated omega-3 and omega-6 lipid bilayers as seen by H-2 NMR and X-ray diffraction. J Am Chem Soc 127:1576–1588 50. Bartlett GR (1959) Phosphorus assay in column chromatography. J Biol Chem 234: 466–468 51. Koynova R, MacDonald RC (2004) Columnar DNA superlattices in lamellar O-ethylphos phatidylcholine lipoplexes: Mechanism of the gel–liquid crystalline lipid phase transition. Nano Lett 4:1475–1479 52. Hammersley AP, Svensson SO, Hanfland M, Fitch AN, Hausermann D (1996) Twodimensional detector software: From real detector to idealised image or two-theta scan. High Pressure Res 14:235–248 53. Kasper JS, Lonsdale K (1985) International tables for X-ray crystallography. Riedel Publis hing Company, Dordrecht, The Netherlands 54. Plotnikov VV, Brandts JM, Lin LN, Brandts JF (1997) A new ultrasensitive scanning calorimeter. Anal Biochem 250:237–244 55. Ross PC, Hui SW (1999) Lipoplex size is a major determinant of in vitro lipofection efficiency. Gene Ther 6:651–659 56. Kennedy MT, Pozharski EV, Rakhmanova VA, MacDonald RC (2000) Factors governing the assembly of cationic phospholipid-DNA complexes. Biophys J 78:1620–1633 57. Almofti MR, Harashima H, Shinohara Y, Almofti A, Li WH, Kiwada H (2003) Lipoplex size determines lipofection efficiency with or without serum. Mol Membr Biol 20:35–43 58. Salditt T, Koltover I, Radler JO, Safinya CR (1997) Two-dimensional smectic ordering of linear DNA chains in self-assembled DNAcationic liposome mixtures. Phys Rev Lett 79:2582–2585

423

59. Artzner F, Zantl R, Rapp G, Radler JO (1998) Observation of a rectangular columnar phase in condensed lamellar cationic lipid-DNA complexes. Phys Rev Lett 81:5015–5018 60. Francescangeli O, Pisani M, Stanic V, Bruni P, Weiss TM (2004) Evidence of an inverted hexagonal phase in self-assembled phospholipid-DNA-metal complexes. Europhys Lett 67:669–675 61. Koynova R, MacDonald RC (2003) Mixtures of cationic lipid O-ethylphosphatidylcholine with membrane lipids and DNA: Phase diagrams. Biophys J 85:2449–2465 62. Koynova R, Rosenzweig HS, Wang L, Wasielewski M, MacDonald RC (2004) Novel fluorescent cationic phospholipid, O-4napthylimido-1-butyl-DOPC, exhibits unusual foam morphology, forms hexagonal and cubic phases in mixtures, and transfects DNA. Chem Phys Lipids 129:183–194 63. Koynova R, Wang L, MacDonald RC (2007) Synergy in lipofection by cationic lipid mixtures: Superior activity at the gel–liquid crystalline phase transition. J Phys Chem B 111:7786–7795 64. Rakhmanova VA, Pozharski EV, MacDonald RC (2004) Mechanisms of lipoplex formation: Dependence of the biological properties of transfection complexes on formulation procedures. J Membr Biol 200:35–45 65. Pantazatos DP, MacDonald RC (1999) Directly observed membrane fusion between oppositely charged phospholipid bilayers. J Membr Biol 170:27–38 66. Pantazatos DP, Pantazatos SP, MacDonald RC (2003) Bilayer mixing, fusion, and lysis following the interaction of populations of cationic and anionic phospholipid bilayer vesicles. J Membr Biol 194:129–139 67. Lei GH, MacDonald RC (2003) Lipid bilayer vesicle fusion: Intermediates captured by highspeed microfluorescence spectroscopy. Biophys J 85:1585–1599 68. Struck DK, Hoekstra D, Pagano RE (1981) Use of resonance energy-transfer to monitor membrane-fusion. Biochemistr y 20: 4093–4099

Chapter 29 Fluorescence Methods for Evaluating Lipoplex-Mediated Gene Delivery Henrique Faneca, Nejat Düzgünes‚ , and Maria C. Pedroso de Lima Abstract The biological activity of cationic liposome/DNA complexes (“lipoplexes”) is strongly dependent on their ability to protect DNA and to interact with cells, including binding to the cell surface, internalization via endocytosis and cytoplasmic delivery of the DNA. In this chapter, we describe a number of methods and procedures to study these processes, based on the use of fluorescent probes. Key words: Cationic liposomes, Lipoplexes, Gene delivery, DNA protection, Binding and uptake of lipoplexes, Intracellular distribution of lipoplexes

1. Introduction Successful gene delivery mediated by lipoplexes requires several cellular obstacles to be surpassed, such as the protection of DNA from degradation by nucleases, binding of the lipoplexes to the cell surface, lipoplex uptake by endocytosis, their escape from the endolysosomal compartment, translocation of DNA across the nuclear envelope, and its dissociation from the cationic lipid (1–3). In an attempt to enhance transfection mediated by lipoplexes, researchers have explored different strategies that incorporate components from biological systems that have naturally evolved the capacity to surpass those barriers (4, 5). For instance, association of human serum albumin (HSA) to lipoplexes seems to be a promising strategy resulting in an enhancement of transgene expression in different types of cells (6–8). Such an increase most likely results from the promotion of lipoplex internalization via endocytosis and escape from the endocytotic pathway.

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_29, © Humana Press, a part of Springer Science+Business Media, LLC 2010

425

426

Faneca, Düzgünes‚ , and de Lima

On the other hand, vinblastine, a chemotherapeutic drug that belongs to the class of microtubule-depolymerizing agents, also has an enhancing effect on the biological activity of lipoplexes (9, 10). This compound binds specifically to tubulin, inhibiting its polymerization, and the subsequent association of microtubules. Microtubules are not only involved in chromosomal segregation, but also play important roles in intracellular dynamics, including the transport of lipoplexes to lysosomes after their uptake through endocytosis (10–12). In this context, vinblastine enhances the transfection capacity of lipoplexes, most probably by helping to overcome the endolysosomal entrapment, and consequently DNA degradation at the lysosomal level (10–12). Fluorometric and confocal microscopy studies allow the evaluation of the effect of these two agents, HSA and vinblastine, on properties that are crucial for the biological performance of lipoplexes both in vitro and in vivo. Such properties include the protection of genetic material, and binding, uptake, intracellular distribution and biological activity of the lipoplexes. The degree of DNA protection conferred by the complexes and the influence of some compounds on this parameter can be evaluated through fluorometric quantification of ethidium bromide (EtBr) access to the DNA associated with the lipoplexes. The principle of this assay is based on the properties of EtBr, a monovalent DNA intercalating agent, whose fluorescence is dramatically enhanced upon binding to DNA, and quenched when displaced by higher affinity compounds or by condensation of the DNA structure. The increase of the EtBr fluorescence, therefore, reflects an enhancement of DNA exposure and consequently a decrease in the degree of DNA protection (7, 12). Studies on the binding and uptake of lipoplexes by cells can elucidate the mechanisms of the transfection-enhancing effect of HSA and vinblastine. In this regard, cells are incubated with rhodamine-phosphatidylethanolamine (PE)-labeled lipoplexes for 2 h under different experimental conditions, and the rhodamine fluorescence associated with the cells is then measured by fluorometry (13). Studies on the intracellular distribution of lipoplexes allow understanding the mechanisms by which albumin-associated lipoplexes mediate the delivery of genetic material into cells and the influence of vinblastine on this process. The hypothesis, that vinblastine helps overcome DNA degradation at the lysosomal level is tested by confocal microscopy. This is achieved by labeling the endolysosomal pathway of the cells with LysoTracker Red that stains acidic compartments in live cells, while the intracellular localization of the lipoplexes is visualized by incorporating carboxyfluorescein-PE into the liposomal membrane (13). The biological activity studies are performed with lipoplexes carrying the pCMVgfp plasmid, encoding green fluorescent

Fluorescence Methods for Evaluating Lipoplex-Mediated Gene Delivery

427

protein (GFP), in order to evaluate the influence of different experimental conditions on their transfection activity, using confocal microscopy (13).

2. Materials 2.1. Cationic Liposomes

1. 1-palmitoyl-2-oleoyl-sn-glycero-3-ethylphosphocholine (EPOPC) (Avanti Polar Lipids, Alabaster, AL) and Cholesterol (Chol) are dissolved at 100 mg/ml and 50 mg/ml, respectively, in chloroform and stored at −20°C. 2. Rhodamine-dioleoylphosphatidylethanolamine (rhodaminePE) (Avanti Polar Lipids) is dissolved at 2 mg/ml in chloroform, protected from the light and stored at −20°C. 3. Carboxyfluorescein-dioleoylphosphatidylethanolamine (carboxyfluorescein-PE) (Avanti Polar Lipids) is dissolved at 2 mg/ml in chloroform, protected from the light and stored at −20°C. 4. Lipid film hydration is performed with deionized water (see Note 1). 5. Lipid films are obtained in a Heidolph VV 2000 rotatory evaporator (Heidolph Instruments GmbH & Co, Schwabach, Germany). 6. Liposome sonication is performed in a Sonorex RK100H sonicator (Bandelin Electronic, Berlin, Germany). 7. Mini-extruder from Avanti Polar Lipids. 8. Polycarbonate membranes of 50 nm pore-diameter (Whatman, Maidstone, UK). 9. Filters of 0.22 mm pore-diameter (Schleicher & Schuell, Dassel, Germany).

2.1.1. Lipid Concentration

1. Inorganic phosphate standard solution: 0.65 mM (KH2PO4) in 50 mM hydrochloric acid solution and stored at 4°C. 2. Acidic hydrolysis: Perchloric acid at 70–72%. 3. Ammonium molybdate solution: 0.22% (w/v) ammonium molybdate in a 2% (v/v) sulphuric acid solution and stored at room temperature (see Note 2). 4. Fiske and Subbarow reagent: 15% (w/v) NaHSO3, 0.5% (w/v) Na2SO3, 0.25% (w/v) 1-amino-2-hydroxy-4-naphthalenesulfonic acid. Protect from light and store at 4°C (see Note 3).

2.2. Cationic Liposome/DNA Complexes

1. HEPES-buffered saline (HBS): 100 mM NaCl, 20 mM HEPES, pH 7.4. Store at 4°C. 2. Plasmid pCMVgfp (Clontech, Mountain View, CA), encoding GFP, solution is prepared at 10 mg/ml in HBS and stored at 4°C. 3. HSA (Sigma) solution:640 mg/ml albumin in HBS. Store at 4°C.

428

Faneca, Düzgünes‚ , and de Lima

2.3. Protection of DNA Carried by the Complexes

1. Ethidium bromide (EtBr) solution: 2.5 mM EtBr in water protected from light and stored at 4°C. 2. Reading buffer: HBS warmed for a working temperature of 37°C. 3. Fluorescence is monitored in a SPEX-Fluorolog 1681 fluorometer (SPEX Industries, Edison, NJ).

2.4. Cell Culture

1. Dulbecco’smodifiedEagle’smedium-highglucose(DMEM-HG) (Gibco/BRL, Bethesda, MD) supplemented with 10% heatinactivated fetal bovine serum (FBS) (Gibco/BRL, Bethesda, MD), penicillin (100 U/ml) and streptomycin (100 mg/ml) (Sigma). Store at 4°C and warm to working temperature of 37°C. 2. Solution of trypsin (0.25%) and ethylenediamine tetraacetic acid (EDTA) (1 mM) from Sigma. Store at 4°C and warm for a working temperature of 37°C. 3. Cells are cultured in 75 cm2 flasks (Corning Costar Corporation, Cambridge, MA, USA). 4. Phosphate-buffered saline, (PBS) (10×): 1.37 M NaCl, 27 mM KCl, 100 mM Na2HPO4, 18 mM KH2PO4 (adjust to pH 7.4 with HCl) and autoclave before storage at room temperature. Prepare working solution by dilution of one part with nine parts water.

2.5. Cell Binding and Uptake of Lipoplexes

1. TSA cells are seeded in 48-well culture plates (Corning Costar Corporation). 2. Transfection medium: Dulbecco’s modified Eagle’s mediumhigh glucose (DMEM-HG) (Gibco/BRL) without serum and antibiotics. Store at 4°C and warm to a working temperature of 37°C. 3. Vinblastine (Sigma) is dissolved at 0.1 mM in HBS and stored at 4°C. 4. Wash buffer: PBS at room temperature. 5. Lysis buffer: 1% (w/v) Triton X-100 in PBS, store at room temperature. 6. Fluorescence is monitored in a SPECTRAmax GEMINI EM fluorometer (Molecular Devices, Union City, CA).

2.6. Intracellular Distribution of Lipoplexes

1. TSA cells are seeded in 12-well culture plates (Corning Costar Corporation), previously covered with sterile coverslips. 2. Staining of acidic compartments: LysoTracker Red DND-99 (Molecular Probes, Eugene, OR), stored at −20°C, is freshly prepared at 200 nM in DMEM-HG medium and warmed to a working temperature of 37°C.

Fluorescence Methods for Evaluating Lipoplex-Mediated Gene Delivery

429

3. Transfection medium: Dulbecco’s modified Eagle’s mediumhigh glucose (DMEM-HG) (Gibco/BRL) without serum and antibiotics. Store at 4°C and warm to a working temperature of 37°C. 4. Vinblastine (Sigma) is dissolved at 0.1 mM in HBS and stored at 4°C. 5. Wash buffer: PBS at room temperature. 6. Nuclear stain: Prepare a 1 mg/ml Hoechst 33342 (Molecular Probes) solution in PBS and store at 4°C. Prepare a working solution at 1 µg/ml by dilution in PBS. 7. Mounting medium: Vecta Shield mounting medium (Vecta Laboratories Inc., Burlingame, CA, USA). 8. Images are taken with a LSM-510 META confocal microscope (Carl Zeiss Inc., Jena, Germany), under the 63× oil immersion objective. 2.7. Biological Activity

1. TSA cells are seeded in 12-well culture plates (Corning Costar Corporation), previously covered with sterile coverslips. 2. Transfection medium: Dulbecco’s modified Eagle’s mediumhigh glucose (DMEM-HG) (Gibco/BRL) without serum and antibiotics. Store at 4°C and warm to a working temperature of 37°C. 3. Vinblastine (Sigma) is dissolved at 0.1 mM in HBS and stored at 4°C. 4. Fixation buffer: 4% (w/v) paraformaldehyde in PBS, store at room temperature. 5. Wash buffer: PBS at room temperature. 6. Nuclear stain: Prepare a 1 mg/ml Hoechst 33342 (Molecular Probes) solution in PBS and store at 4°C. Prepare a working solution at 1 µg/ml by dilution in PBS. 7. Mounting medium: Vecta Shield mounting medium (Vecta Laboratories Inc.). 8. Images are taken with a LSM-510 META confocal microscope (Carl Zeiss Inc.), under the 63× oil immersion objective.

3. Methods Cationic liposomes can be purchased from several companies or easily prepared in the laboratory. In our studies small unilamellar cationic liposomes (SUV) are prepared in the laboratory (14). Regarding lipoplex production, it is relevant to emphasize that the mode of preparation strongly determines their final

430

Faneca, Düzgünes‚ , and de Lima

properties and consequently their biological activity, both in vitro and in vivo. Therefore, critical parameters, such as the composition of the lipoplexes, the structure and size of cationic liposomes, the concentrations of cationic lipid and DNA, the ionic strength and temperature of the suspending medium, the order of addition and the mixing rate of the components, as well as the time of complex formation should be considered and suited to the desired applications (14). Different studies can be performed to evaluate the performance of the lipoplexes for gene delivery applications, including those based on fluorometric and confocal microscopy assays. 3.1. Cationic Liposomes

1. Small unilamellar cationic liposomes (SUV) used are prepared by extrusion of multilamellar liposomes (MLV) composed of 1:1 (mol ratio) mixtures of 1-palmitoyl-2-oleoyl-sn-glycero3-ethylphosphocholine (EPOPC) and cholesterol (Chol). 2. For binding and uptake studies, EPOPC:Chol (1:1) liposomes also contain 1 mol% rhodamine-dioleoylphosphatidylethanolamine (rhodamine-PE), and for intracellular distribution studies, the liposomes are labeled with 0.1 mol% carboxyfluorescein-dioleoylphosphatidylethanolamine (carboxyfluorescein-PE) (see Note 4). 3. The necessary volumes of the stock lipids, dissolved in chloroform at the referred concentration (Subheading 2.1), are mixed at the desired molar ratio in glass tubes previously washed with chloroform. The glass tubes containing the lipid mixtures are placed in the rotatory evaporator (220 rpm) and allowed to dry under vacuum (−0.8 bar) for 1 h at 37°C. 4. The dried lipid films are hydrated with 1 ml of water (to a final lipid concentration of 6 mM) with constant vortexing in order to promote the formation of MLV. 5. The resulting MLV are sonicated for 3 min and extruded 21 times through two stacked polycarbonate filters of 50 nm porediameter, using the referred mini-extruder (Subheading 2.1) to obtain a homogeneous suspension of SUV. The SUV are then diluted three times with water and filter-sterilized utilizing 0.22 mm pore-diameter filters. The suspension is stored at 4°C until use.

3.1.1. Lipid Concentration

1. Cationic liposome concentration is determined by quantification of the phospholipid content (EPOPC) through colorimetric analysis (15, 16) (see Note 5). 2. Standards (0; 32.5; 65; 130; 260 nmol of inorganic phosphate) and samples (volume corresponding to approximately 100 nmol of phosphate from EPOPC) are placed in glass tubes previously washed with chloroform. The standards and

Fluorescence Methods for Evaluating Lipoplex-Mediated Gene Delivery

431

samples are then submitted to acidic hydrolysis with 0.5 ml of perchloric acid at 70–72% for 2 h at 180°C, to promote the conversion of organic phosphate into inorganic phosphate (see Note 6). 3. After hydrolysis, the tubes containing the standards and samples are allowed to cool to room temperature and then 7.5 ml of ammonium molybdate, which converts the inorganic phosphate into phosphomolybdic acid, are added to each tube. After addition of 0.3 ml of Fiske and Subbarow reagent to each tube containing the standards and samples, the phosphomolybdic acid is reduced to molybdate blue upon incubation in a water boiling bath during 15 min. 4. The standards and samples are allowed to cool to room temperature and the absorbance is read at 830 nm. The determination of the phosphate concentration of the liposomal samples is based on the standard curve obtained for the absorbance. 3.2. Cationic Liposome/DNA Complexes

1. Cationic liposome/DNA complexes are prepared in HBS using liposomal suspensions of known concentrations and solutions of DNA and HSA, both prepared in HBS at 10 µg/ ml and 640 µg/ml, respectively (see Note 7). All the solutions are previously filtered under aseptic conditions. 2. Complexes are prepared in sterile polypropylene tubes in order to avoid interactions between DNA and the tube walls. The concentration of DNA is maintained constant (5 µg/ml) independently of the total volume of the complexes, since the physicochemical characteristics of the complexes could change with its concentration. Complexes without HSA, corresponding to 1 µg of DNA, are prepared by using 100 µl of HBS, 100 µl of the DNA solution (10 µg/ml) and the necessary volume of liposomes, which is dependent on the liposome concentration. In the case of lipoplexes containing HSA, half of the HBS volume is replaced by the same volume of HSA solution, in order to obtain 32 µg of HSA/µg of DNA. 3. Lipoplexes are prepared by sequentially mixing the established HBS volume with the necessary volume of liposomes and DNA solution, the latter being added gently to the liposome suspension. After shaking the mixture, this is incubated for 15 min at room temperature to allow the formation of the complexes through the establishment of electrostatic interactions. For lipoplexes containing HSA, liposomes are preincubated with the protein for 15 min to allow interaction between the two components, followed by slow addition of the DNA solution and a further 15 min incubation of the resulting mixture at room temperature. Lipoplexes are used immediately after preparation.

432

Faneca, Düzgünes‚ , and de Lima

3.3. Protection of DNA Carried by the Complexes

1. The degree of DNA protection is evaluated by measuring the EtBr fluorescence in a SPEX-Fluorolog 1681 fluorometer, at excitation and emission wavelengths of 518 and 605 nm, respectively, and using the excitation and emission slits of 1 and 2 mm, respectively (see Note 8). 2. The cationic liposome/DNA complexes are prepared, as described above, by using volumes of the solutions of HBS, albumin and DNA and of the liposomal suspension corresponding to 1.5 µg of DNA. After a period of 15 min for complex maturation, the volume of complexes corresponding to 1 µg of DNA is immediately added to the cuvette containing a volume of HBS (at 37°C) necessary to bring the final volume to 2 ml (see Note 9). The fluorescence of the mixture is measured at 37°C, before (residual fluorescence) and after addition of 10 µl of EtBr (stock concentration: 2.5 mM) to the cuvette. The fluorescence values used in the calculations result from the subtraction of the residual fluorescence from the fluorescence values obtained after EtBr addition. 3. To quantify the access of EtBr to the DNA carried by the complexes, it is also necessary to determine the basal fluorescence of EtBr and that in the presence of naked DNA. The basal fluorescence of EtBr is determined by measuring the fluorescence at 37°C, before and after addition of 10 µl of EtBr to a cuvette containing 2 ml of HBS (37°C). The EtBr fluorescence in the presence of DNA is also measured at 37°C, before and after addition of 10 µl of EtBr to a cuvette containing 1 µg of DNA in 2 ml of HBS (37°C). The amount of DNA available to interact with the probe is calculated by subtracting the values of basal fluorescence of EtBr from those obtained for the complexes and DNA, in order to obtain the fluorescence that is only due to the intercalation of the probe in the double-stranded DNA. The maximal access (100%) of EtBr to the genetic material corresponds to the fluorescence obtained with naked DNA.

3.4. Cells

1. TSA cells (BALB/c female mouse mammary adenocarcinoma cell line) are grown in 75 cm2 flasks and maintained at 37°C, under 5% CO2, in DMEM-HG supplemented with 10% (v/v) heat-inactivated FBS, penicillin (100 U/ml) and streptomycin (100 mg/ml) (see Note 10). 2. When cells reach 80–90% of confluence, the culture medium is removed and the cells are washed with 3 ml of the trypsin solution (see Note 11). After this washing step, cells are detached upon incubation at 37°C, for 2–3 min, with 3 ml of fresh trypsin solution (see Note 12). Cells are then resuspended in 10 ml of DMEM-HG medium and counted in a hemocytometer to determine the cell suspension density.

Fluorescence Methods for Evaluating Lipoplex-Mediated Gene Delivery

433

3. The cell suspension is diluted with DMEM-HG medium to obtain a final density of 50 × 103 cells/ml and the cells are plated in multiwell plates, the number of wells per plate depending on the wanted study. For binding and uptake studies, cells are plated 24 h before the experiments in 48-multiwell plates at 50 × 103 cells/well. For intracellular distribution and biological activity studies, cells are plated 24 h before the experiments in 12-multiwell plates at 120 × 103 cells/well to obtain a cell confluence of 50% at the beginning of the experiments. The multiwells are then placed in the incubator at 37°C, under 5% CO2. All the steps involved in cell preparation are performed under sterile conditions in a bioguard hood. 3.5. Cell Binding and Uptake of Lipoplexes

1. Twenty-four hours before transfection, 0.5 × 105 TSA cells/well are seeded in 1 ml of medium in 48-multiwell plates. At the time of transfection, the culture medium is removed and cells are washed once with DMEM-HG medium without serum and antibiotics, and then covered with 0.3 ml of the same medium. 2. EPOPC:Chol (1:1) liposomes, labeled with 1% rhodamine-PE, are complexed in the presence or absence of HSA with pCMVgfp plasmid at the 4/1 (+/−) cationic lipid/DNA charge ratio. 3. The resulting lipoplexes, corresponding to 1 mg of pCMVgfp plasmid, are incubated with the cells, in the presence or absence of 0.5 µM vinblastine, for 2 h at 4°C (binding) or 37°C (uptake) (see Note 13). 4. Following this incubation, cells are washed twice with PBS to remove the complexes that are not bound or internalized, and lysed with 100 ml/well of lysis buffer to maximally dilute the fluorescent probe contained in the complexes that are attached to or internalized by the cells (see Note 14). Binding and uptake of lipoplexes are immediately monitored in a SPECTRAmax GEMINI EM fluorometer by measuring the fluorescence at excitation and emission wavelengths of 545 and 587 nm, respectively. An example of the results produced is shown in Fig. 1.

3.6. Intracellular Distribution of Lipoplexes

1. Twenty-four hours before transfection, 120 × 103 TSA cells in 2.4 ml of DMEM-HG culture medium are added to each well of a 12-multiwell plate, previously covered with coverslips. 2. At the time of transfection, the culture medium is removed and the cells are incubated for 30 min with 0.5 ml/well of 200 nM LysoTracker Red DND-99 (prepared in DMEM-HG medium), which labels acidic compartments of live cells. Cells are then washed once with DMEM-HG medium, without serum and antibiotics and subsequently covered with 0.5 ml of the same medium prior to addition of lipoplexes. 3. To observe the intracellular distribution of the complexes, EPOPC:Chol liposomes labeled with 0.1% carboxyfluorescein-

434

Faneca, Düzgünes‚ , and de Lima

RFUs/1x106 cells

4500

EPOPC:Chol/DNA (+/–) (4/1)

3750 3000

HSA-EPOPC:Chol/DNA (+/–) (4/1)

2250

HSA-EPOPC:Chol/DNA (+/–) (4/1) + 0.50 µM Vinb

1500 750 0

Binding

Uptake

Fig. 1. Effect of vinblastine on the extent of binding and uptake of EPOPC:Chol/DNA and HSA-EPOPC:Chol/DNA lipoplexes in TSA cells. Liposomes, labeled with rhodamine-PE, were associated or not to HSA before complexation with DNA. Cells were incubated for 2 h with the lipoplexes at 4°C (binding) or 37°C (uptake), in the presence or absence of 0.5 µM vinblastine (Vinb). The data are expressed as RFU (relative fluorescence units) per 106 cells (mean ± S.D.). Association of albumin to lipoplexes resulted in a significant enhancement in the extent of binding and uptake. On the other hand, the presence of vinblastine did not induce any significant change in the extent of binding and uptake of lipoplexes. (Reproduced from ref. (13) with permission from Elsevier Science)

PE and associated to HSA, are complexed with the pCMVgfp plasmid at the (4/1) (+/−) charge ratio. The resulting lipoplexes are incubated with the cells, in the presence or absence of 0.5 µM vinblastine, for 2 h at 37°C. 4. After 2 h of incubation with the complexes containing 2 µg of pCMVgfp plasmid, the cells are washed twice with PBS and the nuclei are labeled through cell incubation with 0.5 ml/well of the fluorescent DNA-binding dye Hoechst 33342 (1 µg/ml), for 5 min at room temperature. Cells are then washed once with PBS and immediately mounted in Vecta Shield mounting medium, and observed under a confocal microscope. Excitation at 577 nm induces LysoTracker Red fluorescence (red emission), excitation at 497 nm induces carboxyfluorescein fluorescence (green emission) and excitation at 350 nm induces Hoechst 33342 fluorescence (blue emission). 3.7. Transfection Activity

1. Twenty-four hours before transfection, 120 × 103 TSA cells in 2.4 ml of DMEM-HG culture medium are added to each well of a 12-multiwell plate, containing a coverslip. At the time of transfection, the culture medium is removed, and the cells are washed once with DMEM-HG medium without serum and antibiotics, and then covered with 0.5 ml of the same medium. 2. EPOPC:Chol (1:1) liposomes with or without HSA are complexed with pCMVgfp plasmid, at the 4/1 (+/−) cationic lipid/ DNA charge ratio. 3. The resulting lipoplexes containing 2 mg of pCMVgfp are incubated with the cells, in the presence or absence of 0.5 µM

Fluorescence Methods for Evaluating Lipoplex-Mediated Gene Delivery

435

Fig. 2. Effect of vinblastine on gene delivery mediated by HSA-EPOPC:Chol/DNA lipoplexes in TSA cells. Representative confocal microscopy images (original magnification: × 630) of GFP expression mediated by (a) HSA-EPOPC:Chol/DNA (+/−) (4/1) lipoplexes and (b) HSA-EPOPC:Chol/DNA (+/−) (4/1) lipoplexes + 0.50 µM vinblastine. TSA cells with Hoechstlabeled nuclei (blue) were transfected with lipoplexes containing the GFP plasmid. The percentage of GFP-expressing cells and the intensity of GFP fluorescence were higher in the presence of 0.5 µM of vinblatine. (Reproduced from ref. (13) with permission from Elsevier Science)

vinblastine, for 4 h at 37°C. Then the medium is replaced with DMEM-HG and the cells are further incubated for 48 h. 4. Cells are washed twice with PBS, fixed by incubation with 4% paraformaldehyde (1 ml/well) for 15 min at room temperature, and washed again once with PBS. The nuclei are labeled with Hoechst 33342 (1 µg/ml; 0.5 ml/well) for 5 min at room temperature. Cells are then washed once with PBS and immediately mounted in Vecta Shield mounting medium, and observed under a confocal microscope. Excitation at 497 nm induces GFP fluorescence (green emission), while excitation at 350 nm induces Hoechst 33342 fluorescence (blue emission). An example of the obtained results is shown in Fig. 2.

4. Notes 1. All solutions should be prepared in water with a resistivity of 18.2 MW cm and total organic content of less than five parts per billion. 2. Since the use of the sulphuric acid solution results in an exothermic reaction this solution should be prepared on ice. 3. The preparation of the Fiske and Subbarow reagent requires continuous stirring of the solution for 4 h. The solution is then allowed to stand overnight and is subsequently filtered.

436

Faneca, Düzgünes‚ , and de Lima

4. Percentage of carboxyfluorescein-PE higher than 0.1 mol% leads to an overload of fluorescence inside the cells, which decreases the quality of the confocal microscopy images. 5. The amount of phospholipids is evaluated by colorimetric analysis of the molybdate blue resulting from the conversion of inorganic phosphate into phosphomolybdic acid. After being reduced, phosphomolybdic acid acquires a blue coloration (molybdate blue), whose intensity is proportional to the amount of inorganic phosphate produced from the acidic hydrolysis of the phospholipids (15, 16). 6. The time of acidic hydrolysis can be reduced to 1 h by increasing the temperature to 250°C or above. An alternative procedure involving sulphuric acid digestion is described in (17). 7. The concentration of the DNA solution is determined by spectrophotometry, at 260 nm, taking into account that one absorbance unit corresponds to 50 µg/ml of double-stranded DNA. 8. The principle of this assay is based on the properties of EtBr, a monovalent DNA intercalating agent, whose fluorescence is dramatically enhanced upon binding to DNA. 9. Fluorescence should be measured immediately after the 15 min maturation of the lipoplexes, since measurements at different times result in different levels of EtBr access for the same lipoplex formulation. The temperature at which the assay is carried out also significantly affects the fluorescence values, and therefore should be maintained constant (37°C). 10. TSA cells grow as a monolayer and should be detached by treatment with a trypsin solution (0.25%) before reaching 100% confluence in order to be maintained in the exponential growth phase. 11. The washing procedure facilitates the complete removal of the culture medium (which inhibits trypsin action due to the presence of the serum), including Ca2+ and Mg2+ which are necessary to the cell attachment. 12. During cell detachment, vigorous shaking of the cells should be avoided in order to prevent the formation of cell clusters, which can complicate the determination of cell density. 13. 4°C (binding) is the temperature at which cells cannot internalize the bound complexes and 37°C (uptake) is the temperature at which endocytosis can occur (14). 14. Since the excitation and emission spectra of rhodamine-PE overlap partially, dilution of the probe results in an increase of fluorescence intensity.

Fluorescence Methods for Evaluating Lipoplex-Mediated Gene Delivery

437

References 1. Wasungu L, Hoekstra D (2006) Cationic lipids, lipoplexes and intracellular delivery of genes. J Control Release 116:255–264 2. Khalil IA, Kogure K, Akita H, Harashima H (2006) Uptake pathways and subsequent intracellular trafficking in nonviral gene delivery. Pharmacol Rev 58:32–45 3. Pedroso de Lima MC, Simões S, Pires P, Faneca H, Düzgünes˛ N (2001) Cationic lipidDNA complexes in gene delivery: from biophysics to biological applications. Adv Drug Deliv Rev 47:277–294 4. Simões S, Filipe A, Faneca H, Mano M, Penacho N, Düzgünes˛ N et al (2005) Cationic liposomes for gene delivery. Expert Opin Drug Deliv 2:237–254 5. Lechardeur D, Lukacs GL (2002) Intracellular barriers to non-viral gene transfer. Curr Gene Ther 2:183–194 6. Faneca H, Cabrita AS, Simões S, Pedroso de Lima MC (2007) Evaluation of the antitumoral effect mediated by IL-12 and HSV-tk genes when delivered by a novel lipid-based system. Biochim Biophys Acta 1768:1093–1102 7. Faneca H, Simões S, Pedroso de Lima MC (2004) Association of albumin or protamine to lipoplexes: enhancement of transfection and resistance to serum. J Gene Med 6:681–692 8. Simões S, Slepushkin V, Pires P, Gaspar R, Pedroso de Lima MC, Düzgünes˛ N (2000) Human serum albumin enhances DNA transfection by lipoplexes and confers resistance to inhibition by serum. Biochim Biophys Acta 1463:459–469 9. Wang L, MacDonald RC (2004) Effects of microtubule-depolymerizing agents on the transfection of cultured vascular smooth muscle cells: Enhanced expression with free drug

and especially with drug–gene lipoplexes. Mol Ther 9:729–737 10. Hasegawa S, Hirashima N, Nakanishi M (2001) Microtubule involvement in the intracellular dynamics for gene transfection mediated by cationic liposomes. Gene Ther 8: 1669–1673 11. Chowdhury NR, Hays RM, Bommineni VR, Franki N, Chowdhury JR, Wu CH et al (1996) Microtubular disruption prolongs the expression of human bilirubin-uridinediphos phoglucuronate-glucuronosyltransferase-1 gene transferred into Gunn rat livers. J Biol Chem 271:2341–2346 12. Faneca H, Simões S, Pedroso de Lima MC (2002) Evaluation of lipid-based reagents to mediate intracellular gene delivery. Biochim Biophys Acta 1567:23–33 13. Faneca H, Faustino A, Pedroso de Lima MC (2008) Synergistic antitumoral effect of nonviral HSV-tk/GCV gene therapy and vinblastine in mammary adenocarcinoma cells. J Control Release 126:175–184 14. Pedroso de Lima MC, Faneca H, Mano M, Penacho N, Düzgünes˛ N, Simões S (2003) Biophysical characterization of cationic liposome-DNA complexes and their interaction with cells. Meth Enzymol 373:298–312 15. Fiske CH, Subbarow Y (1925) The calori metric determination of phosphorus. J Biol Chem 66:375–400 16. Bartlett GR (1959) Phosphorus assay in column chromatography. J Biol Chem 234: 466–468 17. Düzgünes˛ N (2003) Preparation and quantitation of small unilamellar liposomes and large unilamellar reverse-phase evaporation liposomes. Meth Enzymol 367:23–27

Chapter 30 FRET Imaging of Cells Transfected with siRNA/Liposome Complexes Il-Han Kim, Anne Järve, Markus Hirsch, Roger Fischer, Michael F. Trendelenburg, Ulrich Massing, Karl Rohr, and Mark Helm Abstract By monitoring the efficiency of fluorescence resonance energy transfer of dyes attached to the different strands of siRNA, the structural integrity of the latter can be traced inside cells. Here, the experimental details of dye-labeled siRNA construction, tissue culture, and transfection with liposomally formulated siRNAs are given, as well as the conditions for confocal microscopy and an algorithm allowing the visualization of intact siRNA after image data treatment. The method allows rapid screening of different liposomal siRNA formulations, obtained by small scale dual asymmetric centrifugation with high entrapping efficiency. Key words: Fluorescence Resonance Energy Transfer (FRET), siRNA, Cell transfection, Microinjection, Spectral imaging, Dual Assymetric Centrifugation (DAC), Liposome

1. Introduction siRNA mediated RNAi has become an outstandingly powerful tool in fundamental research, with many investigations into therapeutic applications currently being undertaken. siRNAs are short double stranded RNA duplexes of about the length of 21 nucleotides and typically feature two-nucleotide long overhangs on the 3¢-end of each strand. Transfection into cultured eukaryotic cells frequently leads to very efficient gene silencing due to the degradation mRNAs whose sequence is partially encoded in the siRNA duplex. This technique profits from a cellular amplification mechanism, which uses one of the two short complementary RNA strands of siRNAs as an antisense guide to target a catalytic mRNA degradation complex. Thus, only a few molecules out of V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_30, © Humana Press, a part of Springer Science+Business Media, LLC 2010

439

440

Kim et al.

a large dose of applied siRNA actually affect the RNAi after the penetration of the cell, while the fate of the residual large majority is unclear. Insight into the whereabouts of intact and degraded siRNAs may yet lower the effective amount of siRNA required for RNAi in cell culture and in vivo. Also, investigation of the integrity of liposomally formulated siRNA and their pharmacokinetics are of increased interest despite being experimentally challenging. FRET, fluorescence resonance energy transfer, is a distance dependent communication process between two fluorescent dyes which offers a promising approach to these problems. If a suitable pair of dyes is attached to the siRNA duplex, their spatial proximity results in high FRET efficiency, which can be assessed from relative fluorescence emissions of both dyes. Conversely, spatial separation due to dye removal, siRNA degradation, or strand separation entails a breakdown of the FRET effect, which again is easily discernible in the fluorescence emission spectrum from the ratio of red and green emission (R/G ratio). The availability of modern confocal microscopes in conjunction with high grade spectral equipment allows to quantify these effects even inside the cell and at high optical resolution, thus allowing an assessment of the degradation state of a FRET-labeled siRNA population inside the cell and even in certain cellular compartments. The present method is based on a large number of validation experiments which are described in detail in our recent paper (1). These experiments include the construction of several siRNA in which the inter-dye distance was varied. The resulting emission spectra and FRET efficiencies measured inside the cell (see Fig. 1a–c) and in the cuvette were compared and found to be very similar (1), thus providing calibration points which serve to quantify the amount of leftover intact FRET-labeled siRNA. Layered scans of the entire volume of a cell in a confocal microscope at red and green emission wavelengths are processed by an algorithm producing red/green ratio images which visualize populations of intact siRNA inside the cell. In such an image, as displayed in Fig. 1d, volume elements (voxels) with a content of significantly more than 90% intact siRNA are

Fig. 1. (a) True color fluorescence imaging of cells microinjected with intact siRNA showing high FRET efficiency, and a mixture of siRNAs with dyes in separate molecules. This mixture shows no FRET and therefore simulates the spectral properties of an siRNA population where the dyes have been spatially separated by degradation. The upper row (a) shows an image of a cell population, in which one cell is outlined. The enlargement of the outlined cell is shown in (b). Defined regions of interest (ROIs) are outlined in dark grey (nucleus), black (whole cell) and light grey (cytoplasm). The fluorescence emitted from these ROIs upon excitation at 488 nm is shown in 5 nm bin spectral resolution in (c). The high FRET samples are clearly recognizable from the strong red TMR fluorescence emission around 580 nm (left column). No-FRET samples show fluorescein emission around 520 nm (right column). (d) Example of R/G imaging (right panel) from a green fluorescence emission scan (510–540 nm, middle panel) and a red fluorescence scan (570–600 nm, left panel) upon excitation at 488 nm of cells microinjected with FRET labeled siRNAs. In grayscale print, the R/G image (right panel) appears dark gray for red, and light gray for green. The light gray rim thus indicates degraded siRNA populations around a core of intact siRNA populations in dark gray

siRNA FRET in Liposomes

No-FRET

FRET

a

b

Int.

c

2500

600

2000

500 400

1500

300

1000

200

500 0 500

nucleus whole cell cytoplasm

100 550 600 Wavelength [nm]

650

0

500

550 600 Wavelength [nm] 650

d

Red channel image

Green channel image

R/G ratio image

441

442

Kim et al.

colored dark red, while those containing significantly more than 10% degraded siRNA are in green. Yellow indicates voxels with an intermediate R/G ratio that does not allow clear assignment. As Fig. 1d depicts a cell immediately after microinjection of intact siRNAs, the majority of the R/G voxels are red indicating little or no degradation, but over hours and days degradation becomes visible by an increasing number of green voxels. As a whole, this provides a convenient method to assess amount and integrity of siRNA transfected into a cell line by a given method, using e.g., cationic lipid transfection agents such as oligofectamin, or nanoparticles, or liposomes for transfection. While standard transfection agents, such as oligofectamin, are very efficient in cell culture, their inflammatory properties make them unsuitable for in vivo applications. Liposomal formulations of siRNA, on the other hand, have been successfully applied in vivo (2). Their difficult and expensive production for screening purposes is greatly facilitated by small scale production using dual asymmetric centrifugation (DAC) (3, 4). The present paper describes the details of liposomal and other formulations of FRET-labeled siRNA duplexes, their transfection into mammalian cells by either liposomes, or cationic lipid transfection agents, the subsequent confocal microscopy of the transfected cells at different emission wavelengths, and finally the data treatment yielding the so called R/G ratio images for the direct assessment of the siRNA degradation status in the cells.

2. Materials 2.1. siRNA Duplex Preparation

1. Labeled antisense and sense strands of siRNA (IBA, Göttingen, Germany) at the concentration of 100 mM (see Note 1). 2. Phosphate buffered saline (PBS): Prepare 10× stock with 1.4 M NaCl, 27 mM KCl, 15 mM KH2PO4, 80.6 mM Na2HPO4, pH is 6.6. Autoclave before storage at room temperature. Prepare 1× PBS by dilution of one part with nine parts of Milli-Q grade sterile water. After dilution, pH of PBS would be 7.4.

2.2. Cell Culture

1. Growth medium: 45% v/v a-MEM (supplemented with 2.2 g/l NaHCO3) and 45% v/v HAM¢s F10 (Gibco) supplemented with 10% v/v fetal calf serum (FCS, Sigma Aldrich, St. Louis, USA), 100 mg/ml penicillin/streptavidin (PenStrep Mix) (Gibco) and 1 ng/ml basic fibroblast growth factor (bFGF, Invitrogen). bFGF is dissolved at 0.1 mg/ml in PBS and stored in single used aliquots at −20°C up to 6 months. Medium is sterilefiltered and kept at 4°C up to a week.

siRNA FRET in Liposomes

443

2. Freezing medium: 35% v/v a-MEM (supplemented with 2.2 g/L NaHCO3); 35% v/v HAM´s F10, 20% v/v FCS, 10% v/v dimethyl sulphoxide (DMSO). 3. Solution of trypsin (0.25%) and ethylenediamine tetraacetic acid (EDTA) (1 mM) from Gibco. 2.3. Liposomal Formulation

1. Liposomal components: hydrogenated egg phosphatidylcholine (EPC-3), pegylated phosphatidylethanolamine (MPEG2000 – DSPE), cholesterol and a preformed molecular mixture of EPC-3 and cholesterol (55:45 mole/mole). All phospholipids are purchased from Lipoid GmbH (Ludwigshafen, Germany) and cholesterol is purchased from Sigma Aldrich. 2. Dual asymmetric centrifuge (DAC 150 FVZ, Hauschild GmbH & Co KG, Hamm, Germany) with a horizontal insert for 2 ml reaction tubes. 3. Glass beads, 1 mm diameter (Sartorius AG, Göttingen, Germany). 4. Fluorescent labeled siRNA, as described in 2.1. 5. PBS as described in 2.1. 6. Photon correlation spectroscope (PCS; Nicomp submicron particle analyzer model 380, Nicomp Inst Corp, Santa Barbara, USA) with 4 ml disposable cuvettes. 7. Inverted centrifugal spin filters for small volumes; Centrisart I, 100.000 MWCO (Sartorius AG, Göttingen, Germany).

2.4. Transfection

1. Agarose (Invitrogen): Prepare 2% agarose (w/v) in sterile water, heat to 60° and use immediately, while still hot (no autoclaving required). 2. Rat tail collagen, type I (Sigma Aldrich): Prepare stock solution 1 mg/ml by dissolving lyophilized powder in filter-sterilized 0.2% v/v acetic acid. The powder is swirled occasionally and left overnight to dissolve. Dilute with PBS. 3. Glass coverslips (Ø 18 mm, Roth), cleaned in ethanol and autoclaved on a glass petri dish; 12-well culture plates (Corning). 4. Transfection medium: Growth medium without PenStrep and FCS. 5. 3× transfection medium: Growth medium with 30% FCS and without PenStrep. 6. Oligofectamine (Invitrogen), kept at 4°C.

2.5. Microinjection

1. Glass coverslips (Ø 12, Roth), cleaned in ethanol and autoclaved on a glass petri dish; small petri dishes (Ø 35 mm, Nunclon). 2. Microinjection medium: HEPES is dissolved at 25 mM together with carbonate-free a-MEM medium (Sigma Aldrich) in sterile

444

Kim et al.

water and resulting medium with pH 7.4 was sterile-filtered and stored at 4°C. 3. Glass capillaries Femtotips (Eppendorf, 5242 952.008; Westbury, NY). 4. Microloaders (Eppendorf, 5242 952.003; Westbury, NY). 2.6. Fixation of Cultures

1. Formaldehyde (Riedel de Häen): Prepare prior to fixation a 4% (v/v) solution in transfection medium from 36.5% formaldehyde (see Note 2). 2. Nuclear stain: 4,6-diamidino-2-phenylindole 2 mg/ml (7 nM) in PBS. 3. Mounting medium: Fluorescence mounting medium (DAKO Cytomation).

3. Methods 3.1. Preparation of siRNA Duplexes for Transfection and Microinjection

The experimental starting point is the assembly and spectral investigation of double-labeled siRNA after annealing of both strands, each of which contains one fluorophore of a fluorescein-tetramethylrhodamine FRET pair (dye orientation details see Note 1). 1. For transfection of four wells, each with 30 pmol of siRNA duplex, prepare 30 mL of 5 mM siRNA duplex solution by combining 1.5 mL sense siRNA (5 mM final concentration), 1.5 mL antisense siRNA (5 mM final concentration), 3 mL 10× PBS, and 24 mL sterile water. Vortex. 2. Incubate resulting mixture at 90°C for 1 min and subsequently let the strands anneal for 1 h in a hot-block at 37°C (see Note 3). Keep the siRNA on ice until used or store at −20°C as freezing/thawing would not affect duplex stability (see Note 4). 3. For microinjection, prepare 20 mL of siRNA duplex solution (50 mM) by mixing together 10 mL of sense (100 mM) and 10 mL of antisense (100 mM) strands. Reduce the volume of sample to ~15 mL under vacuum (speed-vac) at RT for ~15 min. Add 2 mL of 10× PBS and fill up to 20 mL with sterile water. Heatshock and let siRNA duplexes form as described in step 2. 4. For quality control, fluorescence emission spectra of labeled siRNA duplexes were recorded with a fluorimeter FP-6500 (JASCO, Tokyo, Japan) equipped with ETC-273 temperature controller, a peltier element for heating and an F-25 (Julabo, Seelbach, Germany) cooling unit. Emission spectra (3 nm bandwidth)vrecorded upon excitation at 488 nm (3 nm bandwidth) were corrected for differential wavelength

siRNA FRET in Liposomes

445

PMT sensitivity. Spectra of e.g., 1 mM FRET siRNA duplexes were recorded in 50 ml SUPRASIL cuvettes (HELLMA, Müllheim, Germany) in 1 × PBS (pH 7.4) at 20°C. 5. For determination of the optimal ratio of the fluorophores in siRNA in the 3FL-5TMR siRNA duplexes was determined by annealing the siRNA strands (concentration range 1–3 mM) in a dilution series and measuring FRET spectra as described in step 3. The ratio of the strands which gave maximal FRET was further on applied (see Note 5). 3.2. Cell Culture and Preparation of Cells in 12-Well Plates for Transfection

1. Rat brain endothelial 4 (RBE4) cells (2) are grown in growth medium at 37°C and 5% CO2 atmosphere and are passaged when approaching confluence with trypsin/EDTA to provide maintenance cultures in 75 cm2 flasks. Seeding density for RBE4 is 40,000 cells per cm2 (see Note 6). 2. Prepare 12-well plates: pipet ~600 mL of hot agarose into a well and put on top a sterilized cover slip (Ø 18 mm). To coat with collagen an entire surface of a cover slip in a 3.8 cm2 well at 3 mg/cm2, add a volume of 250 mL of collagen solution (11.4 mL of stock solution is mixed with 238.6 mL of PBS) to a well. Let dry under fume hood for at least 6 h. Wash the wells of the culture plate with PBS and use immediately or store at 4°C for at least up to a week. 3. Two days before transfection, trypsinize 90% confluent cells grown in a 75 cm2 flask and plate them in a 12-well culture plate with the seeding density of 40,000 cells per cm2 in the fresh growth medium (0.7 ml per well) without antibiotics. At the day of transfection, cultures would reach 70% confluence, which corresponds to ~2 × 105 cells per well.

3.3. Small Scale Preparation of Sterile, siRNA Containing Liposomes with High Trapping Efficiency

Small scale preparation of liposomally formulated siRNA with high trapping efficiency was achieved by a dual asymmetric centrifugation (DAC) method (3, 4), especially adapted and optimized for the reproducible preparation of very small batches of (a) conventional liposomes (CL) as well as (b) sterically stabilized liposomes (SSL), containing high amounts of siRNA. The new preparation method resulted in batches of 20 or 60 mg of highly concentrated liposomal formulations (vesicular phospholipid gels, VPG) with siRNA-entrapping efficiencies of 50–70%. The obtained viscous formulations could further be diluted to normal (liquid) sterile siRNA containing liposomal formulations (SSL or CL) with the same high trapping efficiency as for VPG. The siRNA-liposomes can be used in cell culture experiments or for animal studies. 1. For sterile liposome formulations, all of the following preparation steps have to be performed with sterile/autoclaved materials.

446

Kim et al.

2. Preparation of lipid mixtures: For the preparation of conventional liposomes, the commercially available lipid mixture EPC-3/cholesterol is used. For batch sizes of 20 and 60 mg, 8 or 24 mg of the mixture is weighted into 2 ml reaction tubes, respectively. For the preparation of sterically stabilized liposomes (SSL), EPC-3, cholesterol and MPEG2000 in a molar ratio of 56.9:37.9:5.2 were dissolved in EtOH (99% p.a.) in a 2 ml reaction tube. The amount for a 20 (60) mg batch size are: 4.2 (12.6) mg EPC-3, 1.4 (4.3) mg cholesterol and 1.39 (4.2) mg MPEG2000 in 500 mL EtOH (see Note 7). To get a molecular dispersed mixture of the lipids, EtOH is removed by using a Speed-Vac (SC110 Savant, France) at low temperature setting. The process is ready when all EtOH is completely evaporated (typically 4–6 h) (see Note 8). 3. Adding of mixing aids: To the lipid mixtures for preparation of both, CL and SSL, 20 or 60 mg glass beads (1 mm diameter) are added (according to the planned batch size of 20 or 60 mg). 4. Preparation of siRNA solution: Prepare the siRNA sample by mixing sense and antisense siRNA strands of 100 mM stocks and adjust sample to 1× PBS using 10× PBS stock (see Subheading 2.1). The siRNA sample is incubated for 1 min at 90°C and annealed for 1 h at 37°C in a heating-block (see Subheading 3.1). Store on ice or in a freezer until use. 5. Adding liquid phase: Take 13 mL (39 mL) of siRNA sample in case of SSL or 12 mL (36 mL) of siRNA sample in case of CL as liquid phase preparing a 20 mg (or 60 mg) batch. Add the liquid to the lipids in the reaction tube and spin down for 2 min in a bench top centrifuge and incubate for additional 8 min at room temperature to let the lipids swell. 6. Vesicular phospholipid gel (VPG) preparation: Place reaction tube with lipids, glass beads and siRNA solution in the horizontal DAC-insert and spin for total 30 min at 3,540 rpm (six times 5 min). 7. Redispersion of VPG to normal (liquid) liposome formulations: The resulting VPG is redispersed by adding a 200% excess of PBS (40 or 120 mL) and mixing by DAC for two times 1 min in the horizontal DAC insert. Finally, place the reaction tube in a vertical DAC insert and spin down for 30 s. The resulting CL and SSL can be stored at 4°C. 8. Determination of liposome sizes: The diameter of prepared liposomes are to be determined with a photon correlation spectroscope (PCS) by diluting 2–3 mL of the redispersed liposomal formulations in approx. 2 ml PBS which should result in a count rate of about 300 kHz. PCS measurements are performed with standard settings (23°C, liquid viscosity

siRNA FRET in Liposomes

447

0.993, liquid index of refraction 1.333, Gaussian distribution) at a channel width of 10 ms for two times 5 min. The SSL result in a diameter between 110 and 130 nm whereas the diameter CL should be between 60 and 80 nm. 9. Removal of non entrapped siRNA: This can be performed by using inverted centrifugal spin filters (Centrisart I, 100.000 MWCO; see Note 9). For this, place the desired amount of sample in the outer tube of the centrisart device and dilute with PBS to 2.5 ml. Place the inner tube into the outer tube and let stand for 5 min to wet the membrane of the inner tube. Run the spin filters in a centrifuge at 2.500 × g (swingout-rotor) for three times 30 min (depending on the amount of the sample). If possible, perform centrifugation at 4°C. After each of the centrifugation runs take out the liposomes free filtrate from the inner tube to facilitate further filtration. The filtrate may be collected for further analysis. The purification is terminated when the inner tube reaches the bottom of the outer tube. Use a tweezers to take out the inner tube and collect purified sample from the outer tube. Pipetting or vortexing might be necessary to resuspend sedimented liposomes. After purification, a final volume of 100–300 mL should be obtained (depending on used amount of sample). To ensure all free siRNA has been removed, an additional purification step is recommended. The purified sample can be stored at 4°C. Up to 4 month no release of siRNA should occur. 10. Determination of entrapping efficiencies (EE): In general, entrapping efficiency can be determined by measuring the amount of the entrapped molecules and division of this amount by the total amount of molecules used for liposome preparation. Here, a fluorescent labeled siRNA is used, so EE can easily be measured by means of fluorescence spectroscopy. Therefore the liposomes before and after purification need to be cracked and quantified in a fluorescence detector. If available, HPLC analysis is recommended but the measurements are also possible in a micro cuvette (50 mL) of a fluorescence spectrometer. Cracking procedure should be done in a two-step dilution of the pre- and post-purification samples. The initial dilution of 1:3 is performed with a PBS containing 10% TritonX-100 (e.g., 10 mL sample plus 20 mL PBS/Triton) and second dilution of 1:7 with PBS. The final sample should contain approximately 1% TritonX-100 to prevent phospholipids from reassembling to micelles or vesicles. EE can be calculated by the following formula:

EE =

c(purified) × 100% c(not purified)

448

Kim et al.

3.4. Transfection with Liposomal Formulations

The quantities below are calculated for the transfection of 1-well of a 12-well plate. It should be noted that, depending on the liposome composition, the quantity of applied liposomally formulated siRNA may have to be drastically increased in order to observe any transfection at all. 1. Mix 30 pmol of siRNA in liposomes with transfection medium to a final volume of 500 mL. 2. Apply to the well and rock the plate to evenly distribute the mix. 3. Once the cells have been incubated at least for 3 h, replace the transfection medium with 3× transfection medium in order to restore nutrient conditions. Incubate the plate e.g., for 4, 24, 48 h at 37°C in the incubator (see Note 10) before fixation and imaging.

3.5. Transfection with Cationic Lipids

A standard protocol for transfection of HeLa cells (5) with oligofectamin was adapted to RBE4 cells. The following protocol can be adapted for many other adherent cell cultures, e.g., human embryonic kidney (HEK) cells. The quantities below are calculated for the transfection of 1-well of a 12-well plate. 1. Mix 6 mL (30 pmol; 1.8 × 1013 molecules, ~108 molecules per cell) of annealed siRNA with 100 mL of transfection medium. For control, use Oligofectamine instead of siRNA. 2. In a separate tube, mix 6 mL Oligofectamine with 24 mL of transfection medium. Mix the tube gently by inverting. Incubate at room temperature for 5 min. 3. Combine the solutions prepared in steps 1 and 2 and incubate an additional 15 min at room temperature. 4. During that time, replace the growth medium with 0.364 ml of transfection medium. 5. Add 136 mL of lipoplex complexes to one well and rock the plate to evenly distribute the mix. 6. Once the cells have been incubated at least for 3 h, replace the transfection medium with 3× transfection medium in order to restore nutrient conditions. Incubate the plate for 4, 24, 48 h at 37°C in the incubator (see Note 10) before fixation.

3.6. Microinjection

1. One day prior to microinjection, plate ~5 × 105 cells in small droplets in the center of glass cover slips (Ø 12 mm) that were placed in a small petri dish (Ø 35 mm) (see Note 11). 2. Incubate at 37°C until cells attach to the glass (6–8 h). 3. Add carefully 2 ml growth medium to petri dish and let cells grow overnight.

siRNA FRET in Liposomes

449

4. Transfer coverslips with cells into a petri dish filled with at least 2 ml of a carbonate-free a-MEM medium containing 25 mM HEPES (without PenStrep and FCS) prior to injection. 5. For injection, pull capillaries with an automatic, high precision puller (P-87 Brown Flaming Micropipette Puller, Sutter Instrument Company, Novato CA, USA). 6. Centrifuge ~20 mL labeled siRNAs solution for 5 min at 10,000 × g to avoid small particles that might block the capillary. 7. Load ~2 mL siRNA solution from the back into capillary with microloader. 8. In each experiment, 50 cells were injected into the cytoplasm, each cell with 200 fl of the siRNA solution (50 mM, corresponding to ~6 × 106 siRNA molecules per cell) (see Notes 12 and 13). Time needed for precise injection of a sample of 50 cells was ~15 min. The microinjection was performed using an automated computer-assisted Zeiss AIS injection system (Carl Zeiss Jena, Germany) equipped with an Eppendorf 5242 pneumatic injector (Eppendorf, Hamburg, Germany) for precise control of the injection pressure. Working pressure for injection in adherent cells was ~200 hPa and the injection time per individual cell 0.2 s. 9. Fix cells immediately after microinjection procedure (see Subheading 3.7) and store at 4°C in the dark until imaged. 3.7. Cell Fixation After Transfection or Microinjection

1. Before fixation, wash cells on cover slips by dipping cover slips three times in PBS (see Note 14) at room temperature (all following steps at room temperature). 2. Fix the cells in 4% formaldehyde for 15 min. 3. Formaldehyde is discarded into a hazardous waste container and the samples washed three times in PBS 2 min each. 4. Add DAPI for 5 min to obtain nuclear staining. 5. Dip cover slip shortly in PBS, cover cells with a drop of mounting medium and carefully invert onto a microscope slide. Let mounting medium dry in dark until completely dry (~5 h) and view with microscope or store in the dark at 4°C for up to a month.

3.8. Spectral Imaging of the Microinjected Cells

Spectral imaging was performed on a confocal laser scanning system (C1Si, Nikon, Badhoevedorp, The Netherlands) on an inverted microscope (TE2000-E, Nikon), which was equipped with an oil-immersion objective lens (Plan Fluor 40.0×/1.30/0.20).

450

Kim et al.

1. Images were taken with the following settings: excitation with Argon laser at the 12% intensity in standby modus at 488 nm; detection with 32-channel PMTs in the 495–650 nm range at 5 nm wavelength resolution; dichroic mirror 1 20/80 beam splitter was used; spectral gain was 180; pixel dwell 6 ms; pinhole size was medium (M); 3 scans were averaged (average count) (see Note 15). Wavelength sensitivity was corrected automatically. 2. Images were processed with EZ-C1 FreeViewer software (Gold Version 3.20 build 610) which can be used without the C1 set, only the hardware dongle (copy protection device) is needed. Whole cell, cytoplasm and nucleus were marked as areas of interest and corresponding spectra could be viewed and emission values imported to Excel. A True Color image overlay of 32 images in the 495–650 nm range is shown in Fig. 1a. 3.9. Confocal Microscopy for Standard Mode

Higher spatial resolution is obtained in standard mode. Cells were imaged on a DM IRBE TCS MP1 confocal microscope (Leica, Wetzlar, Germany) with an oil-immersion objective (63× magnification; NA 1.32) and an Argon laser for excitation at 488 nm, operated with software version 2.61_Build_1537. A triple dichroic filter (TD 488/544/633) was used as a main excitation beam splitter. The image size width and height were 158.7 mm. 1. The following standard settings were applied to all samples: Gain: 830/Offset: 53.7/Pinhole 228 mm/Excitation: 488 nm/ Laser intensity: 100%/PMT1: 510–540 nm (green channel)/PMT2: 570–600 nm (red channel). In order to avoid pixel saturation, certain additional images were recorded at lower gain values (750, 600 and 500). Examples for red and green channel images are shown in Fig. 1d. 2. Ten images of 512 × 512 8-bit-pixels were recorded with a confocal plane distance of 20 mm. z-stacks were composed from these images. 3. It is advisable to record phase contrast images in addition.

3.10. Calculation of Ratio Images

The cells are imaged at a reference laser gain as used in standard settings (see the previous section). To cope with pixel saturation, additional images are recorded at lower gains. The ratio image is computed based on all images acquired with different laser gains: 1. The intensity values of the red and green channel of an image pair j are added voxel-wise yielding a sum image. 2. For all voxels of the sum image with an intensity value less than a manually set threshold (e.g., T = 200) the corresponding voxels in the red and in the green channel are used to calculate

siRNA FRET in Liposomes

451

the mean background intensity rj of the red channel and gj of the green channel. 3. For each image pair j the ratio for each voxel pair rj (x) and gj(x) at voxel position x is computed as R/Gj(x) = (rj(x)−rj)/ (gj(x)−gj) (see Note 16). 4. The ratio image R/Gj(x) is assembled by choosing for each voxel x the ratio R/Gj(x) from the image pair j with highest gain, where neither rj(x) nor gj(x) have reached the saturation value (see Note 17) 3.11. Visualization of Ratio Images

To enhance the interpretation of a generated ratio image only certain parts of this image are displayed. To this end ratio voxels originating from low intensity values in the images obtained at the reference laser gain are masked out by a ratio mask. Additionally, the region of the corresponding cell is visualized using a cell mask. The ratio mask is computed as follows (see Fig. 2a):

Fig. 2. Flow charts for the computation of (a) the ratio mask, (b) the cell mask, and (c) the colored ratio image

452

Kim et al.

1. The intensity values of the red and green reference channel images are added voxel-wise to yield a sum image. 2. The mean intensity and standard deviation of the sum image are computed. 3. The mean intensity and the standard deviation are added to obtain a threshold value. Thresholding of the sum image is performed by evaluating for each voxel, whether all voxels within a 3 × 3 × 3 neighborhood have an intensity value lower than the threshold value. In this case, the intensity value of a voxel is set to zero. 4. A morphological dilation operation is applied to the thresholded image using a kernel size of 3 × 3 × 3 voxels. The cell mask is computed from the reference channel images as follows (see Fig. 2b): 1. Mean intensity and standard deviation are calculated for the red and green images separately. 2. The red and green channel images are thresholded separately, where the mean plus the standard deviation is used as threshold value. 3. The thresholded images are added voxel-wise yielding a sum image. 4. A median filter is applied to the sum image using a kernel size of 3 × 3 × 3 voxels. 5. A mean filter is applied to the resulting image using a kernel size of 3 × 3 × 3 voxels. 6. The obtained image is thresholded using an intensity value of 10. For the ratio mask, voxels with non-zero ratio values indicate structures to be displayed. For the cell mask, voxels with non-zero values indicate cell structures to be displayed. The final visualization of a ratio image is based on the ratio mask, the cell mask, and the computed ratio values (see also Fig. 2c): 1. If the ratio mask has a non-zero value at a voxel x, this voxel is assigned the ratio value R/G(x) using a color code (from green to red) and lower as well as upper ratio bounds of 1.1 and 2.1 (see Note 18). 2. If the ratio mask has a zero value and the cell mask has a nonzero value at a voxel x, this voxel is assigned the cell mask intensity value using a color code (gray-scale colors in RGB) (see Note 19). 3. If both the ratio mask and the cell mask have a zero value at voxel x, then the intensity of this voxel is assigned the color black (the voxel is interpreted as background).

siRNA FRET in Liposomes

453

Readers interested in using the software for computing ratio images should contact I.-H. K, ([email protected]), K.R., ([email protected]) or M.H. ([email protected]).

4. Notes 1. siRNAs are ordered (or synthesized) which have the antisense oligo labeled with tetramethylrhodamine (TMR or TAMRA) on the 5¢-end, and the sense oligo labeled with fluorescein (FAM) on the 3¢-end. Annealing of such strands consistently produces siRNA whose fluorescence spectrum upon excitation at 488 nm shows emission at 590 nm at least threefold stronger than at 520 nm. 2. Paraformaldehyde could be used instead of formaldehyde. Prepare freshly a 4% (w/v) solution in PBS. The solution should be heated below 60°C on a stirring hot-plate in a fume hood until dissolved (some drops of NaOH would speed up dissolving), then cooled to room temperature. 3. 1× PBS can be used instead of annealing buffer (2×: 200 mM potassium acetate, 4 mM magnesium acetate, 60 mM HepesKOH, pH 7.4; Elbashir et al., 2002). 4. Working with siRNA also the rules for working with RNA do apply (e.g., using gloves to avoid RNase contamination, using single-thawed small aliquots). Work with siRNA (except seed-vac) was performed under sterile hood also because of cell culture application. siRNA is always kept on ice as much as possible to reduce the rate of RNA hydrolysis. Intactness of siRNA single strands could be proved by checking the concentration at 260 nm. 5. The ratio of complementary sense and antisense strands was equimolar (one to one) except 5TMR-As in 1.5 molar ratio to the homologous strand. 6. Cells with passage number 30 were used, although Elbashir warns that to exceed a passage number of 30 after thawing the stock culture might affect siRNA transfection efficiencies (5). 7. Instead of weighing the phospholipids in the reaction tube it is easier and more reproducible to prepare stock solutions of phospholipids and cholesterol in EtOH (99% p.a.). We suggest a 50 mg/ml solutions of MPEG2000, a 100 mg/ml solution of EPC-3 and a 25 mg/ml solution of cholesterol. The volume is adjusted with EtOH to 500 mL. 8. The evaporation in the speed-vac should be performed at low temperature without any heating, as delayed boiling may occur and spread the sample all over the speed-vac.

454

Kim et al.

9. The centrisart I filter devices are not guaranteed to be sterile. Alternatively a sterile purification might be achieved by performing mini-scale dialysis (not tested). 10. It should be tested for each individual cell line, whether transfection could be performed in the presence of serum and/or transfection medium replacement with 3× transfection medium could be omitted. 11. More cover slips with cells were prepared in order to choose the most suitable for microinjection, those with 50–70% of confluence. There should be always enough cells to microinject and cell close to each other are easier to inject. 12. Filled capillary should always remain in the medium as a little liquid on the capillary top can dry very quickly blocking it. From 20 mL of siRNA many capillaries could be filled and with one capillary two or three cover slips could be injected at least until it gets blocked by particles. 13. It is helpful to have some guidance cue on a cover slip (a scratch done with diamond knife with on line crossing) such that one can find the microinjected cells easily and repeatedly. Although injected cells have bright fluorescence and are therefore easily distinguished from uninjected cells, cues can save time. 14. For dipping cover slips into PBS, small beakers were filled with PBS. For fixation and DAPI staining cover slips were placed in 6-well plate wells filled with formaldehyde solution or DAPI solution. Cover slips were taken out with forceps. The plates could be washed and reused. 15. Microscopy settings (Laser intensity, dwell size, pinhole) should be determined by user for the specific application. 16. If a voxel has an intensity value smaller than or equal to the mean value of the background of the respective channel, the result of the subtraction of the intensity value from the mean background value is replaced by a value of 1. 17. If all ratios R/Gj have been computed from voxels with saturated intensity value, the ratio value is chosen from the image pair obtained at the lowest laser gain. 18. The upper and lower bounds are determined from the histograms of images obtained from control experiments. 19. The minimum and maximum intensity values of the thresholded image (not including the value zero) can be used for a conversion to the full gray-scale intensity range. Alternatively, a smaller gray-scale intensity range can be used to adjust brightness and contrast of the displayed structures.

siRNA FRET in Liposomes

455

Acknowledgments M.H. gratefully acknowledges the Heidelberg Rotary Club for funding, Heike Einberger for technical assistance, and the Nikon Imaging Center at the University of Heidelberg. We thank Andres Jäschke for generous support. References 1. Järve A, Müller J, Kim IH, Rohr K, MacLean C, Fricker G, Massing U, Eberle F, Dalpke A, Fischer R, Trendelenburg MF, Helm M (2007) Surveillance of siRNA integrity by FRET imaging. Nucleic Acids Res 35:e124 2. Zimmermann TS, Lee AC, Akinc A, Bramlage B, Bumcrot D, Fedoruk MN, Harborth J, Heyes JA, Jeffs LB, John M, Judge AD, Lam K, McClintock K, Nechev LV, Palmer LR, Racie T, Rohl I, Seiffert S, Shanmugam S, Sood V, Soutschek J, Toudjarska I, Wheat AJ, Yaworski E, Zedalis W, Koteliansky V, Manoharan M, Vornlocher HP, MacLachlan I (2006) RNAi-mediated

gene silencing in non-human primates. Nature 441:111–114 3. Massing U, Cicko S, Ziroli V (2008) Dual asymmetric centrifugation (DAC)–a new technique for liposome preparation. J Control Release 125:16–24 4. Hirsch M, Ziroli V, Helm M, Massing U (2009) Preparation of small amounts of sterile siRNA-liposomes with high entrapping efficiency by dual asymmetric centrifugation (DAC). J Control Release 135:80–88 5. Elbashir SM, Harborth J, Weber K, Tuschl T (2002) Analysis of gene function in somatic mammalian cells using small interfering RNAs. Methods 26:199–213

Chapter 31 Spectral Bio-Imaging and Confocal Imaging of the Intracellular Distribution of Lipoplexes Sebastian Schneider and Regine Süss Abstract The intracellular distribution of nanoparticular drug delivery systems is very complex, but its investigation yields high potential for further development and optimization of these systems. In the following chapter, we introduce the application of fluorescent imaging techniques in order to highlight uptake and cellular processing of nanoparticular drug delivery systems (e.g., liposomal drug delivery systems). We selected a combination of different protocols for the staining of the most important endocytic compartments and organelles. The presented imaging systems are appropriate to detect liposomal drug delivery systems localized in these cellular structures. Key words: Nanoparticular drug delivery systems, Fluorescence microscopy, Endocytosis, Colocalization studies, Intracellular trafficking

1. Introduction In liposome research, investigating the intracellular distribution of liposomal drug delivery systems (e.g., liposomes, lipoplexes) has become an important objective. Following the association to the cell surface, liposomes are mostly processed by endocytic pathways. As the cellular processing determines the intracellular fate of liposomes, investigation of these pathways is a promising strategy for the optimization of liposomal drug delivery systems. In order to track the cellular processing of liposomes, a combination of different fluorescence-based analytical methods is generally accepted: fluorescence-activated cell sorting analysis (flow cytometry) and fluorescence microscopy (1, 2). Fluorescence-activated cell sorting analysis (flow cytometry) provides quantitative information and is convenient to measure V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_31, © Humana Press, a part of Springer Science+Business Media, LLC 2010

457

458

Schneider and Süss

liposomal uptake. In blocking experiments, liposomes are coincubated with endocytic inhibitors that specifically block an endocytic pathway. Flow cytometry allows for the quantification of the change in liposomal uptake and therefore for the assessment of the influence of the respective endocytic pathway. However, a drawback of this method is the difficulty in distinguishing between material that is attached to the cell surface and material that is actually taken up by the cell. Fluorescence microscopy provides information on the spatial position of liposomes in living or fixed cells. Therefore, endocytic compartments (e.g., endosomes, lysosomes, caveosomes) and organelles (e.g., endoplasmic reticulum, golgi apparatus), as well as the liposomes must be stained. In this study, three different staining techniques for the two endocytic pathways “clathrin” and “caveolae” are applied (Fig. 1) (see Note 1): (1) Fluorescent endocytic markers (e.g., transferrin) with well-known cellular distribution are coincubated with fluorescent liposomes. (2) Cells are transfected with reporter genes expressing fluorescent endogenous markers for different endocytic compartments or organelles and incubated with fluorescent liposomes. (3) Endocytic compartments and organelles can be detected and stained with antibodies (immunofluorescence) after incubation with fluorescent liposomes. In colocalization studies, two or more fluorescent signals overlap in the final image due to their proximity within the

Fig. 1. Staining techniques for the two endocytic pathways “clathrin” and “caveolae”. MDCK cells were incubated with fluorescent markers, immunostained with antibodies or transfected with reporter genes as indicated in Table 1

Spectral Bio-Imaging and Confocal Imaging

459

microscopic structure. The accurate performance of colocalization studies demands high spatial and spectral resolution. There are two different imaging techniques employed in this study. The confocal scanning microscope provides high spatial resolution. Fluorescent microscopes often render limited information about spectral features. The SpectraCube™ SD-200 H on the other hand provides high spectral resolution. The spectrum of emitted light for every single pixel in a 2D picture is measured. Using a mathematical pixel classification and comparing the collected spectra with spectra collected in a library it is possible to separate even dyes with overlapping spectra. Thus, the system achieves spectral resolutions of 5–12 nm (depending on the wavelength). But as an inverted fluorescence microscope the spatial resolution is quite low (3). The combination of the two methods therefore delivers images with both high spectral and spatial resolution. These images help understanding uptake and intracellular processing and therefore may lead to more successful strategies to improve liposomal drug delivery systems. This study presents a broad field of different staining procedures and imaging techniques illustrated by the example of cationic liposome/DNA complexes in Madin-Darby canine kidney (MDCK) cells (see Notes 2 and 3).

2. Materials 2.1. Cell Culture

1. Madin-Darby canine kidney cells (MDCK II) (American Type Culture Collection, Manassas, USA, CCL-34) (see Note 2). 2. DMEM (Dulbecco’s Modified Eagle Medium) supplemented with 10% FCS (Biochrom KG, Berlin, Germany). 3. Phosphate buffer saline (PBS) (Biochrom KG, Berlin, Germany). 4. Solution of trypsin and ethylenediamine tetraacetic acid (EDTA) (0.05%/0.02% (w/v)) (Biochrom KG, Berlin, Germany). 5. Gelatine 0.2% (w/v) in PBS (Sigma-Aldrich, Steinheim, Germany). Solution was autoclaved and sterile filtrated prior to use. 6. CellScrub™ washing buffer (Genlantis, San Diego, CA). 7. Albumin bovine essentially fatty-acid-free (DF-BSA) (SigmaAldrich, Steinheim, Germany). 8. 48-well plates (Becton Dickinson, Heidelberg, Germany).

2.2. Lipoplexes ( See Note 3)

1. Transfection medium consists of 25 mM sodium chloride and 250 mM saccharose.

460

Schneider and Süss

2. DC30® consists of DC-Cholesterol (3b [N-(N¢,N¢dimethylaminoethane)-carbamoyl] cholesterol-HCL) and DOPE (dioleoyl phosphatidylethanolamine) 3:7 (w/w) (Avanti Polar Lipids, Birmingham, USA). 3. pEGFP-C1 plasmid (4,7 kb) (BD Clontech, Heidelberg, Germany). 4. Lipoplex staining: Label IT® Cy™3 Nucleic Acid Labeling Kit (Mirus, Madison, WI). 2.3. Fluorescence Microscopy

1. Coverslips (8 mm) and slides (3 × 1 inch) (Langenbrinck, Teningen, Germany). 2. Endocytic markers: (a) Nuclear staining: 10 nM 4,6-diamidino-2-phenylindole (DAPI) in PBS (Molecular Probes, Leiden, Netherlands). (b) Human serum transfer in, Alexa Fluor® 488/594 conjugate (Molecular Probes, Leiden, Netherlands). (c) Cholera Toxin subunit B, Alexa Fluor® 488/594 conjugate (Molecular Probes, Leiden, Netherlands). (d) Bodipy® FL C5-lactosylceramide complexed to BSA (Molecular Probes, Leiden, Netherlands). 3. Immunofluorescence: (a) Primary antibodies: Mouse anti-EEA1 IgG (250 µg/ml) (Abcam, Cambridge, UK); Rabbit anti-LAMP-1 (1 mg/ml) (Sigma-Aldrich, Steinheim, Germany); Mouse anti-Caveolin-1 IgG1 (250 µg/ml) and Mouse anti-Caveolin-1 IgM (250 µg/ml) (Becton Dickinson, Heidelberg, Germany). (b) Secondary antibodies: Goat anti-mouse IgG (H+L) MFP 488 (1 mg/ml), Goat anti-mouse IgG (H+L) MFP 590 (1 mg/ ml) and Goat anti-rabbit IgG (H+L) MFP 488 (1 mg/ml) (Mobitec, Göttingen, Germany); Goat anti-mouse IgM (µ-chain specific) FITC (1 mg/ml) (Sigma-Aldrich, Steinheim, Germany), Goat anti-mouse IgG1 (~1) Alexa Fluor® 594 (2 mg/ml) (Molecular Probes, Leiden, Netherlands). (c) Permeabilization solution: Triton X-100 in PBS 0.1% (v/v) or saponin (1 mg/ml) in PBS (Sigma-Aldrich, Steinheim, Germany) (see Note 4). (d) Blocking solution: PBS containing 10% FCS. Prepare afresh for each experiment. 4. Endocytic reporter genes: (a) Transfection reagent: FuGene® HD (Roche Diagnostics, Mannheim, Germany). (b) –T2.YFP plasmid, a generous gift from K. Simons, MaxPlanck-Institute of Molecular Cell Biology and Genetics, Dresden, Germany.

Spectral Bio-Imaging and Confocal Imaging

461

(c) Caveolin-EGFP plasmid, a generous gift from L. Pelkmans, A. Helenius, ETH, Zürich, Switzerland. 5. Fixation: Paraformaldehyde: Prepare a 4% (w/w) solution in PBS afresh for each experiment. Dissolve in a 50°C water bath for 30 min and then cool down to room temperature for use (adjust to pH 7.4 if necessary) (see Note 5). 6. Mounting: Mounting medium: MoBiGlow (MoBiTec, Göttingen, Germany). Prepare aliquots and store at −80°C. 7. Image acquisition: Images are obtained using either a Zeiss LSM 510 Meta from Carl Zeiss, Jena, Germany with a 63X/1.4 NA or a SpectraCube™ SD-200 H from Applied Spectral Imaging, Migdal HaEmek, Israel with a 100X/1.3 NA. Confocal image recording and image analysis are performed with LSM Image Browse Rel. 4.0. Unmixed images are captured using Spectral Imaging 2.5 software with an acquisition time of 90 s. For analysis of unmixed spectra images are transferred to SpectraView 1.6 software.

3. Methods 3.1. C ell Culture

1. Madin-Darby canine kidney cells (MDCK II) are maintained in DMEM supplemented with 10% FCS in a 37 °C incubator (humidified atmosphere containing 5% CO2). 2. Cells approaching confluency are passaged by treatment with trypsin/EDTA solution. A subcultivation ratio of 1:6–1:10 is recommended with medium renewal at every 2–3 days. 3. Coverslips are sterilized at 200 °C for 2 h. They are placed into 48-well plates, coated with gelatine for 10 min at 37 °C, and allowed to cool down to room temperature. Wash twice with PBS. 4. For uptake and internalization studies, MDCK cells are seeded onto these coated coverslips in 48-well plates at a density of 2–4 × 104 24 h prior to experiment.

3.2. Preparation of Lipoplexes

1. DC30® lipid is hydrated in serum-free transfection medium for 30 min. 2. An equal volume of pEGFP plasmid solution is added to DC30® liposomes resulting in the formation of lipoplexes with a charge ratio of 1.6:1 (cationic lipid:DNA). Dilutions are combined discontinuously by pipetting the plasmid into the liposomes. Mixing is performed by vortexing for 3 s. 3. Lipoplexes are used 20 min after mixing at a concentration of 0.5 µg DNA per 48-well. A final amount of 100 µl lipoplex in a total volume of 500 µl per 48-well is recommended.

462

Schneider and Süss

4. For uptake experiments pEGFP plasmid is labeled with Cy3™ according to the manufacturer´s instructions. Complex formation is performed as described above and cells are incubated with Cy3™-labeled DC30® lipoplexes for 5 h. 3.3. Fluorescence Microscopy 3.3.1. Colocalization with Endocytic Markers

1. Endocytic markers Transferrin Alexa Fluor ® 488/594 conjugate (5 µg/ml cell culture medium), Cholera Toxin subunit B Alexa Fluor® 488/594 (10 µg/ml cell culture medium) and Bodipy® FL C5-lactosylceramide complexed to BSA (1 µM in cell culture medium) are coincubated with DC30® lipoplexes for 5 h. 2. Incubate with 300 µl CellScrub™ washing buffer for 10 min and wash twice with PBS (see Note 6). 3. Subsequently cells are fixed with 500 µl paraformaldehyde solution for 30 min at room temperature. 4. Wash cells twice with PBS. 5. Cells are incubated with 700 µl DAPI solution for 20 min at room temperature in order to stain the nucleus (see Note 7). 6. Wash cells three times with PBS and the samples are ready to be mounted.

3.3.2. Immunofluorescence

1. Cells are incubated with DC30® lipoplexes for 5 h. 2. Subsequently the samples are washed, fixed with paraformaldehyde solution and stained with DAPI as described above (see Notes 4, 6, and 7). 3. The cells are permeabilized by incubation with 500 µl saponin solution for 10 min at room temperature. Alternatively, cells can be permeabilized with 500 µl Triton X-100 solution for 5 min at 4 °C (see Note 4). 4. Before processing for immunofluorescence labeling, cells are incubated with blocking solution for 30 min at room temperature. 5. Dilute primary antibodies in blocking solution: Mouse antiEEA1 IgG at 1:25 (v/v), Rabbit anti-LAMP-1 at 1:20 (v/v), Mouse anti-Caveolin-1 IgG1 at 1:25 (v/v) and Mouse antiCaveolin-1 IgM at 1:25 (v/v). 6. For the staining procedure prepare microscope slides with Parafilm® as follows: stick the Parafilm® tightly on the slide. Remove the coverslips from the 48-well plates and place them cell-side down on a 50–100 µl drop of primary antibody solution deposited on the Parafilm®-covered slide. Protect the staining support from light (e.g., into a Petri dish covered with aluminium foil) (see Note 8). 7. Cells are incubated with primary antibodies at room temperature as indicated (Table 1) (see Notes 9 and 10). Coverslips

Goat anti-rabbit IgG (H+L) MFP 488

Goat anti-mouse IgG1 (ү1) Alexa Fluor® 594

Rabbit anti-LAMP-1

Mouse anti-Caveolin-1 IgG1

48 h

EGFP

YFP

Caveolin-EGFP plasmid

–T2.YFP plasmid

1 µg/well

48 h

30 min

Mouse anti-Caveolin-1 IgM Goat anti-mouse IgM (µ-chain 1:25 specific) FITC 1 µg/well

30 min

60 min

30 min

5 h

5 h

5 h

Time

1:25

1:20

1:25

Goat anti-mouse IgG (H+L) MFP 488/590

Mouse anti-EEA1 IgG

10 µg/ml

1 µM

Alexa Fluor® 488/594

Cholera Toxin (CTX)

5 µg/ml

Concentration/dilution (v/v)

Lactosylceramide (LacCer) Bodipy®

Alexa Fluor® 488/594

Transferrin

Dye conjugate/Sec. antibody

Table 1 Fluorescent markers, antibodies and reporter genes used in this study

The caveolae pathway can target the golgi apparatus (11).

This conjugate is located in the same compartments as the single protein (10).

It is also found in an apical endosomal compartment and at the plasma membrane (9).

Caveolin-1 is a protein localized at the golgi apparatus.

LAMP-1 is a membrane protein associated to late endosomes and lysosomes (8).

EEA1 is a hydrophilic peripheral protein on early endosomes (7).

LacCer internalization is caveolae dependent (6).

CTX can enter cells by numerous modes of endocytosis (5). In cells with high caveolin-1 levels (like MDCK) it is internalized via caveolae pathway (6).

The transferrin protein and its receptor are internalized via clathrin pathway (4).

Endocytic pathway

Spectral Bio-Imaging and Confocal Imaging 463

464

Schneider and Süss

are removed from Parafilm® and returned to a new Petri dish. Wash cells three times with PBS. 8. Dilute secondary antibodies in blocking solution: Goat antimouse IgG (H+L) MFP 488 at 1:25 (v/v), Goat anti-mouse IgG (H+L) MFP 590 at 1:25 (v/v), Goat anti-rabbit IgG (H+L) MFP 488 at 1:25 (v/v), Goat anti-mouse IgM (µ-chain specific) FITC at 1:50 (v/v) and Goat anti-mouse IgG1 (ү1) Alexa Fluor® 594 at 1:25 (v/v). 9. All secondary antibodies are incubated for 60 min at room temperature as described above (see Notes 8, 9, and 10). To check for compatible secondary antibodies see Table 1. 10. After three additional washing steps with PBS the samples are ready to be mounted. 3.3.3. Colocalization with Endocytic Reporter Genes

1. Cells are transiently transfected with –T2.YFP plasmid using FuGene® HD in a ratio of 4:2 (v/w) with 1 µg DNA/well in a 48-well plate. Transfection complexes are formed in transfection medium for 10 min and then diluted in DMEM containing FCS. 2. Transfection reagent/DNA complexes are removed after 3 h and replaced with fresh medium. 3. 24 and 48 h post transfection gene expression is observed by fluorescence microscopy. 4. After 48 h medium is renewed and cells are incubated with DC30® lipoplexes for 5 h. Subsequently the samples are washed, fixed with paraformaldehyde solution and stained with DAPI as described above (see Notes 6 and 7). 5. Alternatively, immunofluorescence is performed as stated above.

3.3.4. Mounting

1. Coverslips are mounted on glass slides with 3 µl MoBiGlow, an antifading substance to reduce photobleaching. The sample can be viewed immediately as soon as the mounting medium is dry or stored at 5 °C in the dark.

3.3.5. Image Acquisition

1. The confocal laser scanning microscope Zeiss LSM 510 Meta is applied for optical sectioning in the Z axis. Capturing a series of images allows for visualizing the entire sample in 3D (full projection). Thus, the exact spatial position is clarified (Fig. 2) (see Note 7). 2. The inverted fluorescence microscope SpectraCube™ SD-200 H is a Fourier-transformation based system which enables the measurement of spectral information from each pixel of an observed image. First, samples with only one dye are imaged and the corresponding spectra acquired. Secondly, based on

Spectral Bio-Imaging and Confocal Imaging

465

Fig. 2. Optical sectioning in the z axis. MDCK cells were incubated with Cy3™-labeled DC30® lipoplexes for 5 h. Fixed cells were stained with Mouse anti-EEA1 IgG as indicated in Table 1. Images show the x–y projection of a confocal stack. (a) Merged image. Single color images of Cy3™ (b) and MFP 488 (c). Corresponding x–z sections, collected at the plane marked by the arrow, are presented below

the generated single spectra the SpectraView software is able to separate fluorescent dyes in a multifluorescent sample (Fig. 3) (see Note 9).

4. Notes 1. At present, one can classify five major mechanisms of endocytosis: phagocytosis, macropinocytosis, clathrin-, caveolae-, and non-clathrin-non-caveolae-dependent endocytosis. We focused on two important endocytic pathways for nanoparticular drug delivery systems, clathrin- and caveolae-dependent endocytosis. 2. In principle this protocol can be adapted to other cell lines. However, it will be necessary to optimize incubation time and concentration for every endocytic cargo (e.g., lipoplex, endocytic marker). Generally, necessary concentration and incubation time of primary and secondary antibodies vary to a minor degree. The presented markers, antibodies and plasmids are just a feasible selection of commercially available substances. 3. Staining procedures and imaging techniques described in these protocols can be adapted to other drug delivery systems in order to investigate their endocytic behavior. 4. Note that immunofluorescence can be antibody– and fixation-dependent. The caveolin-1 antibodies used in this

466

Schneider and Süss

Fig. 3. Unmixing with the SpectraCube™. MDCK cells were incubated with Cy3™-labeled DC30® lipoplexes for 5 h. Fixed cells were stained with DAPI and Mouse anti-Caevolin-1 IgM as indicated in Table 1. (a) Original view. Spectrally decomposed single color images of Cy3™ (c), FITC (d) and DAPI (e) are shown. (b) Overlay of the single color images. Arrows indicate positions of colocalization

study are raised against the same protein but recognize different subcellular pools of the protein. Thus, changing primary antibodies or fixation conditions (e.g., instead of paraformaldehyde/saponin, methanol/acetone or paraformaldehyde/ Triton X-100) leads to the binding of different caveolin-1 epitopes resulting in a different subcellular image (9). 5. Ready-to-use paraformaldehyde solution can be distributed in aliquots and stored at −20 °C for several months. 6. Reduction of out-of-focus fluorescence background is a prerequisite for successful imaging, especially for an inverted fluorescence microscope. Therefore, washing steps have to be carried out carefully. Prior to fixation with paraformaldehyde all serum proteins need to be rinsed out. Lipid fluorescent markers (e.g., Bodipy® LacCer) can be removed by incubating with 5% DF-BSA in PBS. Cell-surface associated lipoplexes are washed out with CellScrub™ washing buffer. 7. In order to assure high spatial resolution an optical thickness of 0.35 µm is chosen for the confocal scanning microscope.

Spectral Bio-Imaging and Confocal Imaging

467

The distinction between emission lights of different dyes is only possible with a negligible degree of spectral overlap. The spatial resolution of the SpectraCube™ SD-200 H is quite low. Here, nuclear staining offers some orientation in the sample. The combination of dyes is not restricted. 8. This staining procedure reduces the necessary incubation volume so that the amount of antibody solution also decreases (150 µl in 48-well compared to a 50 µl drop in the described staining procedure). Keep in mind that the amount of antibody also depends on the cell density. This reduces the antibody saving effect and explains the relatively high antibody concentrations used in this study. Moreover, this staining procedure is confined to immunofluorescence studies with fixed cells. If labeling is performed with living cells (reporter gene, endocytic marker), their application causes toxic effects. 9. In order to reveal the extent of non-specific signal generated by the secondary antibody, the preparation of a control sample without primary antibody is recommended. 10. In the case of double immunofluorescence in one sample, check for cross reactivity beforehand. Antibodies raised in different species can be applied in one mixture.

References 1. van der Aa MA, Huth US, Hafele SY et al (2007) Cellular uptake of cationic polymerDNA complexes via caveolae plays a pivotal role in gene transfection in COS-7 cells. Pharm Res 24(8):1590–1598 2. Huth US, Schubert R, Peschka-Suss R (2006) Investigating the uptake and intracellular fate of pH-sensitive liposomes by flow cytometry and spectral bio-imaging. J Control Release 110(3):490–504 3. Huth U, Wieschollek A, Garini Y, Schubert R, Peschka-Suss R (2004) Fourier transformed spectral bio-imaging for studying the intracellular fate of liposomes. Cytometry A 57(1): 10–21 4. Karin M, Mintz B (1981) Receptor-mediated endocytosis of transferrin in developmentally totipotent mouse teratocarcinoma stem cells. J Biol Chem 256(7):3245–3252 5. Chinnapen DJ, Chinnapen H, Saslowsky D, Lencer WI (2007) Rafting with cholera toxin: endocytosis and trafficking from plasma membrane to ER. FEMS Microbiol Lett 266(2):129–137 6. Singh RD, Puri V, Valiyaveettil JT, Marks DL, Bittman R, Pagano RE (2003) Selective

caveolin-1-dependent endocytosis of glycosphingolipids. Mol Biol Cell 14(8):3254–3265 7. Mu FT, Callaghan JM, Steele-Mortimer O et al (1995) EEA1, an early endosome-associated protein. EEA1 is a conserved alpha-helical peripheral membrane protein flanked by cysteine “fingers” and contains a calmodulinbinding IQ motif. J Biol Chem 270(22): 13503–13511 8. Chen JW, Murphy TL, Willingham MC, Pastan I, August JT (1985) Identification of two lysosomal membrane glycoproteins. J Cell Biol 101(1):85–95 9. Bush WS, Ihrke G, Robinson JM, Kenworthy AK (2006) Antibody-specific detection of caveolin-1 in subapical compartments of MDCK cells. Histochem Cell Biol 126(1):27–34 10. Pelkmans L, Kartenbeck J, Helenius A (2001) Caveolar endocytosis of simian virus 40 reveals a new two-step vesicular-transport pathway to the ER. Nat Cell Biol 3(5):473–483 11. Le PU, Nabi IR (2003) Distinct caveolaemediated endocytic pathways target the Golgi apparatus and the endoplasmic reticulum. J Cell Sci 116(Pt 6):1059–1071

Chapter 32 Techniques for Loading Technetium-99m and Rhenium-186/188 Radionuclides into Pre-formed Liposomes for Diagnostic Imaging and Radionuclide Therapy Beth Goins, Ande Bao, and William T. Phillips Abstract Liposomes can serve as carriers of radionuclides for diagnostic imaging and therapeutic applications. Herein, procedures are outlined for radiolabeling liposomes with the gamma-emitting radionuclide, technetium-99m (99mTc), for non-invasive detection of disease and for monitoring the pharmacokinetics and biodistribution of liposomal drugs, and/or with therapeutic beta-emitting radionuclides, rhenium186/188 (186/188Re), for radionuclide therapy. These efficient and practical liposome radiolabeling methods use a post-labeling mechanism to load 99mTc or 186/188Re into pre-formed liposomes prepared in advance of the labeling procedure. For all liposome radiolabeling methods described, a lipophilic chelator is used to transport 99mTc or 186/188Re across the lipid bilayer of the pre-formed liposomes. Once within the liposome interior, the pre-encapsulated glutathione or ammonium sulfate (pH) gradient provides for stable entrapment of the 99mTc and 186/188Re within the liposomes. In the first method, 99mTc is transported across the lipid bilayer by the lipophilic chelator, hexamethylpropyleneamine oxime (HMPAO) and 99mTc-HMPAO becomes trapped by interaction with the pre-encapsulated glutathione within the liposomes. In the second method, 99mTc or 186/188Re is transported across the lipid bilayer by the lipophilic chelator, N,N-bis(2-mercaptoethyl)-N’,N’-diethylethylenediamine (BMEDA), and 99mTc-BMEDA or 186/188Re-BMEDA becomes trapped by interaction with pre-encapsulated glutathione within the liposomes. In the third method, an ammonium sulfate (pH) gradient loading technique is employed using liposomes with an extraliposomal pH of 7.4 and an interior pH of 5.1. BMEDA, which is lipophilic at pH 7.4, serves as a lipophilic chelator for 99mTc or 186/188Re to transport the radionuclides across the lipid bilayer. Once within the more acidic liposome interior, 99mTc/186/188Re-BMEDA complex becomes protonated and more hydrophilic, which results in stable entrapment of the 99mTc/186/188Re-BMEDA complex within the liposomes. Since many commercially available liposomal drugs use an ammonium sulfate (pH) gradient for drug loading, these liposomal drugs can be directly radiolabeled with 99mTc-BMEDA for non-invasive monitoring of tissue distribution during treatment or with 186/188Re-BMEDA for combination chemoradionuclide therapy. Key words: Radionuclide, Radiolabeling, Liposomes, Scintigraphy, Imaging, Rhenium, Technetium99m, Nanoparticle, Radiopharmaceutical, Nuclear medicine

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_32, © Humana Press, a part of Springer Science+Business Media, LLC 2010

469

470

Goins, Bao, and Phillips

1. Introduction Gamma (photon)-emitting radionuclides such as technetium-99m (99mTc) have properties suitable for non-invasive detection by external scintigraphic gamma cameras and single photon emission computed tomography (SPECT) (1, 2). Liposomes radiolabeled with 99mTc can be used as diagnostic imaging agents (3–18) and as tools for tracking liposomal drugs during development (19). For example, during preclinical and clinical liposomal drug testing, non-invasive imaging has been used for the determination of the blood clearance kinetics and tissue distribution of the radiolabeled liposome vehicle after administration into the body (20–23). More recently, Kleiter et al. described a chemodosimetry technique using scintigraphic imaging where a fraction of 99mTc-liposomes added to the liposomal drug during administration could determine non-invasively and quantitatively how much of a drug is taken up by the targeted site without having to biopsy the tumor (24). Several common gamma-emitting radionuclides including 99m Tc, indium-111 (111In), gallium-67 (67Ga) and iodine-123 (123I) have been used for liposome radiolabeling (9, 25). Each of these radionuclides has different half-lives and photon energies, which can be tailored to a particular radiolabeled liposome application (2). The liposome radiolabeling methods described in this chapter will focus on the use of 99mTc. As the most common clinically used radionuclide, 99mTc is particularly ideal for single photon emission scintigraphic imaging studies because its 141 keV photon energy permits obtaining high quality images with current gamma camera technology. Another favorable feature of 99mTc for liposome radiolabeling is that it can be eluted daily from a commercially available molybdenum-99 (99Mo)-99mTc generator, thus making 99m Tc relatively inexpensive and readily accessible to most nuclear medicine departments. Finally, 99mTc is a pure gamma emitter with physical half-life of 6 h, which leads to minimal radiation absorbed doses to the patient and permits imaging up to 48 h after injection. Although this imaging time frame is adequate for many studies, longer half-life radionuclides such as 111In will need to be used if the physiological process to be monitored is slow and requires delayed imaging 48 h post-injection. In addition, liposomes radiolabeled with several therapeutic radionuclides including alpha- and beta-emitters have been described for use in radionuclide therapy (26–35). This chapter will describe methods for radiolabeling liposomes with therapeutic beta-emitting radionuclides, rhenium-186 (186Re) and rhenium-188 (188Re). The chemical labeling methods for 186Re and 188Re are essentially the same. When describing both rhenium radionuclides at the same time, the designation (186/188Re) will be used. For cancer therapy, an advantage of therapeutic beta-emitting radionuclides

Techniques for Loading Technetium-99m and Rhenium-186/188

471

is their short mm-range or therapeutic radiation distance. Beta particles emitted from the therapeutic radionuclide will effectively treat a radial field of tumor cells surrounding the deposited radionuclide, permitting tumor cell eradication within the therapeutic radiation distance of the beta particles. Normal tissues outside the therapeutic radiation distance will be spared, resulting in less radiation-induced toxicity to normal surrounding tissues. The beta particles from 186Re and 188Re have average ranges of 1.8 and 3.5 mm, respectively (36). Another advantage of using betaemitting radionuclides for cancer therapy is that the beta particles only have to be within range of the cancer cells to be effective, avoiding any complicated release kinetics encountered when using liposome-encapsulated chemotherapeutic agents. As radionuclides for therapy, 186Re and 188Re have many merits including: (1) similar chemical characteristics as 99mTc; (2) appropriate physical half-life of 90.6 h for 186Re and 17.0 h for 188Re for therapy use; (3) maximal beta energies of 1.07 MeV (71.0%) and 0.94 MeV (21.5%) for 186 Re, and 2.12 MeV (71.1%) and 1.96 MeV (25.6%) for 188Re; (4) very low ratio of high-energy gamma rays; (5) 137 keV (9.4%) and 123 keV (0.6%) gamma rays for 186Re and 155 keV (15.1%) for 188Re for monitoring the biodistribution and dose distribution in vivo by using SPECT imaging; (6) low bone affinity; and (7) high non-penetrating-to-penetrating energy ratio of 16.5 for 186 Re and 13.5 for 188Re, resulting in a very high radiation dose to the target but very low dose to normal tissues. An additional advantage of 188Re is the ability to obtain carrier free 188Re by elution from a cost-effective and readily available tungsten-188 (188W)-188Re generator (37). Various methods for stably radiolabeling liposomes with gamma-emitting and therapeutic beta-emitting radionuclides have been reviewed in the literature (3, 4, 8, 25, 38). These methods can be divided into two main categories: (1) radionuclide encapsulation within liposome aqueous interior or intercalation in lipid bilayer at the time of manufacture of the liposomes; and (2) post-labeling of the radionuclide into pre-formed liposomes prepared in advance and after long-term storage. Compared with radiolabeled liposomes prepared during manufacture, post-labeling methods are more convenient and cost-effective. Moreover, since most passive liposome encapsulation methods only result in about 10% encapsulation of starting material, radionuclide encapsulation during liposome manufacture generally requires laboratory personnel to handle higher levels of radioactivity than when post-labeling methods are used. Post-labeling methods reported in the literature include: (1) non-specific association or chelation of the radionuclide to the liposomal surface (33, 39–43) and (2) after-loading of radionuclides within the interior of pre-formed liposomes using a lipophilic chelator to carry the radionuclide across the lipid bilayer and become stably trapped (26, 44–49).

472

Goins, Bao, and Phillips

This chapter will focus on three after-loading methods developed by our research group for the radiolabeling of preformed liposomes with 99mTc and 186/188Re (26, 45, 49). For all liposome radiolabeling methods described, a lipophilic chelator is used to chemically carry 99mTc or 186/188Re across the lipid bilayer of the pre-formed liposomes. Once within the liposome interior, encapsulated glutathione or ammonium sulfate (pH) gradient is used to stably trap the 99mTc and 186/188Re within the liposomes. In the first method, 99mTc is transported across the lipid bilayer by the lipophilic chelator, hexamethylpropyleneamine oxime (HMPAO) and 99mTcHMPAO becomes trapped by interaction with pre-encapsulated glutathione within the liposomes (49). In the second method, 99mTc or 186/188Re is transported across the lipid bilayer by the lipophilic chelator, N,N-bis(2-mercaptoethyl)-N¢,N¢-diethylethylenediamine (BMEDA), and 99mTc-BMEDA or 186/188Re-BMEDA becomes trapped by interaction with pre-encapsulated glutathione within the liposomes (26, 45). In the third method, an ammonium sulfate (pH) gradient loading technique is employed using liposomes with an extra-liposomal pH of 7.4 and an interior pH of 5.1. BMEDA, which is lipophilic at pH 7.4, serves as a lipophilic chelator for 99mTc or 186/188Re to transport the radionuclides across the lipid bilayer (26, 50, 51). Once within the more acidic liposome interior, 99mTcBMEDA or 186/188Re-BMEDA becomes protonated and more hydrophilic. This hydrophilic 99mTc-BMEDA or 186/188Re-BMEDA is stably trapped within the liposomes. Since many commercially available liposomal drugs use an ammonium sulfate (pH) gradient for drug loading, these liposomal drugs can be directly radiolabeled using 99mTc-BMEDA for non-invasive monitoring of tissue distribution during treatment or 186/188Re-BMEDA for combined chemo- and radionuclide therapy (50).

2. Materials 2.1. Preparation of GSH–Liposomes

1. Phospholipids (Avanti Polar Lipids, Pelham, AL). 2. Cholesterol (Calbiochem, San Diego, CA). 3. Dulbecco’s phosphate-buffered saline (PBS) without Mg2+ and Ca2+, pH 7.4 (Tissue culture grade, Sigma-Aldrich, St. Louis, MO) titrated to pH 6.3 with concentrated hydrochloric acid. PBS buffer should be degassed and stored at 4°C under nitrogen gas (see Notes 1 and 2). 4. Sterile water for injection USP (Hospira, Lake Forest, IL). 5. Sucrose, high purity and low endotoxin grade (Ferro Pfanstiehl Laboratories, Waukegan, IL) solution (300 mM) in sterile water for injection.

Techniques for Loading Technetium-99m and Rhenium-186/188

473

6. Reduced glutathione (GSH) (Sigma-Aldrich, St. Louis, MO) solution (200 mM) in PBS, pH 6.3. Store at 4°C and use within 1 week. 7. Chloroform, HPLC grade (Fisher Scientific, Pittsburgh, PA). 8. Methanol, HPLC grade (Fisher Scientific, Pittsburgh, PA). 9. Sucrose solution (75 mM and 300 mM) in PBS, pH 6.3. Store at 4°C. 10. Polycarbonate Filters (Whatman Nucleopore, Florham, NJ). 11. Lipex Thermobarrel 10 ml extruder (Northern Lipid, Vancouver, Canada); alternatively, Microfluidizer (Microf luidics, Newton, MA) or EmulsiFlex (Avestin, Ottawa, Canada). 12. Lyophilizer (Labconco, Kansas City, MO). 13. Ultracentrifuge (Beckman, Fullerton, CA). 14. Rotary evaporator (Buchi, New Castle, DE). 15. Glutathione assay kit (OXIS International, Foster City, CA). 2.2. Preparation of Ammonium Sulfate (pH) Gradient Liposomes

1. Phospholipids (Avanti Polar Lipids, Pelham, AL). 2. Cholesterol (Calbiochem, San Diego, CA). 3. Sterile water for injection USP (Hospira, Lake Forest, IL). 4. Sucrose, high purity and low endotoxin grade (Ferro Pfanstiehl Laboratories, Waukegan, IL) solution (300 mM) in sterile water for injection. 5. Ammonium sulfate (Sigma-Aldrich, St. Louis, MO) solution (300 mM) in sterile water for injection. 6. Chloroform, HPLC grade (Fisher Scientific, Pittsburgh, PA). 7. Methanol, HPLC grade (Fisher Scientific, Pittsburgh, PA). 8. Polycarbonate Filters (Whatman Nucleopore, Florham, NJ).

2.3. Preparation of 99mTc-HMPAO

1. Dulbecco’s phosphate-buffered saline (PBS) without Mg2+ and Ca2+, pH 7.4 (Tissue culture grade, Sigma-Aldrich, St. Louis, MO) titrated to pH 6.3 with concentrated hydrochloric acid. PBS buffer should be degassed and stored at 4°C under nitrogen gas (see Notes 1 and 2). 2. 99mTc-sodium pertechnetate from freshly eluted 99Mo/ 99m Tc-generator (GE Healthcare Medi-Physics, Inc, Arlington Heights, IL) (see Note 3) in 5 ml 0.9% saline (see Note 4). 3. HMPAO Kit, Unstabilized (CeretecTM, GE Healthcare, Little Chalfont, UK) (see Note 5). 4. Sephadex G-25 disposable columns (PD-10, GE Healthcare, Little Chalfont, UK) (see Note 6).

474

Goins, Bao, and Phillips

5. HMPAO quality control chromatography strips (BIODEX Medical Systems, Shirley, NY). 6. Paper chromatography strips, 13 cm × 1 cm (Schleicher and Schuell #589). 2.4. Preparation of 99mTc-BMEDA

1. Sodium glucoheptonate (Sigma-Aldrich, St. Louis, MO). 2. Saline for injection USP (Hospira, Lake Forest, IL) (10 ml/ vial; Degassed and sealed under nitrogen gas) (see Note 1). 3. Stannous chloride (Sigma-Aldrich, St. Louis, MO) should be prepared fresh daily at concentration of 15 mg/ml in 0.9% nitrogen-degassed saline (see Note 1). 4. BMEDA vial containing 10 mg BMEDA sealed under argon gas (ABX Biochemicals, Radeburg Germany). Store at −20°C. 5. 99mTc-sodium pertechnetate from freshly eluted 99Mo/99mTc generator (GE Healthcare Medi-Physics, Inc, Arlington Heights, IL) in 0.9% saline (see Note 3). 6. Sephadex G-25 disposable columns (PD-10; GE Healthcare, Little Chalfont, UK) (see Note 6).

2.5. Preparation of 186/188 Re-BMEDA

1. Sodium glucoheptonate (Sigma-Aldrich, St. Louis, MO). 2. Saline for injection (Hospira, Lake Forest, IL) (10 ml/vial; Degassed and sealed under nitrogen gas) (see Note 7). 3. Stannous chloride (Sigma-Aldrich, St. Louis, MO) should be prepared fresh daily at concentration of 15 mg/ml in 0.9% nitrogen-degassed saline (see Note 7). 4. BMEDA vial containing 10 mg BMEDA sealed under argon gas (ABX Biochemicals , Radeburg Germany). Store at −20°C. 5. 186Re-sodium perrhenate produced by University of Missouri Research Reactor (MURR, Columbia, MO) or 188Re-sodium perrhenate in 0.9% sterile isotonic saline eluted from 188W-188Re generator (IEA-Polatom, Otwock-Swierk, Poland or Oak Ridge National Laboratories, Oak Ridge, TN) (see Note 7). 6. Sephadex G-25 disposable columns (PD-10; GE Healthcare, Little Chalfont, UK) (see Note 6).

3. Methods All liposome labeling methods described in this chapter use lipophilic chelators to transport the radionuclides, 99mTc or 186/188 Re, across the lipid bilayer of the liposomes. As shown in Fig. 1, 99mTc-HMPAO, 99mTc-BMEDA, or 186/188Re-BMEDA

Techniques for Loading Technetium-99m and Rhenium-186/188

475

Fig. 1. Schematic diagram outlining chemical gradient after-loading method for radiolabeling liposomes with 99mTcHMPAO or 99mTc/186/188Re-BMEDA lipophilic chelator. The lipophilic chelators cross the lipid bilayer and become trapped after interaction with glutathione (GSH) encapsulated within the liposome interior

crosses the lipid bilayer and becomes trapped through interaction with pre-encapsulated GSH within the pre-formed liposomes. Example images from a patient acquired 20 h after intravenous injection of 99mTc-pegylated liposomes radiolabeled with the HMPAO method are shown in Fig. 2 (7, 19). The increased uptake of 99mTc-liposomes in the arthritic wrists of this patient can be clearly visualized in the images. These images show the potential of 99mTc-liposomes as a diagnostic imaging agent for the noninvasive detection of inflammatory processes (19). In addition, 99mTc-BMEDA or 186/188Re-BMEDA can be used to radiolabel pre-formed liposomes containing an ammonium sulfate (pH) gradient (Fig. 3) (26, 28, 29, 50, 52). For this labeling procedure, entrapment of 99mTc-BMEDA or 186/188 Re-BMEDA is by protonation of 99mTc/186/188Re-BMEDA at the lower pH environment in the liposome interior. In this method, liposomes containing ammonium sulfate at pH 5.1 within the liposome interior and exterior are prepared and stored. On the day of radiolabeling, a pH gradient between the liposome interior and exterior is created by washing the liposomes in phosphate-buffered saline, pH 7.4. This creates a liposome pH gradient with an interior pH of 5.1 and an exterior pH of 7.4. BMEDA can chelate with 99mTc or 186/188Re radionuclide to form 99m Tc/186/188Re-BMEDA complex. These complexes are more

476

Goins, Bao, and Phillips

Fig. 2. Whole body scintigraphic image and a spot image of the hands of a patient with arthritis at 20 h after intravenous injection of 99mTc-pegylated liposomes radiolabeled using HMPAO after-loading method. (Reproduced from ref. (19) with permission from Elsevier)

Fig. 3. Schematic diagram depicting after-loading method for liposomes containing ammonium sulfate pH gradient radiolabeled with 99mTc-BMEDA and 186/188Re-BMEDA. The lipophilic form of BMEDA at pH 7.4 crosses the lipid bilayer. Once inside the liposome interior, BMEDA becomes protonated at pH 5.1 and trapped within the liposome interior

Techniques for Loading Technetium-99m and Rhenium-186/188

477

lipophilic at pH 7.4 and can carry 99mTc or 186/188Re radionuclide into the liposome aqueous interior. When 99mTc-BMEDA or 186/188 Re-BMEDA is exposed to the lower pH in the liposome interior, 99mTc-BMEDA or 186/188Re -BMEDA is thus protonated using its amine groups. The hydrophilic protonated 99mTcBMEDA or 186/188Re -BMEDA becomes trapped within the liposome interior. By using ammonium sulfate (pH) gradient, the lower pH in the liposome interior can be maintained, since free ammonia NH3 can more readily cross the liposome bilayer than its dissociated counterpart proton ion. Since the ammonium sulfate (pH) gradient is commonly used for high efficiency drug loading in several commercial liposome-based drug products, 99mTc-BMEDA and 186/188Re-BMEDA can be used to directly radiolabel liposome-based drug products such as pegylated liposomal doxorubicin (Doxil® , Johnson & Johnson, Bridgewater, NJ) (50). Figure 4 shows an overlay of microSPECT/computed tomography (CT) images for a rat with a head and neck xenograft acquired 21 h after intravenous

Fig. 4. A set of individual CT (upper panel), SPECT (middle panel) and co-registered SPECT/CT (lower panel) image slices of a rat with head and neck xenograft tumor acquired 21 h after intravenous injection of 99mTc-Doxil® using BMEDA afterloading method. The uptake of 99mTc-Doxil® in the tumor can be clearly visualized in the SPECT images. Image slices are displayed in transaxial (left ), sagittal (middle) and coronal (right ) views. (T tumor; S spleen; L liver; H heart)

478

Goins, Bao, and Phillips

injection of Doxil® radiolabeled with 99mTc-BMEDA. The ability to verify the location and pattern of distribution of 99mTc-Doxil® in the tumor in vivo is useful for understanding the tumor uptake as well as normal organ clearance kinetics and distribution at various time points. Potential applications for 99mTc-Doxil® include pharmacokinetic and distribution imaging studies to verify uptake of a liposomal drug in the target site before administration of a therapeutic dose of Doxil®. BMEDA can also be used to radiolabel Doxil® with 186/188Re, thus leading to possibilities of combined chemo- and radionuclide therapy (26, 50). The liposome radiolabeling procedures described in this chapter require competent personnel to handle hazardous radioactive materials. All laboratory personnel must, therefore, receive proper training in radioactive handling before conducting any experiments and must diligently follow the radiation safety principles of ALARA (As Low As Reasonably Achievable). During radiolabeling experiments, laboratory personnel must wear appropriate personal protective equipment and be enrolled in program to monitor radiation exposure levels. Ideally, a dedicated laboratory area with limited access should be set up for the radiolabeling procedures to prevent radioactive exposure to other laboratory personnel. This area should include a 100% exhaust chemical fume hood or B2 biosafety cabinet with work area shielded with lead bricks (BIODEX Medical Systems, Shirley, NY). Alternatively, a dedicated hot cell can also be used for the radiolabeling experiments. Equipment for measuring radioactivity including dose calibrator (AtomLab™ 100, BIODEX Medical Systems, Shirley, NY) and gamma well counter (Wallac 2480 Wizard, Perkin Elmer Life Sciences, Waltham, MA) should also be available. 3.1. Preparation of GSH–Liposomes

1. Weigh out phospholipids and cholesterol in a round-bottomed flask according to desired lipid concentration and composition of liposome preparation (see Notes 8 and 9). 2. Add chloroform or chloroform-methanol depending on lipid composition to dissolve lipids and form lipid solution. 3. Place lipid solution on rotary evaporator to remove solvent and form lipid thin film. The temperature and evaporation time will depend on lipid formulation. 4. Place lipid thin film in desiccator under vacuum for at least 4 h or preferably overnight. 5. Rehydrate lipid thin film with 300 mM sucrose in sterile water for injection at 120 mM total lipid concentration. Vortex solution vigorously and heat above lipid phase transition temperature until all lipids are in solution. 6. Freeze lipid-sucrose solution and lyophilize until dry powder is formed.

Techniques for Loading Technetium-99m and Rhenium-186/188

479

7. Rehydrate lipid-sucrose powder with 200 mM GSH in PBS, pH 6.3. Vortex solution vigorously and heat above lipid phase transition temperature until all lipids are in solution. 8. These instructions assume the diameter of the liposome particles will be reduced using a lipex thermobarrel 10 ml extruder. However, these procedures can be adapted for other size-reduction methods including microfluidization and homogenization or other extrusion devices. The extruder should be autoclaved before use and assembled according to manufacturer’s recommendations. Depending on the desired liposome diameter, the liposomes should be extruded through a series of polycarbonate filters beginning with a filter with larger pore size and proceeding to filters with smaller pore sizes (see Note 10). 9. After extrusion, the liposomes should be washed to remove extra-liposomal GSH, which will interfere with the labeling procedure by prematurely reacting with the radionuclide complexes and preventing the radionuclide complexes from crossing the lipid bilayer membrane. Extruded liposomes should be diluted 1:2 (v/v) with PBS, pH 6.3 and centrifuged at 244,717 × g (45,000 rpm, Ti 50.2 rotor) for 50 min in an ultracentrifuge. The supernatant should be removed and the liposome pellets resuspended in PBS, pH 6.3 containing 75 mM sucrose. The liposome suspension should be centrifuged again using the same parameters listed above. After repeating the washing procedure three times, the liposome pellet should be resuspended in PBS, pH 6.3 containing 300 mM sucrose (see Note 11). 10. Final liposome product should be stored at 4°C until needed for the radiolabeling procedures. Prior to in vivo imaging studies, the final liposome product should be characterized (53) for liposome diameter by laser light scattering particle sizing, pyrogenicity, sterility, lipid concentration (54), and GSH content using Glutathione assay kit(55). 3.2. Preparation of Ammonium Sulfate (pH) Gradient Liposomes

1. Weigh out phospholipids and cholesterol in a round-bottomed flask according to desired lipid concentration and composition of liposome preparation (see Notes 8 and 9). 2. Add chloroform or chloroform-methanol depending on lipid composition to dissolve lipids and form lipid solution. 3. Place lipid solution on rotary evaporator to remove solvent and form lipid thin film. The temperature and evaporation time will depend on lipid formulation. 4. Place lipid thin film in desiccator under vacuum for at least 4 h or preferably, overnight.

480

Goins, Bao, and Phillips

5. Rehydrate lipid thin film with 300 mM sucrose in sterile water for injection at 60 mM total lipid concentration. Vortex solution vigorously and heat above lipid phase transition temperature until all lipids are in solution. 6. Freeze lipid-sucrose solution and lyophilize until dry powder is formed. 7. Rehydrate lipid-sucrose powder with 300 mM ammonium sulfate in sterile water at 60 mM total lipid concentration. Vortex solution vigorously and heat above lipid phase transition temperature until all lipids are in solution. 8. Freeze the lipid solution with liquid nitrogen and then thaw in water bath set to temperature above the lipid phase transition temperature. Repeat freeze-thaw procedure for three cycles. 9. Set up lipex extruder as described in Subheading 3.1 and extrude liposome sample until desired particle diameter is achieved. 10. After extrusion, final liposome product should be stored at 4°C until needed for the radiolabeling procedures. The liposomes should be characterized for liposome diameter by laser light scattering particle sizing, pyrogenicity, sterility, and lipid concentration (53, 54) before conducting any in vivo imaging studies. 3.3. 99mTc-HMPAO Method for Radiolabeling GSH–Liposomes 3.3.1. Labeling Procedure

1. Pipette 2 ml aliquot of the pre-formed GSH–liposome preparation (see Subheading 3.1) into a sterile vial. Typically in our laboratory, the radiolabeling is performed in a 16 × 100 mm sterile plastic snap cap test tube. 2. Prepare the 99mTc-HMPAO by adding 5 ml of 99mTc-sodium pertechnetate in 0.9% saline to a vial of unstabilized HMPAO. Mix the vial and incubate for 5 min at room temperature (see Notes 4 and 5). 3. To label the GSH–liposomes, draw the 99mTc-HMPAO into a syringe and add 2 ml of the 99mTc-HMPAO to the GSH– liposome sample (2 ml). Incubate the 99mTc-HMPAO/liposome mixture for 15–30 min at room temperature. 4. Immediately after adding the 99mTc-HMPAO to the GSH– liposome preparation, remove a separate aliquot of the 99mTcHMPAO and perform the quality control chromatography as described in the HMPAO kit instructions to make certain that the 99mTc-HMPAO is in the lipophilic form. 5. During incubation of 99mTc-HMPAO/GSH–liposome mixture, prepare a Sephadex G-25 column by passing 25 ml of PBS, pH 6.3, over the column. The columns come in a suspension of water. It is necessary to equilibrate the column material with

Techniques for Loading Technetium-99m and Rhenium-186/188

481

PBS, pH 6.3 buffer before elution of the 99mTc-liposome sample. Prepare 1 column for every 2 ml of 99mTc-liposome mixture that requires separation. 3.3.2. Column Chromatography for Separation of Free 99m Tc-Label from 99m Tc-GSH–Liposomes and Assessment of Percentage Labeling Efficiency

1. Drain any remaining PBS from the Sephadex G-25 column reservoir (see step 5 in Subheading 3.3.1). 2. Measure the 99mTc-activity of the 99mTc-HMPAO/GSH– liposome mixture in a dose calibrator before column separation (Pre-column 99mTc-liposome sample) and record the volume. 3. Pipette 2.0 ml of the 99mTc-HMPAO/GSH–liposome mixture onto the top of the column. Do not load more than 2.5 ml per column. Use multiple columns for larger samples. 4. Once the 99mTc-liposomes have completely entered the column packing, add the PBS elution buffer to the column reservoir. 5. The 99mTc-liposomes will be eluted in the void volume. Collect the 99mTc-liposomes in a new sterile tube as soon as you see a slightly milky eluate coming out of the column and stop when the eluate becomes clear again (post-column 99mTc-liposome sample) (see Note 12). 6. Measure the 99mTc-activity of the post-column 99mTcliposomes in a dose calibrator and record the volume. 7. Calculate the percentage labeling efficiency using the following equation:

=

{(

Tc - activity in post - column 99m Tc - liposome sample ) /

99m

TC - activity in pre - column 99m Tc - liposome sample ) × 100

99m

( 3.3.3. Assessment of Percentage Labeling Efficiency Using Paper Chromatography

% Labeling efficiency

}

1. Using a pencil, mark a line 2 cm from the bottom (origin) and 1 cm from the top (solvent front) of a # 589 paper strip. Also, mark a line 7 cm from the bottom of the strip. 2. Prepare a development tank by adding a small amount of 0.9% saline to cover the bottom of the tank. 3. Apply an aliquot (10 ml) of the 99mTc-HMPAO–liposome mixture (pre-column 99mTc-liposome sample) (see Subheading 3.3.2) on the origin of the strip. 4. Place strip in the tank containing 0.9% saline, making sure that the saline level is below the origin. 5. Allow the saline to rise up the paper until it reaches the solvent front. 6. Once saline has reached the solvent front, remove the paper from the tube with forceps.

482

Goins, Bao, and Phillips

7. Cut the paper at the 7 cm line and place each half in a separate clean tube. 8. Determine the 99mTc-activity (counts per minute (cpm)) of the top and bottom fractions using a gamma well counter. For most liposome formulations, the 99mTc-liposomes will remain at the origin. 9. Calculate percentage labeling efficiency using the following equation: % Labeling efficiency

= {(cpm in bottom portion ) / (cpm in top portion ) + (cpm in bottom portion )}×100

3.4. 99mTc-BMEDA Method for Radiolabeling GSH–Liposomes 3.4.1. Preparation of GSH–Liposomes for Radiolabeling Procedure

3.4.2. Preparation of 99 mTc-BMEDA

1. Prepare a Sephadex G-25 column by passing 25 ml of PBS, pH 7.4, over the column (see step 5 in Subheading 3.3.1). Use 1 column for every 2 ml of GSH–liposomes. 2. Drain any remaining PBS, pH 7.4, from the Sephadex G-25 column reservoir. 3. Pipette 2 ml of the GSH–liposomes onto the top of the column and elute with PBS buffer, pH 7.4. The collected liposomes will be used for 99mTc labeling. 1. Dissolve 3.5 ml BMEDA and 50 mg sodium glucoheptonate (GH) in 5.0 ml nitrogen-degassed saline in a 10 ml sterile glass serum vial. Mix solution on a magnetic stirrer plate for 20 min at room temperature (see Note 13). 2. Add 65 ml of a freshly prepared 15 mg/ml stannous chloride in 0.9% saline solution. 3. Adjust BMEDA–GH–stannous chloride solution to pH 7.0 with 50 mM sodium hydroxide. 4. Remove 1 ml of BMEDA–GH–stannous chloride into a new 10 ml sterile glass serum vial containing 0.6 ml of 60 mCi (2.2 GBq) 99mTc-sodium pertechnetate. Gently shake the vial for 1 min to mix the 99mTc-sodium pertechnetate with BMEDA–GH–stannous chloride. 5. Incubate 99mTc-BMEDA solution at 25°C for 20 min with intermittent gentle shaking. 6. Check labeling efficiency of 99mTc-BMEDA using similar paper chromatography technique described in Subheading 3.3.3 but following the procedures outlined in Bao et al. (45).

3.4.3. Labeling Procedure for 99mTcGSH–Liposomes

1. Add 1 ml of GSH–liposomes (see Subheading 3.4.1) to 1 ml 99m Tc-BMEDA (see Subheading 3.4.2) and incubate mixture for 1 h at 37°C.

Techniques for Loading Technetium-99m and Rhenium-186/188

483

2. During incubation of 99mTc-BMEDA/GSH–liposome mixture, prepare a Sephadex G-25 column by passing 25 ml of PBS, pH 7.4, over the column (see step 5 in Subheading 3.3.1). 3.4.4. Column Chromatography for Separation of Free 99m Tc-Label from 99m Tc-GSH–Liposomes and Assessment of Percentage Labeling Efficiency

1. Drain any remaining PBS, pH 7.4, from the Sephadex G-25 column reservoir (see Subheading 3.4.3). 2. Measure the 99mTc-activity of the 99mTc-BMEDA/GSH– liposome mixture in a dose calibrator before column separation (pre-column 99mTc-liposome sample) and record the volume. 3. Pipette 2.0 ml of the 99mTc-BMEDA/GSH–liposome mixture onto the top of the column. Do not load more than 2.5 ml per column. Use multiple columns for larger samples. 4. Once the 99mTc-liposomes have completely entered the column packing, add the PBS, pH 7.4, elution buffer to the column reservoir. 5. The 99mTc-liposomes will be eluted in the void volume. Collect the 99mTc-liposomes in a new sterile tube as soon as you see a slightly milky eluate coming out of the column and stop when the eluate becomes clear again (post-column 99mTc-liposome sample) (see Note 12). 6. Measure the 99mTc-activity of the post-column 99mTc-liposomes in a dose calibrator and record the volume. 7. Calculate the percentage labeling efficiency using equation in Subheading 3.3.2 (see Note 14).

3.5. 186/188Re-BMEDA Method for Radiolabeling GSH–Liposomes 3.5.1. Preparation of GSH–Liposomes for Radiolabeling Procedure

3.5.2. Preparation of 186/188 Re-BMEDA

1. Prepare a Sephadex G-25 column by passing 25 ml of PBS, pH 7.4, over the column. Use 1 column for every 2 ml of GSH–liposomes. 2. Drain any remaining PBS, pH 7.4, from the Sephadex G-25 column reservoir. 3. Pipette 2 ml of the GSH–liposomes onto the top of the column and elute with PBS, pH 7.4, buffer. The collected liposomes will be used for 186/188Re-labeling. 1. Dissolve 3 ml BMEDA and 50 mg sodium glucoheptonate (GH) in 2.0 ml nitrogen-degassed saline in a 10 ml sterile glass serum vial. Mix solution on a magnetic stir plate for 20 min at room temperature (see Note 13). 2. Add 240 ml of a freshly prepared 15 mg/ml stannous chloride in 0.9% saline solution. 3. Adjust BMEDA–GH–stannous chloride solution to pH 5.0 with 50 mM hydrochloric acid using pH paper.

484

Goins, Bao, and Phillips

4. Remove 1 ml of BMEDA-GH-stannous chloride into a new 10 ml sterile glass serum vial containing 0.2 ml of 100– 150 mCi (3.7–5.55 GBq) 186Re-aluminum perrhenate or 1.0 ml of 50 mCi (1.85 GBq) 188Re-sodium perrhenate. Flush the vial thoroughly with nitrogen gas and seal. 5. Incubate 186/188Re-BMEDA solution at 80°C in the water bath for 1 h. 6. Allow 186/188Re-BMEDA solution to cool down to room temperature. 7. Adjust 186/188Re-BMEDA solution to pH 7.0 by adding 70–90 ml of 50 mM sodium hydroxide. Verify pH using pH paper. 8. Check labeling efficiency of 186/188Re-BMEDA using similar paper chromatography technique described in Subheading 3.3.3 but following the procedures outlined in Bao et al. (26). 3.5.3. Labeling Procedure for 186/188Re-GSH Liposomes

1. Add 1 ml of GSH–liposomes (see Subheading 3.5.1) to 1 ml 186/188 Re-BMEDA (see Subheading 3.5.2) and incubate mixture for 1 h at 37°C. 2. During incubation of 186/188Re-BMEDA/GSH–liposome mixture, prepare a Sephadex G-25 column by passing 25 ml of PBS, pH 7.4 over the column. Prepare 1 column for every 2 ml of 186/188Re-BMEDA/GSH–liposome mixture.

3.5.4. Column Chromatography for Separation of Free 186/188 Re-label from 186/188 Re-GSH–Liposomes and Assessment of Percentage Labeling Efficiency

1. Drain any remaining PBS, pH 7.4, from the Sephadex G-25 column reservoir (see Subheading 3.5.3). 2. Measure the 186/188Re-activity of the 186/188Re-BMEDA/ GSH–liposome mixture in a dose calibrator before column separation (pre-column 186/188Re-liposome sample) and record the volume. 3. Pipette 2.0 ml of the 186/188Re-BMEDA/GSH–liposome mixture onto the top of the column. Do not load more than 2.5 ml per column. Use multiple columns for larger samples. 4. Once the 186/188Re-liposomes have completely entered the column packing, add the PBS, pH 7.4, elution buffer to the column reservoir. 5. The 186/188Re-liposomes will be eluted in the void volume. Collect the 186/188Re-liposomes in a new sterile tube as soon as you see a slightly milky eluate coming out of the column and stop when the eluate becomes clear again (post-column 186/188 Re-liposome sample) (see Note 12). 6. Measure the 186/188Re-activity of the post-column 186/188 Re-liposome fraction (post-column 186/188Re-liposome sample) in a dose calibrator and record the volume.

Techniques for Loading Technetium-99m and Rhenium-186/188

485

7. Calculate labeling efficiency using the following equation: (see Note 14). % Labeling efficiency = {( 186Re - activity in post - column

3.6 99mTc-BMEDA Method for Radiolabeling Ammonium Sulfate (pH) Gradient Liposomes 3.6.1. Preparation of Ammonium Sulfate (pH) Gradient-Liposomes for Radiolabeling Procedure

186/188

Re - liposome sample) / ( 186Re - activity in pre - column

186/188

Re - liposome sample)} × 100

1. Prepare a Sephadex G-25 column by passing 25 ml of PBS, pH 7.4, over the column (see step 5 in Subheading 3.3.1). Use 1 column for every 2 ml of ammonium sulfate– liposomes. 2. Drain any remaining PBS, pH 7.4, from the Sephadex G25 column reservoir. 3. Pipette 2 ml of the ammonium sulfate–liposomes onto the top of the column and elute with PBS, pH 7.4, buffer. The collected liposomes will be used for 99mTc labeling.

3.6.2. Preparation of 99m Tc-BMEDA

1. Prepare 99mTc-BMEDA as outlined in Subheading 3.4.2.

3.6.3. Labeling Procedure for 99mTc-GSH Liposomes

1. Add 1 ml of washed ammonium sulfate–liposomes (see Subheading 3.6.1) to 1 ml 99mTc-BMEDA (see Subheading 3.6.2) and incubate mixture for 1 h at 37°C. 2. During incubation of 99mTc-BMEDA/ammonium sulfate– liposome mixture, prepare a Sephadex G-25 column by passing 25 ml of PBS, pH 7.4 over the column (see step 5 in Subheading 3.3.1).

3.6.4. Column Chromatography for Separation of Free 99m Tc-Label from 99m Tc-GSH–Liposomes and Assessment of Percentage Labeling Efficiency

1. Drain any remaining PBS, pH 7.4, from the Sephadex G-25 column reservoir (see Subheading 3.4.1). 2. Measure the 99mTc-activity of the 99mTc-BMEDA/ammonium sulfate–liposome mixture in a dose calibrator before column separation (pre-column 99mTc-liposome sample) and record the volume. 3. Pipette 2.0 ml of the 99mTc-BMEDA/ammonium sulfate– liposome mixture onto the top of the column. Do not load more than 2.5 ml per column. Use multiple columns for larger 99mTc-liposome samples. 4. Once the 99mTc-liposomes have completely entered the column packing, add the PBS, pH 7.4, elution buffer to the column reservoir. 5. The 99mTc-liposomes will be eluted in the void volume. Collect the 99mTc-liposomes in a new sterile tube as soon as you see a slightly milky eluate coming out of the column and stop when

486

Goins, Bao, and Phillips

the eluate becomes clear again (post-column sample) (see Note 12).

Tc-liposome

99m

6. Measure the 99mTc-activity of the post-column 99mTc-liposome fraction in a dose calibrator and record the volume. 7. Calculate the percentage labeling efficiency using equation in Subheading 3.3.2 (see Note 14). 3.7. 186/188Re-BMEDA Method for Radiolabeling Ammonium Sulfate (pH) Gradient-Liposomes

1. Prepare a Sephadex G-25 column by passing 25 ml of PBS, pH 7.4, over the column. Use 1 column for every 2 ml of liposomes.

3.7.1. Preparation of Ammonium Sulfate– Liposomes for Radiolabeling Procedure

3. Pipette 2 ml of the ammonium sulfate–liposomes onto the top of the column and elute with PBS, pH 7.4, buffer. The collected liposomes will be used for 186/188Re-labeling.

3.7.2. Preparation of 186/188 Re-BMEDA

1. Prepare 186/188Re-BMEDA as outlined in Subheading 3.5.2.

3.7.3. Labeling Procedure for 186/188 Re-Ammonium Sulfate Liposomes

1. Add 1 ml of ammonium sulfate–liposomes (see Subheading 3.7.1) to 1 ml 186/188Re -BMEDA (see Subheading 3.7.2) and incubate mixture for 1 h at 37°C.

3.7.4. Column Chromatography for Separation of Free 186/188 Re-label from 186/188 Re-Ammonium Sulfate–Liposomes and Assessment of Percentage Labeling Efficiency

1. Drain any remaining PBS from the Sephadex G-25 column reservoir (see Subheading 3.7.3).

2. Drain any remaining PBS, pH 7.4, from the Sephadex G-25 column reservoir.

2. During incubation of 186/188Re-BMEDA/ammonium sulfate– liposome mixture, prepare 1 Sephadex G-25 column for every 2 ml of 186/188Re-liposomes.

2. Measure the 186/188Re-activity of the 186/188Re-BMEDA/ ammonium sulfate–liposome mixture in a dose calibrator before column separation (pre-column 186/188Re-liposome sample) and record the volume. 3. Pipette 2.0 ml of the 186/188Re-BMEDA/ammonium sulfate– liposome mixture onto the top of the column. Do not load more than 2.5 ml per column. Use multiple columns for larger samples. 4. Once the 186/188Re-liposomes have completely entered the column packing, add the PBS elution buffer to the column reservoir. 5. The 186/188Re-liposomes will be eluted in the void volume. Collect the 186/188Re-liposomes in a new sterile tube as soon as you see a slightly milky eluate coming out of the column and stop when the eluate becomes clear again (post-column 186/188 Re-liposome sample) (see Note 12).

Techniques for Loading Technetium-99m and Rhenium-186/188

487

6. Measure the 186/188Re-activity of the post-column 186/188 Re-liposome fraction in a dose calibrator and record the volume. 7. Calculate labeling efficiency using Subheading 3.5.4 (see Note 14).

the

equation

in

4. Notes 1. The 99mTc labeling procedure is very sensitive to the presence of oxygen. The buffer or saline used for labeling procedure should be purged with nitrogen for 15–20 min to remove dissolved oxygen. 2. GSH–liposomes prepared in PBS with higher pH (7.4), 0.9% saline or other buffer system can also be labeled with 99mTcHMPAO. 3. A local radiopharmacy can supply the 99mTc-pertechnetate as a single dose by eluting from 99Mo/99mTc generator. If your research facility does not have access to a radiopharmacy, a 99 Mo/99mTc generator can be purchased and eluted in-house. Regardless of the source of 99mTc-pertechnetate, it is very important that the generator has been eluted within the past 24 h before obtaining the 99mTc-pertechnetate for reconstitution with either HMPAO or BMEDA because older 99mTcpertechnetate will contain an excessive amount of technetium-99, which is a decay product of 99mTc that will compete with 99mTc during the 99mTc-HMPAO or 99mTcBMEDA binding reaction (2). 4. For HMPAO labeling, 99mTc-pertechnetate is normally supplied in a syringe containing 10–15 mCi (370–555 MBq) dose of 99mTc-pertechnetate in 5 ml of 0.9% saline. If a higher specific activity is needed, the procedure can be performed with higher doses (30–50 mCi, 1.11–1.85 GBq) of 99mTcpertechnetate or by using a smaller volume (1–2 ml) of saline to elute the 99mTc-pertechnetate from the generator. 5. Carefully read the package instructions before attempting the liposome labeling protocol because HMPAO is only useful for liposome labeling for 30 min after reconstitution. Package instructions for leukocyte labeling using unstabilized product should be followed, not those for brain imaging using the methylene blue stabilized product. 6. Alternatively, columns can be prepared in-house by pouring swollen Sephadex G-25 in columns.

488

Goins, Bao, and Phillips

7. The 186/188Re labeling procedure is very sensitive to the presence of oxygen. The 0.9% saline used for labeling procedure should be purged with nitrogen gas for 15–20 min to remove dissolved oxygen. 8. For sterile liposome product suitable for radiolabeling and in vivo imaging studies, the liposome manufacture is normally performed in a laminar flow hood using aseptic technique and sterile tubes. Sterile disposable syringes, needles and pipettes are also used. 9. An advantage of the liposome radiolabeling procedures described in this chapter is that the radiolabeling is not dependent on the liposome size or lipid composition. The main liposome formulation requirement is the capability of the liposomes to stably maintain either a pre-encapsulated glutathione or ammonium sulfate (pH) gradient under long-term storage and after injection in the body for in vivo imaging studies. A typical liposome formulation used for radiolabeling in our laboratory is distearoyl phosphatidylcholine (DSPC): cholesterol (56:44 molar percentage) at a total lipid concentration of 120 mM for GSH–liposomes and 60 mM for ammonium sulfate (pH) gradient liposomes (51). 10. In our laboratory, a typical extrusion sequence for liposomes of approximately 100 nm in diameter is 2 passes through 2 micron pore filter; 2 passes through 400 nm pore filter; 5 passes through 200 nm pore filter; and 5–10 passes through 100 nm pore filter. 11. For successful labeling, liposomes should be prepared in the presence of GSH. It is important that all unentrapped GSH be removed prior to the labeling procedure. Alternatively, the removal of extra-liposomal GSH after manufacture can be accomplished by passing the liposomes over a Sephadex G-50 column eluted with Dulbecco’s PBS pH 6.3 containing 300 mM sucrose. 12. Occasionally, you will see a tailing of the liposomes toward the end of the collection period. When this occurs, stop the collection to prevent too much dilution of your liposome sample. The column separation normally increases the volume of the loaded sample by 1.4 times the initial volume. 13. If there is any remaining BMEDA in the original vial that will be kept for future use, the vial containing BMEDA should be flushed thoroughly with either nitrogen or argon gas and tightly sealed, and then stored in −20°C freezer. 14. The labeling efficiency should be over 25% for further use. If it is lower than 25%, the labeling procedure is not optimal and the reasons for the poor labeling results should be investigated.

Techniques for Loading Technetium-99m and Rhenium-186/188

489

Acknowledgements The authors would like to thank Anuradha Soundararajan for her help in acquiring the images depicted in Fig. 4 and Jonathan Sumner for his help in preparing the figures displayed in this chapter. This work was supported by NIH National Cancer Institute Cancer Center Specialized Programs of Research Excellence grant 5 P30 CA054174-16. References 1. Carlsson J, Forssell-Aronsson E, Glimelius B, Mattsson S (2002) Therapy with radiopharmaceuticals. Acta Oncol 41:623–628 2. Kowalsky R, Falen SW (2004) Radiopharmaceuticals in nuclear pharmacy and nuclear medicine. American Pharmacists Association, Washington, DC 3. Boerman OC, Laverman P, Oyen WJ, Corstens FH, Storm G (2000) Radiolabeled liposomes for scintigraphic imaging. Prog Lipid Res 39:461–475 4. Boerman OC, Oyen WJ, Corstens FH, Storm G (1998) Liposomes for scintigraphic imaging: optimization of in vivo behavior. Q J Nucl Med 42:271–279 5. Brouwers AH, De Jong DJ, Dams ET, Oyen WJ, Boerman OC, Laverman P, Naber TH, Storm G, Corstens FH (2000) Tc-99m-PEGliposomes for the evaluation of colitis in Crohn’s disease. J Drug Target 8:225–233 6. Dagar S, Rubinstein I, Onyuksel H (2003) Liposomes in ultrasound and gamma scintigraphic imaging. Methods Enzymol 373:198–214 7. Dams ET, Oyen WJ, Boerman OC, Storm G, Laverman P, Kok PJ, Buijs WC, Bakker H, van der Meer JW, Corstens FH (2000) 99mTc-PEG liposomes for the scintigraphic detection of infection and inflammation: clinical evaluation. J Nucl Med 41:622–630 8. Goins B (2008) Radiolabeled lipid nanoparticles for diagnostic imaging. Expert Opin Med Diagn 2:853–873 9. Laverman P, Boerman OC, Storm G (2003) Radiolabeling of liposomes for scintigraphic imaging. Methods Enzymol 373:234–248 10. Laverman P, Brouwers AH, Dams ET, Oyen WJ, Storm G, van Rooijen N, Corstens FH, Boerman OC (2000) Preclinical and clinical evidence for disappearance of long-circulating characteristics of polyethylene glycol lipo-

11. 12. 13.

14.

15.

16.

17. 18.

19.

somes at low lipid dose. J Pharmacol Exp Ther 293:996–1001 Morgan JR, Williams LA, Howard CB (1985) Technetium-labelled liposome imaging for deep-seated infection. Br J Radiol 58:35–39 Osborne MP (1978) Lymph node scanning for breast cancer. Trans Med Soc Lond 95:43–45 Osborne MP, Payne JH, Richardson VJ, McCready VR, Ryman BE (1983) The preoperative detection of axillary lymph node metastases in breast cancer by isotope imaging. Br J Surg 70:141–144 O’Sullivan MM, Powell N, French AP, Williams KE, Morgan JR, Williams BD (1988) Inflammatory joint disease: a comparison of liposome scanning, bone scanning, and radiography. Ann Rheum Dis 47:485–491 Richardson VJ, Ryman BE, Jewkes RF, Jeyasingh K, Tattersall MN, Newlands ES, Kaye SB (1979) Tissue distribution and tumour localization of 99m-technetium-labelled liposomes in cancer patients. Br J Cancer 40:35–43 Richardson VJ, Ryman BE, Jewkes RF, Tattersall MH, Newlands ES (1978) 99mTclabelled liposomes preparation of radio pharmaceutical and its distribution in a hepatoma patient. Int J Nucl Med Biol 5: 118, 121–123 Torchilin VP (2007) Targeted pharmaceutical nanocarriers for cancer therapy and imaging. AAPS J 9:E128–E147 Williams BD, O’Sullivan MM, Saggu GS, Williams KE, Williams LA, Morgan JR (1987) Synovial accumulation of technetium labelled liposomes in rheumatoid arthritis. Ann Rheum Dis 46:314–318 Goins BA, Phillips WT (2001) The use of scintigraphic imaging as a tool in the development of liposome formulations. Prog Lipid Res 40:95–123

490

Goins, Bao, and Phillips

20. Lopez-Berestein G, Kasi L, Rosenblum MG, Haynie T, Jahns M, Glenn H, Mehta R, Mavligit GM, Hersh EM (1984) Clinical pharmacology of 99mTc-labeled liposomes in patients with cancer. Cancer Res 44:375–378 21. Murray JL, Kleinerman ES, Cunningham JE, Tatom JR, Andrejcio K, Lepe-Zuniga J, Lamki LM, Rosenblum MG, Frost H, Gutterman JU et al (1989) Phase I trial of liposomal muramyl tripeptide phosphatidylethanola mine in cancer patients. J Clin Oncol 7: 1915–1925 22. Perez-Soler R, Lopez-Berestein G, Kasi LP, Cabanillas F, Jahns M, Glenn H, Hersh EM, Haynie T (1985) Distribution of technetium99m-labeled multilamellar liposomes in patients with Hodgkin’s disease. J Nucl Med 26:743–749 23. Saari SM, Vidgren MT, Koskinen MO, Turjanmaa VM, Waldrep JC, Nieminen MM (1998) Regional lung deposition and clearance of 99mTc-labeled beclomethasoneDLPC liposomes in mild and severe asthma. Chest 113:1573–1579 24. Kleiter MM, Yu D, Mohammadian LA, Niehaus N, Spasojevic I, Sanders L, Viglianti BL, Yarmolenko PS, Hauck M, Petry NA, Wong TZ, Dewhirst MW, Thrall DE (2006) A tracer dose of technetium-99m-labeled liposomes can estimate the effect of hyperthermia on intratumoral doxil extravasation. Clin Cancer Res 12:6800–6807 25. Goins B (2008) Radiolabeled lipid nanoparticles for cancer diagnosis and treatment. In: Kumar M (ed) Handbook of particulate drug delivery, vol 2. American Scientific Publishers, Stevenson Ranch, CA, pp 65–82 26. Bao A, Goins B, Klipper R, Negrete G, Phillips WT (2003) 186Re-liposome labeling using 186Re-SNS/S complexes: in vitro stability, imaging, and biodistribution in rats. J Nucl Med 44:1992–1999 27. Bard DR, Knight CG, Page-Thomas DP (1985) Effect of the intra-articular injection of lutetium-177 in chelator liposomes on the progress of an experimental arthritis in rabbits. Clin Exp Rheumatol 3:237–242 28. Chang YJ, Chang CH, Chang TJ, Yu CY, Chen LC, Jan ML, Luo TY, Lee TW, Ting G (2007) Biodistribution, pharmacokinetics and microSPECT/CT imaging of 188Re-bMEDAliposome in a C26 murine colon carcinoma solid tumor animal model. Anticancer Res 27:2217–2225 29. Chen LC, Chang CH, Yu CY, Chang YJ, Hsu WC, Ho CL, Yeh CH, Luo TY, Lee TW, Ting G (2007) Biodistribution, pharmacokinetics and imaging of (188)Re-BMEDA-labeled

pegylated liposomes after intraperitoneal injection in a C26 colon carcinoma ascites mouse model. Nucl Med Biol 34:415–423 30. Hafeli U, Tiefenauer LX, Schbiger PA, Weder HG (1991) A lipophilic complex with 186Re/188Re incorporated in liposomes suitable for radiotherapy. Int J Rad Appl Instrum B 18:449–454 31. Hardy JG, Kellaway IW, Rogers J, Wilson CG (1980) The distribution and fate of 131I-labelled liposomes. J Pharm Pharmacol 32:309–313 32. Henriksen G, Schoultz BW, Michaelsen TE, Bruland OS, Larsen RH (2004) Sterically stabilized liposomes as a carrier for alpha-emitting radium and actinium radionuclides. Nucl Med Biol 31:441–449 33. McQuarrie S, Mercer J, Syme A, Suresh M, Miller G (2005) Preliminary results of nanopharmaceuticals used in the radioimmunotherapy of ovarian cancer. J Pharm Pharm Sci 7:29–34 34. Pikul SS 2nd, Parks NJ, Schneider PD (1987) In vitro killing of melanoma by liposomedelivered intracellular irradiation. Arch Surg 122:1417–1420 35. Sofou S, Kappel BJ, Jaggi JS, McDevitt MR, Scheinberg DA, Sgouros G (2007) Enhanced retention of the alpha-particle-emitting daughters of Actinium-225 by liposome carriers. Bioconjug Chem 18:2061–2067 36. Zweit J (1996) Radionuclides and carrier molecules for therapy. Phys Med Biol 41:1905–1914 37. Jeong JM, Knapp FF Jr (2008) Use of the Oak Ridge National Laboratory tungsten-188/ rhenium-188 generator for preparation of the rhenium-188 HDD/lipiodol complex for trans-arterial liver cancer therapy. Semin Nucl Med 38:S19–S29 38. Goins B, Phillips WT (2003) Radiolabelled liposomes for imaging and biodistribution studies. In: Torchilin V, Weissig V (eds) Liposomes: a practical approach. Oxford University Press, Oxford, UK, pp 319–336 39. Ahkong QF, Tilcock C (1992) Attachment of 99mTc to lipid vesicles containing the lipophilic chelate dipalmitoylphosphatidylethanolamine-DTTA. Int J Rad Appl Instrum B 19:831–840 40. Erdogan S, Roby A, Torchilin VP (2006) Enhanced tumor visualization by gammascintigraphy with 111In-labeled polychelatingpolymer-containing immunoliposomes. Mol Pharm 3:525–530 41. Hnatowich DJ, Friedman B, Clancy B, Novak M (1981) Labeling of pre-formed liposomes

Techniques for Loading Technetium-99m and Rhenium-186/188

42.

43.

44.

45.

46.

47.

48.

with Ga-67 and Tc-99m by chelation. J Nucl Med 22:810–814 Laverman P, Dams ET, Oyen WJ, Storm G, Koenders EB, Prevost R, van der Meer JW, Corstens FH, Boerman OC (1999) A novel method to label liposomes with 99mTc by the hydrazino nicotinyl derivative. J Nucl Med 40:192–197 Richardson VJ, Jeyasingh K, Jewkes RF, Ryman BE, Tattersall MH (1977) Properties of (99mTc) technetium-labelled liposomes in normal and tumour-bearing rats. Biochem Soc Trans 5:290–291 Awasthi VD, Goins B, Klipper R, Phillips WT (1998) Dual radiolabeled liposomes: biodistribution studies and localization of focal sites of infection in rats. Nucl Med Biol 25:155–160 Bao A, Goins B, Klipper R, Negrete G, Mahindaratne M, Phillips WT (2003) A novel liposome radiolabeling method using 99mTc“SNS/S” complexes: in vitro and in vivo evaluation. J Pharm Sci 92:1893–1904 Gabizon A, Huberty J, Straubinger RM, Price DM, Papahadjopoulos D (1988) An improved method for in vivo tracing and imaging of liposomes using a gallium-67-desferoxamine complex. J Liposome Res 1:123–135 Harrington KJ, Mohammadtaghi S, Uster PS, Glass D, Peters AM, Vile RG, Stewart JS (2001) Effective targeting of solid tumors in patients with locally advanced cancers by radiolabeled pegylated liposomes. Clin Cancer Res 7:243–254 Mougin-Degraef M, Jestin E, Bruel D, Remaud-Le Saec P, Morandeau L, FaivreChauvet A, Barbet J (2006) High-activity radio-iodine labeling of conventional and

491

stealth liposomes. J Liposome Res 16: 91–102 49. Phillips WT, Rudolph AS, Goins B, Timmons JH, Klipper R, Blumhardt R (1992) A simple method for producing a technetium-99m-labeled liposome which is stable in vivo. Int J Rad Appl Instrum B 19:539–547 50. Bao A, Goins B, Klipper R, Negrete G, Phillips WT (2004) Direct 99mTc labeling of pegylated liposomal doxorubicin (Doxil) for pharmacokinetic and non-invasive imaging studies. J Pharmacol Exp Ther 308:419–425 51. Bao A, Phillips WT, Goins B, Zheng X, Sabour S, Natarajan M, Ross Woolley F, Zavaleta C, Otto RA (2006) Potential use of drug carriedliposomes for cancer therapy via direct intratumoral injection. Int J Pharm 316:162–169 52. Wang SX, Bao A, Herrera SJ, Phillips WT, Goins B, Santoyo C, Miller FR, Otto RA (2008) Intraoperative 186Re-liposome radionuclide therapy in a head and neck squamous cell carcinoma xenograft positive surgical margin model. Clin Cancer Res 14:3975–3983 53. Zuidam NJ, de Vrueh R, Crommelin DJA (2003) Characterization of liposomes. In: Torchilin VP, Weissig V (eds) Liposomes: a practical approach. Oxford University Press, Oxford, UK, pp 31–78 54. Stewart JC (1980) Colorimetric determination of phospholipids with ammonium ferrothiocyanate. Anal Biochem 104:10–14 55. Medina LA, Calixto SM, Klipper R, Li Y, Phillips WT, Goins B (2006) Mediastinal node and diaphragmatic targeting after intracavitary injection of avidin/99mTc-bluebiotin-liposome system. J Pharm Sci 95:207–224

Chapter 33 Fluorescence Correlation Spectroscopy for the Study of Membrane Dynamics and Organization in Giant Unilamellar Vesicles Ana J. García-Sáez, Dolores C. Carrer, and Petra Schwille Abstract Fluorescence correlation spectroscopy (FCS) is a powerful technique to study the lateral organization of membranes. It measures fluorescence intensity fluctuations in the single molecule regime and allows the determination of diffusion coefficients. When applied to lipid membranes, their fluidity and lipid phase can be estimated from the diffusion rates of fluorescent particles partitioned to the membrane. Here, we describe the theoretical basis of FCS and discuss the z-scan approach for measurements on lipid membranes. We also list the materials necessary for a FCS experiment on giant unilamellar vesicles (GUVs). Finally, we present simple protocols for the preparation of GUVs and the acquisition and analysis of FCS data on the vesicles, so that diffusion coefficients of fluorescent probes within lipid membranes can be estimated. Key words: Fluorescence correlation spectroscopy, Giant unilamellar vesicles, Diffusion coefficient, Membrane dynamics, Liquid-ordered phase, Liquid disordered phase, Membrane diffusion

1. Introduction The organization and dynamic properties of biological membranes are intimately related to their function. However, their molecular basis still remains poorly understood. Since the fluid mosaic model proposed by Singer and Nicholson in the 1970s (1), the importance of lateral heterogeneities in lipid membranes has become evident. The raft hypothesis, which predicts a functional role for the formation of sphingolipids-enriched microdomains in cell membranes, has matured during the last decade (2). According to it, cells exploit the coexistence of liquid- disordered (Ld) and liquid-ordered (Lo) lipid phases for processes like cell signaling and membrane trafficking. Both Ld and Lo phases are fluid, but they do differ in their organization: Lo phases are usually formed by saturated lipids and cholesterol and exhibit tightly

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_33, © Humana Press, a part of Springer Science+Business Media, LLC 2010

493

494

García-Sáez, Carrer, and Schwille

packed acyl chains, while lipids in Ld phases are usually unsaturated and exhibit disordered acyl chains (3). Fluorescence correlation spectroscopy (FCS) is a powerful technique to measure lateral diffusion of particles within lipid membranes and, hence, to study their lateral organization (4). Based on single-molecule sensitivity, it analyses the fluctuations in fluorescence intensity inside a tiny detection volume (see Fig. 1) (5, 6). The intensity variations in the fluorescence signal originate from changes in the number or the molecular brightness of the fluorophores inside the detection volume. They can be a result of probe diffusion or transport in and out of the detection volume, chemical reactions and/or photophysical processes. As a consequence, diffusion coefficients, dye concentrations, particle sizes a

b

Autocorrelation curve 0.04

Fluorescence trace

0.03

G (τ)

diffusion time (ms) 0.02

0.01

0.00 10

c

100

τ (ms)

1000

10000

15

τD[ms]

10

5

0

–1

–0.8

–0.6

–0.4

–0.2

0

0.2

0.4

0.6

0.8

1

z[µm]

Fig. 1. Fluorescence correlation spectroscopy. The changes with time of the fluorescence intensity emitted by the sample in the focal volume produce an intensity trace (a) that is converted by the autocorrelation function (Eq. 1) into an autocorrelation curve (b). This shows the correlation of the fluorescence signal with itself as a function of time. By fitting of this curve to a model function (see Table 1), the diffusion time can be extracted (see Subheadings 3.1 and 3.4). For the z-scan method, plotting the diffusion time and the particle number as a function of the distance to the membrane plane gives rise to parabolic distributions (c). Fitting these parabolas with Eqs. 4 and 5 allows for the calibration-free calculation of the diffusion coefficient of the membrane fluorophore and the waist of the focal plane

Fluorescence Correlation Spectroscopy for the Study of Membrane Dynamics

495

and conformational changes, and molecular interactions can be assessed by FCS. During the last few years, FCS has been successfully employed for the study of membrane dynamics and protein/lipid interactions (7–12). For practical reasons, FCS measurements have to be performed on membranes several orders of magnitude larger than the detection volume. For example, analysis of the diffusion properties of fluorescent probes within coexisting phases in giant unilamellar vesicles (GUVs) can be used to characterize the lipid phases as Ld or Lo, since they differ in fluidity (13, 14). In addition to model membranes, FCS has contributed to the characterization of the lateral organization of the plasma membrane in living cells (15). Some practical considerations for the measurement of FCS on membrane are reviewed in (16). Currently, there are a number of FCS setups commercially available. Because of their visualization capabilities, those combined with a confocal microscope are preferred for measurements of lipid membranes. Here, we describe a protocol to determine the concentration and the diffusion coefficient of a reporter fluorophore on the membrane of GUVs by FCS. 1.1. Fluorescence Correlation Spectroscopy

The FCS technique relies on the temporal autocorrelation of fluorescence intensity fluctuations. The signal is acquired for a tiny focal volume (in the order of fL) and detected with singlemolecule sensitivity. The autocorrelation analysis measures the self-similarity of the signal with time. It is a mathematical tool that relates the fluorescence signal with itself at different correlation times. The autocorrelation of the fluorescence intensity is calculated with the following expression:

G (τ) =

δF (t )·δF (t + τ) 2 F (t )

(1)

where G is the autocorrelation function, F is the fluorescence intensity as a function of time t, t is the correlation time. The angular brackets refer to time averaging, so that dF(t) = F(t)−áF(t)ñ. To obtain the physical parameters of interest, the correlation curve obtained is fitted with a mathematical model that is based on the characteristics of the system that give rise to the fluorescence fluctuations in the detection volume. Table 1 shows some model functions that are used depending on the processes that take place in the sample under study, like probe diffusion, or photophysical processes like blinking and triplet state. This model also considers the size and shape of the focal volume, the molecular brightness of the fluorophore, and its concentration as a function of position and time. For a detailed derivation of the fitting functions, see (6, 17). For example, in the simple case of pure diffusion in 2D in a plane perpendicular to the optical axis at the equator of the focal volume,

496

García-Sáez, Carrer, and Schwille

Table 1 Model functions for fitting of FCS data Type of diffusion

Model function

3D diffusion

1 

τ 

−1

G3D (τ) = 1 + τ  N  D 2D diffusion

1 G 2D (τ) = N

2D diffusion for two components

2D diffusion with triplet

 τ  1 + τ  D

G 2D+ 2C (τ ) =

1 N total

1 2 τ 1 + τ0 ωD

−1

q f2U f G Df (τ) + q s2U sG Ds (τ) (q f U f + q sU s )2

  τ  −1 G 2D + T (τ) = 1 + T (1 − T ) exp  −   G D (τ)  τT    G 2D + bl (τ) = 1 + c dark / c bright exp (−kbl τ ) G D (τ )t

2D diffusion with blinking 2D with elliptical-Gaussian profile

G 2DG (τ) =

1 N

 τ  1 + τ  D

−1/ 2

1 1 + τ / S 2 τD

Here G is the autocorrelation curve, N is the average number of particles in the detection volume, t is the lag time and tD is the diffusion time. tT corresponds to the triplet time, w0 is the waist of the detection volume, q is the molecular brightness of the f (fast) and s (slow) diffusing components, Y refers to their molar fraction, T is the fraction of fluorophores in the triplet state within the detection volume, c is the concentration of the dark and bright species, kbl is the blinking rate, s is the structure parameter and s = wz/w0. The terms introduced to correct for two components, triplet and blinking are also valid for 3D diffusion.

which approximates the diffusion of a fluorescent probe within the membrane, the autocorrelation function can be modeled as:

1 G D (τ) = N

−1  τ  1 + τ  D

(2)

where N is the average number of fluorescent particles in the detection area, N = C τ 02t , C is the area concentration and w0 the waist radius of the laser focus. Then, the diffusion time tD is related to the diffusion coefficient D through the expression

τD =

τ 02 4D

(3)

Fluorescence Correlation Spectroscopy for the Study of Membrane Dynamics

497

The waist of the focus w0 can be estimated from the same expression by measuring the diffusion time under the same experimental conditions for a reference dye of known diffusion coefficient. In this way, FCS can be used to determine the relative diffusion coefficient of the particles of interest. 1.2. z-Scan Approach

FCS measurements on GUVs are very sensitive to the relative axial position of the membrane plane with respect to the focal volume. In addition, simple FCS acquisitions need calibration of the shape of the focal volume for the estimation of the diffusion coefficient of the probe. This can be a problem in membranes because the calibration measurements are usually performed in solution and the presence of the membrane plane may alter the shape of the focal volume, leading to artefacts in data analysis. Fortunately, specific solutions to perform calibration-free FCS measurements on membranes have been developed during the last years. Some examples are scanning FCS and the z-scan strategy (18–20); simple implementations that also solve the problem of positioning of the membrane plane. Here, we describe the z-scan approach, because it can be performed on any FCS setup without modifications. The z-scan method consists of the collection of several autocorrelation curves at several z-positions at, above and under the membrane plane. This helps with the positioning of the detection volume, since the curve with the smallest diffusion time and number of particles corresponds to the best alignment of focal and membrane planes. In addition, the diffusion coefficient and area concentration can be inferred without calibration from the analysis of the estimated diffusion times and particle numbers at the different z-positions (19, 20). For a Gaussian beam, the following expressions are used:  λ 2 ∆z 2  N = N 0  1 + 20 2 4  p n w0   τD =

ω 02  λ 02 ∆z 2  1 + 4D  π 2n 2 ω 04 

(4)t (5)

where l0 is the excitation wavelength, n is the refractive index of the medium, and Dz is the distance between the membrane and focal planes.

2. Materials 2.1. Instrumentation

A variety of commercial research-grade FCS setups is currently available in the market; for example, the ConfoCor2, recently upgraded to ConfoCor3, from Zeiss (Jena, Germany) that we

498

García-Sáez, Carrer, and Schwille

have used in this case. These setups combine a confocal microscopy system with fluorescence correlation/cross-correlation capabilities and are developed as user-friendly platforms. They include several lasers for excitation and several detection channels with fiber-coupled avalanche photo diodes (APDs) as detectors. Figure 2 shows a schematic view of an FCS setup. In general, an FCS setup consists of an inverted microscope in combination with a high-numerical aperture objective like for example the 40× NA 1.2 UV-VIS-IR C-Apocromat water-immersion objective from Zeiss. The incoming laser light is reflected by the dichroic mirror and focused by the microscope objective to a spot of approximately 0.3–0.5 mm diameter. The fluorescence light emitted by the sample passes the dichroic and the emission filter that removes residual Rayleigh and Raman scattered light. A pinhole is inserted into the image plane to enforce axial resolution. Alternatively, an optical fiber of the same diameter can be used. For detection, APDs with single-photon sensitivity are usually used. The fluorescent trace can then be correlated with a hardware or a software correlator.

Fig. 2. A confocal FCS setup. The laser light is reflected on a dichroic filter and focused on the sample through a high numerical aperture objective. Emitted fluorescence is collected by the objective, passes the emission filter and the pinhole, and is detected by the APD, which is connected to a computer

Fluorescence Correlation Spectroscopy for the Study of Membrane Dynamics

2.2. Samples 2.2.1. GUV Preparation using Pt Wires

499

1. Lipid mixtures: stock solutions prepared by dissolving lyophilized lipids form a reliable source (Avanti Polar Lipids, Alabaster, AL, USA) into chloroform or a 2:1 (v:v) mixture of chloroform and methanol. 2. Home-made teflon and Pt chambers (see Fig. 3). 3. Home-made Al heating block (Fig. 3). 4. A Hamilton syringe. 5. An electric heating plate. 6. A function generator, for example TTi TG315 from TurlbyTandar Instruments Ltd, Huntingdon, Cambs, England. 7. Connecting clamps and cables, for example BNC Male Minigrabber Clip Cable from Pomona Electronics, sold by Farnell, www.farnell.de. 8. Observation chambers, for example Lab-Tek cambered coverglasses from Nalge Nunc Intl., Rochester, New York, USA, www.nuncbrand.com.

2.2.2. Using ITO Coverslips

1. Lipid mixtures as in Subheading 2.2.1. 2. ITO coverslips, for example from GeSiM, Grosserkmannsdorf, Germany, www.gesim.de. 3. Vacuum grease, for example Glisseal from Borer Chemie, Zuchwil, Switzerland,

a

d

c

c b a b

Fig. 3. The Pt wires chamber. A home-made Pt-and-teflon chamber with the following dimensions is shown in (a): a) 15 mm, b) 8 mm, c) 4 mm, d) 30 mm. The connection of the chamber to the function generator and the heating up of the whole setup is shown in (b) and (c) respectively

500

García-Sáez, Carrer, and Schwille

a b

c spacer

tubing holder Cu tape

ITO glass

silicon grease holder

Fig. 4. The ITO coverslips chamber. The metal holders, plastic spacer and ITO coverslips are shown in (a). The assembled chamber is shown in (b). A schematic drawing of the assembled chamber is shown in (c)

4. Polyethylene tubing matching the size of the inlets in the acrylic spacer (see step 6 and Fig. 4), for example 1.57OD × 1.14ID from Warner Instruments, www.warneronline.com. 5. Copper tape, adhesive on one side, for example from SPI Supplies, West Chester, PA, USA, www.2spi.com. 6. Home-made acrylic spacer with three inlets as shown in Fig. 4. 7. Home-made metal holder (see Fig. 4). 8. Crocodiles clamps and cables (see Subheading 2.2.1) 9. Function generator like in Subheading 2.2.1.

3. Methods 3.1. Preparation of Giant Unilamellar Vesicles 3.1.1. Using Pt Wires

1. Clean the Pt chamber with water, ethanol and chloroform and let it dry. Perform these actions under the hood and wear appropriate gloves to protect you from the solvents. 2. Take the stocks of your desired lipid mixtures at 1 mg/mL from the freezer and let them reach room temperature. Since FCS works in the single-molecule regime, include 0.01% of a lipophylic fluorescent dye (see Note 4 on suitable dyes for FCS measurements on lipid membranes).

Fluorescence Correlation Spectroscopy for the Study of Membrane Dynamics

501

3. With a Hamilton syringe, take 6 mL from the lipid solution and spread 3 mL on each of the Pt wires. 4. Put the Pt wires under vacuum at room temperature for 2 h to allow complete evaporation of the solvents. 5. Prepare a sucrose solution of the desired osmolarity (see step 10). For example, you can use concentrations between 10 and 300 mM. Warm up a small fraction of the solution above the transition temperature of the lipid mixture under study. The sucrose solution can be stored for further use. 6. Fill the electroformation chamber with 350 mL of warmed sucrose solution and close it by screwing the cap with the Pt-wires. 7. Place the chamber in a heating block above the transition temperature of the lipid mix and connect the cables of the function generator to each of the two Pt-wires avoiding contact between them (Fig. 3). 8. Start the function generator with a sinus wave at a voltage of 2V (RMS) and a frequency of 10 Hz. Allow electroformation in these conditions for around 1.5 h. Then, decrease the frequency to 2 Hz for around 15 min, so that the GUVs formed gently detach from the Pt wires. Turn off the function generator and disconnect the cables. 9. Fill the wells of an 8-well observation chamber with 700 mL of a blocking solution of 2 mg/mL of BSA (bovine serum albumin) and incubate for around 30 min. Rinse the wells of the observation chamber extensively with Milli-Q water. 10. Add 350 mL of a glucose solution of matching osmolarity to the sucrose solution prepared in step 5, to a well in the observation chamber. If desired, buffers with matching osmolarity to the sucrose solution can be used, and also proteins or other molecules of interest can be added to the chamber (see Note 6). 11. Cut the end of a micropipette tip to widen the opening and gently take the GUV suspension from the electroformation chamber. Add it slowly on top of the glucose solution on the observation chamber and let it sediment for around 5 min. 3.1.2. Using ITO Coverslips

1. Clean two ITO-coverslips with water, ethanol and chloroform and dry them. 2. Take the stocks of your desired lipid mixtures at 10 mg/mL from the freezer and let them reach room temperature. 3. Determine the conductive side of the coverslips by measuring resistance with a voltage tester and attach the end of a 5 cm adhesive copper strip to it. 4. Place an ITO coverslip on a heating block above the transition temperature of the lipid mixture to be used.

502

García-Sáez, Carrer, and Schwille

5. With a Hamilton syringe, take 5 mL from the lipid solution and spread them slowly in a zig-zag over the surface of the coverslip, so that a lipid film is formed after solvent evaporation. 6. Put the ITO coverslip under vacuum for around 2 h to allow complete evaporation of the solvents. 7. Cut three pieces of tubing of around 10 cm and insert them into the outlets of the plastic spacer. 8. Spread grease evenly on both sides of the plastic spacer, taking care not to block the inlet holes, but covering the entire surface to ensure complete sealing of the chamber once it is assembled (Fig. 4). 9. Carefully attach an ITO-coverslip to each side of the plastic spacer, taking care to ensure the conductive side of the coverslips face the inside of the spacer, the cupper strips coincide with the corresponding gaps and the coverslip containing the lipids is positioned at the bottom of the chamber. Press gently on each side to ensure proper sealing with the grease. 10. Assemble the two coverslips with the spacer and the tubing into the metallic holder and fix it with the screws. 11. Warm the assembled chamber above the transition temperature in an oven or a heating block. 12. Prepare a sucrose solution of the desired osmolarity (see Note 5). Warm it up at the same temperature as the electroformation chamber and degas it by sonication in a bath. 13. Take 2 mL of water or sucrose solution with a pre-warmed syringe with an 18G blunt needle and connect it into the tubing of one of the inlets. Inject the solution through the bottom inlet of the chamber inclined around 45° to avoid the formation of bubbles inside the chamber. 14. Inject solution until it exits the chamber through the other two remaining outlets and then seal them with clamps. 15. Place the chamber in an oven or a heating block above the transition temperature of the lipid mixture and connect the cables of the function generator to each of the two cupper strips carefully avoiding contact between them. 16. Start the function generator with a sinus wave at a voltage of 1.4 V (RMS) and a frequency of 10 Hz. Allow electroformation in these conditions for around 1.5 h. 17. Turn off the function generator and let the chamber equilibrate to room temperature before use. 3.2. Setup Alignment and Calibration

To ensure optimal conditions during data acquisition, the FCS setup should be aligned prior to use. In order to check for proper alignment, a calibration measurement in solution with a dye spectrally similar to the fluorophore that will be measured in the

503

Fluorescence Correlation Spectroscopy for the Study of Membrane Dynamics

membrane and with a well-known FCS behavior is recommended. These steps should be performed for the microscope beam path configuration that fits the spectral properties of the fluorophore to be measured. 1. Turn on the excitation laser around 30 min prior to the measurements, to allow stabilization of the light source. 2. Place 200 mL of 50 nM dye solution in an observation chamber similar to the one that will be used for the GUVs samples. For example, free Alexa-488 is a well-known dye in the green range of the spectrum. For other examples of dyes that could be used as references, see Table 2. 3. Select the FCS configuration of the microscope that best fits the spectral properties of the fluorophore under study. Select laser power at 2% (around 20 mWatt in the 488 channel) and place the focal volume well inside the reference dye solution (for example, 100 mm above the coverslip surface). 4. Set the size of the pinhole in front of the detector to 70–100 mm and align it with respect to the beam path. This step involves selecting the position of the pinhole in X and Y directions so that the fluorescence count rate arriving to the detector is maximized. 5. In some setups, it may also be necessary to adjust the position of the collimator to reach maximum count rate. However, modern systems like Confocor 3 are stable in this regard and do not need realignment. 6. While monitoring the counts per molecule or the count rate, turn the correction collar of the objective to correct for the glass thickness of the observation chamber. Select the position that gives the maximum counts per molecule or count rate.

Table 2 Diffusion coefficient of fluorophores commonly used as references in FCS measurements Fluorophore

Diffusion coefficient (mm2 s−1)

References

Rhodamine 6G

426, 414 ± 1*

(23, 24)

Alexa488

435

(23)

eGFP

95

(23)

Alexa543

341

(23)

Atto655

426 ± 8*

(25)

Diffusion coefficients calculated in water solutions at 22.5°C, except for those marked with *, which were obtained at 25°C

504

García-Sáez, Carrer, and Schwille

7. Select the laser power that will be used later on during the sample measurements, usually 0.3–1%, and perform a calibration FCS measurement of the reference dye. As an example, perform 10 acquisitions of 10 s each. 8. Save the data for further analysis. In case of using a mathematical program, export of the data in ASCII format may be required (see step 1 on Subheading 3.4). 9. Check that the autocorrelation curve obtained is similar to the ones published in the literature for that dye and to the curves obtained in previous measurements under the same conditions. Following the steps described below (Subheading 3.4), fit the autocorrelation curve and estimate the diffusion time and the structure parameter. If the setup is properly aligned, the results should be reproducible and in agreement with the bibliography. 3.3. Data Acquisition using z-Scan

Commercially available setups usually provide online data acquisition, which allows monitoring the FCS data while measuring. This has the advantage that one can check instabilities in the fluorescence signal, photobleaching effects, as well as alterations in the shape of the autocorrelation curve. If any of the latter effects occur, the corresponding FCS data should be discarded. In case a single measurement on the membrane plane is to be done, follow steps 1–8 only. 1. Place the observation chamber containing the GUVs on the sample holder of the microscope. 2. In the imaging mode, focus the sample and look for a GUV of a size bigger than around 50 mm, if possible (see Notes 7 and 8). 3. In the imaging mode, focus approximately on top of the vesicle and center the FCS focal volume on it. 4. In the FCS mode, scan in z-direction for the focal plane with maximum fluorescence intensity. Under such conditions, the plane of the membrane is contained within the focal volume. Set the stage height at 0 in that position. 5. Select a laser power that gives maximum counts per molecule without significant photobleaching during the acquisition time. Depending on the microscope and the sample, values around 0.3–1% are normal. 6. Start a test FCS measurement on the focal plane with maximum fluorescence intensity. Check the stability of the system to photobleaching (diffusion times should be compared at different laser powers) and membrane undulations and movements and optimize the measuring time. For statistical reasons, the measuring time has to be of the order of 10,000 times longer than the diffusion time of the fluorophore under study.

Fluorescence Correlation Spectroscopy for the Study of Membrane Dynamics

505

For example, for a probe that shows a diffusion time in the membrane of around 3 ms, the measuring time should be 30 s. If the stability of the systems allows it, perform 3–5 repetitions of the measurement. 7. Again in the imaging mode, select another GUV of similar characteristics and focus on the top plane of the vesicle. 8. Change to FCS mode and scan in z-direction for the focal plane that contains the lipid membrane. 9. Start a FCS measurement under the conditions set for the prior test GUV. 10. Save the data obtained. If necessary, export them to ASCII format (see step 1 in Subheading 3.4). 11. For the z-scan, move the stage and perform FCS measurements above and below the 0 position in steps of 0.1–0.2 mm, until distances of ±0.7–0.8 mm from the 0 position (Fig. 5). If possible, perform the FCS acquisitions with the same duration and repetitions at each of the z positions. For a reliable z-scan analysis, data at a minimum of 7–9 different z positions should be acquired. 12. Select another GUV and measure FCS on it by repeating steps 6–9. Measure on a representative number of vesicles in the sample. 3.4. Data Analysis

1. Open the data obtained with specific software, usually implemented in commercial systems, or on a program with mathematical fitting options, such as Origin or Matlab. 2. Plot the autocorrelation curves obtained for a z-position and discard those with distorted shapes. Repeat the same procedure with the curves obtained at the other z positions.

b laser focal volume

a Lo domain

Ld domain

Fig. 5. Measuring FCS on a GUV with Ld/Lo phase separation. A confocal 3-D reconstruction of the top pole of a GUV is shown in (a). The cross shows the best place to position the focal volume. A schematic drawing of the GUV and the focal volume is shown in (b). The solid arrows show the relative movement of the focal volume when performing a z-scan

506

García-Sáez, Carrer, and Schwille

3. Calculate the average curve for each of the z positions. For qualitative comparisons, the shape of the autocorrelation curve is related to the processes causing the fluorescence fluctuations. Among other, the decay of the curve depends on the type of particle motion, such as transport or diffusion. In addition, the amplitude of the autocorrelation curve is a function of the area concentration of fluorescent particles in the focal plane. 4. For quantitative analysis, fit the autocorrelation curves with an appropriate model function (see Table 1). Usually, a nonlinear least-square fitting algorithm is used. A plot of the fitting residuals provides information about the goodness of fit. Successful fitting of the data will provide an estimation of the diffusion time tD and number of particles N in the focal volume. 5. For the z-scan method, fit simultaneously the calculated values of diffusion time and particle number at the different z positions to Eqs. 4 and 5. See Fig. 1. From this fitting, the values for the waist of the focal volume w0 and the diffusion coefficient D are estimated.

4. Notes 1. Avoid contact of solutions in organic solvents with plastics. Use glassware and organic-solvent resistant material for their handling. 2. Let lipid stocks reach room temperature before opening the vial and minimize the time that it is open to avoid water condensation and changes in lipid concentration due to solvent evaporation. Before closing the vial again, flux an inert gas like N2 or Ar into the vial and seal it with parafilm. 3. Wear appropriate gloves to protect you from organic solvents and work under the hood. 4. An area concentration of fluorescent dye between 1 and 1,000 molecules/mm2 is usually desired for FCS on membranes (16). For the conditions described above, we suggest the use of 0.05% (mol). The family of long-chain dialkylcarbocyanines, like DiI, DiO and DiD (Invitrogen, Oregon), are lipid analogs that span a wide range of wavelengths. They have been successfully used to measure the lateral organization of lipids in model membranes, both in GUVs (13, 21). Another possibility is the use of fluorescently labeled lipids, like rhodamine phosphatidylethanolamine (Avanti Lipids, Alabaster) or BODIPY-cholesterol (Invitrogen). In that case, he fluorophore bound to the lipid molecule can change the behavior of

Fluorescence Correlation Spectroscopy for the Study of Membrane Dynamics

507

the lipid. This happens for the cholesterol derivative, which behaves as a non-raft marker in model membranes (22). 5. Differences in osmotic pressure in- and outside the vesicles can cause them to shrink or swell, even to burst. To avoid this, care should be taken that the sugar solutions and buffers in- and outside the GUVs are isoosmolar. For example, if the GUVs are electroformed in 300 mM sucrose, they should be sedimented on 300 mM glucose, or PBS (phosphate saline buffer), which have the same osmolarity. 6. Always pipette and mix GUV suspensions gently, with slow movements and tips with wide openings to avoid shear flows that destroy the vesicles. 7. Usually, with the electroformation methods described above large GUVs with sizes between 20 and 300 mm are obtained. For FCS measurements, larger vesicles are preferred because they approximate better to a plane in the detection area and cause less photobleaching problems. Depending on the sample composition, it may be difficult to get large-enough vesicles, and in that case, the difficulties just mentioned could become crucial. 8. The selection of the adequate GUV is also important because it will affect the quality of the data and, therefore, of the results obtained. For this reason, we strongly recommend measuring on truly unilamellar, tensed, homogeneous and immobile vesicles.

References 1. Singer SJ, Nicolson GL (1972) The fluid mosaic model of the structure of cell membranes. Science 175:720–731 2. Simons K, Ikonen E (1997) Functional rafts in cell membranes. Nature 387:569–572 3. Simons K, Vaz WLC (2004) Model systems, lipid rafts, and cell membranes. Annu Rev Biophys Biomol Struct 33:269–295 4. Garcia-Saez AJ, Schwille P (2007) Single molecule techniques for the study of membrane proteins. Appl Microb Biotechnol 76: 257–266 5. Eigen M, Rigler R (1994) Sorting single molecules – application to diagnostics and evolutionary biotechnology. Proc Natl Acad Sci U S A 91:5740–5747 6. Magde D, Webb WW, Elson E (1972) Thermodynamic fluctuations in a reacting system

7.

8.

9.

10.

– measurement by fluorescence correlation spectroscopy. Phys Rev Lett 29:705 Chiantia S, Ries J, Kahya N, Schwille P (2006) Combined AFM and two-focus SFCS study of raft-exhibiting model membranes. Chemphyschem 7:2409–2418 Cordeaux Y, Briddon SJ, Alexander SP, Kellam B, Hill SJ (2008) Agonist-occupied A3 adenosine receptors exist within heterogeneous complexes in membrane microdomains of individual living cells. FASEB J 22(3):850–60 Doeven MK, van den Boggart G, Krasnikov VV, Poolman B (2008) Probing receptor– translocator interactions in the oligopeptide ABC transporter by fluorescence correlation spectroscopy. Biophys J 94(10):3956–3965 Kahya N, Scherfeld D, Bacia K, Poolman B, Schwille P (2003) Probing lipid mobility of

508

11.

12.

13.

14.

15.

16.

17.

García-Sáez, Carrer, and Schwille raft-exhibiting model membranes by fluorescence correlation spectroscopy. J Biol Chem 278:28109–28115 Smithers NP, Hodgkinson CP, Cuttle M, Sale GJ (2008) Insulin-triggered repositioning of munc18c on syntaxin-4 in GLUT4 signalling. Biochem J 410:255–260 Swift JL, Burger MC, Massotte D, Dahms TE, Cramb DT (2007) Two-photon excitation fluorescence cross-correlation assay for ligand-receptor binding: cell membrane nanopatches containing the human micro-opioid receptor. Anal Chem 79:6783–6791 Bacia K, Schwille P, Kurzchalia T (2005) Sterol structure determines the separation of phases and the curvature of the liquid-ordered phase in model membranes. Proc Natl Acad Sci U S A 102:3272–3277 Kahya N, Scherfeld D, Bacia K, Schwille P (2004) Lipid domain formation and dynamics in giant unilamellar vesicles explored by fluorescence correlation spectroscopy. J Struct Biol 147:77–89 Lenne PF, Wawrezinieck L, Conchonaud F, Wurtz O, Boned A, Guo XJ, Rigneault H, He HT, Marguet D (2006) Dynamic molecular confinement in the plasma membrane by microdomains and the cytoskeleton meshwork. EMBO J 25:3245–3256 Ries J, Schwille P (2008) New concepts for fluorescence correlation spectroscopy on membranes. Phys Chem Chem Phys 10(24): 3487–3497 Petrov EP, Schwille P (2008) State of the art and novel trends in fluorescence correlation spec troscopy. Springer Ser. Fluoresc. 6:145–197

18. Ries J, Schwille P (2006) Studying slow membrane dynamics with continuous wave scanning fluorescence correlation spectroscopy. Biophys J 91:1915–1924 19. Benda A, Benes M, Marecek V, Lhotsky A, Hermens WT, Hof M (2003) How to determine diffusion coefficients in planar phospholipid systems by confocal fluorescence correlation spectroscopy. Langmuir 19:4120–4126 20. Humpolickova J, Gielen E, Benda A, Fagulova V, Vercammen J, Vandeven M, Hof M, Ameloot M, Engelborghs Y (2006) Probing diffusion laws within cellular membranes by Z-scan fluorescence correlation spectroscopy. Biophys J 91:L23–L25 21. Ries J, Schwille P (2006) Studying slow membrane dynamics with continuous wave scanning fluorescence correlation spectroscopy. Biophys J 91:1915–1924 22. Chiantia S, Kahya N, Schwille P (2007) Raft domain reorganization driven by short- and long-chain ceramide: a combined AFM and FCS study. Langmuir 23:7659–7665 23. Petrasek Z, Schwille P (2008) Precise measurement of diffusion coefficients using scanning fluorescence correlation spectroscopy. Biophys J 94:1437–1448 24. Culbertson CT, Jacobson SC, Michael Ramsey J (2002) Diffusion coefficient measurements in microfluidic devices. Talanta 56:365–373 25. Dertinger T, Pacheco V, von der Hocht I, Hartmann R, Gregor I, Enderlein J (2007) Twofocus fluorescence correlation spectroscopy: a new tool for accurate and absolute diffusion measurements. Chemphyschem 8:433–443

Chapter 34 Liposome Biodistribution via Europium Complexes Nathalie Mignet and Daniel Scherman Abstract The drug-delivery field needs tools to follow vector biodistribution. Radioactive tracers and conventional fluorophores are widely used. We propose here to use europium complexes. Use of pulsed light source time-resolved fluorimetry takes into account the fluorescence decay time of the lanthanide chelates to gain sensitivity in biological media. The method was developed to follow liposome biodistribution. Octadecyl–DTPA.Eu compound has been prepared and incorporated into liposomes without alteration of its fluorescence signal. The method has been validated by comparison with fluorophore-labelled liposomes. The way to proceed to use this method for liposomes or other vectors is detailed. Key words: Liposomes, Europium, Time-resolved fluorimetry, Fluorescence, Pharmacokinetics

1. Introduction The field of vectorisation represents a major goal in formulating drugs, modifying their pharmacokinetic profile or reducing their toxicity. Hence, following the distribution of the vector is of major interest; however, only a few methods are available for this purpose. Unless the intrinsic properties of the vector allow detecting it, a tracer should be incorporated into the vector. This tracer should be stable, not released in biological media and not be exchangeable with lipidic membranes, in case of a tracer-bearing lipid. The more sensitive tracers so far are radiolabelled (14C, 1H), and have been used for liposome biodistribution studies. These tracers are also advantageous, since they do not interfere with the liposome bilayer such as fluorophores could do. However, performing experiments with radioactive tracers require legal authorisation and experimental constraints such as safety cautions and waste processing are to be tackled.

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_34, © Humana Press, a part of Springer Science+Business Media, LLC 2010

509

510

Mignet and Scherman

Spectrofluorescence is an easier method to handle in the laboratory. Incorporating commercially available labelled lipids into liposomes allowed following them after systemic injection in mice (1). We used this method by incorporating phosphatidylethanolamine–fluoresceine or phosphatidylethanolamine–rhodamine in cationic lipid/DNA complexes and improved the extraction process to recover the whole lipoplex content injected in the blood (2). Though this method is simple and reproducible, it requires extraction of the lipids with organic solvents from the biological media to improve signal to noise ratio. We developed an original method that combines sensitivity in biological media and versatility, as it does not require any treatment of the biological tissue in organic solvent. Lanthanides are an interesting class of compounds thanks to their intrinsic properties, large stokes shift, narrow emission peak and optimal emission wavelength which allow using time-resolved fluorimetry, useful for biological media. Lanthanides suffer from low molar extinction coefficients but chelates of lanthanides have been developed to transfer energy to the lanthanide ions and limit this drawback. The ability to measure delayed fluorescence in biological media allows getting rid of the medium background and solely measuring the lanthanide-chelate signal. The use of lanthanidechelates for protein detection or to follow interleukin-2 biodistribution has been reported (3, 4). Lipid-based chelates have also been shown to amplify the lanthanide signal (5, 6). Based on these previous results, we chose to use lipid-based lanthanidechelate and incorporate them into liposomes. The method to measure directly these liposomes in biological media was worked out. Then, biodistribution of these liposomes post-systemic injection was evaluated and compared to rhodamine labelled liposomes to validate the assay.

2. Materials 2.1. Abbreviations used

Polyethylene glycol (PEG); Egg-phosphatidylcholine (EPC); diethylenetriaminepentaacetic acid (DTPA); Magnetic Resonance Imaging (MRI); Europium (Eu); Dulbecco’s Modified Eagle’s Medium (DMEM); and Foetal bovine serum (FBS).

2.2. Chemicals Provided or Synthesized

EPC, phosphatidylethanolamine rhodamine, Phosphatidylethanolamine-Poly(Ethyleneglycol)45 and filters for the extruder were purchased from Avanti Polar Lipids. Cholesterol, chloroform, Hepes, DMEM and FBS were purchased from Sigma. The centrifugal concentrator was purchased from Millipore. The octadecyl-DTPA used for this study was prepared in three steps from stearic acid as described in Mignet et al. (7).

Liposome Biodistribution via Europium Complexes

2.3. Equipment

511

Liposomes were prepared by the film method on a rotary evaporator Heidolph, VWR, equipped with a vacuubrand CVC2 to control the pressure. Avanti Mini-Extruder from Avanti Polar Lipids, Inc. Size and zeta potentials measurements were performed on a Zeta Sizer NanoSeries from Malvern Instruments equipped with a 632.8 nm helium neon laser, 5 mW power, and an angle detection at 173° (non-invasive back scattering). The centrifugator used is a megafuge 1.0 from Heraeus Sepatech. Time-delayed fluorescence was measured on a multilabel plate reader Wallac Victor2 1420 Multilabel Counter, Perkin Elmer, France. TRF program for europium performs 1,000 pulses/s with an excitation light at 320 nm. In the period between flashes, 615 nm fluorescence of the sample is measured for 400 ms that allows the short-time fluorescence to decay. The photons counted during one-second were recorded and expressed as counts per second (cps). Phosphorescence spectra were performed on a Varian cary Eclipse Fluorescence Spectrophotometer using the phosphorescence mode.

3. Methods The DTPA is a widely used chelate for a lanthanide such as europium (8). We chose to use a lipid-bearing the DTPA moiety in order to easily incorporate it into liposomes via hydrophobic interactions between the carbon chains. The DTPA moiety is a big molecule and insertion of this lipid, as for all kinds of labelled lipids, into liposome bilayer might modify the liposome structure and a first study on the liposome integrity has to be performed. For this purpose, we did evaluate different kinds of lipids-bearing DTPA, such as cholesterol-DTPA and octadecyl–DTPA (7) to follow conventional liposomes. We had also previously developed a cationic spermine–DTPA (9) to introduce it into cationic lipoplexes and follow the biodistribution of these particles by MRI. Indeed, the DTPA moiety is able to condense different ions, such as gadolinium which is a contrast agent used in MRI and also europium, useful for time-resolved fluorimetry (10). After lipid–DTPA.Eu insertion into liposomes, size of the liposome formed was first evaluated by dynamic light scattering, then excitation and emission phosphorescence measurements were performed to insure that the europium signal was not modified by the lipid insertion (Fig. 1). A liposome concentration range in buffers and biological media was searched to insure that this label would be useful for in vivo experiments. The influence of the biological media on the signal of rhodamine

512

Mignet and Scherman

a

1000 Intensity (a.u)

30

800 Intensity (a.u)

50

35

900 700 600

40

25

30

350

15

20

300

10

250

0

200

10 0 300

400

400

20

5

500

450

350 Wavelength (nm)

300

400

150

200

100

100

50

0 200

250

300

350

400

450

0 500

Wavelength (nm) Octadecyl-DTPA

b

Octadecyl-DTPA lipo

20

30

18 25

Intensity (a.u)

16 14

20

12 10

15

8 10

6 4

5

2 0 550

600

650

700

0 750

Wavelength (nm) Octacedyl-DTPA

Octacedyl-DTPA lipo

Fig. 1. Influence of the insertion of the octadecyl–DTPA.Eu in the liposomes as compared to the free octadecyl–DTPA.Eu on the excitation (a) and emission (b) spectra. The emission spectrum is unchanged when the lipid is inserted in the liposome bilayer. In opposite, the excitation spectrum presents some differences. The characteristic signal of europium (10) at 320 and 396 remain at the same position but their level is increased upon lipid incorporation. Reproduced with permission from (7)

and europium-labelled liposomes was also evaluated. We could show that the loss on fluorescence was a factor 25 while the loss on time-delayed fluorescence was only a factor 3 when liposomes were incubated in hepes or medium-containing serum (Fig. 2). Finally, the method was validated for biodistribution experiments by comparing the biodistribution of conventional liposomes labelled with a rhodamine lipid and identical liposomes labelled with the octadecyl–DTPA.Eu compound.

Liposome Biodistribution via Europium Complexes

513

Medium influence on Europium labelled liposome detection

100000

CPS

10000 1000 100 10 0,001

1000000

Hepes DMEM+FBS

0,01 0,1 1 Liposome concentration (µM)

10

Medium influence on Rhodamine labelled liposome detection

CPS

100000 10000 1000 Hepes DMEM+FBS

100 10 0,001

0,01 0,1 1 Liposome Concentration (µM)

10

Fig. 2. Influence of the medium on the detection of the liposomes. The liposomes were incubated in Hepes 10 mM or DMEM+10% FBS and fluorescence read, either direct fluorescence to read the rhodamine or delayed fluorescence to read the europium content. One can see that for the same concentration of liposome, bearing the same amount of lipidrhodamine or lipid DTPA.Eu, the level is lower when the liposome contains the lipid–DTPA.Eu. One can also notice that the fluorescence level is far less influenced by the medium if the liposome is labelled with the lipid-DTPA.Eu. The fluorescence loss between Hepes and DMEM+SVF was measured as a factor 3 ± 1 for the Eu labelled liposomes and a factor 26 ± 5 for the rhodamine labelled liposomes

The data obtained for both liposomes were very similar and correlated with a similar liposome isotopically labelled as previously reported (Fig. 3). 3.1. L ipid Labelling

1. The octadecyl–DTPA is dissolved in H2O (1.5 µmol, 1 mg/ mL) (see Notes 1 and 2). 2. Add EuCl3.6H2O (7.5 µmol, 2.7 mg) to the lipid and stir overnight. 3. Separate the EuCl3 . 6H2O in excess by centrifugal filtration. Load the complex formed on a centrifugal concentrator equipped with a 5 kD PES membrane. Add 4 mL of H2O. Centrifuge at 2,000 × g for 2 min (see Note 3) and repeat the operation until no Europium could be detected in the filtrate (see Note 4).

514

Mignet and Scherman Liposome Eu / 2H

80

Liposome Eu 24H 70

Liposome Rhod / 2H Liposome Rhod 24H

% of the injected dose

60 50 40 30 20 10 0 Blood

Liver

Spleen

Lung

Fig. 3. Comparaison of the biodistribution of rhodamine and europium labelled liposomes at 2 and 24 h post systemic injection into C57/Bl6 mices. The levels obtained are comparable to what was previously reported for similar isotopically labelled liposomes (11)

4. Check that the lipid is well labelled by taking a sample and reading the associated level of fluorescence as compared to the non-labelled lipid and the background of EuCl3⋅6H2O (see Note 5). 3.2. Preparation of the Conventional Liposomes Labelled with Europium

1. Dissolve separately the lipids Egg-phosphatidylcholine (15 mmol, 12 mg), Cholesterol (10 mmol, 3.8 mg), the octadecyl–DTPA.Eu (1.5 mmol, 1.5 mg) and phosphatidylethanolamine-Poly(Ethyleneglycol)45 (1.5 mmol, 4 mg) in chloroform (400 mL for the lipids, 200 mL for the PEG–ipid). Take care that the lipids are well dissolved separately before mixing them (see Note 6). 2. Mix them into a round-bottomed flask (10 mL) (see Note 7) 3. Put the flask at the evaporary evaporator to remove the solvent in a pressure-controlled manner. First, reduce the pressure from 1,000 to 200 mbar in approximately 15 min with a middle rotation speed. When the drop forms, increase the rotation speed at its maximum level to drag the drop into the film. Then, reduce the pressure from 200 to 5 mbar in 30 min and leave the film under reduced pressure for an additional hour (see Note 8). 4. The film being dry, add 1 mL (to afford a final concentration of 28 mM) milliQ filtered (0.22 µm) H2O and leave the flask under gentle rotation overnight at room temperature (see Note 9). 5. Mix gently the mixture on a vortex if the film is not fully detached from the wall.

Liposome Biodistribution via Europium Complexes

515

6. Extrude the liposome 10 times using successively polycarbonate filters of 0.44 and 0.22 µm. 7. Control the size by dynamic light scattering. For measurements on a nanoZS (Malvern Instruments), dilute 5 mL of the particles obtained in a 500 mL cuve, start the measure in the automatic mode. The liposomes should have a size around 100 nm with a poor polydispersity index (see Note 10). 3.3. Preparation of the Conventional Liposomes Labelled with Rhodamine 3.4. Phosphorescence Spectra

The same protocol as in Subheading 3.1 was used. Only the octadecyl–DTPA.Eu was replaced by phosphatidylethanolamine rhodamine in the same proportions. 1. Prepare a cuve containing the octadecyl–DTPA.Eu lipid and another containing europium-labelled liposomes. Put the same amount of Europium in both cuves. 2. In the phosphorescence mode of the system, chose lem = 616 nm, delay time: 0.1 ms, gate time: 5 ms, band pass fixed at 10 nm (see Note 11) and scan an excitation spectrum. 3. In the phosphorescence mode of the system, chose lem = 396 nm, delay time: 0.1 ms, gate time: 5 ms, band pass fixed at 10 nm and scan an emission spectrum. 4. Compare the excitation and emission scans of the lipid and the liposome.

3.5. Test in Buffers

1. Prepare a concentration range of liposomes to evaluate a straight concentration range useful for quantification (see Note 12). 2. Dilute in the medium of interest for the time of interest. For the measurements presented in Fig. 34.2, liposomes were diluted in hepes 10 mM, and DMEM+FBS 10% and measured without any incubation. 3. Read the fluorescence. For the phosphorescence either using lex = 396 nm, lex = 616 nm, delay time: 0.1 ms, gate time: 5 ms, band pass fixed at 10 nm on the Varian Eclipse spectrophotometer or lex = 320 nm, lex = 615 nm, delay time: 0.4 ms, gate time: 0.4 ms, 1,000 pulses/s on a multilabel plate reader Wallac Victor2 1420 Multilabel Counter, (Perkin Elmer) (Last program used for the measurements presented in Fig. 2). For the fluorescence, rhodamine level was read using lex = 550 nm and lex = 590 nm band pass fixed at 5 nm on the Varian Eclipse spectrophotometer or with filters lex = 535 ± 10 nm and lex = 570 ± 10 nm on the multilabel plate reader Wallac Victor2 (Perkin Elmer).

516

Mignet and Scherman

3.6. Biodistribution In vivo ( see Note 13)

1. Mice administration. Anesthetize the C57Bl/6 mice (Janvier) by intraperitoneal injection of a mix of Ketamine (85.8 mg/ kg; Centravet) and Xylazine (3.1 mg/kg; Bayer) diluted in 150 mM NaCl. 2. Inject europium-labelled liposomes (2 mmol, 200 ml) or rhodamine-labelled liposomes (2 mmol, 200 ml) into the mouse-tail vein. 3. Euthanize the first group of mice, 2 h post-injection, the second group 24 h post-injection. 4. Collect the plasma by cardiac puncture. 5. Collect and weigh the liver, spleen, and lungs. 6. Homogenize the organs in pH 7.4 PBS using an Ultra Thurax (Diax 600, Heidolph, Fisher). 7. For the Europium-labelled lipids, read directly the level of fluorescence in the plasma by TRF with a spectrofluorometer (Victor, Perkin Elmer). 8. For the Europium-labelled lipids in tissue, incubate the tissue homogenates at 2.5 mL/g in lysis buffer (Roche) 1× (1/1, v/v). Mix vigorously overnight. Centrifuge (1,000 rpm, 5 min). 9. Assay the fluorescence intensity on the supernatant by TRF. 10. Perform a calibration cuve in each different tissue with the labelled liposomes (see Note 14). 11. Calculate the amount of liposome in the plasma or tissue homogenates with the calibration performed in the tissues of interest and expressed as the remaining percentage of injected dose. 12. For the rhodamine-labelled lipids, extract the blood or tissues with 3 mL methanol/Chloroform (1/1, v/v) by vigorous mixing for 30 min in the blood and overnight for the tissue homogenates. Centrifuge at 3,000 rpm during 10 min. Assay the supernatant for rhodamine using lex = 550 nm and lex = 590 nm wavelengths. Perform a calibration curve in each different tissue with the labelled liposomes. Calculate the amount of liposome in the plasma or tissue homogenates with the calibration cuve performed in the tissues of interest and expressed as the remaining percentage of injected dose.

4. Notes 1. All buffers and water used should be filtered on 0.22 µm filters since any dust might interfere with light-scattering experiments.

Liposome Biodistribution via Europium Complexes

517

2. Pay attention to the pH. The DTPA–lipid will precipitate while the pH remains too acidic. pH has to be adjusted to pH 7 by dropwise addition of 15 mL NaOH (0.1 M). 3. Adjust the centrifugation speed and length by checking that the lipid is not dry, a minimum of 200 mL should remain to recover it properly. 4. Take a sample (200 mL) of the filtrate and load it on a 96-well plate (it could be 1 mL sample in a cuve) to read the europium content in the filtrate. To increase the signal, you could add EDTA in the well or in the cuve to complex Eu. 5. To give an idea of the level range, with the spectrofluorimeter used (Victor, perkin Elmer), we obtained around 3,000 CPS at 5 µM lipid. 6. Solubility of the lipids should be checked with intensive care since presence of non-soluble entities will appear in the film and reduce particle homogeneity after hydration. 7. The ratio between the volume to be reduced (or the amount of lipids) and the round-bottomed flask is important since the film should occupy as much flask wall as possible. The surface of the flask occupied by the film will depend on the number of layers in the liposomes. 8. Make sure that the film is not crackled by a too rapid pressure reduction. If so, dissolve again the lipids in 1 mL CHCl3 and start again part 3. It is always preferable to obtain an homogeneous film along the flask wall, it will provide more homogeneous liposome size after the hydration step. 9. Evaporation and hydration time are usually reported as shorter, but we have found that taking time to do these steps is required to form homogeneous liposome sizes. 10. The polydispersity index reflects the homogeneity of the suspension. It is less than 0.1 for standards such as latex, it should be lower than 0.2 to have homogeneous liposomes. 11. Adjust the band pass according to the signal (dependent on the amount put in the cuve). If the signal saturates, reduce the band pass or dilute the sample. If the signal is too low, increase the band pass or concentrate the sample. 12. A straight concentration range of interest for the in vivo studies was determined from 10 nM to 1 µM for europium and rhodamine-labelled liposome (Fig. 2). 13. Experiments should be conducted following the NIH recommendations and in agreement with a regional ethic committee for animal experimentation. A number is provided upon project presentation, in pour case, n°P2.PB003.04.

518

Mignet and Scherman

14. Signal of the europium complexes is highly dependent on the environment, a calibration curve should be performed in each medium of interest. References 1. Takeuchi H, Kojima H, Yamamoto H, Kawashima Y (2001) Evaluation of circulation profiles of liposomes coated with hydrophilic polymers having different molecular weights in rats. J Control Release 75:83–91 2. Nicolazzi C, Mignet N, de la Figuera N, Cadet M, Torero R, Seguin J, Bessodes M, Scherman D (2003) Anionic poly(ethyleneglycol) lipids incorporated into cationic lipoplex increase their circulation time in the blood. J Control Release 88:429–443 3. Laukkanen ML, Orellana A, Keinänen K (1995) Use of genetically engineered lipidtagged antibody to generate functional europium chelate-loaded liposomes: application in fluoroimmunoassay. J Immunol Methods 185:95–102 4. Neville M, Richau K, Boni L, Pflug L, Robb R, Popescu M (2000) A comparison of biodistribution of liposomal and soluble IL-2 by a new method based on time-resolved fluorimetry of europium. Cytokine 12:1702–1711 5. Roy B, Santos M, Mallik S, Campiglia A (2003) Synthesis of metal-chelating lipids to sensitize lanthanide Ions. J Org Chem 68:3999–4007 6. Marchi-Artzner V, Brienne MJ, GulikKrzywicki T, Dedieu JC, Lehn JM (2003) Selective complexation and transport of europium ions at the interface of vesicles. Chem Eur J 10:2342–2350

7. Mignet N, le Masne de Chermont Q, Randrianarivelo T, Seguin J, Richard C, Bessodes M, Scherman D (2006) Liposome biodistribution by time resolved fluorimetry of lipophilic europium complexes. Eur Biophys J 35:155–161 8. Weinmann H, Brasch R, Press W, Wesbey G (1984) Characteristics of gadolinium–DTPA complex: a potential NMR contrast agent. Am J Roentgenol 142:619–624 9. Leclercq F, Cohen-Ohana M, Mignet N, Herscovici J, Scherman D, Byk G (2003) Design synthesis and evaluation of gadolinium cationic lipids as tools for biodistribution studies of gene delivery complexes. Bioconj Chem 14:112–119 10. Elster A, Jackels S, Allen N, Marrache R (1989) Dyke Award. Europium-DTPA: a gadolinium analogue traceable by fluorescence microscopy. Am J Neuroradiol 10:1137–1144 11. Parker D, Dickins R, Puschmann H, Crossland C, Howard J (2002) Being excited by lanthanide coordination complexes: aqua species, chirality, excited-state chemistry, and exchange dynamics. Chem Rev 102:2389–2403 12. Parr M, Ansell S, Choi L, Cullis P (1994) Factors influencing the retention and chemical stability of poly(ethylene glycol)–lipid conjugates incorporated into large unilamellar vesicles. Biochim Biophys Acta 1195:21–30

Chapter 35 Biosensor-Based Evaluation of Liposomal Binding Behavior Gerd Bendas Abstract Biosensors can be regarded as analytical devices that transform biologically given facts, such as the appearance of physiological substrates, or biological recognition processes of ligands and receptors into detectable signals without the need of further labeling. This chapter introduces acoustic wave sensors as mass-sensitive tools to investigate the liposomal binding behavior onto simulated biological surfaces. These sensors do not only allow for quantification of the liposomal binding intensity, but further analytical readings give insight into the liposomal appearance at the binding site, e.g., deformation or fusion. Since the liposomal behavior at the target binding site might have strong impact on therapeutic effects, a prediction of liposomal appearance and a controlled modulation thereof appear possible with the help of biosensors. Here, the function of a quartz crystal microbalance (QCM) and the bio-functionalization of quartz sensors are reported for a series of liposomal binding experiments. Liposomes containing biotin as model ligands were selected to evaluate their binding to avidin-modified sensors. The data, representing binding intensity and liposome deformation, are explained with respect to the role of binding strength and lipid composition for liposomal behavior. Key words: Biosensors, Quartz crystal microbalance (QCM), Liposome binding, Fusion, Liposome deformation

1. Introduction Successful liposomal binding and accumulation at the target site is the basis for therapeutic targeting approaches. However, therapeutic effects will result finally from the mode of liposomal interaction with the target cells. With respect to the liposomal drug load or the pharmacological approach, it might be anticipated that liposomes will either be taken up by cells for generating intracellular effects, or stay at the target cells and disrupt the release of the drug load locally outside the target cells. Both options will V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_35, © Humana Press, a part of Springer Science+Business Media, LLC 2010

519

520

Bendas

overlap in practice and can partly be directed by cellular epitopes, such as targeting internalizing receptors, or by the liposomal composition. Although the prediction of the liposome–cell interactions would greatly increase therapeutic benefit, most analytical techniques can neither detect the liposome–cell interactions directly in the physiological situation nor in adequate model systems. The development of biosensor analytics during the last decade and successful biosensor application in a wide variety of analytical approaches might be an option to get further insight into the liposomal binding characteristics. Biosensors are analytical devices of different analytical principles, design, and application, which allow label-free detection of biological substrates or processes by transforming the information into detectable readouts. Biosensors can be classified according to the mode of signal transduction. An early established and the most popular biosensor in biosciences is an optical sensors using surface plasmon resonance (SPR) based technique (1, 2), known as BIAcore systems. We refer to acoustic wave sensors that are based on piezoelectric properties of the sensor materials, in most cases quartz crystals. It was in 1959 when Sauerbrey showed the mathematical correlation that the shift in resonance frequency of a thickness-shear mode resonator is proportional to the mass load on its surface (3), which given in modified form in Eq. (1), where ∆f is the change in frequency, f0 the oscillation frequency, A the sensing area, rq the density of the resonator, mq the shear-modulus, and ∆m the bound mass on the surface.

∆f = −

2 f 02 ∆m = −Cf ∆m A mq rq

(1)

This was the basis for the development of piezoelectric, masssensitive devices (4). Besides surface acoustic wave (SAW) sensors (5), quartz crystal microbalance (QCM) is the most popular piezoelectric sensor (6). In principle, an electrical excitation of a thin quartz sensor results in an oscillation of the resonance frequency (most commonly between 5 and 10 MHz), which indirectly correlates with the mass load on its surface. Consequently, changes in mass can be detected in real time by monitoring frequency shifts. To apply the QCM as a biosensor, the quartz electrodes have to be functionalized with the relevant bio-component. QCM has successfully been applied in a wide spectrum of analytical approaches, such as polymer surface characterization, ligand– receptor recognitions, protein adsorption onto membranes, or DNA/RNA interaction with complementary strands. Biosensors have also been applied in the field of liposomes. In most cases, liposome adhesion and the formation of supported

Biosensor-Based Evaluation of Liposomal Binding Behavior

521

planar bilayers have been studied extensively as a model for cellular membranes in biophysical studies and biomedical application. In combination with other techniques, spreading or vesicle fusion onto different surface materials was investigated (7) and interpreted with respect to the lipid species (8). The binding of biotinylated liposomes to surface bound avidin was investigated under static conditions, and the liposomal deformation was confirmed finally by atomic force microscopy (AFM) (9). Several studies have used the liposomal presentation of ligands to amplify the ligand binding signals onto the immobilized receptors at the QCM sensors (10, 11), and others immobilized intact liposomes at the quartz surface to investigate release kinetics (12) or protein–liposome interactions (13). Besides a quantification of liposome binding by detecting the frequency shifts, QCM devices can have further analytical readings, namely monitoring of the dissipation energy (QCM-D) or alternatively damping analysis (Fig. 1). Both provide information on the flexibility of surface-bound layers. These parameters are most useful when given as a ratio with the frequency drop (D/f ratio). Thus, differences between intact (higher damping) or flattened liposomes (lower damping) can be detected. On a biophysical background, liposome fusion onto different materials was investigated with respect to liposomal size and concentration (14, 15). Patel and Frank illustrated rapid formation of dense surface bilayers using osmotically stressed vesicles in a QCM dissipation study (16). In a recent report (17), a QCM damping study was performed to investigate the liposomal behavior under simulated target binding conditions. It was probed whether the liposomal behavior at the target site is predictable. Liposomes that differed

Free oscillation of the quartz sensor

Damped oscillation of the sensor in aqueous medium

Increased damping of the oscillation by viscoelastic bound mass (liposomes)

Fig. 1. Schematic illustration of oscillating quartz sensors and the damping increase by viscoelastic masses at the sensor surface (upper cartoons), and the corresponding oscillation amplitudes (lower cartoons)

522

Bendas

B & W IN PRINT

in lipid composition, amount, and affinity of the coupled homing device were analyzed with respect to binding intensity and deformation onto immobilized target molecules. Damping analysis, e.g., the D–f plot, appeared as meaningful parameter. Biotinylated liposomes deformed upon binding to avidin and the deformation correlated with biotin concentration. Liposomal deformation can trigger fusion and content release when liposomes contain dominant fractions of unsaturated phosphatidylethanolamine (PE). In contrast, PEGylated liposomes displayed high damping, indicating conformational stability. It was concluded that liposome behavior can be controlled by certain parameters and be predicted with therapeutic relevance. This chapter refers to the QCM device used in (17). The following instructions assume the application of a QCM biosensor with damping analysis, such as a Liqui Lab21 model (ifak e.V. Barleben, Germany), to investigate the liposomal binding under simulated shear force conditions given in a flow chamber system, as illustrated in Fig. 2. Although the principle of measurement appears to be comparable to similar QCM devices, the absolute changes in frequency and damping frequency are related to this model and may deviate in quantity and dimensions using other QCM biosensors.

Fig. 2. (a) Schematic illustration of the gold-covered quartz sensors (diameter 14 mm, the resulting sensitive area A 0.282 cm2, f0 10 MHz, mq the shear-modulus 2.947 × 1011 g/cm s2, density rq 2.648 g/cm3. According to Eq. (1), these parameters result in a constant Cf of 1.24 ng/Hz. (b) The quartz sensor is integrated into the schematically depicted flow chamber, where 1 is top of chamber with tube connections for medium flow, 2 is a rubber gasket, 3 is a conductive gasket and 4 the bottom with contacts to the oscillator. The chamber has a volume of about 100 mL

Biosensor-Based Evaluation of Liposomal Binding Behavior

523

2. Materials 2.1. Liposome Preparation

The following lipids were dissolved as 10 mM chloroform (Riedel de-Haen, Seelze, Germany) stem solutions, kept at −20°C, and equilibrated to room temperature for pipetting and use. 1. Soy phosphatidylcholine (SPC) from Lipoid AG Ludwigshafen (Germany). 2. Methoxy-polyethyleneglycol-phosphatidylethanolamine (PEG-PE 2000) from Avanti Polar Lipids (Alabaster, AL, USA). 3. N-biotinyl-dipalmitoyl-phosphatidylethanolamine Polar Lipids).

(Avanti

4. N-[biotinyl(polyethylenglycol)-2000]-dipalmitoylphosphatidylethanolamine (Avanti Polar Lipids). 5. Phosphate buffer (150 mM), pH 7.4 prepared according to standard protocols and kept at 7°C until use. 2.2. Biosensor Activation

1. 6-Mercaptohexan-1-ol (Fluka, Neu-Ulm, Germany) was dissolved in chloroform to obtain a 1 mM solution, which was kept at −20°C before use. 2. Cyanuric chloride (Sigma, Deisenhofen, Germany) was kept dry at −20°C before use and freshly dissolved in chloroform at 1% (mass/volume). 3. Borate buffer (10 mM), pH 8.8 was prepared according to standard protocols and kept at 7°C until use. 4. Avidin (Sigma, Deisenhofen) was freshly dissolved in borate buffer at 0.2 mg/mL for immediate use.

3. Methods

The following instructions refer to the investigation of large unilamellar vesicles (LUVs) (diameter about 100 nm), PEGylated or non-PEGylated with different concentrations of biotin–PE or biotin–PEG–PE as ligand for binding sensor-immobilized avidin. Taking biotin–avidin interaction as model for ligand–receptor recognition, the impact of binding affinity and ligand concentration on liposome binding intensity and liposomal deformation should be evaluated. Detailed results can be found in (17). In principle, these instructions can be transferred to other liposomal targeting approaches, e.g., immunoliposomes, taking the immobilization of the corresponding biological recognition structures as prerequisite.

524

Bendas

3.1. Liposome Preparation

1. Suitable aliquots of the chloroform lipid stem solutions (10 mmol) were pipetted into a 25-mL round bottom flask. 2. Chloroform was removed by a rotary evaporator for 30 min in a 35°C water bath, followed by drying the resulting lipid film for 16 h in a vacuum exsciccator. 3. The lipid film was hydrated by adding 1 mL phosphate buffered saline (PBS, pH 7.4) to obtain 10 mM multilamellar liposome vesicle (MLV) dispersion after vortexing for about 10 min, followed by mechanical shaking for about 16 h. 4. Unilamellar liposomes were prepared with a mini-extruder (Avanti Polar Lipids) from multilamellar vesicles extruded 19 times through a 200-nm polycarbonate membrane, 19 times through a 100-nm polycarbonate membrane, and 10 times through a 50-nm polycarbonate membrane (Whatman Nuclepore 18 mm polycarbonate membrane, Richmond, USA). 5. The liposome size has to be controlled by photon correlation spectroscopy (PCS).

3.2. Activation of the Sensor Surface

1. For preparing the cleaning solution of the sensors (piranha solution), conc. H2O2 and conc. H2SO4 were mixed at a volume ratio of 1:3 in a cooling water bath (see Note 1). 2. When the mixture is cooled down to room temperature (RT), quartz sensors can be dipped in the solution with the help of a pair of tweezers (or similar) for about 10 s followed by rinsing the sensor with demineralized water. 3. The cleaning procedure is repeated six times to remove any remaining bound residues at the gold surface. 4. The sensor has to be rinsed with absolute ethanol and dried under a stream of nitrogen. 5. The cleaned quartz sensors were immersed into a 1 mM chloroform solution of 6-mercaptohexan-1-ol for about 12 h at RT. Due to the thiol–gold reaction, a kind of monolayer is formed with terminal hydroxyl groups. 6. After rinsing the quartz crystals briefly with ethanol and drying them under a stream of nitrogen, the sensors were put into a chloroform solution of cyanuric chloride (see Note 2) for 2 h at RT, followed by ethanol washing and drying. Cyanuric chloride is a reactive linker with three binding sites (with successive reactivity). Attached to the mercaptohexanol, it is able to react with nucleophils, resulting in protein immobilization. Reactivity can be obtained in aqueous media under mild basic conditions. 7. A solution of 20 mL avidin (1 mg/mL in 10 mM borate buffer, pH 8.8) was placed on the upper gold surface of the quartz

Biosensor-Based Evaluation of Liposomal Binding Behavior

525

Fig. 3. AFM image of an avidin-coated quartz sensor. Avidin was coupled to 6-mercaptohexan-1-ol via cyanuric chloride; dimension of the image is 15 mm

sensors. After 2 h at RT, the quartz crystals were rinsed once again with demineralized water and dried under a stream of nitrogen. The resulting dense layer of immobilized avidin is illustrated in Fig. 3, as seen by AFM. 8. To control binding specificity, the coupling step in 7 can be performed with a bovine serum albumin (BSA) solution of identical concentration. 3.3. Biosensor Measurements of the Liposomal Binding Behavior

1. The functionalized biosensor is inserted into the flow chamber system as illustrated in Fig. 2, taking care of sealing the chamber by proper screwing down. 2. The flow chamber is connected to the oscillator by pins and attached with tubes to a peristaltic pump (via other flow chambers, since up to four flow chambers can be detected simultaneously) (Fig. 4). Tubes can be arranged to allow a closed circuit flow system or (as used here) flow from a buffer reservoir for a single liposome passage along the sensors. 3. The transparent cover of the biosensor has to be closed to allow thermostatic conditions (see Note 3). A constant flow of PBS (volume flow rate of 270 mL/min) is started to equilibrate the quartz crystals under flow conditions, and the

526

Bendas

Fig. 4. The QCM biosensor device used for this study. Four flow chamber systems, each connected to an oscillator, are combined within a peristaltic pump flow system and inserted into a thermostatic device. Data were collected and evaluated by a computerized system

frequency is monitored in the corresponding computerized system. 4. When equilibration has occurred, and no trend in frequency is evident, liposomal solution can be introduced into the flowing medium by a three-way valve, resulting in a final liposomal concentration of 0.5 mM (see Note 4). 5. Liposome binding to the quartz surface is evident by a significant drop in frequency, while the damping frequency is increased. The binding of the biotinylated liposomes reaches saturation within 4 min. The data are monitored in real time (Fig. 5) and stored for subsequent evaluation. 6. For comparison of the binding quantity, the maximum frequency drop can be used and calculated in bound mass by Eq. (1). Insight into the liposomal binding behavior can be obtained by analyzing damping changes. Whereas intact liposomes induce a higher damping by viscoelasticity (Fig. 5a),

Biosensor-Based Evaluation of Liposomal Binding Behavior Hz

Adhesion of intact liposomes

400 300 200 100 0 –100 –200

0

500

1000

1500

Spreading and fusion of liposomes

500 Change in frequency Damping frequency change

Change in frequency Damping frequency change

Hz

500

2000 time [s]

527

400 300 200 100 0 –100 –200

0

500

1000

1500

2000 time [s]

Fig. 5. Schematic illustration of a frequency and damping frequency course of liposome binding onto the quartz sensor surface. Although liposome binding might induce a similar drop in frequency in (a) and (b), indicating similar mass increase at the sensor surface, the damping change refers to the binding of intact liposomes in (a) or the deformation and flattening in (b)

flattening and fusion of liposomes is evident by lower damping (Fig. 5b). 7. The data obtained with the QCM device described here can be used for the evaluation of the liposomal appearance. D/f ratios of 4–3 indicate nondeformed or less deformed, e.g., PEGylated, liposomes and further decrease in the D/f ratio indicates stronger deformation, while D/f ratios of 1 and lower refer to fused liposomes and bilayer formation. In case of the biotinylated liposomes, liposomes with up to 5 mol% biotin–PE bind with low deformation, higher biotin–PE concentrations (up to 20 mol%) induce much stronger flattening, while 30 mol% induced liposomal disruption and bilayer formation (17). Liposomes with biotin coupled onto the PEG terminus tend also to be deformed.

4. Notes 1. When preparing the cleaning solution (Subheading 3.2, step 1), take proper care when mixing the reagents, because the solution is very corrosive and gets hot and has to be cooled. Do not clean the sensors in the hot solution. The cleaning of the sensor surface is often accompanied with the appearance of small pinholes. Long-term contact with the sensors can destroy the gold electrodes. 2. For successful biosensor activation (Subheading 3.2, step 5.), the use of dried chloroform is essential, since saturation of chloroform with air humidity might impede the coupling reaction of cyanuric chloride. A protocol for drying chloroform can be taken from chemical textbooks. For example,

528

Bendas

phosphorus pentoxide is added to commercial chloroform for about 12 h followed by a distillation procedure. The freshly distilled chloroform should be kept in closely stoppered flasks. 3. Oscillation frequency is highly sensitive to temperature changes within the QCM device or the flow chambers. Although the QCM device is equipped with a thermostatic system, direct sunlight, airflow or non-thermostatic solutions can impede the detection. 4. When inserting the liposomal dispersion into the flow medium, air bubbles have to be avoided.

Acknowledgments The author would like to thank Dr Matthias Höpfner for expert application and optimization of the QCM device for liposomal binding studies and for designing the illustrations, and Professor Udo Bakowsky for performing the AFM investigations. References 1. Masson L, Mazza A, Brousseau R (1994) Stable immobilization of lipid vesicles for kinetic studies using surface plasmon resonance. Anal Biochem 218:405–412 2. Besenicar M, Macek P, Lakey JH, Anderluh G (2006) Surface plasmon resonance in protein– membrane interactions. Chem Phys Lipids 141:169–178 3. Sauerbrey G (1959) Verwendung von Schwingquarzen zur Wägung dünner Schichten und zur Mikrowägung. Zeitschrift für Physik 155:206–222 4. Janshoff A, Galla HJ, Steinem C (2000) Piezoelectric mass-sensing devices as biosensors – an alternative to optical biosensors? Angew Chem Int Ed 39:4004–4032 5. Martin F, Newton MI, McHale G, Melzak KA, Gizeli E (2004) Pulse mode shear horizontal–surface acoustic wave (SH-SAW) systems for liquid based sensing applications. Biosens Bioelectron 19:627–632 6. Janshoff A, Steinem C (2005) Label-free detection of protein–ligand interactions by the quartz crystal microbalance. Methods Mol Biol 305:47–64 7. Keller CA, Kasemo B (1995) Surface specific kinetics of lipid vesicle adsorption measured with a Quartz crystal microbalance. Biophys J 75:1397–1402

8. Richter R, Mukhopadhyay A, Brisson A (2003) Pathways of lipid vesicle deposition on solid surfaces: a combined QCM-D and AFM Study. Biophys J 85:3035–3047 9. Pignataro B, Steinem C, Galla HJ, Fuchs H, Janshoff A (2000) Specific adhesion of vesicles monitored by scanning force microscopy and quartz crystal microbalance. Biophys J 78:487–498 10. Alfonta L, Willner I, Trockmorton DJ, Singh AK (2001) Electrochemical and quartz crystal microbalance detection of the cholera toxin employing horseradish peroxidase and GM-1functionalized liposomes. Anal Chem 73:5287–5295 11. Yun K, Kobatake E, Haruyama T, Laukkanen ML, Keinänen K, Aizawa M (1998) Use of quartz crystal microbalance to monitor immunoliposome–antigen interaction. Anal Chem 70:260–264 12. Brochu H, Vermette O (2007) Liposome layers characterized by quartz crystal microbalance measurements and multirelease delivery. Langmuir 23:7679–7686 13. Morita S, Nukui M, Kuboi R (2006) Immobilization of liposomes onto quartz crystal microbalance to detect interaction between liposomes and proteins. J Colloid Interface Sci 298:672–678

Biosensor-Based Evaluation of Liposomal Binding Behavior 14. Reimhult E, Höök F, Kasemo B (2003) Intact vesicle adsorption and supported biomembrane formation from vesicles in solution: influence of surface chemistry, vesicle size, temperature, and osmotic pressure. Langmuir 19:1681–1691 15. Reimhult E, Höök F, Kasemo B (2006) A multitechnique study of liposome adsorption on Au and lipid bilayer formation on SiO2. Langmuir 22:3313–3319

529

16. Stalgren JJR, Claesson PM, Wärnheim T (2001) Adsorption of liposomes and emulsions studied with a quartz crystal microbalance. Adv Colloid Interface Sci 89–90: 383–394 17. Höpfner M, Rothe U, Bendas G (2008) Biosensor-based evaluation of liposomal bahavior in the target binding process. J Liposome Res 18:71–82

Chapter 36 Use of Liposomes to Study Vesicular Transport Kohji Takei, Hiroshi Yamada, and Tadashi Abe Abstract Liposomes have been utilized for variety of membrane transport studies including clathrin-mediated endocytosis. Here we introduce clathrin-coated structures that are generated on large unilamellar liposomes by incubating with clathrin coat proteins. Large unilamellar liposomes are also used to reconstitute vesicle formation in a cell-free system, and the vesicle formation can be quantified by using dynamic light scattering (DLS). Furthermore, phagocytosis assay using liposome-conjugated styrene beads is demonstrated. Key words: Vesicular transport, Membrane traffic, Endocytosis, Phagocytosis

1. Introduction Intracellular vesicular transport is mediated by coated vesicles. In vitro reconstitution of coated vesicles was first developed by Rothman and colleagues (1), and the method led to the identification of coat protein complex I (COPI) (2) and coat protein complex II (COPII) (3). All the coated vesicles and coated structures with clathrin coat, COPI, and COPII were generated on liposomes instead of biological membranes (4–6). Vesicle formation from liposomes mimicking endocytosis was reconstituted in vitro, and the vesicle formation was quantified using dynamic light scattering (DLS) (7, 8). Utilization of liposomes in vesicular transport study made it possible to clarify therole of membrane phospholipids in vesicle formation and elucidate the protein–lipid interactions involved in the process. Foreign bodies, bacteria, and apoptotic cells are eliminated by phagocytosis (9). In vertebrates, macrophages and testicular Sertoli cells are the predominant phagocytes. Binding of these objects to the receptors on the surface of phagocytes, such as Fcg receptors or phosphatidylserine (PS) receptors, initiates phagocytosis. V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_36, © Humana Press, a part of Springer Science+Business Media, LLC 2010

531

532

Takei, Yamada, and Abe

Phagocytosis in Sertoli cells is initiated by the recognition of PS on the germ cell surface (10), and the PS-dependent phagocytosis is analyzed using PS-coated styrene beads (11).

2. Materials 2.1. Lipids

1. Rat total brain lipid extract from bovine brain (Sigma, St. Louis, MO). 2. Phosphatidylinositol (4,5) bis phosphate [PI(4,5)P2] (Sigma, MO). 3. Phosphatidylserine (PS) (Avanti Polar Lipids, AL). 4. Phosphatidylethanolamine (PE) (Avanti Polar Lipids, AL). 5. Phosphatidic acid (PA) (Avanti Polar Lipids, AL). 6. Phosphatidylinositol (PI) (Avanti Polar Lipids, AL). 7. Phosphatidylcholine (PC) (Avanti Polar Lipids, AL). 8. Cholesterol (Avanti Polar Lipids, AL). 9. Biotinylated phosphatidylethanolamine (Molecular Probe, CA).

(biotinylated-PE)

10. Lipids solubilized in chloroform at the concentration of 10–50 mg/ml to make the stock solution, which can be stored at −20°C (see Note 1). 2.2. Preparation of Bovine Brain Cytosol

1. Brain sauce: 320 mM sucrose, 25 mM Tris–HCl, pH 7.4. 2. Breaking buffer: 500 mM KCl, 10 mM MgCl2, 250 mM sucrose, 25 mM Tris–HCl, pH 8.0. Add 2 mM ethylenediaminetetraacetic acid (EDTA) or ethylene glycol tetraacetic acid (EGTA), 1 mM dithiothreitol (DTT), and protease inhibitors (phenylmethylsulfonyl fluoride (PMSF) and protease inhibitors cocktail, see below) just before use. 3. PMSF (1,000× Stock): 0.4 M PMSF dissolved in EtOH 4. Protease inhibitors cocktail (250× Stock): Benzamidine 1 mg/ ml H2O, Leupeptin 1 mg/ml H2O, antipain 1 mg/ml EtOH, pepstatin A 1 mg/ml MeOH, and aprotinin 1 mg/ml H2O 5. Dialysis buffer: 50 mM KCl, 25 mM Tris–HCl, pH 8.0. Add 1 mM DTT immediately before use; omit where indicated.

2.3. Preparation of Clathrin Coat Proteins from Rat Brain

1. Homogenization buffer: 320 mM sucrose, 4 mM HEPES/ NaOH, pH 7.4. 2. Tartrate buffer: 0.1 M K2–tartrate, pH 7.3. 3. Buffer A: 10 mM glucose, 5 mM KCl, 140 mM NaCl, 5 mM NaHCO3, 1 mM MgCl2, 1.2 mM Na2HPO4, 20 mM HEPES, pH 7.4. 37°C.

Use of Liposomes to Study Vesicular Transport

533

4. 10× stock of buffer B: 100 mM MES/NaOH, pH 6.5, 1 mM EGTA, 0.5 mM MgCl2. 5. Ficoll buffer (100 mM MES/NaOH, pH 6.5, 1 mM EGTA, 0.5 mM MgCl2, 12.5% sucrose (w/v), 12.5% ficoll (w/v). 6. Buffer B for density gradient (BBD): 100 mM MES/NaOH, pH 6.5, 1 mM EGTA, 0.5 mM MgCl2, 8% sucrose (w/v). 2.4. Preparation of Cytosolic Buffer

1. Cytosolic buffer, in which intracellular cytosolic ionic conditions are mimicked, is used for in vitro reconstitution experiments. Final concentrations are as follows. 25 mM HEPES–KOH, 25 mM KCl, 2.5 mM magnesium acetate, 100 mM K–glutamate, pH 7.2. 2. Concentrated cytosolic buffer (×10) can be prepared and stored at −20°C.

2.5. Cell Culture for Phagocytic Assay

1. Cells: TM-4 (mouse Sertoli cell line; ATCC No. CRL-1715). 2. Cell Culture: Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 2.5% fetal bovine serum (FBS) and 5% horse serum (HS). 3. Cytocharasin D (Sigma-Aldrich, MO) is dissolved at 2 mM in dimethylsulfoxide (DMSO) and stored at −30°C. 4. Dynasore (Sigma-Aldrich, MO) is dissolved at 40 mM in DMSO and stored at −30°C. 5. Microscope cover slips coated with collagen type I (12 mm diameter). 6. Phosphate buffered saline (PBS (+)): 150 mM NaCl, 10 mM phosphate, pH 7.4, 1.5 mM CaCl2, and 1 mM MgCl2. 7. Beads labeling buffer; Rhodamine–avidin (Pierce Biotech., IL) in PBS (+) at 4°C. 8. Paraformaldehyde: Prepare a 2% (w/v) solution in PBS (+) fresh for each experiment. 9. Permeabilization solution: 100 mM digitonin in PBS (+). 10. Nuclear staining: 300 nM DAPI (4’,6’-diamino-2-phenylindole) in PBS (+). 11. Mounting medium: PermaFluor Aqueous Mounting Medium (Beckman Coulter Inc., CA).

3. Methods In vitro reconstitution of endocytic structures on liposomes involves the preparation of large unilamellar liposomes, and cytosolic protein fractions. Clathrin-coated pits can be generated by incubating liposomes with clathrin coat proteins (12) in the

534

Takei, Yamada, and Abe

Fig. 1. Electron micrograph of clathrin-coated buds formed on liposomes. Liposomes of total brain lipid extract were incubated with clathrin coat proteins in vitro. Bristle-like structures typical of clathrin coats are evident. Calibration bar represents 100 nm. (Reproduced from ref. (6) with permission from Cell Press)

presence of adenosine triphosphate (ATP) and GTPgS (guanosine 5’-O-[g-thio]triphosphate), which arrest fission process of the coated pits, and they can be morphologically observed by electron microscopy (6) (see Fig. 1). Incubation of liposomes with brain cytosol in presence of ATP and GTP results in the formation of small vesicles, which can be quantified by DLS (7, 8). As particles in the sample randomly move by Brownian movement, which largely depends on their sizes, when light is irradiated onto the moving particles, the wavelength of the scattered light is altered by Doppler effect, and the qualitative changes of the reflected light called “fluctuation” is detected by DLS. Thus, DLS allows us to determine not only the diameters of particles but also the relative numbers and weight distribution at each size range of the vesicles (13). Phagocytosis is triggered by stimulating PS receptors on the surface of testicular Sertoli cells (10). To detect PS-dependent phagocytosis, unilamellar small liposomes containing PS and biotinylated-PE are prepared, and then the liposomes are conjugated with streptavidin-coated styrene beads (2 mm in diameter). Finally, PS-conjugated beads are used for phagocytosis assay to determine the kinetics of phagocytosis and immuno-localization of functional proteins (11, 14). 3.1. Large Unilamellar Liposomes for In Vitro Reconstitution

1. Eighty microliters of the lipid stock solution, containing 2 mg lipid (rat total brain lipid extract, or mixture of defined composition), is transferred to a 10-ml glass tube and diluted with 0.5 ml of the same organic solvent. All the processes are carried out at room temperature, unless otherwise noted. 2. Evaporate the solvent by passing a stream of nitrogen gas through the tube. During this process, the test tube is rotated manually so that the lipid spreads evenly forming opaque thin

Use of Liposomes to Study Vesicular Transport

535

lipid film on the glass wall. The lipid film is further dried in a lyophilizer for 2 h (see Note 2). 3. Hydrate the dried lipid film by blowing a stream of watersaturated nitrogen gas over the dried lipid film until opacity of the lipid film is slightly lost by the hydration, for approximately 20 min. 4. Two microliters of degassed 300 mM sucrose in distilled water is gently poured into the tube. Nitrogen gas was flushed into the tube without agitating the solution, and then the tube is sealed and left undisturbed for 2 h at 37°C to allow spontaneous formation of liposomes. 5. Gently swirl the glass tube to resuspend the liposomes. Large aggregates of lipids are removed by quick centrifugation. Properties of the prepared liposomes are checked by negative staining EM (Fig. 2a). The liposomes are mostly unilamellar spheres with large diameters, some exceeding 1 mm. The prepared liposomes can be stored at 4°C for several days.

c

80

Distribution (%)

a

20.0

94.6

447.2

2114.7

10000

Diameter (nm)

b 80

Distribution (%)

d

20.0

94.6

447.2

2114.7

10000

Diameter (nm) Fig. 2. Formation of small vesicles from liposomes (a, b) and quantification of the vesicle formation by dynamic light scattering assay (c, d). Electron micrograph of negatively stained large unilamellar liposomes composed of 20% cholesterol and 80% brain lipid extract (w/w) (a). Incubation of the large unilamellar liposomes with brain cytosol in the presence of ATP and GTP leads to massive formation of small vesicles (b). Calibration bar represents 200 nm. Distributions of prepared large nunilamellar liposomes (c) and vesicles formed by the incubation as above (d). Note an evident peak of small vesicles in (d) (Reproduced from ref. (7) from PNAS Publishing)

536

Takei, Yamada, and Abe

3.2. Preparation of Bovine Brain Cytosol

1. Collect two fresh brains from a slaughterhouse, and keep in Brain sauce. All the processes are carried out at 4°C. 2. Remove brain stem. Use the forebrain and cerebellum (600 g). Remove the clotted blood and meninges if present. 3. Rinse and mince the tissue coarsely in 1,500 ml of Breaking buffer. 4. Homogenize with a blender (30 s), and additionally with a Potter-Elvehjem glass–Teflon homogenizer (two strokes). 5. Centrifuge at 5,000 × g for 1 h. 6. Centrifuge the supernatant at 10,000 × g for 1 h. 7. Centrifuge the supernatant at 171,000 × g for 2 h. 8. Dialyze the supernatant against dialysis buffer (20 L for 2 h × 2 times, using Spectrum Spectra/Por 3 membrane). The second dialysis may proceed overnight. Save the second dialysis buffer. 9. Centrifuge the dialyzed sample at 171,000 × g for 2 h. 10. Add ammonium sulfate to 60% saturation slowly over 20–30 min. Allow to stand for 30 min. 11. Remove the precipitate by centrifugation at 9,500 × g for 30 min. 12. Resuspend the pellet in dialysis buffer to one-tenth of the volume dialyzed in step 11. 13. Dialyze against dialysis buffer (20 L for 1 h × 2 times, using Spectra/Por 3). 14. Dialyze for 2 h against 20 L dialysis buffer without DTT. 15. Centrifuge the dialyzed sample at 100,000 × g for 2 h. 16. Recover, mix, and freeze the supernatant in liquid nitrogen, and store at −80°C.

3.3. Preparation of Clathrin Coat Proteins from Rat Brain

This preparation involves two parts; purification of clathrin coated vesicles (CCV) (steps 1–19) and removal of the coat proteins from CCV (steps thereafter). 1. Take out 30 fresh rat brains. Wash twice with ice-cold homogenization buffer. All processes are carried out at 4°C unless otherwise noted. 2. Homogenize the brain in total 300 ml of ice-cold homogenization buffer at 600 rpm, 10 strokes. Centrifuge the homogenate at 3,000 × g for 2 min. 3. Centrifuge the post-nuclear supernatant at 14,500 × g for 12 min. 4. Resuspend the pellet in 25 ml ice-cold tartrate buffer and centrifuge at 5,000 × g for 4 min.

Use of Liposomes to Study Vesicular Transport

537

5. Remove inner, red core part of the pellet (see Note 3) with a Pasteur pipette, and resuspend remaining white pellet in 25 ml of tartrate buffer. Centrifuge at 5,000 × g for 4 min. 6. Resuspend the pellet in 50 ml of buffer A. Incubate at 37°C for 15 min with stirring. 7. Dilute the incubated sample with twofold tartrate buffer, and centrifuge at 5,000 × g for 4 min. 8. Resuspend pellet in 20 ml of tartrate buffer. 9. Add tenfold ice-cold H2O, and homogenize immediately on ice with three strokes at 2,000 rpm. 10. Add 1/10 vol. of a 10× stock of buffer B and mix. Add protease inhibitors (final 1 mM PMSF, 1 mM benzamidine, 1 mg/ml pepstatin A). 11. Centrifuge for 20 min at 20,000 × g. 12. Centrifuge the supernatant at 55,000 × g for 1 h. 13. Resuspend the pellet in 10 ml buffer B by homogenization (three strokes at 2,000 rpm). Pass the suspension once through a 27-gauge needle. 14. Mix with 10 ml of ficoll buffer, and centrifuge for 40 min at 40,000 × g. 15. Take the supernatant and dilute with fivefold buffer B. 16. Centrifuge at 100,000 × g for 1 h. 17. Resuspend the pellet in 15 ml buffer B, and centrifuge at 20,000 × g for 20 min. 18. Layer the supernatant on top of 15 ml BBD in an SW28 tube, and centrifuge for 2 h at 82,700 × g. 19. Resuspend the CCV pellet in 0.3 ml of buffer B. 20. Dilute 40 ml CCV suspension into 1 ml of 300 mM Tris–HCl, pH 9.0, and rotate for 1 h at 37°C. 21. Centrifuge at 120,000 × g for 15 min. 22. Resuspend the pellet in 40 ml of buffer B. 3.4. In Vitro Formation of Clathrin-Coated Pits on Liposomes

In vitro vesicle formation from liposomes is performed by incubation of liposomes with brain cytosol (final 6 mg/ml) and nucleotides in cytosolic buffer. Usually, the incubation mixture volume is 500 ml. 1. Dilute brain cytosol in 10× cytosolic buffer (25 mM HEPES– KOH at pH 7.2, 25 mM KCl, 2.5 mM magnesium acetate, 100 mM potassium glutamate). 2. Add nucleotides, 2 mM ATP, 200 mM GTPgS. Samples containing ATP are also supplemented with an ATP regenerating

538

Takei, Yamada, and Abe

system consisting of 18.1 mM creatine phosphate and 18.1 IU/ml creatine phosphokinase. In control samples omitting ATP, an ATP depleting system consisting of 5 U/ml of hexokinase and 10 mM glucose is supplemented (see Note 4). 3. Add 200 mg of liposomes and incubate mixtures for 15 min at 37°C. 3.5. In Vitro Formation of Vesicles from Liposomes

In vitro vesicle formation from liposomes is performed by incubation of liposomes with brain cytosol (final 6 mg/ml) and nucleotides in cytosolic buffer. Usually, the incubation mixture volume is 500 ml. 1. Dilute brain cytosol in 10× cytosolic buffer (25 mM HEPES– KOH at pH 7.2, 25 mM KCl, 2.5 mM magnesium acetate, 100 mM potassium glutamate). 2. Add nucleotides 2 mM ATP, 200 mM GTP (see Note 5). Samples containing ATP are also supplemented with an ATP regenerating system consisting of 18.1 mM creatine phosphate and 18.1 IU/ml creatine phosphokinase. In samples with brain cytosol and GTP but not with ATP, an ATP depleting system consisting of 5 U/ml of hexokinase and 10 mM glucose is supplemented. 3. Add 200 mg of liposomes and incubate mixtures for 15 min at 37°C.

3.6. Quantification of Vesicle Formation by DLS

1. DLS assay – The sizes, relative numbers, and weight distribution in each size of vesicles formed are measured by DLS with a DLS-7000 AR-III spectrophotometer (Otsuka Electronics Co., Osaka, Japan). 2. To prepare the sample for the DLS assay, 500 ml of the incubation mixture is diluted with 1 ml of 300 mM sucrose. As for liposomes, 200 ml of the liposome suspension is diluted with 1.2 ml of 300 mM sucrose. 3. Measurement range for DLS is set from 20 to 10,000 nm to avoid protein particles (see Note 6). As negative controls, a mixture of cytosol, ATP, and GTP without liposomes is measured and no peak is detected at this setting.

3.7. Phagocytic Assay

1. Liposomes composed of 30% PS (w/w), 60% PC (w/w), and 10% biotinylated-PE (w/w) are mixed in a fresh glass tube, and messed up to 200 ml with chloroform. The mixture is dried under nitrogen (see Note 7) and further under vacuum overnight. 2. The lipid mixtures are incubated with serum-free DMEM at 37°C for more than 2 h. 3. The glass tubes are gently rotated manually to remove the lipid membrane from the surface of the tube.

Use of Liposomes to Study Vesicular Transport

539

4. The mixture is gently vortexed six times. Then, small liposomes are made by sonication (see Note 8). 5. Two milligrams of the liposomes is incubated with 80 ml of streptavidin-coated styrene beads (2 mm in diameter, Poly sciences, Inc., PA) with gentle rotation for 2 h at room temperature. 6. The slurry is centrifuged at 5,000 × g for 5 min and the beads are resuspended in 4 ml of serum-free DMEM. 7. TM4 cells, mouse Sertoli cell line, are plated on a 6-well plate containing four collagen type I-coated cover glasses. 8. After 48 h culture, the cells (1 × 104 cells/well) are incubated with drug (ex. dynasore) at the desired concentration, 2 mM cytochalasin D for negative control, or with the equivalent amount of DMSO through the experiments at 37°C in culture medium. 9. To measure phagocytosis, cells are incubated with diluted bead suspension (1:3) previously prepared with culture medium on each cover glass and incubated at 37°C for the various time points (see Note 9). 10. Cells are gently washed four times with PBS (+). 11. To distinguish between internalized beads and those attached to the cell surface, the latter are labeled with diluted rhodamine–avidin (1:300, Pierce Biotech., IL) with PBS (+), which specifically binds to biotinylated-PE, at 4°C for 30 min before fixation. 12. Cells are gently washed three times with PBS (+) at 4°C. 13. Then cells are fixed with 2% paraformaldehyde (PFA) in PBS (+). 14. Cells are permeabilized with 100 mM digitonin and stained with 300 nM DAPI for 10 min at room temperature (see Note 10). 15. The samples are then ready to be mounted. If they are on a cover glass, then the slip is carefully inverted onto a drop of mounting medium on a microscope slide (see Note 11). The samples are left at room temperature for 2 h and stored in the dark at 4°C before observation. 16. Excitation at 543 nm induces the rhodamine fluorescence (red) for the lipid-coated styrene beads, which are not internalized to cells, while excitation at 364 nm induces DAPI fluorescence (blue). Suitable software can be used to overlay the phase contrast and fluorescence images. Examples of the signals for non-phagocytosed beads and phagocytosed beads are shown in Fig. 3 (14). 17. To quantify phagocytosis, the rhodamine-positive and negative beads on the cell surface are counted using phase contrast and fluorescent microscopy. Phagocytic indices are presented as the number of rhodamine-negative beads per Sertoli cell.

540

Takei, Yamada, and Abe

a Control

+ Dynasore

+ Cyt D

Rhodamine-avidin

Overlay

b Phagocytosis (beads per cell)

DIC

n=120 3

2 n=149

**

n=289

**

1

n=146

**

0

0

10

40

Cyt D

Dynasore (mM)

Fig. 3. Phagocytosis assay using PS-coated beads. (a) TM4 cells were incubated with PS-coated beads in the presence of dynasore at the indicated concentrations for 6 h at 37°C, and then washed. As a negative control, cells were incubated with 2 mM cytochalasin D (Cyt D). Rhodamine-avidin positive beads represent those attached to the cell surface but not internalized. Nuclei are labeled with DAPI. Bar: 10 mm. Quantitation of phagocytosis in control or dynasore-treated cells (b). Rhodamine-negative beads, which represent internalized beads, were counted using phase contrast and conventional fluorescent microscopy. All results are reported as means ± SEM from three experiments. Statistical significance was determined by Student’s t tests (**p < 0.01). (Reproduced from ref. (14) with permission from Elsevier Science)

4. Notes 1. In order to avoid oxidation during the storage, nitrogen gas is flushed in stock vials, and sealed with aluminum foil. 2. In case a lyophilizer is not available, sample can be dried in a desiccator connected with a vacuum pump. 3. The red core part of the pellet is enriched with mitochondria, and surrounding white part contains synaptosomes, isolated nerve endings, form which CCV are to be isolated. As mitochondria and synaptosomes are sedimented together, the former are segregated at this step. 4. GTPgS, the unhydlolyzable analogue of GTP, is often used to block processes that require GTP hydrolysis. As dynamin GTPase is implicated in the fission process of clathrin-mediated endocytosis, the presence of GTPgS results in the massive formation of clathrin-coated pits (6). 5. Both ATP and GTP are required to promote vesicle formation from liposomes in vitro (7). 6. Dilute samples with 300 mM sucrose to adjust the light scattering intensity to approximately 10,000.

Use of Liposomes to Study Vesicular Transport

541

7. It is important to make lipid film uniformly on the surface of the glass tube. 8. Fifteen seconds of sonication at medium power is enough. 9. Since the kinetics of phagocytosis in Sertoli cell is relatively slow (10), cells are incubated for 6 h to ensure internalization of the beads. 10. The cells can be immunostained after the permeabilization with digitonin at this step. 11. We usually use a mounting medium, which is not required for sealing the cover glasses. In either case, nail varnish is used to seal the sample. The sample can be viewed immediately after the varnish is dry, or may be stored in the dark at 4°C for up to a month.

Acknowledgement The work was supported in part by grants from the Ministry of Education, Science, Sports and Culture of Japan. References 1. Balch WE, Dunphy WG, Braell WA, Rothman JE (1984) Reconstitution of the transport of protein between successive compartments of the Golgi measured by the coupled incorporation of N-acetylglucosamine. Cell 39:405–416 2. Orci L, Glick BS, Rothman JE (1986) A new type of coated vesicular carrier that appears not to contain clathrin: its possible role in protein transport within the Golgi stack. Cell 46:171–184 3. Barlowe C, Orci L, Yeung T, Hosobuchi M, Hamamoto S, Salama N, Rexach MF, Ravazzola M, Amherdt M, Schekman R (1994) COPII: a membrane coat formed by Sec proteins that drive vesicle budding from the endoplasmic reticulum. Cell 77:895–907 4. Spang A, Matsuoka K, Hamamoto S, Schekman R, Orci L (1998) Free in PMCCoatomer, Arf1p, and nucleotide are required to bud coat protein complex I-coated vesicles from large synthetic liposomes. Proc Natl Acad Sci U S A 95:11199–11204 5. Matsuoka K, Schekman R (2000) The use of liposomes to study COPII- and COPI-coated vesicle formation and membrane protein sorting. Methods 20:417–428

6. Takei K, Haucke V, Slepnev V, Farsad K, Salazar M, Chen H et al (1998) Generation of coated intermediates of clathrin-mediated endocytosis on protein-free liposomes. Cell 94:131–141 7. Kinuta M, Yamada H, Abe T, Watanabe M, Li S-I, Kamitani A et al (2002) Phosphatidylinositol 4, 5-bisphosphate stimulates vesicle formation from liposomes by brain cytosol. Proc Natl Acad Sci U S A 99:2842–2847 8. Kinuta M, Takei K (2002) Utilization of liposomes in vesicular transport studies. Cell Struct Funct 27:63–69 9. Greenberg S, Grinstein S (2002) Phagocytosis and innate immunity. Curr Opin Immunol 14:136–145 10. Shiratsuchi A, Umeda M, Ohba Y, Nakanishi Y (1997) Recognition of phosphatidylserine on the surface of apoptotic spermato genic cells and subsequent phagocytosis by Sertoli cells of the rat. J Biol Chem 272:2354–2358 11. Yamada H, Ohashi E, Abe T, Kusumi N, Li SA, Yoshida Y et al (2007) Amphiphysin 1 is important for actin polymerization during phagocytosis. Mol Biol Cell 18: 4669–4680

542

Takei, Yamada, and Abe

12. Maycox PR, Link E, Reetz A, Morris SA, Jahn R (1992) Clathrin-coated vesicles in nervous tissue are involved primarily in synaptic vesicle recycling. J Cell Biol 118: 1379–1388 13. Berne BJ, Pecora R (1976) Dynamic light scattering with applications to chemistry,

biology and physics. John Wiley and Sons, New York 14. Otsuka A, Abe T, Watanabe M, Yagisawa H, Takei K, Yamada H (2009) Dynamin 2 is required for actin assembly in phagocytosis in Sertoli cells. Biochem Biophys Res Commun 378:478–482

Index A AChRs. See Nicotinic acetylcholine receptor Actin assembly.................................................................... 100 microfilaments............................................................ 95 motility of organelles.................................................. 95 Actin polymerization SDS-PAGE...................................................97, 98, 100 Western blotting............................................97, 98, 101 Affinity chromatography................. 46, 70, 87, 88, 300, 301 Alpha-emitter................................................................. 470 Amcon® ultracentrifuge filter........................................... 85 Amino acid transport................................................. 55–67 Ammonium sulfate gradient liposomes..........473, 479–480, 485–488 Amyloid protein............................................................... 70 Anesthetics general anesthetics............................................ 295, 311 local anesthetics................. 304, 305, 309, 311, 312, 314 Annexin............................................................................ 69 Antibodies.....................................29, 92, 101, 102, 108, 111, 135, 352, 458, 460, 462–467 Anti-COX antibody......................................................... 87 Antioxidant lipophilic antioxidant................................................ 170 water-soluble antioxidant................................. 170, 172 ANTS/DPX assay...................................213–214, 222–223 Aquaporin...................................................................... 352 ASCT2. See Glutamine/amino acid (ASCT2) transporter Atomic force microscopy (AFM)...........................182, 320, 351–360, 521, 525 Autoradiography............................................................. 136 Autoxidation................................................................... 169 Avanti Mini Extruder System.................................. 23, 511 Avanti Polar Lipids, Inc........................ 2, 15, 37, 70, 71, 81, 85, 97, 102, 117, 137, 210, 211, 213–215, 235, 252, 274, 297, 298, 322, 353, 401–403, 427, 460, 472, 473, 499, 510, 511, 523, 524, 532 Avidin.............................................. 521–525, 533, 539, 540

B Baculovirus................................................ 43, 84, 87, 88, 94 BCA protein assay............................... 86, 99, 297, 301, 353 Beef heart muscle............................................152–154, 156 Benzocaine..............................................298, 304, 305, 311

Beta-emitter................................................................... 470 Betaine transport........................................................ 22–27 BIAcore® system............................................................ 520 BioBeads SM-2, 15–17 Bioneb® cell disruption system................................... 84, 89 Biosensor................................................................ 519–528 Biosynthesis............................................................ 127–144 Biotin biotinylated amino acids........................................... 143 biotinylated liposomes.......................521, 522, 526, 527 BODIPY-PC..................................................386, 387, 390 Bradford assay............................................................ 40–42 Brush border membrane........................................57, 58, 61

C Calcein calcein leakage............................................................ 74 Capacitation..................................................................... 70 Capillary viscometer.........370, 374, 376, 377, 380, 382, 383 Carboxyfluorescein carboxyfluorescein-PE............... 426, 427, 430, 434, 436 Caveosomes.................................................................... 458 Cdc42....................................................................... 96, 100 Cell-free protein expression kit.............................. 135–137 Chelator hexamethylpropyleneamine oxime............................ 472 N,N-bis(2-mercaptoethyl)-N’, N’-diethylethylenediamine................................. 472 Chemodosimetry............................................................ 470 Cholate dialysis........................................................ 14, 156 Cholesterol cholesteryl-oleate...................................................... 2, 5 cholesteryl-oleoyl-ether.....................................2, 5, 7, 8 Circular dichroism...................................235–236, 240, 241 Clathrin................................... 458, 465, 531–534, 536, 540 Clathrin-coated pits................................533, 537–538, 540 Cluster aggregation diffusion-limited aggregation....................193–195, 197 reaction-limited aggregation......................193, 195, 197 C6-NBD-PC..................................................119, 121–124 Coat protein complex..................................................... 531 Coenzyme Q.......................................................... 148–150 Colloids aggregation....................................................... 189–197 stability......................................................191, 192, 375 Colocalization studies............................................. 458, 459

543

Liposomes 544 Index

Concave fracture............................................................. 343 Convex fracture.............................................................. 338 Corynebacterium glutamicum.......................................... 22 Critical micelle concentration (CMC)..........13, 17, 18, 235 Cryofixation........................................................... 334, 336 Cyclooxygenase (COX).............................................. 83–94 Cyclosporin A............................................................ 3, 7, 8 Cys-loop receptors.................................................. 292, 293 Cytochomes b562...........................................................148–150, 157 b566.................................................................. 148–150 bc1 complex................................................147–157, 162 cytochrome C............................................. 17, 148–162 energy transduction.......................................... 147, 149 eukaryotic cytochromes............................................ 147 respiratory cytochromes.................................... 147–163 Cytochrome P450 monooxygenase............................ 11–19 Cytosol.......................................... 96, 98–100, 150, 532–538

D DC30®lipid.................................................................... 461 Debye length.......................................................... 192, 196 Dehydration–rehydration method........... 37, 39, 46, 47, 324 Detergent solubilization................................13, 16, 18, 234 Diagnostic imaging................................................ 469–488 Dialkylcarbocyanines...................................................... 506 4,6-Diamidino-2-phenylindole (DAPI).................444, 449, 454, 460, 462, 464, 466, 533, 539, 540 Dicetylphosphate (DCP)........................................... 2, 3, 8 Diene formation............................................................. 169 Differential free energy of association............................ 314 Differential scanning calorimetry (DSC)..........77, 406, 417 Diffusion coefficient........................ 256, 494–497, 503, 506 1,2-Dioleoyl-sn-glycero-3-phosphocholine (DOPC)............38, 72, 73, 85, 89, 93, 121, 123–125, 132, 134, 235, 238, 240–242, 311, 315, 322, 402 1,2-Dioleoyl-sn-glycero-3-[phospho-L-serine] (DOPS).................................... 85, 89, 93, 121, 124, 125, 402, 414–416 Diphenylhexatriene propionic acid (DPH-PA)........... 170, 171, 174–175, 179–181, 184 Discrimination by oxygen transport............................... 249 DLVO model......................................................... 192, 193 DNA binding assay.................................................... 395–397 from calf thymus....................................................... 386 from Herring.............................................386, 404, 419 labeling......................................................386, 394, 397 from salmon.............................................................. 386 DOT method. See Discrimination by oxygen transport Double labeling...............................................203, 205, 444 Doxil®..................................................................... 477, 478 Drug metabolism........................................................... 11–19

transfer...................................................................... 1–9 uptake...........................................................................1–9 Drug carrier system............................................................ 1 Drug/lipid ratio.............................................................. 364 Dual asymmetric centrifugation (DAC)..........442, 443, 445, 446 Dynamic light scattering................... .... 162, 218–220, 286, 355, 404, 407, 511, 515, 531, 535

E Egg phosphatidylcholine (EPC)........................2, 3, 8, 153, 173, 175, 178–184, 215, 322, 443, 446, 453, 510, 514 Ejaculation....................................................................... 70 Electroformation..............105–113, 119, 120, 501, 502, 507 Electron paramagnetic resonance spectroscopy...... 291–316 Electron spin-lattice relaxation time....................... 254, 260 Electron transfer..................................................... 147–152 Electrophoretic mobility......................................... 162, 196 Electrophorus species..................................................... 292 Electroswelling............................................................... 130 EmrE transport protein...................................233, 235–242 Endocytosis....................................... 69, 199, 225, 226, 425, 426, 436, 465, 531, 540 Endolysosomal compartment......................................... 425 Endoplasmic reticulum (ER)...............................12, 15, 458 Environmental scanning electron microscopy (ESEM)............320–321, 323, 325–329 Epithelial cells.................................................................... 1 EPR spectroscopy................................................... 291–316 EPR spin labeling....................................247–267, 297–298 Erythrocyte ghosts inside out (IO) ghosts........................106, 109, 110, 113 right-side out (RSO) ghosts..............106, 108–110, 113 Escherichia coli.............................................................32, 234 Ethanol injection method................................130, 132, 134 Europium............................................................... 509–518 Extruder............................................... 4, 23, 71, 80, 85, 89, 137–139, 144, 158, 173, 202, 215, 218, 219, 238, 286, 355, 427, 430, 473, 479, 480, 510, 511, 524

F Flotation assay............................................................ 71–73 Fluorescence anisotropy..........................................172, 182, 183, 209 quenching..............................74, 79, 159, 172, 181, 183, 223, 230, 390, 396 Fluorescence correlation spectroscopy (FCS).........201, 202, 204–206, 442, 443, 449, 459–461, 464, 493–507 Förster radius.......................................................... 396–398 Fourier-transform infrared spectroscopy (FTIR)............. 77 Fractal dimension............................................193–194, 197 Free energy........................................................84, 189, 314 Freeze-fracture electron microscopy (FFEM)........ 333–347

Liposomes 545 Index

FRET-labelled siRNA............................................ 440, 442 FRET spectroscopy.......................................................... 38 FuGene®................................................................ 460, 464 Fusion................................................33, 34, 38–41, 45, 116, 151, 159, 160, 162, 185, 190, 199, 209, 212, 214, 220–225, 229–231, 394, 397, 420, 521, 522, 527

G Gallium-67..................................................................... 470 Gel-liquid crystal transition temperature....................... 324 Giant unilamellar vesicles (GUV)........................47, 49, 50, 105–113, 115–126, 493–507 Glutamine/amino acid (ASCT2) transporter........56, 57, 60 Glutathione sepharose................................................ 34, 40 Gold labeling........................................................ 87, 92–94 Golgi apparatus (isolated membranes)..................... 96, 458 Green fluorescent protein (GFP).....................50, 129, 132, 133, 135, 136, 426, 427, 435 Gut..................................................................................... 1

H Halobacterium halobium................................................ 346 Hamaker constant.......................................................... 192 Heat capacity curve........................................................ 406 Hepatitis B surface antigen............................................ 322 Hexagonal phase.............................. 400, 401, 405, 411–414 High-pressure extrusion......................................... 210, 211 His-tag....................... 34, 36, 38–41, 43, 45, 84, 87, 88, 134 HPTLC/FID......................................................... 363–367

I Ibuprofen.........................................................322, 323, 326 Immunogold labeling................................................. 88, 93 Immunostaining..............................................108, 111, 112 Indium-111, 470 Inner monolayer mixing assay................................ 214, 224 Intracellular delivery........................210, 215–216, 225–227 Intracellular trafficking................................................... 457 Intrinsic protein fluorescence..................................... 72–74 Iodine-123...................................................................... 470 Iron sulfur protein...........................................148, 149, 157 Isothermal titration calorimetry......................233–244, 271 ITC. See Isothermal titration calorimetry ITO slides......................................... 49, 116, 118–120, 125

L Labeling efficiency..................................394–397, 481–488 Lamellar-nonlamellar phase transition................... 400, 416 Lamellar phase........................................400, 405, 413, 416 Lanthanide-chelate........................................................ 510 Large unilamellar liposomes...................210–211, 216–218, 220, 222, 224, 225, 227–229, 533–535 Large unilamellar vesicle (LUV)..........................71–76, 78, 80, 81, 132, 176–179, 181–183, 185, 286, 523

Laser-Doppler microelectrophoresis.............................. 162 Lens lipid................................ 251, 254, 256, 258–263, 266 Ligand-gated ion channels............................................. 292 Lipex Biomembranes, Inc......................................... 71, 173 Lipid bilayer.............................................................13, 17, 31, 32, 74, 83, 84, 152, 158–160, 170–172, 181, 201, 209, 234, 236, 247, 250–252, 254, 258, 259, 261, 264, 266, 273, 323, 334, 345, 352, 354, 357, 360, 363, 379, 393, 397, 400, 407, 408, 417, 471, 472, 474–476, 479 domain..............................................105–113, 115–126, 295, 311, 343–346 exchange........................................................... 199, 420 mixing assay...............................................230, 395, 397 organization...................................................... 247–267 raft............................................................................ 343 Lipid–protein interface............................. 32, 293, 295, 306, 308–312, 314, 315 Lipid-stabilized gas bubbles........................................... 339 Lipofectamine.................................................................. 43 Lipophilic drug................................................................... 2 Lipoplex cell binding....................................................... 428, 433 endocytosis....................................................... 425, 426 internalization................................................... 403, 425 intracellular distribution...........................426, 428–430, 433–434, 457–467 preparation................................. 276, 385, 395, 404–405 LiposoFast Extruder™................................................... 215 Liposomal binding behavior................................... 519–528 Liposomal membrane............................. 2, 7, 56, 57, 96, 97, 99, 157, 352, 354, 426 Liposome cell binding............................................................... 203 cell interaction...........................................199–207, 520 deformation...................................................... 521–523 fusion.........................................................209–231, 521 intracellular homing.................................................. 199 uptake........................................................200, 205, 210 Liquid-disordered domain.............................................. 116 Liquid disordered phase.................. 123, 124, 250, 252, 493 Liquid ordered domain........................................... 112, 115 Liquid scintillation counting......................................... 6, 7, 203, 204 Liver endothelial cells........................................................ 200 Low-density lipoprotein (LDL)..................................... 168 LysoTracker Red.....................................426, 428, 433, 434

M Macropinocytosis........................................................... 465 Madin–Darby canine kidney cells (MDCK II)............. 458, 459, 461, 463, 465, 466

Liposomes 546 Index

Malonaldehyde............................................................... 169 Mechanosensitive ion channel.................................... 31–52 Mechanosensitivity........................................................... 32 Mechanotransduction....................................................... 32 Membrane anisotropy..................................................172, 182, 183 asymmetry assay............................................... 108, 111 dynamic..................................... 251, 257, 259, 493–507 fluidity.......................................................259–260, 495 fusion......................................... 210, 229–231, 397, 521 integrity......................................... 70, 74, 111, 112, 125 translocation..................................................... 271–288 Membrane protein integral membrane protein...........................83–94, 116, 234, 244, 250, 314 peripheral membrane protein................................... 115 Membranes blebbing............................................................ 106, 207 lateral structure................................................. 105, 352 mechanical properties............................................... 105 Mesomorphic phase............................................... 400, 414 Microinjection.................................442–445, 448–449, 454 Microsomal lipid.............................................................. 12 Mini column method..................................................... 6, 9 Minimal genome............................................................ 128 Monocytic leukemia cells....................................... 214, 224 Monotopic integral membrane protein...................... 83–94 Multidrug transport protein................................... 233–244 Multilamellar lipoplexes................................................. 390 Multilamellar liposomes.................................176, 210–211, 218–219, 224, 225, 229, 253, 264, 430, 524 Multilamellar vesicle (MLV)..................................... 71, 78, 130–133, 176, 185, 276, 277, 285, 286, 319, 323–326, 329, 340, 355, 524

N NADPH cytochrome P450 reductase.............................. 12 Nanoparticular drug delivery system.............................. 465 Nano-resolution scale............................................. 333–347 NBD-PE/Rhodamine-PE assay.............223, 394, 395, 406 n-Dodecyl-b-D-maltopyranoside (DDM)............... 23, 25, 34, 35, 38, 40, 42, 47, 51, 235–237, 242 Nephelometry......................................................... 190, 192 Neurotransmitter.................................................... 292, 293 Nicotinic acetylcholine receptor (AChRs).............. 291–316 Nigericin................................................................. 152, 157 Niosomes.................................................320–322, 324–328 N-NBD-DPPE.......................................................... 72, 73 Non-ionic vesicles.......................................................... 328 Nonlamellar phase...................................400, 403, 414, 416 Nonsteroidal anti-inflammatory drugs (NSAID)............. 88 Nuclear medicine............................................................ 470 Nucleopore® polycarbonate membrane.................. 238, 286

O Octadecyl-DTPA europium............................512, 514, 515 Octyl-beta-D-gluco-pyranoside............................. 353, 354 Oligofectamin.................................................442, 443, 448 Osmolality osmoregulation........................................................... 21 osmosensor........................................................... 21–30 osmotic stress...........................................21, 22, 24, 135 O-substituted phosphatidylcholine................................. 401 Oxygen permeation........................................................ 247 Oxygen transport domain (Fast, FOT; Slow, SLOT).......................................257, 266, 267

P 1-Palmitoyl-2-(4-doxylpentanoyl)-sn-glycero3-phosphocholine (SL-PC)............................ 71, 76 Parkinson’s disease.................................................... 70, 116 Partition coefficient........................... 13, 172, 180–181, 305 Passive diffusion................................................................. 1 Patch clamp...............................................32, 37–38, 49–51 PDC-109........................................................70–77, 79, 80 Penetratin................................ 273, 280, 282–284, 288, 471 Peroxide formation................................................. 169, 184 Peroxyl radical.................................................170–172, 184 Phagocytic assay..............................................533, 538–540 Pharmacokinetics............................................440, 478, 509 Phase transition.......................................... 24, 94, 110, 125, 134, 184, 219, 253, 285, 337, 400, 403, 404, 406, 408–410, 413, 416–418, 478–480 Phenylmethanesulphonylfluoride (PMSF)................ 34, 35, 40, 42, 96, 97, 162, 315, 532, 537 Phosphate assay............................... 202, 396, 403, 404, 419 Phospholipid assay......................................................... 202 Phosphor imager............................................................ 135 Photobleaching................................ 112, 287, 464, 504, 507 Photodynamic therapy (PDT)........................................ 150 Piezoelectric sensor........................................................ 520 Planar lipid bilayers........................................................ 352 Procaine...........................................298, 304–308, 310–313 Propidium iodide.............................................215, 225, 226 Propofol...........................................................322, 323, 327 Prostaglandin endoperoxide H2 synthase.................... 83, 94 Protein protein expression......................... 33, 35–36, 43, 50, 88, 129, 130, 133–137, 237 protein quantification.................... 86, 90, 236, 242–243 protein sorting.................................................. 115–126 Protein-expressing liposomes......................................... 129 Protein-membrane interaction............................78, 79, 119 Proteoliposomes....................................... 22, 25–30, 47, 48, 57, 60, 62–67, 85–88, 90, 92–94, 149, 156, 157, 238, 337, 345, 351–360 P-selectin.................................................352–354, 356–358

Liposomes 547 Index

Puresystem®........................................... 130, 133, 134, 136, 138, 139, 143

Q Qiagen Inc........................................................................ 85 Quartz crystal microbalance (QCM)..................... 520–522, 526–528

R Radicals hydroxyl radicals....................................................... 167 peroxyl radicals..........................................170–172, 184 radical inducer.................................................. 173, 174 Radioactive liposome marker.......................................... 202 Radiolabeling.................................. 131, 132, 135, 140, 143, 470–472, 475, 478–488 Radiolabelled gel.................................................... 135, 142 Radionuclide therapy.............................................. 469–488 Radiopharmaceutical...................................................... 487 Reactive nitrogen species (RNS).................................... 167 Reactive oxygen species (ROS)....................................... 167 Real-time fluorescence analysis.............................. 139–140 Reconstitution actin polymerization........................................... 95–102 cell-free reconstitution................................................ 96 Resonance energy transfer............ 50, 74–76, 230, 393–398, 406–407, 440 Reverse phase evaporation.......................210, 217–218, 229 Rhenium-186/188................................................. 469–488 Riboosomes.............................................128, 129, 134, 143 Rotary shadowing........................................................... 340

S Saturation-recovery EPR................. 250, 251, 254–257, 265 Scavenger................................................................ 171, 172 Scintigraphy.................................................................... 469 Seminal plasma protein.................................................... 69 Semiquinone................................................................... 149 Sertoli cells..............................................531–534, 539, 541 Signal transduction................................................. 115, 520 Single photon emission computed tomography (SPECT).........................470, 471, 477 siRNA siRNA duplex............................ 439, 440, 442, 444–445 siRNA/liposome complexes.............................. 439–454 Small-angle X-ray diffraction..................400, 405, 408–410 Small unilamellar liposomes...................................211, 216, 219–220, 222, 227–229 Smoluchoswski equation................................................ 196 Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE).................. 36–37, 44–46, 70, 72, 73, 78, 97, 98, 100–101, 131, 132, 135–143, 315

Sonication method............................................................. 6 SpectraCube™ SD-200H.......................459, 461, 464, 467 Spectral bioimaging................................................ 457–467 Sperm cells....................................................................... 70 Spin label.................................... 2, 71, 76–77, 81, 247–267, 297, 298, 302–316 Spin-labeled lipids................... 71, 76–77, 81, 305, 313, 314 Spodoptera frugiperda insect cells (SF21)............35, 84–85, 88–89 Steady-state fluorescence spectroscopy................... 182, 183 Stern layer...................................................................... 193 Stern–Volmer equation................................................... 181 Streptavidin-linked enzymes.......................................... 143 Streptomycetes nodosus................................................. 344 Styrene beads...................................................532, 534, 539 Submitochondrial particle.............................................. 149 Succinate-cytochrome c reductase.......................... 152–154 Succinate dehydrogenase........................................ 154–156 Sulfo-NHS............................................................. 353, 356 Surface acoustic wave (SAW) sensor.............................. 520 Surface modification....................................................... 351 Surface plasmon resonance (SPR).................................. 520 Surfactant vesicles.......................................................... 328 Synchrotron radiation......................................237, 240, 420 Synthetic biology.................................................... 129, 134 Synuclein........................................... 70, 116, 118, 122–124

T Tb/DPA assay........................................................ 212–213 TCA precipitation.......................................................... 143 Technetium-99m.................................................... 469–488 Tetracaine................................................298, 305, 311, 313 Tetramethylrhodamine............................118, 122, 444, 453 Tetraphenyphosphonium bromide................................. 236 Thermoplasma acidophilum........................................... 343 Titanium probe.................................................................. 5 Torbedo species.............................................................. 292 Transbilayer movement.................................................. 230 Transcription–translation machinery...............128–130, 138 Transferrin................................ 44, 341, 353–356, 358–359, 419, 458, 462, 463 Transgene expression...................................................... 425 Transmembrane peptides.................................118, 119, 121 Transmission electron microscopy...................87, 88, 92–94 Trapping efficiency......................................................... 445 tRNA...............................................................128, 129, 143 Trolox..........................................................170, 173–179, 184 Tubulin................................................................... 101, 426 Turbidimetry...................................................151, 160–161 Turgor pressure................................................................. 22

U Ubiquinone-binding protein.................................. 155, 156 Ultrasonic disintegrator...................................................... 5

Liposomes 548 Index

Unilamellar liposomes.......................... 85, 88, 89, 210–211, 216–220, 222, 224, 225, 227–229, 264, 524, 533–535

V Valinomycin................................. 23, 26, 149, 152, 156–158 Van der Waals forces............................................... 189, 192 Vesicles multilamellar vesicles............................ 71, 78, 130, 132, 133, 276, 319, 329, 340, 355, 524 oligolamellar vesicles................................................ 130 unilamellar vesicles.........................................71, 78, 99, 105–113, 115–126, 132, 133, 176, 277, 286, 352, 493–507, 523 vesicle morphology........................................... 129, 133 vesicle size.................................................133, 286, 355 Vesicular phospholipid gel (VPG).......................... 445, 446 Vesicular reconstituted system (VRS)......................... 13–19 Vesicular transport.................................................. 531–541 Vinblastine...................................... 426, 428, 429, 433–435

Virus budding................................................................. 115 Viscometric analysis............................................... 369–383 Viscosity non-dimensional reduced viscosity................... 371, 375 relative viscosity.........................................371, 372, 383 specific viscosity................................................ 371, 377 Vitamin E....................................................................... 184

W Wiscott–Aldrich syndrome protein (WASP)................... 96

X X-ray diffraction.............. 400, 404–406, 408–410, 413, 419

Z Zero shear viscosity.........................................373, 377–378 Zeta-potential.............................................5, 162, 172, 183 Z-scan.....................................................494, 497, 504–505 Zwittergent............................................3–14, 386, 387, 396